Mathematical Modeling for

Intelligent Systems
Mathematical Modeling for Intelligent Systems: Theory, Methods, and Simulation aims
to provide a reference for the applications of mathematical modeling using intelligent tech-
niques in various unique industry problems in the era of Industry 4.0. Providing a thor-
ough introduction to the field of soft-computing techniques, this book covers every major
technique in artificial intelligence in a clear and practical style. It also highlights current
research and applications, addresses issues encountered in the development of applied sys-
tems, and describes a wide range of intelligent systems techniques, including neural net-
works, fuzzy logic, evolutionary strategy, and genetic algorithms. This book demonstrates
concepts through simulation examples and practical experimental results.
Key Features:

• Offers a well-balanced mathematical analysis of modeling physical systems

• Summarizes basic principles in differential geometry and convex analysis as
needed
• Covers a wide range of industrial and social applications and bridges the gap
between core theory and costly experiments through simulations and modeling
• Focuses on manifold ranging from stability of fluid flows, nanofluids, drug deliv-
ery, and security of image data to pandemic modeling, etc.

This book is primarily aimed at advanced undergraduates and postgraduate students

studying computer science, mathematics, and statistics. Researchers and professionals
will also find this book useful.
Mathematical Modeling for
Intelligent Systems
Theory, Methods, and Simulation

Edited by
Mukesh Kumar Awasthi
Babasaheb Bhimrao Ambedkar University, India
Ravi Tomar
University of Petroleum & Energy Studies, India
Maanak Gupta
Tennessee Tech University, USA
First edition published 2023
by CRC Press
6000 Broken Sound Parkway NW, Suite 300, Boca Raton, FL 33487-2742

and by CRC Press

2 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN

© 2023 selection and editorial matter Mukesh Kumar Awasthi, Ravi Tomar and Maanak Gupta; individual chapters,
the contributors

CRC Press is an imprint of Taylor & Francis Group, LLC

Reasonable efforts have been made to publish reliable data and information, but the author and publisher cannot
assume responsibility for the validity of all materials or the consequences of their use. The authors and publishers
have attempted to trace the copyright holders of all material reproduced in this publication and apologize to copyright
holders if permission to publish in this form has not been obtained. If any copyright material has not been acknowledged
please write and let us know so we may rectify in any future reprint.

Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmitted, or
utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including
photocopying, microfilming, and recording, or in any information storage or retrieval system, without written
permission from the publishers.

For permission to photocopy or use material electronically from this work, access www.copyright.com or contact the
Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. For works that are
not available on CCC please contact [email protected]

Trademark notice: Product or corporate names may be trademarks or registered trademarks and are used only for
identification and explanation without intent to infringe.

ISBN: 9781032272252 (hbk)

ISBN: 9781032272269 (pbk)
ISBN: 9781003291916 (ebk)

DOI: 10.1201/9781003291916

Typeset in Palatino
by codeMantra
Contents

Preface ............................................................................................................................................ vii

Editors ..............................................................................................................................................ix
Contributors ....................................................................................................................................xi

1 Sensors, Embedded Systems, and IoT Components........................................................1

Ananya Sharma and Niharika Singh

2 Security and Privacy in Mobile Cloud Computing ....................................................... 17

A. Katal

3 A Review on Live Memory Acquisition Approaches for Digital Forensics ............. 35

Sarishma Dangi and Diksha Bisht

4 IoMT and Blockchain-Based Intelligent and Secured System for

Smart Healthcare .................................................................................................................. 61
Hitesh Kumar Sharma, Ravi Tomar, and Preeti

5 AI-Enabled Cloud-Based Intelligent System for Telemedicine .................................. 75

Hitesh Kumar Sharma, Ravi Tomar, Preeti, and Prashant Ahlawat

6 Fuzzy Heptagonal DEMATEL Technique and Its Application .................................. 85

A. Felix, PP. Ajeesh, S. Karthik, and R. Dinesh Jackson Samuel

7 A Comparative Study of Intrapersonal and Interpersonal Influencing Factors

on the Academic Performance of Technical and Nontechnical Students ............... 101
Deepti Sharma, Rishi Asthana, and Vaishali Sharma

8 PIN Solar Cell Characteristics: Fundamental Physics ................................................ 119

Aditya N. Roy Choudhury

9 A New Approximation for Conformable Time-Fractional Nonlinear

Delayed Differential Equations via Two Efficient Methods ..................................... 133
Brajesh Kumar Singh and Saloni Agrawal

10 Numerical Treatment on the Convective Instability in a Jeffrey Fluid

Soaked Permeable Layer with Through-Flow .............................................................. 159
Dhananjay Yadav, Mukesh Kumar Awasthi, U. S. Mahabaleshwar, and
Krishnendu Bhattacharyya

11 Computational Modeling of Nonlinear Reaction-Diffusion Fisher–KPP

Equation with Mixed Modal Discontinuous Galerkin Scheme ............................... 171
Satyvir Singh

v
vi Contents

12 A Numerical Approach on Unsteady Mixed Convection Flow with

Temperature-Dependent Variable Prandtl Number and Viscosity.......................... 185
Govindaraj N, Iyyappan G, A. K. Singh, S. Roy, and P. Shukla

13 Study of Incompressible Viscous Flow Due to a Stretchable Rotating

Disk Through Finite Element Procedure....................................................................... 197
Anupam Bhandari

14 Instability of a Viscoelastic Cylindrical Jet: The VCVPF Theory............................. 215

Mukesh Kumar Awasthi, Sudhir Kumar Pundir, Manu Devi,
Dhananjay Yadav, Vivek Kumar, and A. K. Singh

15 Evaporative Kelvin–Helmholtz Instability of a Porous Swirling

Annular Layer...................................................................................................................... 231
Shivam Agarwal and Mukesh Kumar Awasthi

Index.............................................................................................................................................. 245
Preface

The modeling of various engineering problems using mathematics is an important tech-

nique that is being used by many engineers and scientists around the globe. The real-life
problems can also be modeled through mathematics, and by using analytical/numerical
methods, one can solve these mathematical models. Bio-mathematical engineering can be
used for modeling various diseases and the impact of different parameters involved. The
security system can also be handled through mathematical techniques. Encryption and
decryption play a vital role in the security system. The use of number theory and cryp-
tography makes the study easier. The exact solution of the Navier–Stokes equation is not
known to date and, therefore, the numerical methods are very helpful for the computation
of various properties of fluid flows.
In recent years, scientists have been trying to solve practical problems using mathemati-
cal models and they are also getting success. As computing technologies have grown tre-
mendously, the possibilities of solving complex problems have also increased. Most of the
engineering systems are very complex, and it takes a lot of time and money to find their
solutions. To design an engineering system and understand its processes, it is very impor-
tant to understand and analyze all the parameters used in that system and mathematical
models are very helpful in understanding and analyzing such parameters.
This book is intended as reference material for educators, researchers, and everyone who
wishes to learn and have a grasp of the various advances in the fields of computer science,
applied mathematics, computing, and simulation. This book will be a readily accessible
source of information for the researchers in the area of applied computing as well as for
the professionals who want to enhance their understanding of theoretical and application
studies. This book is also beneficial for the researchers and professionals who are involved
in the research on Intelligent Systems. This book is written for the students, professionals,
researchers, and developers who are involved or have a keen interest in the domain of
computational intelligence. One of the unique features of this book is it covers the various
case studies and future directions in the domain of mathematical modeling.
This book is intended to be multidomain applications, which will help researchers and
scholars to think beyond one specific field and work on cohort domain in mathematics and
computer science. This book can be an idea for an audience working in the domain of model-
ing and simulations. It will present the reader with comprehensive insights about the vari-
ous kinds of mathematical modeling and numerical computations for the problems arising
in various branches of engineering starting from computer Science engineering, mechanical
engineering, Electrical, and Electronic Engineering, Civil Engineering, etc.
This book contains 16 chapters, each of which presents commonly observed topics and
problems which somewhere deep inside are solved through mathematical computation
and usage of the latest computational intelligence. This book starts with discussing prac-
tical aspects and components of Intelligent systems and followed with the theoretical
approaches to solve other day-to-day problems.
Chapter 1 is about Sensors, Embedded Systems, and IoT Components as the title suggests,
the chapter presents a comprehensive understanding of the components of IoT and smart
computing. Chapter 2 discuss the very critical aspect of Security and Privacy in Mobile
Cloud Computing, followed by Chapter 3 about a Review of Live Memory Acquisition
Approaches for Digital Forensics. The chapter discusses a practical approach around

vii
viii Preface

acquiring memory dump for different memory acquisition tools. Further, Chapter 4 is dedi-
cated to discussing the need and applicability of the latest technologies like IoMT with
Blockchain-Based Intelligent and Secured System for Smart Healthcare. Chapter 5 com-
plements it by the implementation of an AI-Enabled Cloud-Based Intelligent System for
Telemedicine.
The expansion of The DEMATEL method under the uncertain linguistic situation with
heptagonal fuzzy numbers is given in Chapter 6, while the academic performance of techni-
cal and nontechnical persons is discussed in Chapter 7. Chapter 8 presents the fundamental
aspects of the modeling of a PIN solar cell. Chapter 9 proposes the solution of conformable
time-fractional nonlinear partial differential equations with proportional delay.
Chapters 10–15 are based on the fluid flows in various geometries. Chapter 10 is about
the convective instability in a Jeffrey fluid-soaked permeable layer and Chapter 11 presents
a numerical solution of the two-dimensional nonlinear Fisher–Kolmogorov–Petrovsky–
Piscounov reaction–diffusion equation. The impact of transient fluid flow on a vertical
plate in the laminar boundary layer with variable viscosity Prandtl number is given in
Chapter 12. Chapter 13 complements the flow of viscous fluid due to a rotating and stretch-
ing disk. Chapters 14 and 15 discuss the interfacial instability in cylindrical geometries.
We wish the readers a successful study of the material presented, leading to new inspira-
tion, a deepening understanding of the described concepts, and also fruitful applications
to the contemporary challenges of science and engineering.

Dr. Mukesh Kumar Awasthi

Dr. Ravi Tomar
Dr. Maanak Gupta
Editors

Dr. Mukesh Kumar Awasthi is working as an Assistant

Professor in the Department of Mathematics at Babasaheb
Bhimrao Ambedkar University (A Central University),
Lucknow, India. He completed his doctorate with a major
in Mathematics from the Indian Institute of Technology
Roorkee, Roorkee, India in 2012. He has qualified for
various national-level competitive exams in the area of
Mathematical Sciences. He has been a recipient of a research
fellowship under the Council of Scientific and Industrial
Research, India during his Ph.D. His research areas include
Computational Fluid Dynamics, Heat transfer, Applied
Mathematical Modeling, and Numerical analysis. He loves
to teach applied mathematics courses to undergraduate,
postgraduate, and Ph.D. students. He has been indifferent to
academic and administrative positions during his academic profession. He has more than
100 publications (90 SCI indexed, 92 Scopus indexed) to his credit in the high impact factor
journals of international repute. He has attended many symposia, workshops, and confer-
ences in mathematics as well as fluid mechanics. He has got research awards consecutively
four times from the University of Petroleum & Energy Studies, Dehradun, India. He has
also received the start-up research fund for his project “Nonlinear study of the interface in
multilayer fluid system” from UGC, New Delhi.

Dr. Ravi Tomar is currently working as Associate Professor

in the School of Computer Science at the University of
Petroleum & Energy Studies, Dehradun, India. Skilled in
Programming, Computer Networking, Stream processing,
Core Java, J2EE, RPA and CorDapp, his research interests
include Wireless Sensor Networks, Image Processing, Data
Mining and Warehousing, Computer Networks, big data
technologies, and VANET. He has authored more than 51
papers in different research areas, filled four Indian patent,
edited 5 books, and have authored 5 books. He has deliv-
ered Training to corporates nationally and internationally
on Confluent Apache Kafka, Stream Processing, RPA, CorDapp, J2EE, and IoT to clients like
KeyBank, Accenture, Union Bank of Philippines, Ernst & Young, and Deloitte. Dr. Tomar
is officially recognized as Instructor for Confluent and CorDapp. He has conducted vari-
ous International conferences in India, France, and Nepal. He has been awarded a young
researcher in Computer Science and Engineering by RedInno, India in 2018, Academic
Excellence and Research Excellence Award by UPES in 2021 and Young Scientist Award
by UCOST, Dehradun.

ix
x Editors

Dr. Maanak Gupta is an Assistant Professor in the

Department of Computer Science at Tennessee Tech
University, United States. He received his Ph.D. in Computer
Science from the University of Texas at San Antonio and has
worked as a Postdoctoral Research Fellow at the Institute for
Cyber Security. He also holds an M.S. degree in Information
Systems from Northeastern University, Boston. His primary
area of research includes security and privacy in cyber-
space focused on studying foundational aspects of access
control and their application in technologies including
cyber-physical systems, cloud computing, IoT, and Big Data.
Dr.  Gupta has worked in developing novel security mechanisms, models, and architec-
tures for next-generation smart cars, smart cities, intelligent transportation systems, and
smart farming. He is also interested in machine learning–based malware analysis and
AI-assisted cyber security solutions. His scholarly work is regularly published at top peer-
reviewed security venues including ACM SIGSAC conferences and refereed journals. He
was awarded the 2019 computer science outstanding doctoral dissertation research award
from UT San Antonio. His research has been funded by the US National Science Foundation
(NSF), NASA, the US Department of Defense (DoD), and private industry.
Contributors

Saloni Agrawal Diksha Bisht

Department of Mathematics Department of Computer Science and
Babasaheb Bhimrao Ambedkar University Engineering
Lucknow, India Graphic Era Deemed to be University
Dehradun, India
Shivam Agarwal
Department of Mathematics Sarishma Dangi
Babasaheb Bhimrao Ambedkar University Department of Computer Science and
Lucknow, India Engineering
Graphic Era Deemed to be University
Dehradun, India
Prashant Ahlawat
Department of Computer Science Manu Devi
GL Bajaj Institute Department of Mathematics
Greater Noida, India Motherhood University
Roorkee, India
PP. Ajeesh
Mathematics Division A. Felix
School of Advanced Sciences Mathematics Division
Vellore Institute of Technology School of Advanced Sciences
Vellore, India Vellore Institute of Technology
Vellore, India
Rishi Asthana
School of Engineering and Technology Govindaraj N
BML Munjal University Department of Mathematics
Gurgaon, India Hindustan Institute of Technology and
Science
Mukesh Kumar Awasthi Chennai, India
Department of Mathematics
Babasaheb Bhimrao Ambedkar University Iyyappan G
Lucknow, India Department of Mathematics
Hindustan Institute of Technology and
Anupam Bhandari Science
Department of Mathematics Chennai, India
School of Engineering
University of Petroleum & Energy Studies S. Karthik
Dehradun, India Department of Mathematics
Vel Tech Rangarajan Dr. Sagunthala R&D
Krishnendu Bhattacharyya Institute of Science and Technology
Department of Mathematics Chennai, India
Institute of Science
Banaras Hindu University
Varanasi, India

xi
xii Contributors

Avita Katal Deepti Sharma

School of Computer Science School of Management
University of Petroleum & Energy Studies BML Munjal University
Dehradun, India Gurgaon, India

Vivek Kumar Vaishali Sharma

Department of Mathematics School of Management
Shri Guru Ram Rai (PG) College BML Munjal University
Dehradun, India Gurgaon, India
U. S. Mahabaleshwar
Department of Mathematics P. Shukla
Davangere University Mathematics Division
Davangere, India School of Advanced Science
VIT University
Preeti Chennai, India
School of Computer Science and
Engineering Niharika Singh
University of Petroleum & Energy Studies Department of Computer Science
Dehradun, India University of Petroleum & Energy Studies
Dehradun, India
Sudhir Kumar Pundir
Department of Mathematics Brajesh Kumar Singh
S. D. (P.G.) College Department of Mathematics
Muzaffarnagar, India Babasaheb Bhimrao Ambedkar University
Aditya N. Roy Choudhury Lucknow, India
Department of Physics
Techno India University Satyvir Singh
Kolkata, India School of Physical and Mathematical
Sciences
S. Roy Nanyang Technological University
Department of Mathematics Singapore
Indian Institute of Technology Madras
Chennai, India A. K. Singh
Mathematics Division
R. Dinesh Jackson Samuel School of Advanced Science
Department of Mathematics VIT University
Oxford Brookes University Chennai, India
Oxford, UK

Ananya Sharma Ravi Tomar

Department of Computer Science School of Computer Science and
University of Petroleum & Energy Studies Engineering, University of Petroleum &
Dehradun, India Energy Studies
Dehradun, India
Hitesh Kumar Sharma
School of Computer Science and Dhananjay Yadav
Engineering, University of Petroleum & Department of Mathematical & Physical
Energy Studies Sciences
Dehradun, India University of Nizwa
Nizwa, Oman
1
Sensors, Embedded Systems, and IoT Components

Ananya Sharma and Niharika Singh

University of Petroleum and Energy Studies

CONTENTS
1.1 Introduction ............................................................................................................................1
1.2 Introduction to Embedded Computing Systems ..............................................................2
1.2.1 Usage of Sensors and Actuators in Embedded System ........................................ 3
1.2.2 Embedded Computing System: Reliability Analysis ...........................................5
1.3 Internet of Things: Introduction, Applications, and Trends ...........................................5
1.4 Applications of IoT .................................................................................................................6
1.4.1 Data Privacy, Adaptability, Sensibility, and Security ...........................................6
1.5 Introduction to Sensors .........................................................................................................7
1.5.1 Environmental Monitoring ......................................................................................8
1.5.2 Control Applications .................................................................................................8
1.5.3 Trends in Sensor Technology ...................................................................................9
1.5.4 Future Scope ............................................................................................................. 10
1.6 Artificial Intelligence and Internet of Things .................................................................. 10
1.6.1 Intro to Artificial Intelligence (AI) ........................................................................ 11
1.6.2 Convergence of AI and IoT ..................................................................................... 12
1.7 Conclusion ............................................................................................................................ 13
1.8 Summary............................................................................................................................... 13
References....................................................................................................................................... 14

1.1 Introduction
The Internet of Things (IoT) defines a network of embedded computing devices that are
linked together enabling communication and interaction within themselves and the exter-
nal environment. According to internet usage worldwide statistics (https://www.statista.
com/statistics/617136/digital-population-worldwide/) as of 2021, there were 4.95 billion
internet users estimated across the world. The growth in the use of the internet in the
past few years has been exponentially high, which has led to the development of IoT. The
unifying thread running through these various IoT visions is that “things” are expected
to become active participants in industries, data retrieval, and social processes (Hassan,
2018). IoT has a wide range of applications, including the digital transformation of indus-
tries, quality control, logistics, and supply chain optimization. Fitness wearables, smart
home applications, health care, and smart cities have all been made possible by the IoT. A
home automation system can be a great example as it portrays integration between devices
and interoperability of all functions resulting in gathering information and conveying it to
DOI: 10.1201/9781003291916-1 1
2 Mathematical Modeling for Intelligent Systems

FIGURE 1.1
IoT infrastructure.

the user as per requirement. RFID (Radio-Frequency Identification) systems can be used
to monitor objects in real-time; this enables mapping the real world into the virtual one
(Atzori et al., 2010). Figure 1.1 illustrates the complete infrastructure of an IoT environment
making it easier to understand how the IoT device and gateway are linked in an embedded
system. The data is stored in the cloud over a network in order to reduce response time and
fasten the data processing.
The sensors collect the data and send it to a control center where a decision is taken and
an associated command is sent to the actuator in its response. According to the total data
volume generation statistics, as of 2020, approximately 64.2 zettabytes (1 zettabyte = 1012
gigabytes) of data were created, consumed, and stored in that particular year, which is
increasing day by day at exponential rates. The introduction of artificial intelligence (AI)
and its subfields like machine learning (ML) and deep learning have paved the way for
extracting useful insights from big data. The convergence of IoT and AI has enabled the
viability of numerous real-life applications of IoT. AI proves to be helpful by avoiding
expensive and unplanned downtime, increasing operational efficiency, enhancing risk
management, and enabling new and improved products and services. According to the
forecast by Lee et al. (2009), with a compound annual growth rate (CAGR) of 30.8%, India’s
AI spending would rise from US$300.7 million in 2019 to US$880.5 million in 2023. This
shows that AI will help “all effective” IoT endeavors and data from deployments will have
“limited value” without it. The applications of AI incorporated with IoT include IoT data
mining platforms and tools, 5G-assisted IoT systems and applications, monitoring robots,
and self-driving cars. The convergence of AI and IoT enables the systems to be predictive,
prescriptive, and autonomous. AI can be applied to embedded systems. By making use of
AI-embedded systems, developers can look to deliver efficient solutions.

1.2 Introduction to Embedded Computing Systems

The combination of software and hardware, which is specifically designed to perform
a certain task, is known as an embedded system. These are controlled by the software
and are responsible for meeting users’ needs. The composition of an embedded system is
shown in Figure 1.2. The software component consists of
Embedded Systems and IoT Components 3

FIGURE 1.2
Composition of embedded systems.

• Application software: It performs the required functions and interacts with the user.
• The mediator device: It acts as a medium of communication between different
layers of the software part.
• An operating system: It frames rules for the required function to be performed.
• Firmware: It is a program written for certain hardware.

The hardware comprises memory, I/O (input/output) devices, memory buses, sensors,
actuators, and peripherals (cameras, printers, scanners, and keyboards). The applications
of embedded systems include embedded Linux, Android, and Windows-embedded com-
pact (Windows CE). It is also used in appliances such as washing machines, digital cam-
eras, televisions, and refrigerators. The increasing global demands for communication
devices and different electronics equipped with embedded systems are the fundamental
cause driving the growth of the market of embedded systems. The embedded systems
these days require less space and have low power consumption due to their small size.

1.2.1 Usage of Sensors and Actuators in Embedded System

Sensors and actuators are essential components of embedded systems. Sensors are devices
that measure a physical characteristic of the environment and transform it into output
signals. Essentially, sensors are needed to gain an understanding of the system. Sensor
data can be used for a variety of purposes, including system operation and control, pro-
cess monitoring, experimental modeling, product testing and qualification, product qual-
ity evaluation, fault prediction and diagnosis, warning generation, and surveillance. The
different sensors that are being developed to pace up with developing devices include
microminiature sensors, intelligent sensors with integrated preprocessing, and hierar-
chical sensory architectures. Sensor classification can range from extremely simple to
complex and is designed on the basis of different requirements. The multimedia wireless
sensor networks (WSNs) (Lee et al., 2009) are proposed to enable tracking and extracting
data in the form of multimedia such as imaging, video, and audio. Figure 1.3 shows the
overview of the battlefield, which is based on the principle of the multimedia network. It
includes nodes that are deployed by using aircraft in order to capture the images of dif-
ferent scenarios such as the movement of the enemy and their equipment. These nodes
are embedded with video cameras and microphones in order to capture the audio and
4 Mathematical Modeling for Intelligent Systems

FIGURE 1.3
Multimedia network in a battlefield.

video of the enemies. The nodes are interconnected over a wireless connection for data
compression, retrieval, and correlation. Sensors have their applications in real-life situa-
tions as well as in industries. They can be used in monitoring patients in medical applica-
tions; weighing scales; agriculture; and heating, ventilation, and air conditioning (HVAC)
technologies.
Actuators, on the other hand, are devices that enable movement and are responsible for
controlling the mechanism or a system. The actuators used in industries include electric
motors, hydraulic motors, and pneumatic control waves. Sensors are being embedded at
an exponential rate in almost every IoT device that is leading to a vast increase in their
usage. Sensors and actuators are essential in an embedded system. The embedded system
also contains a controller that generates control signals and manages the process that is
being controlled by making use of sensors and actuators (Sarishma et al., 2021). Table 1.1
gives a comprehensive overview of the characteristics of sensors and actuators.

TABLE 1.1
Sensors and Actuators
Sensors Actuators
Inputs and outputs They look at the input and initiate a They track the outputs of devices.
particular action.
Electrical signaling They use signaling to analyze the They measure the heat to determine the
environmental conditions to perform output.
a particular task.
Dependency They are dependent on each other. Actuators rely on sensors to do their job.
Conversion It converts physical attributes to It changes electrical signals to physical
electrical signals. action.
Location It is present at the input port. It is present at the output port.
Application They are used to measure temperature, They are used in operating dampers,
vibration, and pressure. valves, and couplings.
Embedded Systems and IoT Components 5

1.2.2 Embedded Computing System: Reliability Analysis

The use of embedded systems is growing exponentially in order to pace up the develop-
ment of IoT devices, starting from everyday requirements like mobile phones, televisions,
to complex healthcare applications. The systems that used 16-bit and 8-bit processors in
the past are now being replaced with processors that complete the assigned tasks with
less power consumption and response time. Security-sensitive embedded systems include
aircraft flight control systems, surgical robotics, and patient monitoring systems used in
hospitals, as well as nuclear power plant instrumentation and control systems. A chal-
lenge faced in embedded system development is the choice of hardware, as the cost of
the product depends on the size of hardware, which is another challenge faced during
development. Power consumption is another issue in the development of battery-powered
applications. While in nonbattery-powered applications, excessive heat is released due to
power consumption. A solution to this problem is to make the system run slowly, but this
will lead us to missed deadlines as output will be delivered late due to increased run-
ning time. The fault-tolerant software, compiler, and instruction set design for embedded
processors is explained in Azarpeyvand et al. (2010). The proposed method leads to time-
and cost-effective fault-tolerant strategies since compiler-level fault-tolerant techniques are
usable even after processor manufacture. Brute way of completing the task before a given
deadline is to speed up the hardware so that the program runs faster, but this factor even-
tually makes the system expensive. Another main challenge is often software debugging
because to really debug software, it needs to be run on the embedded hardware it’s been
designed for, and if the hardware design slips by a few months and the product launch
date remains fixed, then the software debugging stage gets squeezed very often. While the
hardware is capable of doing the task on paper, making it happen in software is generally
the crucial problem. In Glaß et al. (2012), a flexible framework for cross-level Compositional
Reliability Analysis (CRA) is proposed, which allows smooth integration of multiple reli-
ability analysis approaches across different levels of abstraction. Reliability is always an
important factor while selling a product, especially a safety-critical system like aircraft
flight control, weapons, or nuclear systems. These are certain challenges that the develop-
ers must take care of while designing reliable embedded systems.

1.3 Internet of Things: Introduction, Applications, and Trends

“In a few decades, computers will be interwoven into nearly every industrial product,”
declared Karl Steinbuch, a German computer scientist, in 1964. These words came to real-
ity when the concept of IoT was proposed in 1999 by the MIT Auto-ID Center (Liu & Lu,
2012). John Romke created a toaster in 1990 that functioned its power using the inter-
net. This toaster was the very first IoT device. Later that year, some students at Carnegie
Mellon University uploaded a Coca-Cola distribution machine to the internet in order
to determine which of the machine’s columns had the most iced coke. However, the
whole technology was not given any name yet. When Kevin Ashton created the term
“Internet of Things” in 1999, he gave a presentation about connecting RFID tags in sup-
ply chain management to the internet. The IoT is a world of interconnected things (smart
things or smart objects) that are capable of sensing, actuating, and communicating
with one another and with the environment, as well as sharing information and acting
6 Mathematical Modeling for Intelligent Systems

autonomously in response to real/physical world events, as well as triggering processes

and creating services with or without direct human intervention (Hassan, 2018). However,
a significant milestone that triggered the contemporary IoT craze occurred in 2008–2009,
when more items were connected to the internet than there were individuals. By the year
2020, we had more than 50 billion devices connected to the IoT. IoT is expected to be a
­$19-trillion industry. It is also one of the six “Disruptive Civil Technologies” identified by
the US National Intelligence Council as having the potential to alter US national power
(Conference Report CR 2008). Most of the technologies built until now can be mentioned
as the internet of people. So, whether it is a CRM application or an e-commerce applica-
tion, largely it is believed that there is a person at the end of it. John Chambers, former
CEO of CISCO assumed that there are going to be 500 billion things connected to the
internet in the future which is nearly 100 times the world’s population. The other benefit
is that things can be present where people aren’t. For all these reasons, IoT is dramatically
different from the internet for people, and it will ultimately reshape all the fundamental
industries that run the planet, power, water, agriculture, healthcare, construction, trans-
portation, etc. These are a few ways by which the internet has changed our standard of
living by making things to communicate with people by making use of IoT. It has the
potential to provide this process a new dimension by allowing communication with and
among smart things, opening the way for the vision of “anytime, anywhere, any medium,
any object” communications (Atzori et al., 2010).

1.4 Applications of IoT
Today, IoT is one of the few technologies whose influence will extend across all businesses
and directly into our homes. IoT deployment has already begun in areas such as logistics,
energy monitoring, military, and industrial automation. Furthermore, IoT has an impact
on all parts of the technical stack in terms of hardware and software innovation. But there
are fields that are not able to make enough use of IoT standards for their development.
Industries like manufacturing, transportation, logistics, mining, and agriculture are the
physical industries (Section 1.4, Atzori et al., 2010). The mining industry also has a lot of
interest and opportunities. Another application area is agriculture where advanced tech-
nology through feedback and coordination with sensor technologies such as electrochemi-
cal sensors is used to determine the level of nutrients in the soil. Its usage has shown a 23%
reduction in the amount of fertilizer needed, 15% reduction in the amount of seed needed,
and overall yield has increased to 27% by the automation of industrial IoT. The advantages
of advanced tracking and precision agriculture are huge. Additional uses include loca-
tion sensing and sharing, environmental sensing, remote control, ad hoc networking, and
encrypted communication (Chen et al., 2014).

1.4.1 Data Privacy, Adaptability, Sensibility, and Security

The ever-growing number of devices can be treated as an incredible opportunity that
brings immense risks. There are multiple challenges that IoT technology faces these days.
The most important among them are data privacy, adaptability, sensibility, and secu-
rity. Security stands for the protection of users’ information from any type of malicious
access such as preventing theft, corruption, and other types of damage while allowing the
Embedded Systems and IoT Components 7

successful exchange of the data in the most efficient way, whereas privacy is the right of
individuals to have control of when and how their personal information is collected and
processed in the IoT environment. The Open Web Application Security Project (OWASP)
(https://owasp.org/www-project-internet-of-things/) has identified the most common
issues in IoT and demonstrated that many of these flaws are caused by a lack of under-
standing of well-known security applications such as encryption, authentication, access
control, and role-based access control. All of these applications are explained in Bertino
(2016). We can consider an example of location privacy. Most of the devices nowadays are
equipped with GPS (global positioning system) receivers or other sensors that allow deter-
mining the position of the device. However, different applications require various lev-
els of location accuracy, so the user can decide based on the collected information as to
whether to disclose the coordinates or not. Personal health indicators like heart rate and
no of steps taken, user whereabouts, payment credentials, calendar information, and con-
tacts in the smartphone are all examples of data acquired by IoT services. In addition, the
privacy impact assessment methodology of RFID serves as a prototype for a risk manage-
ment–based approach that can be used in IoT scenarios. At the design stage, data privacy,
security, and protection should be addressed. Meanwhile, IoT devices don’t always have
enough processing capacity to execute all of the necessary security layers and function-
alities (Anuj Kumar et al., 2012). Usually, IoT devices have the functions to preserve users’
privacy, access control and authentication, secure network protocols, cryptography, and
edge computing, which means computation and data storage on the local device. If the
data is transferred to the cloud, the techniques that can be applied to the datasets to avoid
users profiling and linkage include data analysis and summarization, digital forgetting,
and the concept of differential privacy. Some ways to protect data are to use trusted con-
nectivity, install trusted software, check privacy settings, and study the privacy policy that
details what the company is going to do with your data as one has the right to understand
what happens to the information and whether the company shares any personal data
with third parties. Still, there are loopholes in matters of data security because of which
data adaptability and sensibility are considered difficult to handle. Security and privacy
include challenges (Shukla, 2021) that need to be resolved for a secure digital future.

1.5 Introduction to Sensors

Sensors are devices that measure a physical characteristic of the environment like humid-
ity, temperature, and pressure and transform it into output signals. Sensors that can see,
hear, smell, and even taste are now available. Our life at home and at work would be much
more difficult if we didn’t have sensors. Sensors embedded in the road regulate the traf-
fic lights at an intersection. When you arrive at the intersection, these sensors detect your
presence. A sensor also causes the door to open automatically as you approach the grocery
shop. A sensor is a device that measures physical input from its surroundings and provides
an electrical output in reaction to that input in the realm of instrumentation and process
control. Sensors detect level, temperature, flow, pressure, speed, and position, among other
physical qualities. Sensors can be classified as either passive or active in terms of process
control. An active sensor does not require an external source of electricity to function, but
a passive sensor does. A thermocouple is an active sensor if it operates without the use of
an external power supply. When a thermocouple is exposed to an increase in temperature,
8 Mathematical Modeling for Intelligent Systems

the voltage across it increases. A piezoelectric sensor is another type of active sensor. A
passive sensor is a resistance temperature detector, or “RTD.” It’s a gadget, the resistance
of which changes with change in temperature. An external supply or an excitation circuit
is required to take advantage of this change in resistance by causing a voltage change.
A strain gauge is an example of a passive sensor. Because the output of a sensor must be
conditioned or amplified, almost every sensor used in process control will be connected
to a transmitter.

1.5.1 Environmental Monitoring
Environmental monitoring is one of the most compelling industrial IoT applications driv-
ing operational transparency and efficiency. Across a variety of industries such as manu-
facturing, mining, oil and gas, and agriculture, monitoring and managing environmental
conditions such as air quality, water quality, and atmospheric hazards is critical to prevent
adverse conditions that may impact production process, product quality equipment, and
worker’s safety. However, the accessibility of environmental data requires the sensor net-
work enabled by robust, scalable, and cost-effective connectivity. Low-power wide-area
networks (LPWAN) are an excellent alternative for linking environmental sensor net-
works because they provide the wide coverage and indoor penetration required to cover
large regions and reach previously unreachable locales. Compared to industrial alterna-
tives like mesh networks, LPWAN solutions deliver much higher energy efficiency along-
side easy installation and management. Three separate wireless sensors were created,
built, and assessed in order to create IoT-based environmental monitoring systems (Mois
et al., 2017). Another beneficial alternative is the ultra-low power design and simple star
topology that has low device and network costs, fewer infrastructure requirements, and
reduces setup and maintenance complexity. All contribute to a low total cost of ownership.
Critical ambient conditions such as temperature, humidity, air quality, and pressure can
be monitored and controlled to improve quality; measure workplace conditions such as air
quality, heat, humidity, radiation, and noise to improve worker safety; ensure regulatory
compliance and maintain quality assurance, and optimize energy consumption by con-
necting an LPWAN-enabled environmental sensor network. Many such algorithms and
protocols are explained as well as analyzed in the work by Prabhu et al. (2018) along with
their merits and demerits. Energy efficiency, overall system cost, sensor module response
time, system accuracy, adequate signal-to-noise ratio, and radio frequency interference/
electromagnetic interference (RFI/EMI) rejection during varying atmospheric conditions
and in inhomogeneous environments are all requirements of a WSN for environmental
monitoring (Prabhu et al., 2018). It addresses the complete procedure of the development of
WSN for IoT devices from scratch considering all aspects of the required platform.

1.5.2 Control Applications
Sensors are characterized as low cost, low power, and highly reliable networks that collect
data and make it available for processing (Atzori et al., 2010; Tomar & Tiwari, 2019). There
are many types of sensors available for use in industrial applications. Some of the common
sensors used are listed in Table 1.2.
Sensors play an important role by providing intelligence to the “things” so that they
respond to all the tasks without any human interference. There is a sensor for almost
every single requirement. Some of them include proximity, position, occupancy, motion,
Embedded Systems and IoT Components 9

TABLE 1.2
Commonly Used Sensors
Photoelectric sensors They work by emitting a visible or infrared light beam from their light-emitting
element. They can work differently depending on the type of sensor. Some may
reflect their light back to themselves for detection, while others may emit their
light to a separate element that watches for breaks in the signal.
Magnetic sensors They work by detecting changes and disturbances in a magnetic field such as flux,
strength velocity, and direction.
Hall effect sensors These can be thought of as being in a similar category as magnetic sensors since
they are devices that are attracted by external magnetic fields.
Presence sensors They work by enabling or disabling devices based on whether presence is detected
nearby. They are commonly used in buildings that turn the lights on or off based
on motion in the room.
Flow sensors These are devices that are used to measure the flow rate of something typically a
fluid. These sensors are often part of a larger device such as a flow meter. Similar
devices are also available for air and gas measurement but flow sensors are
commonly used in various applications, including industrial processes, and HVAC.

velocity, optical, and chemical sensors (Sehrawat & Gill, 2019). The development of
­IoT-based a­ pplications relies heavily on these sensors.

1.5.3 Trends in Sensor Technology

In the developing world of the IoT, most of the things are sensors. Experts estimate that
there will be as many as a trillion sensors connected to the IoT during this decade. The
sensors are changing and evolving constantly and huge technology and market trends are
driving them. Besides the fact that there are a whole bunch of sensors available probably
joining things, social media groups on the IoT. Several trends can be seen in sensor devel-
opment and they are primarily driven by the users and their requirements. Five major
trends that can be observed include:

• Miniaturization: The demand for portable devices with various features, like
improved security and cheap prices, has been increasing lately. Different kinds of
designs and ideas are being incorporated across the sensor products to give them
a smaller device with a lot more capability. Some applications of this trend include
sports and fitness watches. The advantages are that these sensors can fit into small
packages.
• Digitization: Almost all the sensors introduced in the present day are available
in both analog and digital output. The common digital output protocols that can
be seen in the digital interfaces globally are I2C (interintegrated circuit) and SPI
(serial peripheral interface). The advantages when we digitize the sensors is we
get better performance in trends of resolution and accuracy and digital sensors
draw less power. These are more compatible with IoT applications. Digitization
can be seen across many industries. Examples include pneumatic control systems
for factories that control air pressure and monitor various air systems that might
be used for the machinery in a factory. There are sensors that can handle both
low- and high-pressure systems. The ability to fit into different kinds of systems
is made much easier through digital sensors. HVAC control systems that consider
humidity, temperature, and airflow rate are all going digital. So, it makes it a lot
easier to apply sensors in these kinds of applications.
10 Mathematical Modeling for Intelligent Systems

• Low power sensors: The Low power sensor emphasize more on the running time
of a sensor on a single charge. For e.g., TSYS02 and HTU21D are low power con-
suming temperature and humidity sensor(s), MS5837 is an low power altimeter
sensor. These sensors consumes as low as 180 µA of power.
• Multisensor modules: The demand to put multisensor capabilities into single
packages is also increasing. The advantages are space and power savings as sev-
eral sensing elements are powered to a single power source. Some examples that
include these combinations are HTU21D, which is a relative humidity and tem-
perature sensor that takes information and measures both relative humidity and
temperature, and MS8607, which goes into home appliances and automotive appli-
cations. One good example is that this sensor finds its way into an inlet manifold
in a car and monitors relative humidity, air temperature, and barometric pressure,
and all this information goes into the engine control unit in the vehicle. The fuel
is adjusted to make sure of its proper usage and does not produce carbon monox-
ides and nitrous oxides. It monitors all three of those things, is a small package
that fits into the intake manifold so that it doesn’t restrict the air, and it is very low
powered. HVAC systems are also examples of such kinds putting multiple sensing
capabilities in the devices and creating better efficiencies.
• Harsh environment sensors: Many locations where sensors are needed are not so
“friendly” to electronic devices and are termed harsh environments. One solution
to this problem is the inclusion of sensors in packaging that can withstand some
of these very harsh environments. A lot of sensors are finding their way into areas
where there are very high EMI and RFI fields. So sensors are imbibed with some fea-
tures to protect them. The MS5837-02BA pressure sensor is an example of this kind.

1.5.4 Future Scope
IoT and its devices are making progress rapidly with time. None of this would be pos-
sible without communicative, diligent data-collectors at the lowest field level called sen-
sors. These sensors utilize the globally standardized I/O-link communication protocol
that allows them to be configured via the system controller and flexibility adjusted to suit
each new production job. The advantage of using small, disposable sensors is that the
measurements can be done directly on the spot. Electrochemical and biochemical sensors
are expected to become a commodity in the future because there are so many different
possibilities with these kinds of sensors such as detecting viruses, safety, and explosive
detection. In the work by Bogue (2007), the future use of microelectromechanical systems
(MEMS) is discussed, as well as its capacity to create sensors at very low unit prices for big
volume applications. The primary characteristics of any chemical sensor are specificity,
accuracy, precision, resolution, and detection level. Its performance is thoroughly exam-
ined in the work by Prien (2007).

1.6 Artificial Intelligence and Internet of Things

IoT is the interconnectivity of devices, sensors, and systems. Some examples of IoT include
sensors that can determine when something is either in proximity or as it moves through
an environment and so forth. Smart homes and buildings are a perfect example of the
Embedded Systems and IoT Components 11

convergence of artificial Intelligence (AI) and IoT where if someone has not been in a room
for a certain amount of time, it actually shuts the lights off. It will set the air conditioner
down or do whatever it needs to make sure that the building is operating efficiently. This is
known as proactive or predictive maintenance and quality. AI simulates the ability of the
computer to perform the assigned task by analyzing the data and giving desired outputs
after processing the data. It can be observed in things like process automation and robot-
ics. Nowadays, a lot of chatbots are embedded with ML. Things like smart assistants help
with specific tasks that are also called decision optimization. But the intelligence of these
devices increases eventually with time. The architectural alternatives for embedding AI
in IoT are discussed in the work by Calo et al. (2017). The system developed controls the
data flow rate and extracts the required data implementing the concept of edge computing.

1.6.1 Intro to Artificial Intelligence (AI)

AI is a technology that allows machines to sense, grasp, act, and learn how to accomplish
a task in a way that is similar to human intelligence. Almost every company today uses AI
in its goods. AI can be classified as either weak or strong. Weak AI is built and educated
for a specific task, such as voice-activated assistants. It can either respond to your query
or carry out a preprogrammed instruction, but it won’t work unless you interact with it.
Strong AI has generalized human cognitive abilities, which means it can solve problems
and find answers without the need of human participation. A self-driving automobile is
an example of powerful AI that employs a combination of computer vision, image recogni-
tion, and deep learning to operate a vehicle while keeping in a specific lane and avoiding
unforeseen impediments such as pedestrians. Comparative analysis of weak and strong
AI says weak AI becomes so strong while “strong AI” is almost as weak as it was decades
ago (Liu, 2021). Healthcare, education, finance, law, and manufacturing are just a few of
the fields where AI is being used to benefit both businesses and customers. AI in vari-
ous modules is developed for the purpose of education and engineering, which include
intelligent tutoring systems (Beck et al., 1996) and fuel shuffling in a nuclear powerpoint
(Faught, 1986). Many technologies, such as automation, ML, machine vision, natural
language processing, and robotics, incorporate AI. The use of AI creates legal, ethical,
and security concerns. Hackers are gaining access to sensitive networks by employing
advanced machine-learning technologies. Despite the dangers, the usage of AI tools is
governed by a few restrictions. Experts, on the other hand, guarantee us that AI will only
improve products and services. Data economy is one of the factors behind the emergence
of AI. It refers to how much data has grown over the past few years and how much more
it can grow in the coming years. The proliferation of data has spawned a new economy,
and there is a perpetual war for data ownership among businesses seeking to profit from
it. With the increase in data, the volume has given rise to big data that helps manage huge
amounts of data. Data science helps analyze that data. As a result, data science is moving
toward a new paradigm in which robots may be taught to learn from data and generate
a range of useful insights, giving rise to AI. AI mimics human and animal intelligence. It
involves intelligent agents, autonomous entities that perceive their environment, and take
actions that maximize their chances of success at a given goal. AI is a method for com-
puters to imitate human intellect through logic. It’s a program with the ability to detect,
reason, and act. By giving greater personalization to users and automating procedures, AI
is transforming industries. ML is at the heart of AI. The use of algorithms to find mean-
ing in random and unordered data is the first part of ML, and the second half is the use of
learning algorithms to uncover the relationship between that knowledge and improve the
12 Mathematical Modeling for Intelligent Systems

FIGURE 1.4
AI applications.

learning process. So the overall goal for ML is to improve the machines’ performance on a
specific task such as predicting the stock market. Emerging industries that might require
AI in the coming years are shown in Figure 1.4.

1.6.2 Convergence of AI and IoT

The digital world is interconnected through the internet. But onboarding hardware into
the network has presented myriad challenges over the years. With research and develop-
ment, the past decade has seen about 50 billion devices being connected to the internet and
the number is growing expansively. Institutions and researchers including municipalities,
governmental bodies, and multinational corporations are harnessing IoT via collecting
dynamic data points through various sources like cameras, multiple sensors, and personal
devices. This is one of the factors behind the emergence of AI. AI has made it easier, in con-
gruence with big data, to gather crucial insights from these data points. It involves intel-
ligent agents, the autonomous entities that perceive their environment, and takes actions
that maximize their chances of success at a given goal. AI is a method for computers to
imitate human intellect through logic. It’s a program with the ability to detect, reason, and
act. It is transforming industries by allowing users to be more personalized and by auto-
mating procedures. The basic components of a neural network include an artificial neuron,
and this neuron is connected to all other neurons and the interconnections are represented
by a number called weight. The influence of an artificial neuron on other neurons depends
directly on the value of weight. Three types of layers make up the usual architecture of a
neural network:

• Input layer: The input layer receives data from the physical environment, such as
a smartphone camera feed.
• Hidden layer: The real network’s compute engine is the hidden layer. As a result,
there could be a lot of hidden layers.
• Output layer: The outcome is produced by the output layer.
Embedded Systems and IoT Components 13

ML is the backbone of AI. The first half of ML entails utilizing algorithms to find meaning
in random and disordered data, while the second half entails using learning algorithms to
discover the relationship between that knowledge and improve the learning process. As
a result, the overarching purpose of ML is to increase the computers’ performance on a
certain task, such as stock market prediction. More than 80% of commercial IoT initiatives
will have an AI component by 2022 (Internet of things (IoT), 2019). A self-driving car is an
example of advanced AI that uses a combination of computer vision, image recognition,
and deep learning to drive a vehicle while staying in a defined lane and avoiding obstacles
like pedestrians. Healthcare, education, finance, law, and manufacturing are just a few of
the fields where AI is being used to benefit both businesses and customers. AI is used in
a variety of technologies, including automation, ML, machine vision, natural language
processing, and robotics. AI raises ethical, legal, and security concerns. If an autonomous
vehicle is involved in an accident, for example, it is difficult to determine who is to blame.
In addition, advanced ML technologies are being used by hackers to obtain access to sensi-
tive networks. Despite the dangers, the usage of AI tools is governed by few restrictions.
Experts assure us, however, that AI will simply improve products and services in the near
future, not replace us, the people.

1.7 Conclusion
The IoT is paving the way for digital transformation at an exponential rate. Technical as
well as physical industries are adopting IoT to digitize their products. The embedded sys-
tems give enhanced results because of their small size and low power consumption capa-
bility. The future scope of embedded software testing can be improved by the adoption
of technologies that can process huge amounts of data with a reduced response time. The
data collected from sensors such as liquid levels, distance measures, smoke detectors, soil
moisture levels, temperature, humidity, and light data, as well as location’s longitude and
latitude, is used in burgeoning IoT applications. Apart from different control applications
of sensors, they have also been helpful in environmental monitoring techniques. The data
perceived from sensors becomes difficult to manage, so AI is incorporated in IoT. This
combination not only automates the industries but also results in devices with optimized
results. AI techniques can be used in order to embed human intelligence into devices and
industrial products. AI models train the data for enhanced results. Overall, the impor-
tance and future scope of sensor technologies, embedded systems, and the need for IoT in
the industrial and business sectors have been covered.

1.8 Summary
The IoT has been described here as the emerging resource for a future where all physical
and technical devices will be connected through the internet. The importance of embed-
ded systems has been covered with the supporting statements of their features. The lack
of some characteristics in these systems leading to less reliable behavior of embedded
systems has also been discussed. The sensors that collect huge amounts of data also play
14 Mathematical Modeling for Intelligent Systems

an important role in the study. The factors that make the technology of sensors more
advanced and useful have been stated. The adoption of techniques discussed will result
in desired results depending on the expectations of the consumer. Applications of IoT and
its resulting outcomes have been studied to achieve a deeper understanding of the study.
AI incorporated with IoT has given new hope in the development of a world of digitized
things.

References
Atzori, L., Iera, A., & Morabito, G. (2010). The internet of things: A survey. Computer Networks, 54(15),
2787–2805. https://doi.org/10.1016/j.comnet.2010.05.010
Azarpeyvand, A., Salehi, M. E., Firouzi, F., Yazdanbakhsh, A., & Fakhraie, S. M. (2010). Instruction
reliability analysis for embedded processors. In: 13th IEEE Symposium on Design and Diagnostics
of Electronic Circuits and Systems (pp. 20–23). https://doi.org/10.1109/DDECS.2010.5491824
Beck, J., Stern, M., & Haugsjaa, E. (1996). Applications of AI in education. XRDS: Crossroads, the ACM
Magazine for Students, 3(1), 11–15. https://doi.org/10.1145/332148.332153
Bertino, E. (2016). Data security and privacy in the IoT. In: EDBT. https://doi.org/10.5441/002/
edbt.2016.02
Bogue, R. (2007). MEMS sensors: Past, present and future. Sensor Review, 27(1), 7–13. https://doi.
org/10.1108/02602280710729068
Calo, S. B., Touna, M., Verma, D. C., & Cullen, A. (2017). Edge computing architecture for applying
AI to IoT. In: IEEE International Conference on Big Data (Big Data), 2017 (pp. 3012–3016). https://
doi.org/10.1109/BigData.2017.8258272
Chen, S., Xu, H., Liu, D., Hu, B., & Wang, H. (2014). A vision of IoT: Applications, challenges, and
opportunities with China perspective. IEEE Internet of Things Journal, 1(4), 349–359. https://
doi.org/10.1109/jiot.2014.2337336
Faught, W. (1986). Applications of AI in engineering. Computer, 19(7), 17–27. https://doi.org/10.1109/
MC.1986.1663273
Glaß, M., Yu, H., Reimann, F., & Teich, J. (2012). Cross-level compositional reliability analysis for
embedded systems. In F. Ortmeier & P. Daniel (Eds.) Computer Safety, Reliability, and Security.
Lecture Notes in Computer Science. SAFECOMP 2012 (vol. 7612, pp. 111–124). Springer. https://
doi.org/10.1007/978-3-642-33678-2_10
Hassan, Q. F. (2018). Introduction to the internet of things. In: The Internet of Things A to Z: Technologies
and Applications (pp. 1–50). IEEE. https://doi.org/10.1002/9781119456735.ch1
https://www.idc.com/getdoc.jsp?containerId=prAP46899220
https://www.idc.com/tracker/showproductinfo.jsp?containerId=IDC_P33198
https://owasp.org/www-project-internet-of-things/
https://www.maximizemarketresearch.com/market-report/global-industrial-sensors-market/
34465/
https://www.statista.com/statistics/273018/number-of-internet-users-worldwide/
https://www.statista.com/statistics/871513/worldwide-data-created/
https://www.wired.com/brandlab/ 2018/05/bringing-power-ai-internet-things/
Internet of things (IoT). (2019). HI, road map AU. https://eps.ieee.org/images/files/HIR_2019/
HIR1_ch03_iot.pdf
Lee, I., Shaw, W., & Fan, X. (2009). Wireless multimedia sensor networks. In: Misra, S, Woungang, I.
& Misra, S. (eds). Computer Communications and Networks (pp. 561–582). Springer. https://doi.
org/10.1007/978-1-84882-218-4_22
Liu, B. (2021). ‘Weak AI’ is likely to never become “strong AI”, so what is its greatest value for us?
https://arxiv.org/pdf/2103.15294.pdf
Embedded Systems and IoT Components 15

Liu, T., & Lu, D. (2012). The application and development of IoT. In: International Symposium on
Information Technologies in Medicine and Education, 2012 (pp. 991–994). https://doi.org/10.1109/
ITiME.2012.6291468
Mois, G., Folea, S., & Sanislav, T. (2017). Analysis of three IoT-based wireless sensors for environmen-
tal monitoring. IEEE Transactions on Instrumentation and Measurement, 66(8), 2056–2064. https://
doi.org/10.1109/TIM.2017.2677619
National Intelligence Council (1663). Disruptive civil technologies—six: Technologies with potential
impacts on the US interests out to 2025. https://www.dni.gov/. Conference Report CR 2008,
07(April) p. 2008.
Prabhu, K., Dhineshkumar, P., Gobisankar, S., Gowrisankar, M., & Suryapranesh, M. B. (2018).
A comparative study of algorithms for measuring HVAC in IoT. https://doi.org/10.15680/
IJIRSET.2018.0702023
Prien, R. D. (2007). The future of chemical in situ sensors. Marine Chemistry, 107(3), 422–432. https://
doi.org/10.1016/j.marchem.2007.01.014
Sarishma, T. R., Kumar S., Awasthi M.K. (2021) To beacon or not?: Speed based probabilistic adap-
tive beaconing approach for vehicular ad-hoc networks. In: Paiva S., Lopes S.I., Zitouni R.,
Gupta N., Lopes S.F., Yonezawa T. (eds) Science and Technologies for Smart Cities. SmartCity360°
2020. Lecture Notes of the Institute for Computer Sciences, Social Informatics and Telecommunications
Engineering, vol. 372. Springer, Cham. https://doi.org/10.1007/978-3-030-76063-2_12
Sehrawat, D., & Gill, N. S. (2019). Smart sensors: Analysis of different types of IoT sensors. In: 3rd
International Conference on Trends in Electronics and Informatics (ICOEI), 2019 (pp. 523–528).
https://doi.org/10.1109/ICOEI.2019.8862778
Shukla, A. (2021). A study on Internet of things: Industrial application and difficulties. https://
easychair.org/publications/preprint/5TVQ
Tomar R., & Tiwari R. (2019) Information delivery system for early forest fire detection using inter-
net of things. In: Singh M., Gupta P., Tyagi V., Flusser J., Ören T., Kashyap R. (eds) Advances in
Computing and Data Sciences. ICACDS 2019. Communications in Computer and Information Science,
vol. 1045. Springer, Singapore. https://doi.org/10.1007/978-981-13-9939-8_42
2
Security and Privacy in Mobile Cloud Computing

A. Katal
University of Petroleum and Energy Studies

CONTENTS
2.1 Introduction .......................................................................................................................... 18
2.1.1 General Architecture ............................................................................................... 18
2.1.2 Features ..................................................................................................................... 20
2.2 Mobile Cloud Service Models ............................................................................................ 20
2.3 Green Mobile Cloud Computing ....................................................................................... 21
2.4 Introduction to Security and Privacy in Mobile Cloud Computing ............................22
2.5 Security and Privacy Challenges in Mobile Cloud Computing ................................... 23
2.5.1 Challenges Related to Data Security ..................................................................... 23
2.5.2 Offloading and Partitioning Security Challenges .............................................. 23
2.5.3 Virtualization ........................................................................................................... 24
2.5.4 Challenges Related to Mobile Cloud Applications ............................................. 24
2.5.5 Security Challenges Related to Mobile Devices .................................................. 25
2.5.6 Security of Mobile Network ................................................................................... 25
2.5.7 Problems Related to Mobile Terminal .................................................................. 25
2.5.7.1 Malware ...................................................................................................... 25
2.5.7.2 Vulnerabilities in Software ...................................................................... 26
2.5.8 Mobile Cloud ............................................................................................................ 26
2.5.9 Privacy Challenges .................................................................................................. 27
2.6 Current Solutions for the Security and Privacy Issues .................................................. 27
2.6.1 Data Security ............................................................................................................ 27
2.6.2 Partitioning and Offloading ................................................................................... 27
2.6.3 Virtualization ........................................................................................................... 28
2.6.4 Mobile Cloud Applications..................................................................................... 28
2.6.5 Mobile Devices ......................................................................................................... 29
2.6.6 Mobile Network ....................................................................................................... 29
2.6.7 Mobile Terminal ....................................................................................................... 29
2.6.7.1 Antimalware .............................................................................................. 29
2.6.7.2 Software Vulnerabilities ..........................................................................30
2.6.8 Mobile Cloud ............................................................................................................30
2.6.9 Privacy .......................................................................................................................30
2.7 Discussion and Conclusion ................................................................................................30
References....................................................................................................................................... 31

DOI: 10.1201/9781003291916-2 17
18 Mathematical Modeling for Intelligent Systems

2.1  Introduction
Since 2007, MCC has been an important field of research for the technological and engi-
neering sectors. Cloud computing is usually characterized as a collection of services made
available through a Web cluster system. Such cluster structures are defined as a collection
of reduced cost servers or Personal Computers (PCs) that arrange the machines’ various
resources as per a specific strategy as well as provide services for data processing that are
secure, reliable, rapid, efficient, and consistent to customers. Meanwhile, smartphones have
come to symbolize a wide range of mobile devices since they have become increasingly
internet-connected as wireless network technology has progressed. Growing popularity
and portability are two essential aspects of next-generation networking, which provides
a variety of personalized communication networks via a number of network endpoints
and access methods. The basic technology of Cloud Computing (CC) is the centralization
of computers, services, and specialized applications as a utility to be sold to consumers in
the same way that water, gas, or electricity are. As a result of the combination of ubiquity
mobile networks and CC, a new computing paradigm, termed MCC, is created.
Instead of typical local PCs or servers, resources in MCC networks are virtualized and
deployed to a group of multiple independent systems as development of cloud comput-
ing, and are made available to smartphones like cellphones, mobile terminals, and so
on. Mobile Cloud Computing (MCC) has seen a faster expansion in terms of research as
mobile phones have become a need for human life. The introduction of MCC represents a
fundamental shift in computer science technologies as well as phone developers.
MCC is a relatively new subset of cloud computing in which mobile apps are created
or driven by cloud computing technologies. It is constantly developing and is commonly
used for data storage and processing in mobile devices. Mobile cloud technologies pro-
vide processing power and data management from smartphones and the cloud to a sig-
nificantly greater population of cellular customers, delivering applications and mobile
computing to nonsmartphone consumers (Noor et al., 2018). The number of mobile device
users has increased dramatically in a very short period. Almost every person possesses
at least one Smartphone Device (SMD), and as the number of SMDs has grown so has
the use of the internet in SMDs. Because of the widespread usage of the internet, MCC is
anticipated to play a significant role in the IT sector. A mobile cloud approach enables pro-
grammers to construct mobile user–specific apps which are not constrained by the phone’s
mobile operating system, computation, or storage abilities. MCC has attracted the interest
of businessmen as a lucrative and beneficial solution provider that minimizes mobile app
creation and maintenance costs while letting mobile consumers to easily acquire the on-
demand technology.
While MCC will make and is making a significant contribution to our everyday lives, it
will and has introduced a slew of new obstacles and concerns. In order to overcome the
constraint of limited resources and processing capacity of mobile devices and to attain
good access to CC equivalent to that of regular computers and laptops (Al_Janabi &
Hussein, 2019), it is necessary to study MCC paradigm in detail.

2.1.1  General Architecture

Mobile phones are becoming increasingly commonplace in most people’s daily lives. As
a result, businesses have been compelled to develop programs that can be shared. Access
is made easier by mobile phones. However, due to restricted resources, mobile devices
Mobile Cloud Computing 19

FIGURE 2.1
Architecture of mobile cloud computing.

provide certain design problems to mobile application developers. Cloud computing is

being utilized to solve these issues. Today’s lifestyle frequently necessitates staying in
touch via mobile communication devices. Data transmission and reception are getting
more convenient. MCC’s architecture is depicted in Figure 2.1.

• Mobile user layer: This tier is made up of many different mobile cloud service
consumers who use mobile devices to access cloud services.
• Mobile network layer: This layer is made up of several mobile network providers
who handle mobile user queries and deliver information via base stations. These
user requests are handled by the home agent’s mobile network services, such as
Authentication, Authorization, and Accounting (AAA). Mobile network carriers
assist in identifying the subscribers’ data held in their databases via their Home
agent (HA).
• Cloud service provider layer: This layer is made up of numerous CC service pro-
viders who offer a variety of CC services. These CC services are elastic, meaning
they may be scaled up or scaled down depending on what CC service consumers
20 Mathematical Modeling for Intelligent Systems

require. CC offers services to consumers, especially to those with mobile phones,

who may access cloud services over the Internet.

2.1.2 Features

• Elasticity: Users can quickly gain access to provide computer resources without
requiring human intervention. User capabilities may be supplied fast and elasti-
cally, in certain cases dynamically, to easily scale out or up.
• Infrastructure scalability: With minimal modifications to an infrastructure
design and software, new nodes in data centers may be added or withdrawn from
the network. Based on demand, mobile cloud architecture may rapidly expand
horizontally or vertically.
• Access to a broad network: The network makes user skills and abilities accessible,
and they may be retrieved via standard techniques that encourage utilization by
diverse platforms such as mobile phones, PCs, and PDAs.
• Location independence: Another feature of MCC is its independence from loca-
tion. There is a distinct feeling of location independence, in which the user has had
no influence over or awareness of the actual location of the delivered resources.
However, indicating location at different levels of abstraction beyond nation, state,
or data center may be conceivable.
• Cost and scale effectiveness: Mobile cloud deployments, regardless of the deploy-
ment strategy, prefer to be as big as feasible and make it more efficient.

2.2 Mobile Cloud Service Models

MCC service models are more properly categorized based on the responsibilities of com-
puting objects within its project architect, whereas MCC service models may be catego-
rized depending on the roles and relationships among cellular objects. Existing MCC
services may be divided into three primary types based on this viewpoint: mobile as a
service consumer (MaaSC), mobile as a service provider (MaaSP), and mobile as a service
broker (MaaSB). Figure 2.2 depicts the MCC service models.
MaaSC evolved from the conventional client–server paradigm in the early stages by
incorporating virtualization, fine-grained authentication, as well as other cloud-based
technologies. Smartphones may outsource memory and communication activities to the
cloud in order to enhance speed and program abilities. Under this design, the operation is
one-way from the cloud to smartphones.
MaaSP differentiates with MaaSC in that the function of a smart phone is switched from
that of a prospective customer with that of a provider. Smartphones with built-in sensors,
for instance, can detect data from gadgets and their surroundings and then give sensing
services to other smartphones through the cloud. Depending on their sensing and pro-
cessing capabilities, smartphones provide a wide range of services.
MaaSB is a MaaSP extension that offers connectivity and packet forwarding capabili-
ties to many other smartphones or edge devices. In certain cases, MaaSB is required since
mobile devices often have poor sensing capabilities when compared to sensors devoted to
particularly developed functionality and sensing sites. Mobile phones, for example, may
be used to gather Nike Fuelband users’ (Nike. Just Do It. Nike IN, n.d.) physical activity.
Mobile Cloud Computing 21

FIGURE 2.2
Service models of mobile cloud computing.

MaaSB extends the cloud boundaries to phones and wireless sensors. As a consequence,
a smartphone may be configured to act as a gateway or proxy, providing network access
via different communication protocols such as 3/4G, Bluetooth, and Wireless internet.
Furthermore, the proxy mobile device may preserve the confidentiality and privacy of
their interfaced sensors.

2.3 Green Mobile Cloud Computing

Green Computing (Green Chemistry) is responsible for creating, producing, utilizing,
and disposing of computer systems as well as their equipment such as displays, print-
ers, storage systems, and networking and communication systems to consume quickly
22 Mathematical Modeling for Intelligent Systems

and productively with no impact on the environment. The goal of Green Computing is
to reduce the usage of hazardous materials, utilize energy-efficient products to extend
their lifespan, and promote the recyclability or biodegradability of obsolete materials and
manufacturing waste. Low carbon footprints must be used to reduce carbon emissions.
MCC is a new and popular mobile technology that makes use of completely accessible
cloud computing services and capabilities. Users, network operators, and cloud service pro-
viders may all benefit from this technology’s extensive computational functions. However,
because of user mobility and intensive computing processes, energy consumption is a big
concern. There is a “capacity crunch” in mobile networks, which indicates that service
providers are straining to meet requests for mobile services. To provide Quality-of-Service
to the user, the service provider must provide a complete package of capabilities, such
as increasing network efficiency and decreasing energy consumption depending on the
user’s mobility behaviors. In the scope of MCC, energy conservation must be addressed
not just on the cloud side but also on the mobile side. The former is utilized to overcome the
latter’s existing constraint by utilizing a distant resource provider rather than processing/
storing data locally. Externalization of data and apps is considered a viable option for con-
serving energy in mobile devices. This, however, is dependent on the MCC design under
consideration.

2.4 Introduction to Security and Privacy in Mobile Cloud Computing

In the MCC scenario, the security barrier achieves the concealed sections in the protec-
tion of smartphone app architectures. Data protection, user authentication, integrity
protection, privacy protection, risk assessment, securely shared data governance, and
safe transportation are among the security issues. The general security criteria for MCC
may be obtained from the ITU [ITU-T Technical Report XSTR-SEC-MANUAL Security
in Telecommunications and Information Technology (7th Edition), n.d.] and US National
Security Agency (IA Guidance, n.d.) security standards, which are outlined here.

• Confidentiality: Confidentiality is a key need that relates to keeping the data of

mobile users hidden in the cloud. It is a major barrier for mobile consumers to use
different cloud services. Because data is sent and collected via public internet and
stored or analyzed on public cloud infrastructure to deliver cloud services, there
is a danger that it will be revealed to unauthorized individuals.
• Integrity: Data computations and storage are handled by the service provider in
MCC. In this case, the integrity must assure the correctness and consistency of
the data provided by users. In other terms, data integrity avoids undetected data
change by unauthorized individuals or systems. Integrity infringement has an
impact on commercial and other losses of mobile users.
• Availability: MCC ensures that all cloud solutions are always accessible at all
times and in all circumstances as requested by phone users. Keeping services
available includes preventing several forms of unavailability threats that lead to
delays, modifications, or disruptions in network connectivity.
• Authentication and access control: The act or action of verifying the user’s iden-
tity, participant’s data, or program is known as authentication. Following the
Mobile Cloud Computing 23

successful authentication procedure, it is necessary to identify which resources

are authorized to acquire and also what activities, like read, execute, change, or
remove, are allowed. This is known as access control.
• MCC privacy needs: Safety goals such as secrecy, consistency, and identification
encourage anonymity, and these goals explicitly or implicitly preserve the privacy
of cloud service customers on smartphones.

2.5 Security and Privacy Challenges in Mobile Cloud Computing

2.5.1 Challenges Related to Data Security
The most serious data security issue happens as a result of phone devices’ data being
stored and analyzed in clouds located at network operator terminals. Information con-
cerns include loss of data, data leakage, backup and recovery, data localization and data
protection. The loss and compromise of data violates two security standards: consistency
and privacy. Loss of data in this sense refers to user information that was damaged or lost
through any physical force while process, transport, or storage. In a data breach incident,
individuals’ information is acquired, copied, or utilized by intruders. These two can be
caused by malevolent insiders or by malicious apps from outside. Another issue to be
concerned about is data recovery. However, because the users’ data is kept on the prem-
ises of the service providers in cloud service models, firms need to ensure wherever the
information is hosted or kept; thus, data localization is also a challenge. One customers’
information should also be kept distinct from the other. When one user’s data is mixed,
combined, or confounded with the data of other users, it becomes considerably more sus-
ceptible. Whenever information is exported to cloud servers to improve storage space,
phone devices lose physical control of the information at the same time. As a result, in the
cloud storage scenario, one of the issues for mobile users is the accuracy of the data. Even
though data centers are considerably more reliable and efficient than portable devices,
they suffer a plethora of threats and vulnerabilities to data security.

2.5.2 Offloading and Partitioning Security Challenges

Offloading in MCC necessitates the use of wireless networks, over which there is no
control, posing even additional danger. The integrity challenge will be breached if the
offloaded content does not provide the desired outcome. Jamming assaults, availability
attacks, and harmful content threats are among the other concerns. During the offloading
process, wireless networks must be used to connect to the cloud. Furthermore, because
content executions are offloaded to cloud or edge servers rather than mobile devices, there
is a risk of compromising the privacy and consistency of exported data. The security prob-
lem occurs when mobile phones cannot easily check the authenticity of the outcomes after
processing outsourced content if the result is erroneous or altered. The availability of cloud
services can be impacted by jamming attacks that can occur between the information and
the smartphone while segmentation, as well as between the smartphone and the cloud
while offloading. Furthermore, the existence of harmful material between the partitioning
and offloading stages might have an influence on the users’ data confidentiality as well as
breach mobile users’ privacy.
24 Mathematical Modeling for Intelligent Systems

2.5.3 Virtualization
Cloud service providers in MCC deliver cloud services to mobile consumers using virtu-
alization methods. A replica of the portable smartphone’s virtual machine (VM) is pre-
installed in the cloud, as well as the smartphone’s tasks are outsourced to the VM for
execution. This VM is also known as a thin VM or a phone clone. VMs are preloaded in
the cloud so that duties may be offloaded to the VM for execution, which is known as vir-
tualization. The primary goal of virtualization is to provide several VMs that run on the
same physical system. The implementation of this procedure in mobile apps may provide
numerous problems, such as illegal user access and VM-to-VM attack. The primary goal of
virtualization is to offer several VMs that run on the same physical system or mobile device
while being separated from one another. A hypervisor, often known as a VM Monitor
or Manager (VMM), is software that facilitates the creation, operation, and management
of VMs and other virtual subsystems. Once implemented to MCC, even so, virtualiza-
tion technologies create several security threats, such as security issues inside the VMs,
unauthorized users, VM-to-VM strike, security mechanisms within the virtual environ-
ment, security issues inside the Hypervisors, and data security (Sgandurra & Lupu, 2016).
Efficiency, coverage, complexity, security, and resilience are five essential properties of
virtualization-based security techniques (Annane & Ghazali, 2019; Sahu & Pandey, 2018).
The amount of rogue VMs that properly colocate with a targeted (victim VM) based on the
number of VMs created by the hacker is approximately described as efficiency. Security
relates to the secrecy of the VMs and their confidential material, which means no one
must be able to delete or recover the characteristics unless they have the digital signature.
Likewise, resilience relates to the way of tolerance to any sort of manipulation. Any new
technique must achieve a balance between these five fundamental criteria. Many research-
ers have begun proposing frameworks, strategies, and techniques to address these types
of issues in order to assure mobile user security.

2.5.4 Challenges Related to Mobile Cloud Applications

Cloud-based mobile app level can endanger the confidentiality and reliability of both infor-
mation and programs via a number of means, such as the inclusion of malware (Pokharel
et al., 2017; Prokhorenko et al., 2016). Malwares such as viruses, worms, Trojans, rootkits,
and botnets (Qamar et al., 2019) are undesirable, programs, or protocols that are inva-
sive or bothersome. These malwares are intended to run with malicious intent on mobile
devices or to connect to apps without the consent of the user. As a consequence, mobile
app performance can be adjusted. An intruder will often select a target system, insert
malicious software into this one, and then relaunch it. The easiest solution to discover
security threats will be to download and configure security software and antivirus pro-
grams on smartphones. Mobile gadgets, on the other hand, have computation and energy
constraints; defending them from these attacks may be more challenging than securing
traditional PCs. Several techniques for moving threat detection and security procedures to
the cloud have been explored. Before mobile consumers may use a particular application,
it must be subjected to some kind of danger assessment. Instead of running antivirus or
penetration testing programs onsite, smartphones perform simple tasks such as transmit-
ting execution logs to cloud safety servers. Giving out sensitive data, such as your current
location and the user’s vital information, raises privacy problems. Threats to private infor-
mation might be reduced by choosing and assessing company requirements and requiring
just specific services to be bought and migrated to the cloud.
Mobile Cloud Computing 25

2.5.5 Security Challenges Related to Mobile Devices

Physical hazards to mobile devices exist. If mobile devices are forgotten, lost, or stolen,
data or apps may be lost, leaked, accessed, or accidentally disclosed to unauthorized users.
Although password or pattern-based locking mechanisms are available, many smartphone
users may not make use of them. Moreover, the identity module card contained within the
smartphone could be stolen and accessed by unauthorized persons, and the number of
users’ phones lacks an antiattack feature. Intruders can use a variety of available attack
techniques, such as sending a huge volume of unwanted activity and big texts to target
smartphones in addition to making them underutilized or reducing their capabilities. The
authors of Alakbarov (2021) address the constraints that arise in mobile device resources
(energy usage, processing and memory capacities, etc.) as well as the delays that occur in
communication channels. Racic et al. (2006) address this type of assault. They demonstrate
that as a result of this attack. The power of a portable smartphone’s batteries depletes up to
22 times faster than would under normal mode, rendering the gadget unusable in a short
period. Malware developers and hackers are concentrating their efforts due to the popu-
larity of smartphones and digital services. Malware allows attackers to get root access to
mobile devices and manipulate them, which may then directly compromise the computa-
tional integrity of mobile platforms and apps.

2.5.6 Security of Mobile Network

Due to its mobility, a mobile network, based on a standard network, increases the network
node and the access path of users. The networking module is expanded to smartphones
and tablets and so on; smartphones can join the system in a number of ways, for example,
phone users can access cell service, messaging, and other online services via 3G technology.
As a result, additional security concerns, such as sensitive information leakage or mali-
cious assault, would be introduced. For example, many types of public places (such as
cafes, restaurants, and airports) can provide free Wi-Fi. In this situation, there is a risk of
information leaking. Aside from this type of public Wi-Fi, even private Wi-Fi is vulner-
able to security threats due to a flaw in Wi-Fi encryption process. Communication among
smartphones and digital cloud providers is frequently accomplished using a variety of
methods. It gives rise to security concerns.

2.5.7 Problems Related to Mobile Terminal

In general, mobile terminals have the following characteristics: an open OS; support for
third-party applications; “personalization”; and wireless Internet connectivity at any time
and from any location. Because of this, security concerns in mobile terminals are quite
significant.

2.5.7.1 Malware
The mobile terminal’s openness and adaptability constantly attract the attention of attack-
ers. Some malware can be downloaded and carried without the user’s knowledge, along-
side helpful programs and systems. This allows the malware to get unauthorized access to
personal information without the user’s intervention. In order to combat malware, several
security firms have created antivirus software for mobile devices. However, as harmful
assaults get more complicated, antimalware solutions should perform a similar role to that
26 Mathematical Modeling for Intelligent Systems

of the desktop. They have limited capacity and resources, necessitating antimalware solu-
tions that need considerable compute resources difficult to achieve.

2.5.7.2 Vulnerabilities in Software

• Application software: At the moment, the smartphone is the primary mobile ter-
minal and the majority of smartphone customers are used to controlling their
phones via smartphone management system, which manages the information
inside the device through data syncing between the phone and the PC. This pro-
cedure is often carried out via FTP (File Transfer Protocol). The username and
password of FTP are transmitted across the network and recorded in clear text in
the configuration file. This leads to unauthorized access to the cellular telephone
via FTP from PCs on the very same connection, resulting in personal data theft as
well as unauthorized access, intentional deletion, and harmful manipulation. In
reality, application software vulnerabilities are quite frequent. Since the software
is somewhat lax, attackers can infiltrate the mobile phone via a flaw in the applica-
tion program.
• Operating system: The OS is in charge of managing and controlling software
and hardware resources. And because the software is so sophisticated, it may
have coding flaws. In certain cases, attackers will utilize these flaws to destroy the
mobile phone.
• Others: Furthermore, security concerns in mobile terminals are still caused by
mobile users. Mobile users frequently lack security knowledge; second, mobile
users themselves may make mistakes. As a result, companies must identify and
prohibit unusual user activity.

2.5.8 Mobile Cloud

• Platform reliability: Because of the huge concentration of user information

resources, the cloud platform is vulnerable to assault. The dangerous suspect’s
objective, on the one hand, is to acquire important information and services. Such
attacks could come from malicious outsiders, a genuine CC user, or a part of the
CC operator’s team. The hostile attacker’s goal, on the other hand, is to shut down
the cloud service. For example, a DOS (denial of service) assault will damage the
platform’s availability and force the cloud’s service to be shut down. Users risk
data loss if they submit all of their data to Cloud Service Providers (CSPs) before
subscribing for the more costly backup and recovery option. Such incidents have
occurred on occasion at cloud providers in recent years. As a result, the cloud pro-
vider should incorporate modern security technology to assure the service, and
customers must not be overly reliant on the CSP.
• Privacy and data protection: Privacy and security preservation are major prob-
lems in the mobile cloud. To start, inside the clouds, user information manage-
ment and governance are segregated, which causes users’ concerns over their own
information resource to become a significant impediment to the widespread adop-
tion of MCC. As a result, consumers’ sensitive information is at a greater danger of
being exposed (Feng et al., 2011).
Mobile Cloud Computing 27

2.5.9 Privacy Challenges
One of the key problems is privacy, which arises when mobile users’ sensitive data or apps
are processed and sent from mobile devices to heterogeneously dispersed cloud servers
when using various cloud services. Such computers are located in diverse locations owned
and managed completely by the providers. In this scenario, consumers cannot personally
be monitoring the preservation of their data, so problems regarding data privacy issues
are in the hands of service suppliers, and consumers are not held accountable for privacy
violations. The use of cloud storage and processing in various places raises privacy con-
cerns (Gupta et al., 2017a, b). Furthermore, it is critical for consumers to obtain information
about the cloud hosting location because the legislation varies from one nation to another.
There are several mobile applications accessible that may be dangerous owing to having
different functionalities, gathering unintentionally users’ personal information such as
hobbies and whereabouts, and spreading unlawfully. Unwanted advertising emails, often
known as trash emails, can infringe on consumers’ privacy (Gupta et al., 2018). One of the
key differences between mobile apps and PCs is context awareness, which is provided via
sensors on mobile devices. Moreover, many systems involve and gather location informa-
tion from users, which may then be used to directly market to clients depending on their
whereabouts. As a result, location-based services pose privacy concerns since they must
collect, store, and process user information.

2.6 Current Solutions for the Security and Privacy Issues

2.6.1 Data Security
In Alqahtani & Kouadri-Mostefaou (2014), the authors propose a system based on dis-
tributed multicloud storage, encryption keys, and compression algorithms techniques to
ensure data protection in MCC. At the mobile phone end, data was originally divided
into discrete segments based on user preference, and the sections are then secured and
reduced. Finally, it keeps data on a decentralized multicloud platform. However, in order
to increase security, the user can keep one section on its own storage. A nearly identical
technique is given in Abdalla & Pathan (2014). The information is updated to a CSP end
using a data security administrator that is split into two portions. The first one is a data
fragmenter that splits information and sends it to many sites for preservation. The second
is data merging, which combines data as the user downloads from the stored data. Wang
et al. (2014) suggest a security strategy for protecting mobile users’ data in the media cloud.
The authors in (Lai et al., 2013) proposed an attribute-based data strategy. This technique
is the combination of ciphertext attribute-based authentication (CP-ABE) and encryption
algorithm. It is only suitable for CC and not for resource-constrained smartphones, as the
current CP-ABE has severe performance constraints due to its large cipher size and pro-
cessing power.

2.6.2 Partitioning and Offloading

Al-Mutawa and Mishra (2014) present a data-splitting technique for preventing cellu-
lar user data vulnerability while transferring to a remote and trusted or untrustworthy
28 Mathematical Modeling for Intelligent Systems

business. This concept is divided into three stages. First, based on the user’s preferences,
the data is categorized as nonsensitive or sensitive. The sensitive element is subsequently
handled on the smartphone or by a trusted distant entity, whereas the nonsensitive part
is transferred to another entity. TinMan is proposed by the authors of Xia et al. (2015) to
maintain anonymity and to carry out security-oriented shifting from portable devices to
the cloud. To limit vulnerability, critical data from mobile devices is isolated from mobile
applications, which are subsequently downloaded and kept on a neighboring device. A
reliable node is installed on a virtual computer in a reliable cloud. The TinMan is in charge
of determining how efficiently and cleanly this unloading process is carried out. Saab
et al. (2015) suggest a cloud-based smartphone app offloading that is both secure and
energy-efficient. A profiling engine, a deception engine, and an offloading engine are all
part of this system. The decision-making and profiling algorithms are used for continuous
categorization of SMDs in order to enhance power consumption and safety. Following the
deception, the entire or partial component is offloaded to the cloud using the offloading
engine. Langdon et al. (2015) propose and test a method for offloading mobile applications
while maintaining privacy. The privacy of the user information is maintained within the
mobile device during unloading in this technique. The experiment results demonstrate
that the proposed technique protects privacy while conserving energy throughout the
offloading process.

2.6.3 Virtualization
Vaezpour et al. (2016) offer SWAP, a security-aware provisioning and migration approach
for cellphone clones. Hao et al. (2015) explain SMOC, a secure mobile cloud platform. This
technique helps phone devices to safely copy their OS and applications to cloud-based
VMs. To maintain the protection of user information on smartphones, a hardware virtu-
alization technique is used that enables data and application functionality to be separated
from the mobile service device’s operating system. Paladi et al. (2017) provide a security
architecture for cloud infrastructure. Before deploying guest VMs, the trusted VM launch-
ing protocol is utilized.

2.6.4 Mobile Cloud Applications

Tan et al. (2015) offer a paradigm called STOVE (Strict, Observable, Verifiable Data, and
Execution) for safely executing untrusted mobile applications. Once an unknown software
starts and runs, it is not capable of causing harm to other programs or the OS, and it does
not access any unauthorized programs or data. This model consists of three stages. In the
first stage, the STOVE strictly restricts and completely isolates untrusted programs from
other applications in the environment in which it operates. STOVE then uses formal veri-
fication methods to validate the tightly isolated code in the second stage. Finally, STOVE
performs all data access in favor of the untrusted program so that all data access is visible.
Tang et al. (2016) provide a token-based three-factor strong Application Programming
Interface (API) security architecture. These are (a) fundamental identification, like API
registration process with robust strong passwords; and (b) contemporary security tech-
niques, such as messaging security requirements, online signatures, and internet cryptog-
raphy; and (c) as a third prevention measure, the main protein inside API and its backward
compatibility.
Mobile Cloud Computing 29

2.6.5 Mobile Devices
While customers retain and use confidential material, app developers should pay atten-
tion to the two issues: developers can implement a layer of security as an added step in
the application level, and then they can ensure that customers’ sensitive information also
isn’t kept in the identification module card. Moreover, cloud backup solutions are critical
for customers because if a customer’s mobile is stolen or their information is deleted, the
user can restore his data from the cloud. At the moment, unique security mechanisms are
integrated into mobile devices. There are numerous security programs for mobile devices
available, each with its own set of functions such as detecting and preventing mobile
viruses, controlling unwanted access, and protecting privacy.

2.6.6 Mobile Network
The two ways to protect mobile networks are data encryption and security protocol. In
data encryption, only encrypted information is reasonably safe while transmission over
the mobile network, regardless of how the mobile terminals access the network. The inves-
tigation of the security protocol is essential for all types of access methods. Rupprecht
et al. (2018) aim to compile security information on mobile phone networks into a complete
overview and to generate pressing open research topics. To accomplish this methodically,
they create a framework that categorizes known attacks based on their purpose, sug-
gested countermeasures, underlying reasons, and root causes. In addition, they evaluate
the impact and efficacy of each assault and response. This technique is then applied to
existing research on attacks and responses in all three network generations. This allows
them to identify ten causes and four root causes of assaults (Tomar, 2020; Fore et al., 2016).
Mapping the assaults to suggested countermeasures and 5G specification recommenda-
tions are helpful in identifying outstanding research topics and obstacles for the construc-
tion of next-generation mobile networks. The NEMESYS method to smart mobile network
security is presented by the authors of Gelenbe et al. (2013). The NEMESYS project seeks to
deliver novel security mechanisms for seamless network services in the intelligent phone
environment and to enhance cellular data security through a better knowledge of the lat-
est threats. It will gather data on the nature of cyber-attacks on smart devices and the
network architecture in order to execute appropriate measures for protection.

2.6.7 Mobile Terminal
2.6.7.1 Antimalware
To safeguard the mobile terminal from attack, two steps must be taken: the first step is
to investigate and eliminate virus. Virus monitoring might be relocated to the cloud to
avoid the resource restrictions of mobile terminals. If virus is detected, lawful application
form of the cloud can be allocated to a mobile terminal and run to remove it. CloudAV is
a perfect example of antimalware software. CloudAV (CloudAV: N-Version Antivirus in
the Network Cloud, n.d.) is a novel paradigm for malware detection on mobile devices
that is built on offering antivirus as an in-cloud network service. CloudAV offers the fol-
lowing significant advantages: improved identification of dangerous software; the effect
of antivirus loopholes is reduced; previously affected systems are detected retroactively;
forensics abilities are enhanced; and modifiability and administration are improved. It
also includes bridge hosting agents and a data network featuring ten antivirus programs
30 Mathematical Modeling for Intelligent Systems

and two behavioral recognition algorithms that improves identification substantially. The
other is antimalware protection. Users must be cautious of their actions in order to avoid
malicious software from being installed on their mobile devices.

2.6.7.2 Software Vulnerabilities
Consumers must give importance to smartphones OS version notifications and download
patches or revised versions from of the OSs research and development firm as soon as
possible. In the meantime, they must be cautious while installing third-party programs.
On the other hand, in order to decrease software vulnerabilities, they need to implement a
number of technological steps. Checking the validity and integrity of software, for exam-
ple, is a crucial practice before using it.

2.6.8 Mobile Cloud
The MCC platform’s dependability and availability are critical for both cloud providers
and consumers. To begin with, CSPs should use contemporary security mechanisms such
as VPN technology, authentication and authorization, cryptography, and other technical
approaches to assure high supply from different attacks. Second, CSPs should provide
comprehensive recovery and backup solutions in order to retrieve users’ data in the event
of a major assault. By using these methods, the cloud platform may improve service qual-
ity and boost customer confidence. Encryption technology is required for sensitive data
to survive the transition from storage to transportation. However, because encryption
reduces data usage, emphasis shifts to effectively interpreting and processing the cipher-
text. The privacy homomorphism algorithm is the current focus of ciphertext processing
research. Meanwhile, key management is a critical task for corporate users.

2.6.9 Privacy
Pasupuleti et al. (2016) suggest a method that protects the confidentiality of smartphone
data that has been transferred to the cloud. A stochastic public-key cryptography method
and a ranking keyword searching engine are employed in this situation. After creating an
indexing for file collecting, the phone node secures both the information and the search
while transmitting it to be hosted in the system. Bahrami and Singhal (2015) describe a
lightweight authentication algorithm for smartphones to store information on clouds. This
method is dependent on the effect of pseudo-random permutation. Since it is compact, the
permutations process is done on a smartphone instead of the cloud to protect information
confidentiality.

2.7 D iscussion and Conclusion

Private cloud software solutions can be utilized to safeguard your privacy and security.
Secured cloud-based services offer smartphones access control, identity management,
on-demand cryptography, intrusion prevention, identification, and authorization in addi-
tion to confidentiality and protection in order to provide a secured channel of interaction
between the cloud and the smartphone. Virtualization increases cloud resource efficiency
Mobile Cloud Computing 31

but raises additional security concerns owing to the absence of complete isolation of VMs
housed on a single server. VM safe monitoring, mirroring, and migration can assist to
address some of the security concerns faced by virtualization. Smartphone users should
be allowed to verify the level of security of managed services in order to establish a visible
cloud environment. Cloud storage monitoring can be used to aid with the audits. The
resource-constrained mobile devices pose a major threat to the rapid expansion of mobile
computing. However, by integrating mobile computing with cloud computing, the expan-
sion of mobile computing may be accelerated, resulting in the emergence of a new com-
puting paradigm known as MCC. This chapter provides details about the introduction,
architecture, and service models of MCC. It covers the details in depth about the security
and privacy challenges in different domains of MCC and concludes with the solutions to
the existing problems in MCC.

References
Abdalla, A. K. A., & Pathan, A. S. K. (2014). On protecting data storage in mobile cloud comput-
ing paradigm. IETE Technical Review (Institution of Electronics and Telecommunication Engineers,
India), 31(1), 82–91. https://doi.org/10.1080/02564602.2014.891382
Alakbarov, R. G. (2021). Challenges of mobile devices’ resources and in communication channels
and their solutions. International Journal of Computer Network and Information Security, 13(1),
39–46. https://doi.org/10.5815/IJCNIS.2021.01.04
Al_Janabi, S., & Hussein, N. Y. (2019). The reality and future of the secure mobile cloud com-
puting (SMCC): Survey. Lecture Notes in Networks and Systems, 81, 231–261. https://doi.
org/10.1007/978-3-030-23672-4_18
Al-Mutawa, M., & Mishra, S. (2014). Data partitioning: An approach to preserving data privacy in
computation offload in pervasive computing systems. In Q2SWinet 2014- Proceedings of the 10th
ACM Symposium on QoS and Security for Wireless and Mobile Networks, pp. 51–60. https://doi.
org/10.1145/2642687.2642696
Alqahtani, H. S., & Kouadri-Mostefaou, G. (2014). Multi-clouds mobile computing for the secure
storage of data. In 2014 IEEE/ACM 7th International Conference on Utility and Cloud Computing,
pp. 495–496. https://doi.org/10.1109/UCC.2014.68
Annane, B., & Ghazali, O. (2019). Virtualization-based security techniques on mobile cloud comput-
ing: Research gaps and challenges. International Journal of Interactive Mobile Technologies (IJIM),
13(4), 20–32. https://doi.org/10.3991/IJIM.V13I04.10515
Bahrami, M., & Singhal, M. (2015). A light-weight permutation based method for data privacy in
mobile cloud computing. In 2015 3rd IEEE International Conference on Mobile Cloud Computing,
Services, and Engineering, 189–196. https://doi.org/10.1109/MOBILECLOUD.2015.36
CloudAV: N-Version Antivirus in the Network Cloud (n.d.). https://www.usenix.org/legacy/event/
sec08/tech/full_papers/oberheide/oberheide_html/index.html
Feng, D. G., Zhang, M., Zhang, Y., & Xu, Z. (2011). Study on cloud computing security. Ruan Jian Xue
Bao/Journal of Software, 22(1), 71–83. https://doi.org/10.3724/SP.J.1001.2011.03958
Fore, V., Khanna, A., Tomar, R., & Mishra, A. (2016). Intelligent supply chain management system.
In International Conference on Advances in Computing and Communication Engineering (ICACCE),
pp. 296–302. https://doi.org/10.1109/ICACCE.2016.8073764
Gelenbe, E., Görbil, G., Tzovaras, D., Liebergeld, S., Garcia, D., Baltatu, M., Lyberopoulos, G., Gelenbe,
E., Görbil, G, Görbil, G., Tzovaras, D., Liebergeld, S., Garcia, D., & Baltatu, M. (2013). NEMESYS:
Enhanced network security for seamless service provisioning in the smart mobile ecosystem.
In Information Sciences and Systems 2013 Lecture Notes in Electrical Engineering, vol. 264, pp. 369–
378. https://doi.org/10.1007/978-3-319-01604-7_36
32 Mathematical Modeling for Intelligent Systems

Gupta, M., Benson, J., Patwa, F., & Sandhu, R. (2017a). Dynamic groups and attribute-based access
control for next-generation smart cars. In Proceedings of the Ninth ACM Conference on Data and
Application Security and Privacy, pp. 61–72. 2019.
Gupta, M., Patwa, F., & Sandhu, R. (2017b). Object-tagged RBAC model for the Hadoop ecosystem. In
IFIP Annual Conference on Data and Applications Security and Privacy, pp. 63–81. Springer, Cham.
Gupta, M., Patwa, F., & Sandhu, R. (2018). An attribute-based access control model for secure big data
processing in Hadoop ecosystem. In Proceedings of the Third ACM Workshop on Attribute-Based
Access Control, pp. 13–24.
Hao, Z., Tang, Y., Zhang, Y., Novak, E., Carter, N., & Li, Q. (2015). A secure SMOC mobile cloud
computing platform. In 2015 IEEE Conference on Computer Communications (INFOCOM),
pp. 2668–2676. https://doi.org/10.1109/INFOCOM.2015.7218658
IA Guidance (n.d.). https://apps.nsa.gov/iaarchive/library/ia-guidance/
ITU-T Technical Report XSTR-SEC-MANUAL Security in telecommunications and information
technology (7th edition) (n.d.).
Lai, J., Deng, R. H., Guan, C., & Weng, J. (2013). Attribute-based encryption with verifiable out-
sourced decryption. IEEE Transactions on Information Forensics and Security, 8(8), 1343–1354.
https://doi.org/10.1109/TIFS.2013.2271848
Langdon, R. J., Yousefi, P. D., Relton, C. L., & Suderman, M. J. (2015). Privacy-preserving offloading of
mobile app to the public cloud. Clinical Epigenetics. https://doi.org/10.2/JQUERY.MIN.JS
Nike. Just Do It. Nike IN. (n.d.). https://www.nike.com/in/
Noor, T. H., Zeadally, S., Alfazi, A., & Sheng, Q. Z. (2018). Mobile cloud computing: Challenges and
future research directions. Journal of Network and Computer Applications, 115, 70–85. https://doi.
org/10.1016/J.JNCA.2018.04.018
Paladi, N., Gehrmann, C., & Michalas, A. (2017). Providing User Security Guarantees in Public
Infrastructure Clouds. IEEE Transactions on Cloud Computing, 5(3), 405–419. https://doi.
org/10.1109/TCC.2016.2525991
Pasupuleti, S. K., Ramalingam, S., & Buyya, R. (2016). An efficient and secure privacy-preserving
approach for outsourced data of resource constrained mobile devices in cloud comput-
ing. Journal of Network and Computer Applications, 64, 12–22. https://doi.org/10.1016/J.
JNCA.2015.11.023
Pokharel, S., Choo, K. K. R., & Liu, J. (2017). Mobile cloud security: An adversary model for light-
weight browser security. Computer Standards and Interfaces, 49, 71–78. https://doi.org/10.1016/J.
CSI.2016.09.002
Prokhorenko, V., Choo, K. K. R., & Ashman, H. (2016). Web application protection techniques: A
taxonomy. Journal of Network and Computer Applications, 60, 95–112. https://doi.org/10.1016/J.
JNCA.2015.11.017
Qamar, A., Karim, A., & Chang, V. (2019). Mobile malware attacks: Review, taxonomy & future
directions. Future Generation Computer Systems, 97, 887–909. https://doi.org/10.1016/J.
FUTURE.2019.03.007
Racic, R., Ma, D., & Chen, H. (2006). Exploiting MMS vulnerabilities to stealthily exhaust mobile phone’s
battery. In 2006 Securecomm and Workshops. https://doi.org/10.1109/SECCOMW.2006.359550
Rupprecht, D., Dabrowski, A., Holz, T., Weippl, E., & Popper, C. (2018). On security research towards
future mobile network generations. IEEE Communications Surveys and Tutorials, 20(3), 2518–
2542. https://doi.org/10.1109/COMST.2018.2820728
Saab, S. A., Saab, F., Kayssi, A., Chehab, A., & Elhajj, I. H. (2015). Partial mobile application offloading
to the cloud for energy-efficiency with security measures. Sustainable Computing: Informatics
and Systems, 8, 38–46. https://doi.org/10.1016/J.SUSCOM.2015.09.002
Sahu, I., & Pandey, U. S. (2018). Mobile cloud computing: Issues and challenges. In 2018 International
Conference on Advances in Computing, Communication Control and Networking (ICACCCN),
pp. 247–250. https://doi.org/10.1109/ICACCCN.2018.8748376
Sgandurra, D., & Lupu, E. (2016). Evolution of attacks, threat models, and solutions for virtualized
systems. ACM Computing Surveys (CSUR), 48(3), 1–38. https://doi.org/10.1145/2856126
Mobile Cloud Computing 33

Tan, J., Gandhi, R., & Narasimhan, P. (2015). STOVE: Strict, observable, verifiable data and exe-
cution models for untrusted applications. In Proceedings of the International Conference on
Cloud Computing Technology and Science, CloudCom, pp. 644–649. https://doi.org/10.1109/
CLOUDCOM.2014.116
Tang, L., Ouyang, L., & Tsai, W. T. (2016). Multi-factor web API security for securing Mobile
Cloud. In 2015 12th International Conference on Fuzzy Systems and Knowledge Discovery (FSKD),
pp. 2163–2168. https://doi.org/10.1109/FSKD.2015.7382287
Tomar R. (2020). Maintaining trust in VANETs using blockchain. ACM SIGAda Ada LettersAda. 40,
91–96. https://doi.org/10.1145/3431235.3431244
Vaezpour, S. Y., Zhang, R., Wu, K., Wang, J., & Shoja, G. C. (2016). A new approach to mitigating
security risks of phone clone co-location over mobile clouds. Journal of Network and Computer
Applications, 62, 171–184. https://doi.org/10.1016/J.JNCA.2016.01.005
Wang, H., Wu, S., Chen, M., & Wang, W. (2014). Security protection between users and the
mobile media cloud. IEEE Communications Magazine, 52(3), 73–79. https://doi.org/10.1109/
MCOM.2014.6766088
Xia, Y., Liu, Y., Tan, C., Ma, M., Guan, H., Zang, B., & Chen, H. (2015). Tinman: Eliminating confiden-
tial mobile data exposure with security oriented offloading. In Proceedings of the 10th European
Conference on Computer Systems, EuroSys 2015. https://doi.org/10.1145/2741948.2741977
3
A Review on Live Memory Acquisition
Approaches for Digital Forensics

Sarishma Dangi and Diksha Bisht

Graphic Era Deemed to be University

CONTENTS
3.1 Introduction .......................................................................................................................... 36
3.2 Related Work ........................................................................................................................ 37
3.3 Technical Background ......................................................................................................... 38
3.3.1 Direct Memory Access ............................................................................................ 39
3.3.2 IEEE 1394 ................................................................................................................... 39
3.3.3 System Management Mode .................................................................................... 40
3.3.4 x86 Architecture ....................................................................................................... 41
3.3.5 Virtualization Support ............................................................................................ 41
3.4 Review of Memory Acquisition Techniques ....................................................................42
3.4.1 Hardware-Based Memory Acquisition.................................................................42
3.4.1.1 Dedicated Hardware ................................................................................42
3.4.1.2 Special Hardware Bus ..............................................................................42
3.4.1.3 Cold Booting ..............................................................................................42
3.4.1.4 Limitations of Hardware-Based Techniques ........................................43
3.4.2 Software-Based Memory Acquisition ...................................................................43
3.4.2.1 User Level ...................................................................................................43
3.4.2.2 Kernel Level ............................................................................................... 45
3.4.2.3 Virtualization Based ................................................................................. 46
3.4.2.4 Limitations of Software-Based Techniques........................................... 46
3.4.3 Firmware-Based Memory Acquisition ................................................................. 46
3.5 Memory Dump Formats ..................................................................................................... 46
3.5.1 Raw Memory Dump ................................................................................................ 47
3.5.2 Windows Crash Dump ........................................................................................... 47
3.5.2.1 Complete Memory Dump ........................................................................ 47
3.5.2.2 Kernel Memory Dump ............................................................................. 47
3.5.2.3 Small Memory Dump ............................................................................... 48
3.5.3 Windows Hibernation File ..................................................................................... 48
3.5.4 Expert Witness Format............................................................................................ 48
3.5.5 AFF4 Format ............................................................................................................. 48
3.5.6 HPAK Format ........................................................................................................... 49
3.6 Memory Acquisition Tools ................................................................................................. 49
3.6.1 Magnet RAM Capture............................................................................................. 50
3.6.2 Belkasoft Live RAM Capturer ............................................................................... 50

DOI: 10.1201/9781003291916-3 35
36 Mathematical Modeling for Intelligent Systems

3.6.3 FTK Imager ............................................................................................................... 50

3.6.4 DumpIt ...................................................................................................................... 50
3.7 Implementation .................................................................................................................... 50
3.8 Performance Comparison ................................................................................................... 52
3.9 Conclusion ........................................................................................................................... 59
References....................................................................................................................................... 59

3.1 Introduction
Digital forensics is defined as the collection, preservation, examination, and presentation
of digital evidence in a coherent, verifiable manner at a court of law using scientifically and
technologically driven methods (Beebe, 2009). In the case of digital forensics, data present
in digital systems is of great significance as it serves as strong baseline evidence in relation
to cybercrimes being committed. Recently, Google suffered a major outage on December
14, 2020 that lasted for about 45 minutes, amounting to serious monetary and service loss
(2020_Google_services_outages @ En.Wikipedia.Org, n.d.).
Since 2000s, malicious attackers have become sophisticated as well as complex in their
approach to attacking and stealing. Various obfuscation based techniques (Harichandran
et al., 2016), hiding via antiforensics based (Palutke et al., 2020) techniques, are now being
actively used. Advanced malware can hide and is often encrypted, making it much more diffi-
cult for the incident response team to determine the true nature of the malware (Rathnayaka &
Jamdagni, 2017). Memory analysis techniques can be used to acquire and examine certain
malware behaviors in their native form while they are running in memory. Live memory
forensics aids in acquiring and analyzing such activities within the digital systems.
Memory forensics is a specialized realm of acquiring, recovering, interpreting, and
analyzing evidence present in the volatile memory of digital systems by using various
approaches and techniques. Fast-paced evolution of cybercrime landscape has necessi-
tated the presence of this area. Memory forensics is used to analyze physical or volatile
memory; more commonly known as RAM (Random Access Memory) in order to collect
the evidence by retrieving data from the confiscated device that was used at the commis-
sion of a crime. It is about acquiring the memory first and then identifying as well as ana-
lyzing it depending upon the case being investigated at hand. It apprehends the memory
contents that is a prominent tool for malware analysis and digital forensic investigators.
Acquisition is a process of collecting evidence from the confiscated device through
which the crime is committed by using various techniques. There are two types of meth-
odologies that are followed for acquiring the image of digital evidence and these are Live
Acquisition and Dead Acquisition.
The data stored on the hard disk drive and within the file system is not lost because the
storage is nonvolatile in nature. However, for volatile memory such as RAM, the data is
lost as soon as the power is cut off. Once the power has been cut off for a period of time,
the state of the data becomes completely unknown. This traditional approach is more com-
monly known as “pulling the plug” where the focus of analysis is on nonvolatile stor-
age. This is now more commonly referred to as Dead Memory Acquisition. It refers to the
powered-off or dead system, which is turned off and no data processing is taking place. It
is performed by pulling the plug methodology to avoid any malicious process from run-
ning and potentially deleting evidence. There are fewer chances of data modification. It
has practical, legal constraints and lacks standardized issues.
In the early 2000s, most investigators simply used to capture the picture of the screen of
what was open at the moment and would pull the plug out of the system. In cases where
Digital Forensics 37

the system was live and encrypted, the investigator easily lost the mere opportunity to
access the system state (Beebe, 2009). If pull the plug methodology is followed, it allows
little or no insight into what operations the processor was executing at the time when the
power was cut off. This severely limits the investigative process.
As the processor’s computational ability and memory increased, the need to analyze it
also increased since many applications started to execute only in-memory leaving no trace
behind. Live Acquisition on the other hand refers to live systems that are up and running
where information may be altered as data is continuously processed. It is performed by
using specialized tools and techniques to extract volatile data from the computer before
turning it off. It can easily retrieve volatile information. The acquisition of RAM can further
be performed via two different ways, i.e., hardware- and software-based memory acquisi-
tion. Sometimes firmware-based memory acquisition is also possible (Stüttgen et al., 2015).
The key contributions of this work are outlined as follows:

• To provide a critical and comprehensive review of the existing memory acquisi-

tion approaches.
• To present a detailed overview of memory acquisition tools prominent in the
market and outline their limitations.
• To compare the performance of prominent memory acquisition tools on two
different machines operating on Windows platform with 4 and 8GB RAM.
• To assess the performance of the machine when the tools are executing in terms of
processor utilization, disk utilization, and time taken for acquisition.

The rest of this paper is categorized as follows: Section 3.2 outlines the related work in
this area. Section 3.3 provides the technical background which is necessary to give a con-
text for further discussion. Section 3.4 gives a comprehensive review of various hardware-
and software-based acquisition approaches for memory acquisition. Section 3.5 discusses
the various memory dump formats that are provided by using memory acquiring tools.
Section 3.6 includes various prominent memory acquisition tools along with their detailed
comparison. Section 3.7 provides the testbed setup and methodology required for imple-
menting memory acquisition tools in the system. Lastly, we conclude our work.

3.2 Related Work
Memory forensics was highlighted and worked upon after the initial impetus provided
by the DFRWS challenge in 2005 (Reust & Friedburg, 2006). A trend was observed in the
research community toward acquisition, and analysis of volatile memory in particular
post this challenge. Acquisition techniques faced some setbacks as discussed further,
whereas analysis techniques continue to face resistance and are still worked on.
Figure 3.1 represents the research contribution in the area of volatile memory acquisi-
tion over the last decade. It is clear that the contributions started to significantly increase
once the DFRWS challenge was launched. This impetus fed not only the progress of the
research community but fed directly into the industry-based development. The result was
a steep rise in the presence of write blockers and encryption alternatives which prevented
reliable memory acquisition. This led to a downfall in research interest in the lieu of lim-
ited support from industry and government-backed organizations.
38 Mathematical Modeling for Intelligent Systems

FIGURE 3.1
Year-wise contribution in research publications.

In this section, we analyze and review different research works that are done in the
area of volatile memory acquisition for live systems. The methodologies, techniques,
research works, and anything other than memory acquisition-based are out of the
scope of this discussion. In particular, we have excluded analysis-based discussion
from this work.
Brian D. Carrier and Grand (2004) presented a procedure for acquiring volatile memory
using a hardware expansion card that can copy memory to an external storage device
(Carrier & Grand, 2004). With the flow of passing time, Acquisition of Random-Access
Memory (RAM) was described into two types, i.e., hardware acquisition and software
acquisition as mentioned in the work of Vidas (2007). Vömel and Freiling (2011) in their
work proposed various main memory acquisition approaches on the basis of two fac-
tors, atomicity and availability, and also provided analysis techniques for the Windows
Operating system family. Classification of memory accessibility levels on the basis of pre-
or postincident deployment and terminating vs nonterminating acquisition was provided
in their work by Latzo et al. (2019). Through the years, many research works have been
done in this field. Table 3.1 provides the review of these works with a focus on the purpose
of work and the implementation, if any. Table 3.2 provides a review of the advantages and
limitations of these works.

3.3 Technical Background
In this section, we provide the necessary technical background required for better under-
standing of rest of the work. Some generic concepts such as DMA, IEEE 1394, SMM,
Virtualization support are described in this section. We have used Intel’s x86-64 architec-
ture to demonstrate these principles as it is widely deployed and well documented.
Digital Forensics 39

TABLE 3.1
Review of Existing Work on Memory Acquisition
Author’s Name
and Year Purpose of the Work Proposed and Implementation (If Any)
Carrier and Procedure for obtaining volatile memory with a hardware expansion card, as well as general
Grand (2004) memory acquisition tool specifications. Tribble: The Proof-of-Concept device was designed to
ensure that the protocol was successful
Vidas (2007) The advantages and disadvantages of conventional incident management approaches versus
an augmented model were discussed, as well as a basis for evaluating the captured memory.
On systems with 512MB of RAM and an IBM ThinkPad R51, tests were performed.
Vömel & Freiling An overview of the prevailing techniques and methods to collect and analyze a computer’s
(2011) memory. No specific implementation is provided.
Balogh & Mydlo Proposed a new method for activating DMA, as well as a new approach for memory
(2013) acquisition via DMA solution based on a custom NDIS protocol driver. The NDIS driver and
Network Interface Card were used to implement a method that permitted the memory
content to be dumped.
Stüttgen et al. Variety of firmware-based manipulation techniques and methods are presented for identifying
(2015) firmware-level threats. Two plug-ins for the Volatility and Rekall forensic frameworks were
created for inspection of ACPI environment.
McDown et al. Evaluated widely used seven shareware or freeware/open-source RAM acquisition forensic
(2016) tools that are compatible to work with the latest 64-bit Windows Operating systems. Internet
connection was not present while conducting the experiments on 2 and 4GB of physical
memory.
Parekh & Jani Analyzing the memory that stores relevant data, collection of evidences, extraction of crucial
(2018) information using various memory forensic tools. No practical implementation was done.
The tools are defined and compared theoretically.
Sokolov et al. Proposed a modern instrumental approach for retrieving the contents of a volatile memory, as
(2018) well as general implementation guidelines for retrieving data from memory. It just offers a
theoretical framework for investigating security incidents, with no practical implementation.
Latzo et al. (2019) A taxonomy of acquisition methods that generalizes the principle of ring-based privilege
separation and is based on a well-defined partial order. It provides the techniques for
memory acquisition and the process to achieve this.

3.3.1 Direct Memory Access

Traditionally, the processor was entirely used to initiate, execute, and terminate the mem-
ory access operations. This led to overloading as well as unavailability of the processor
during the entire duration when memory transfer was being conducted, thus severely
limiting the processor’s ability to execute other workloads. As a solution to this issue,
Direct Memory Access (DMA) was introduced. It’s a method of transferring data between
peripheral devices and the system’s main memory without involving CPU. Since the CPU
is only used to initiate the transfer and not to wait for it to finish, it is not used for the
entire duration of read and write operations. The DMA controller sends an interrupt to the
CPU to signal the end of the transfer. DMA requires direct access to the physical memory.
Carrier and Grand (2004) proposed “Tribble,” a Proof-of-Concept approach that creates a
copy of physical memory using DMA operations (Vömel & Freiling, 2011).

3.3.2 IEEE 1394
IEEE 1394 (Witherden, 2010) also known as Firewire is a hardware bus potentially used
for accessing physical memory. It is a serial expansion bus that allows one device on the
bus to directly read/write from the physical memory to another. Most operating systems
40 Mathematical Modeling for Intelligent Systems

TABLE 3.2
Advantage & Limitation of the Existing Work
Author’s Name and Year Advantage and Limitation of the Work
Carrier and Grand (2004) Provides general process of volatile memory acquisition and more reliable
hardware-based evidence than software-based solution.
Prior to an incident, the hardware-based acquisition system must be installed.
Vidas (2007) Techniques for RAM acquisition and Time Sliding Window are provided along
with an overview of RAM analysis.
Operating System level malicious activity detection can be improved; current
tools only support specific runtime environments.
Vömel & Freiling (2011) Classification of volatile memory acquisition is provided on the basis of atomicity
and availability.
The work is oriented toward traditional approaches without providing the path
for future of acquisition
Balogh & Mydlo (2013) Study and implementation of direct memory access that satisfies the
requirements for rootkit detection and for forensic analysis.
The installation of the modified driver can change the contents of the memory
and could not be used if a sophisticated rootkit is already installed.
Stüttgen et al. (2015) Rootkits and other malicious applications ability to exploit x86 systems at the
firmware level is explored.
Software-based memory forensic methods are unable to identify certain firmware
rootkits.
McDown et al. (2016) Memory acquisition software usage, user interface, processing time, loaded
DLLs, and Windows registry keys are retrieved.
The implementation is limited to 2 and 4GB of physical memory, while existing
industry standards call for 16–32GB
Parekh & Jani (2018) Comparison between different tools and techniques for memory acquisition is
provided.
Different tools have been compared but only theoretically and no practical
implementation is done.
Sokolov et al. (2018) Theoretical foundation for examining information security incidents, as well as a
novel approach for accessing a hardware copy of the contents of a virtual
machine.
De-energizing the target computer results in the complete or partial destruction
of information.
Latzo et al. (2019) Classification on the basis of three factors: privilege level, pre- or postincident,
and potential termination.
On embedded systems, memory acquisition is still a hassle.

automatically configure 1394 devices upon insertion, and this can be done without requiring
administrator privileges on the system. Even if the system is locked, it will function. The prob-
lem is that an experienced intruder with administrator access can easily change permissions.

3.3.3 System Management Mode

System Management Mode (SMM) is a mode in which system architectures contain specific
operating modes that are entirely independent of all other aspects of the system’s operation.
Low-level management functions such as power management and legal interface emula-
tion are handled by SMM. The unmaskable System Management Interrupt (SMI) initiates
SMM. Following the trigger, the CPU is transferred to another address space known as
SMRAM, where it stores its current context. A corresponding SMI handler is then exe-
cuted as a result of this operation. SMI handlers are installed to SMRAM until and unless
Digital Forensics 41

D_ LAC bit in the System Management Control Register is unlocked (Latzo et al., 2019).
Prior to a system’s boot operation, BIOS performs this task. Once the installation is com-
plete, the BIOS can lock the SMRAM, preventing any more modifications.
In SMM, only 4GB of RAM can be accessed because the address mode is equivalent to
16-bit real mode. The SMRAM can be physically located in system memory or a separate
RAM. The saved context can be restored and resumed from SMRAM using the RSM instruc-
tion. The SMM’s execution is fully transparent to the Operating System, which can detect
its execution through time discrepancies. Interrupts are completely disabled in SMM.

3.3.4 x86 Architecture
The Intel IA-32 architecture, which supports 32-bit computing, is referred to as x86 archi-
tecture. It defines Intel’s 32-bit processors instruction set and programming environment.
Byte addressing is used in this architecture, and software running on this processor may
have a linear address space of up to 4GB and a physical address space of up to 4GB. Virtual
memory, paging, privilege levels, and segmentation are all supported by the IA-32’s opera-
tional protected mode.
Fast memory registers are used by the CPU for temporary storage during processing in
the IA-32 architecture. For logical and arithmetic operations, each processor core has eight
32-bit general-purpose registers. The EIP register, also referred to as the program counter,
stores the linear address of the next instruction to be executed. This architecture also has
five control registers that define the processor’s configuration as well as the characteristics
of the task being executed.
Segmentation and paging are two memory management mechanisms implemented in
the IA-32 architecture. The 32-bit linear address space is divided into several variable-
length segments using segmentation. Memory references are addressed using 16-bit
segment selector that identifies memory-resident data structure known as a segment
descriptor, and a 32-bit offset into the specified segment. The ability to virtualize the lin-
ear address space is known as paging. Every 32-bit linear address space is divided into
fixed-length parts known as pages, which can then be mapped into physical memory in
any order. This design also supports pages up to 4MB in size, with only a page directory
needed for translation.
The Intel x86-64 architecture was introduced, which was identical to the IA-32 architec-
ture but with 64-bit registers. It has a linear address space of up to 264 bytes and can accept
64-bit linear addresses. Only 48-bit linear addresses are used in the current version, which
does not accommodate the full 64 bits. On these systems, virtual addresses are in canonical
format, which means that 63:48 can be set to all 1s or all 0s depending on the status of bit
47. It also supports an additional level of paging structures called page map level 4 (PML4).

3.3.5  Virtualization Support

A virtual machine is a technology that simulates a computer environment, allowing an
operating system and other programs to run inside of it. Since the operating system and
applications do not often realize that they are within an emulated virtual environment
when virtualization support is used, they can continue to run normally. If a virtual system
is corrupted, suspending the virtual machine will preserve the memory and disk contents.
Some virtual machines save the contents of the disk and memory in a raw file, while oth-
ers save them in a proprietary format. The machine state is saved using trusted software
in this case.
42 Mathematical Modeling for Intelligent Systems

3.4 Review of Memory Acquisition Techniques

In this section, we provide the various techniques used for acquiring volatile memory.
Primarily, these can be categorized into hardware-based, software-based, and firmware-
based techniques.

3.4.1 Hardware-Based Memory Acquisition

Hardware-based memory acquisition techniques refer to the techniques that can directly
access the system’s main memory for acquiring the memory via hardware support. Due to
the fact that it directly accesses the memory, it is regarded as more secure and reliable. It is
capable of accessing memory without the use of operating system interfaces. These tech-
niques do not use system memory while they are executing, thereby reducing the footprint
and tampering. Currently, there are two hardware-based techniques, i.e., dedicated hard-
ware and special hardware bus. Another means to acquire memory via hardware support is
via cold booting. We cover it in this section since it does not fall under the software category.

3.4.1.1 Dedicated Hardware
Using a special hardware card was one of the first techniques for acquiring a forensic rep-
resentation of a system’s RAM storage. Carrier and Grand (2004) introduced the “Tribble”
Proof-of-Concept, which uses DMA operations to construct a replica of physical memory.
The PCI card is mounted as a dedicated hardware unit that can save volatile data to a stor-
age medium connected to it. When an external switch is pressed, the card is triggered, and
the imaging process begins. The contents of RAM are not altered by using this technique
to create a copy.
This approach has a few disadvantages, such as the need to configure a PCI card prior
to use and the fact that hardware cards can be reprogrammed to establish a different view
of physical memory. Advanced attackers can prevent PCI cards from accessing the entire
physical RAM.

3.4.1.2 Special Hardware Bus

As an alternative to PCI cards, another technique was proposed using the special hard-
ware bus support available in the chipset. It includes reading of volatile memory via the
IEEE 1394 (Firewire) bus (Witherden, 2010). Any hardware bus can potentially be used for
physical memory access. In this technique, the interface is present by default in the major-
ity of systems.
The use of this technique can cause random system crashes or malfunction. Also, this
technique requires specific hardware requirements. This technique cause inconsistency
after comparing images, hence it is not reliable to obtain a precise copy of a system’s RAM.

3.4.1.3 Cold Booting
As volatile information is not immediately erased after powering off a machine, instead
it gradually fades away. The information can be recovered for a nonnegligible amount of
time as data stays present in the processor for some duration post power cut-off. This pres-
ence of data for a negligible period (nearly 10 seconds) is called data remanence (Vasisht &
Digital Forensics 43

Lee, 2008). It is achieved by artificially cooling down the RAM modules using liquid nitro-
gen to increase the data remanence from 10 to nearly 30 minutes. To access the retained
memory, the target machine can be restarted, i.e., cold booted with a custom kernel.

3.4.1.4 Limitations of Hardware-Based Techniques

Hardware-based acquisition is relatively secure and reliable. However, it still faces some
grave limitations. The first limitation of using a hardware-based acquisition system is that
it must be set up before an incident occurs. It can seem costly and inefficient to mount a
PCI card into several systems before an incident occurs. Second, since memory is continu-
ously changing, verifying that the image is an exact copy and that data was not overwrit-
ten to target memory during acquisition would be difficult. By and large, hardware-based
memory acquisition techniques seem infeasible for a majority of standard environments
due to negligible attention being given to incident response with respect to cybercrimes.

3.4.2 Software-Based Memory Acquisition

Software-based memory acquisition techniques refer to the techniques that depend upon
the functions and interfaces provided by the operating system and its associated privilege
levels. These techniques are entirely dependent on the operating system kernel, which
schedules processor access and manages all data flow to the storage locations. It is divided
into three parts, i.e., User-level, Kernel-level, and Virtualization-based techniques, which
are further categorized as follows.

3.4.2.1 User Level
User level is the least privileged level in most operating environments. Privilege separa-
tion is achieved by operating system by using different rings and the user operates in the
least privileged ring. The user, however, retains complete control over the applications
and targets that are operating on the same privilege level. Using the user-level privileges,
memory dump can be acquired by using different methods such as Software via software
emulators, via data dumper (dd) utility, via PMDump, via Processor Dumper utility (PD).
All of these methods are outlined below (Figure 3.2).
User level is the least privileged level in most operating environments. Privilege separa-
tion is achieved by the operating system by using different rings and the user operates in
the least privileged ring. The user, however, retains complete control over the applications
and targets that are operating on the same privilege level. Using the user-level privileges,
memory dump can be acquired by using different methods such as Software via software
emulators, via dd utility, via PMDump, and via PD. All of these methods are outlined below.

3.4.2.1.1 Software Emulators
Software emulators can be used to acquire memory via user-level privileges. One of the emu-
lators is Bochs that is an open-source x86-64 emulator (Hilgers et al., 2014). No special ker-
nel driver is required for Bochs. An emulator attempts to replicate the original machine as
closely as possible. It fetches, decodes, and executes instructions one by one, mapping them
from source to target and vice-e-versa. The emulator, which is implemented as some kind of
array, provides guest memory, which enables Bochs to generate a dump efficiently. It has a
built-in debugger that can read memory from both virtual and physical addresses. Because of
the significant performance overhead associated with instruction-wise interpretation, typical
44 Mathematical Modeling for Intelligent Systems

FIGURE 3.2
Privilege rings.

programs are not recommended for execution in an emulated environment. Emulators are
also mostly used in software creation and debugging.

3.4.2.1.2 Data Dumper
Garner (2009) proposed Data Dumper, also known as dd, which is part of the Forensic
Acquisition Utilities (FAU) suite. It employs the \\.\Device\PhysicalMemory section object
to construct a full memory dump of the target machine when run with administrative
privileges. Access to this object at the user level has been constrained for security reasons.
This approach has very little value as compared to modern computing platforms.

3.4.2.1.3 PMDump
PMDump, or Process Memory Dump, is a method that enables you to save the address
space of a single process and dump the memory contents of that process into a file. It
dumps process memory with its header information from /proc//maps file. Data is dumped
either to the file or through the network to a designated storage location. This technique
is a closed source in nature that makes the development of additional parsing tools more
difficult. It also has an effect on the target host’s level of contamination.

3.4.2.1.4 Processor Dumper Utility

PD was proposed by Klein (2006); Vömel and Freiling (2011). It obtains the environment
and state of a process by retrieving an executable’s data and code mappings. By default,
all collected data is redirected to standard performance. This approach makes it more dif-
ficult to create additional parsing methods because it uses a proprietary data format. It has
an impact on the level of contamination on the target host, and it also requires the specifi-
cation of a process ID, which involves the use of a process listing utility in the first place.
Digital Forensics 45

3.4.2.2 Kernel Level
Kernel in the system is responsible for maintaining volatile memory and its associated
operations along with managing and segregating privileges of different levels. Kernel-
based interfaces and utilities can be used for memory acquisition purposes. To monitor the
debuggee and access its memory, software debuggers use kernel functionality. It has many
different alternatives which can be used to take a memory dump such as: kernel drivers,
hibernation files, software crash dumps, and software debuggers.

3.4.2.2.1 Kernel Drivers
Kernel drivers are commonly used to implement memory acquisition tools that run at
the kernel level. Kernel drivers are more reliable and easier to deploy. Even after a poten-
tial incident has occurred, investigators can initiate the acquisition process. Pmem is one
of the physical memory dumping techniques used with kernel-level access. Physical
memory can be acquired by either requesting kernel support or manually mapping
physical frames on its own. It can acquire the physical memory of both user processes
and the kernel.

3.4.2.2.2 Hibernation Files
Hibernation is a built-in feature that is being adopted by most of the recent operating
system kernels. The hibernation file is known as hiberfil.sys which is located in the root
directory. It contains runtime-specific information for running processes in the volatile
memory. The system state is frozen and a snapshot of the current working set is retained
when the operating system is about to reach an energy-saving sleep mode and is sus-
pended to disk. The hiberfil.sys file allows investigators to capture a consistent image of
volatile data.

3.4.2.2.2.1 Software Crash Dumps When a system crashes, modern operating systems have
the ability to save crash-related data in the form of a crash dump. The dump may be con-
fined to the kernel’s address space, an application’s address space, or the entire physical
address space. The system state is frozen in the event of a critical failure, and the main
memory and CPU details are saved in the system’s root directory. The user can use the built-
in CrashOnCtrlScroll feature or a third-party utility like NotMyFault.exe to force a crash. The
contents of memory will be preserved to a file if the machine is programmed for a complete
crash dump.

3.4.2.2.3 Software Debuggers
A kernel’s debugging support is used by most software debuggers. Debugging system
calls is included in most modern operating systems. The debugger can fully access the
debuggee’s memory. The debuggee is not terminated by a debugger, but it may be inter-
rupted for a random period. Since windows is proprietary in nature, a dedicated debugger
is provided by Windows itself i.e., Windbg. Windbg can be used to understand the memory
dump and map it as process and user level.

3.4.2.2.4 Operating System Injection

Schatz (2007) developed Body-Snatcher, a Proof-of-Concept application that injects an
independent operating system into a target machine’s kernel (Schatz, 2007). It is accom-
plished by freezing the host computer’s state and relying on acquisition operating sys-
tem functions to establish a stable snapshot of the volatile data. Because of its low-level
approach and high complexity, the prototype is highly platform-specific.
46 Mathematical Modeling for Intelligent Systems

3.4.2.3 Virtualization Based
Virtualization is a process of emulating complete, isolated, and reliable system environ-
ments known as virtual machines on top of a host computer. Virtual machine monitor
(VMM) is a special software layer that is responsible for sharing, managing, and restrict-
ing access to the available hardware resources. A virtual machine has the ability to be sus-
pended, which means it can interrupt its execution process. In the case of VMware-based
machine, all volatile data is saved to a.vmem file located in the virtual machine’s working
directory (Shaw et al., 2014). The use of virtualization makes memory acquisition proce-
dure comparatively easier.

3.4.2.4 Limitations of Software-Based Techniques

The software-based acquisition procedures rely on the underlying operating environment
which may be untrustworthy in nature. Since it manages the scheduling of processor access
and also maintains control over all data flow to storage locations, there is no way to stop
using the base kernel for a software solution. Another disadvantage of using a software
solution is that it will still require process and kernel memory to function. This poses the
challenge of overwriting data in the memory that may serve as potential evidence in itself.

3.4.3 Firmware-Based Memory Acquisition

Firmware-based memory acquisition combines a commercial PCI network card with the
SMM to remotely acquire the contents of memory, CPU registers, and peripheral state. SMM
is primarily designed for firmware or basic input–output system (BIOS). A driver is required
for a commercial PCI network card to acquire the system’s physical memory. So, drivers for
PCI network cards are placed in the SMM because the code in the SMM is trusted and
locked. As a result, the malware does not pose a modification threat to the driver.
After acquiring the system’s physical memory, the system converts it to virtual memory,
which is used by the operating system, which determines the memory’s semantics. The
CPU registers are obtained using a firmware (BIOS)-based method that reads the registers
using SMM code from the BIOS. It retains the CPU registers used by the OS in protected
mode because the hardware saves them automatically before switching to SMM.
Another technique is used only when one wants to check a few variables that do not
require the complete memory dump. The system is stopped, memory locations or par-
ticular registers are checked, and the system is then resumed. SMM is used to bring the
machine to a halt and retrieve the contents of the memory (Wang et al., 2011). One disad-
vantage is that the debugger cannot freeze the target OS in SMM mode for an extended
period. Another issue is that performing this technique efficiently on machines with a
large amount of memory is difficult.

3.5 Memory Dump Formats

Memory dump formats are the file formats in which the memory dump acquired from
the system is saved when the system crashes or memory dump of the system is required
for analysis. It is the bridging factor between acquisition and analysis of the main
Digital Forensics 47

memory. They can have binary, octal, or hexadecimal data in them. It is of great signifi-
cance as the investigators analyze the memory dump file to identify the cause of the
crash or abnormal termination of the system while using different analysis techniques.
Memory dump can be used to figure out what was going on in the machine when it
crashed and one can find the appropriate solutions to fix the problem. Memory dumps
are useful for diagnosing device bugs and examining memory contents in the event of
a program failure.

3.5.1 Raw Memory Dump

It is the most frequently used and widely supported memory dump format by modern
analysis tools. Memory acquisition tools like Magnet RAM Capture and DumpIT use this
file format to acquire the memory. There are no headers, metadata, or magic values in
these raw memory dumps. Any memory ranges that were deliberately missed or that
could not be read by the acquisition tool are padded in the raw format, which aids in
preserving spatial integrity. RAW memory dump can be converted into Windows crash
dump using raw2dmp plugin.

3.5.2 Windows Crash Dump

When Windows crashes, it frequently generates a memory dump file. This file contains
significant details that will help you determine why your system stopped working.
The DMP_HEADER or DMP_HEADER64 structure appears first in these crash dumps.
In order to create a memory dump in Windows, the device needs a paging file on the
boot volume that is at least 2 MB. A memory dump will not operate even if there is a
big paging file on another disk but none on the boot volume. When the system is run-
ning Microsoft Windows 2000 or later, every time the system abruptly stops, a new file
is generated. dumpchk.exe is a Microsoft application that checks memory dump files
for data. Memory dumps are available in three different formats in Windows crash
dump, as follows:

3.5.2.1 Complete Memory Dump

The largest memory dump file is the complete memory dump, which comprises everything
that was in physical memory. When the machine abruptly stops working, it preserves all of
the contents of system memory. If the Complete memory dump option is chosen, the boot
volume must have a paging file large enough to contain all of the physical RAM plus 1MB.
It doesn’t have anything to do with the platform firmware’s use of physical memory. The
previous file is overwritten only if any other problem occurs during acquisition or if any
other complete memory dump (or kernel memory dump) file exists.

3.5.2.2 Kernel Memory Dump

Kernel memory dump records all the contents that were used by the kernel at the time of
capturing the memory. Since it only includes the content used by the kernel, this memory
dump is considerably smaller than the full memory dump. When your machine suddenly
stops functioning, it speeds up the process of recording information in a log. To accom-
modate kernel memory, the page file must be large enough. Kernel memory for 32-bit sys-
tems is usually between 150MB and 2GB. Unallocated memory and memory allocated to
48 Mathematical Modeling for Intelligent Systems

user-mode programs are not included in these memory dumps. It omits only those parts
of memory that aren’t likely to have played a significant role in the issue. The previous file
is overwritten only if the “Overwrite any existing file” option is selected, or if any other issue
arises, or if another kernel memory dump file (or a complete memory dump file) exists.

3.5.2.3 Small Memory Dump

In Windows crash dump, this is the smallest memory dump file that can be generated. It
saves the tiniest amount of details that can be used to figure out why the machine sud-
denly shuts down. This choice specifies that Windows 2000 and later build a new paging
file any time the device stops unexpectedly and needs a paging file of at least 2MB on the
boot volume. These files’ record is kept in a separate folder. When space is restricted, this
type of dump file can be useful. If some other problem arises or a second small memory
dump file is made, the previous file is protected (unlike other formats). The following
information is normally included in a small memory dump file:

• The error or stop message and its parameters.

• The processor context (PRCB) for the processes that caused Windows to stop work-
ing normally.
• The process information and kernel context (EPROCESS) for the processes that
have been stopped.
• The process information and kernel context (ETHREAD) for the thread that stopped.
• The kernel-mode call stack for the thread hat halted the execution of the operation.
• A list of loaded drivers.

3.5.3 Windows Hibernation File

Hibernation is the method of turning a machine off while maintaining it in its current
state. The contents of RAM are saved to the hard disk or other nonvolatile storage, and
when the process is resumed, the machine is in the same state as before hibernation. When
the machine enters hibernation mode, a hibernated file is generated in the root system
folder, containing the contents of a full memory dump. The hibernation file on Windows
systems is called hiberfil.sys and is located in the root directory.

3.5.4 Expert Witness Format

Encase Forensics uses the Expert Witness Format, or EWF, to acquire a memory for
EnCase-based software. EnCase has become one of the standardized file formats as
a result of its popularity in forensic investigation. For analyzing these memory files,
there are only a few resources available. Investigators should be experienced with the
EWFAddressSpace, Mounting with EnCase, and Mounting with FTK Imager methods for con-
verting EWF memory dumps.

3.5.5 AFF4 Format
The Advanced Forensics File Format 4 (AFF4) is an open-source format for digital evidence
and data storage. Multiple image sources, arbitrary data, and storage virtualization are
all integrated into the standard forensic file format. It supports features such as writing
and reading zip file style volumes, writing and reading AFF4 image streams, writing and
Digital Forensics 49

reading RDF metadata, logical file acquisition, etc. AFF4 format is believed to have a huge
impact on the future of memory acquisition to create a unified storage format for forensic
evidence present worldwide.

3.5.6 HPAK Format
HBGary software cooperation developed the HPAK file format. HPAK allows the physical
memory and page file(s) of a target device to be embedded in the same output file. Since
this is a proprietary format, the only tool that can create an HPAK file is FastDump. This
file format requires the HPAK command-line option; otherwise, the memory dump is cre-
ated in the default file format, raw, which excludes the page file (s).

3.6 Memory Acquisition Tools

In this section, we outline the memory acquisition alternatives which are available in the
market in a comparative manner. The comparison is provided in Table 3.3. We also briefly
describe the four leading acquisition tools which are further used for performance com-
parison in our work.

TABLE 3.3
Comparison of Memory Acquisition Tools
Belkasoft Live RAM
Tools/Features Magnet RAM Capture Capture FTK Imager DumpIT
Release year 2014 2014 2013 2007
Latest version v1.2.0 v1.0 v4.5.0.3 v1.3.2
Copyright Magnet Forensics Inc. Belkasoft Evidence Access Data Group Inc. Moonsols
Center
Developer Magnet Forensics Belkasoft Evidence Access Data Moonsols
Center
Acquisition time 75 seconds 60 seconds 79 seconds 65 seconds
(4GB)
Acquisition time 54 seconds 31 seconds 36 seconds 34 seconds
(8GB)
Average memory 5.7MB 7.6MB 6.8MB 11.5MB
usage (4GB)
Average memory 6.8MB 7.8MB 7.0MB 12.3MB
usage (8GB)
GUI/CLI GUI GUI GUI CLI
Source go.magnetforensics.com belkasoft.com accessdata.com qpdownload.
com
Portable software Single portable executable Portable executable Requires installation Portable
folder folder executable
folder
Supported image .raw .mem .mem .raw
format
Cost effective Yes Yes Yes Yes
OS supported Windows Vista, XP, 2003, Windows Vista, XP, Windows Vista, XP, 2003, Windows Vista,
2008, 2012, 7, 8, 10 2003, 2008, 7, 8, 10 2008, 2012, 7, 8, 10 XP, 7, 8, 10
50 Mathematical Modeling for Intelligent Systems

3.6.1 Magnet RAM Capture

Magnet RAM Capture is a free imaging tool that enables investigators to capture the phys-
ical memory of a target system, allowing them to recover and analyze valuable objects
that are often only present in memory. Since Magnet RAM Capture has a small memory
footprint, investigators can use it while reducing the amount of data that is overwritten
in memory. It exports captured memory data in Raw (DMP) format, which can be easily
imported into prominent analysis tools like Internet Evidence Finder. Processes and pro-
grams running on the device, network links, proof of malware infiltration, registry hives,
username and passwords, decrypted files and keys, and evidence of operation not usually
stored on the local hard disk are all examples of evidence found in RAM (ram-capture-
twitter @ go.magnetforensics.com, n.d.).

3.6.2 Belkasoft Live RAM Capturer

Belkasoft Live RAM Capturer is a small free forensic tool that enables you to retrieve all of
the contents of a computer’s volatile memory, even if it’s protected by an active antidebug-
ging or antidumping scheme. Separate 32-bit and 64-bit builds are available to keep the
tool’s footprint as small as possible. Live RAM Analysis can be used to analyze memory
dumps captured with Belkasoft Live RAM Capturer in Belkasoft Evidence Center (ram-
capturer @ belkasoft.com, n.d.).

3.6.3 FTK Imager
FTK Imager is a data preview and imaging tool for forensically sound data acquisition by
producing copies of data without tampering the original evidence. Following the creation
of an image, the Forensic Toolkit (FTK) is used to conduct a comprehensive forensic analy-
sis and generate a report of the findings (ftkimager3 @ marketing.accessdata.com, n.d.). FTK
Imager makes it simple to create forensic images, display files and directories, preview
their contents, mount an image for read-only viewing, recover deleted files, generate hash
reports, and more.

3.6.4 DumpIt
DumpIT is a free memory forensics tool created by Matthieu Suiche from MoonSols. This
tool is used to create a physical memory dump of Windows system. It is compatible with
both 64-bit and 32-bit versions of Windows operating systems. It is a very easy tool that
can be used by any user with administrative privileges. Only a confirmation question
is prompted before the raw memory dump is created in the current directory. Since it
is a very simple-to-use tool, it does not possess any analysis capabilities. Tools such as
Mandiant Redline can be used for analysis purpose.

3.7 Implementation
Acquiring the memory dump of a system requires specific tools which can be used to
extract and dump the volatile memory onto the physical memory of the system. In this
Digital Forensics 51

section, we provide the testbed setup details and further compare four leading memory
acquisition tools. The four memory acquisition tools for Windows, as described in the
previous section, were implemented on two different physical machines with different
architecture and RAM.

When these tools were executed in the system, it was assured that the same processes were
functioning to have a common standard for analysis of the tool’s performance. At the time
of acquiring memory, the system was connected to the Internet. We provide the methodol-
ogy for executing these tools, in order to assist others as follows:

• Methodology for Magnet RAM Capture

1. Visit the site and enter your personal details and click on download.
2. Download the tool from the link sent in the entered email account.
3. A single portable executable file will be downloaded in the system.
4. Users will need to specify two choices prior to start acquisition i.e., path to
save the acquired memory data, and whether the segmentation is needed
or not.
5. Run the executable file named “MRCv120” in the system.
6. Select the segment size if segmentation is to be done. By default, the segmenta-
tion is set as off.
7. Select the location to save the captured memory dump file and click on “Start”
button.
8. It will begin acquiring the memory dump of the system.
• Methodology for Belkasoft Live RAM Capture

52 Mathematical Modeling for Intelligent Systems

• Methodology for DumpIT

1. Visit the site and download the DumpIT tool latest version.
2. A portable executable zip folder will be downloaded into the system.
3. Extract the files from the zip folder.
4. Run the executable file named as DumpIt in the system.
5. It will ask for the confirmation then enter either yes or no according to your
choice.
6. If yes, then it will begin acquiring the memory dump of the system.

3.8 Performance Comparison
Performance comparison of the abovementioned tools was conducted and results were
analyzed on the basis of following:

The abovementioned tools are thoroughly compared in Table 3.3 with respect to certain
features. These tools were run on two Windows-based systems with varying configura-
tions. The average processor utilization done by the tools over the acquisition procedure
is graphically represented in Figures 3.3 and 3.4. Figures 3.5 and 3.6 represent processor
utilization by the tools with respect to time. The average disk utilization done by the tools
over the acquisition procedure are graphically represented in Figures 3.7 and 3.8. Figures
3.9 and 3.10 represent disk utilization by the tools with respect to time. The average time
taken by the tools over the acquisition procedure is graphically represented in Figures 3.11
and 3.12. Another detailed comparison with respect to acquisition time taken by the tools
in different runs is represented in Figures 3.13 and 3.14.
The time was calculated manually using a stopwatch. Belkasoft Live RAM Capture was
the fastest for the acquisition of both 4 and 8GB RAM, with DumpIT just a couple of seconds
behind. While the tools are being executed to acquire the memory, the tools themselves
will also utilize the processor and disk space. When it comes to 4GB RAM, FTKImager
Digital Forensics 53

FIGURE 3.3
Average processor utilization during the entire acquisition process (4GB).

FIGURE 3.4
Average processor utilization during entire acquisition process (8GB).
54 Mathematical Modeling for Intelligent Systems

FIGURE 3.5
Processor utilization with respect to time (4GB).

FIGURE 3.6
Processor utilization with respect to time (8GB).

utilizes the least percentage of the processor, whereas Magnet RAM Capture utilizes
the highest percentage of the processor. As soon as the RAM size increases, the results
were completely changed as now FTKImager utilizes the highest percentage of processor,
with DumpIT being the best because it utilizes the least percentage of processor. When
Digital Forensics 55

FIGURE 3.7
Average disk utilization during the entire acquisition process (4GB).

FIGURE 3.8
Average disk utilization during the entire acquisition process (8GB).

compared according to disk utilization in 4GB RAM, Belkasoft Live RAM Capture utilizes
the minimal space while DumpIT utilizes the maximum space in the disk. Whereas in 8GB
RAM, Magnet RAM Capture utilizes minimal space while the rest of the tools have only
the slightest differences in their outcomes in utilizing the space in the disk. The variation
in result of processor and disk utilization are not constant as it might increase, decrease, or
remain constant. The result depends upon the various processes going on while executing
the acquisition tools. The result also depends upon the configuration of windows as vary-
ing configurations will result in varying outcomes.
56 Mathematical Modeling for Intelligent Systems

FIGURE 3.9
Disk utilization with respect to time (4GB).

FIGURE 3.10
Disk utilization with respect to time (8GB).
Digital Forensics 57

FIGURE 3.11
Average time taken to acquire entire volatile memory (4GB).

FIGURE 3.12
Average time taken to acquire entire volatile memory (8GB).
58 Mathematical Modeling for Intelligent Systems

FIGURE 3.13
Time taken by tools in different runs to acquire memory (4GB).

FIGURE 3.14
Time taken by tools in different runs to acquire memory (8GB).
Digital Forensics 59

3.9 Conclusion
In this paper, we have reviewed various memory acquisition techniques that include
hardware-based, software-based, and firmware-based techniques. Hardware-based tech-
niques include dedicated hardware card, special hardware bus, and cold booting, whereas
software-based techniques include user-level, kernel-level, and virtualization-based.
Further, we have also defined various memory dump formats that include raw memory
dump, windows crash dump, and AFF4 Format. We have selected four memory acquisi-
tion tools as they are reliable, easy to use, easily available, and famous in the market. The
methodology for each of these tools is outlined in its own section. These four tools were
implemented in two Windows systems with different configurations and results obtained
were analyzed on the basis of different parameters such as memory acquisition time, pro-
cessor utilization, and disk utilization. The result depends upon the various factors such
as the processes going on while executing the acquisition tools or the windows configu-
rations, as varying configurations will result in varying outcomes. Belkasoft Live RAM
Capture appears to be the best in terms of memory acquisition time, but the variation in
the result of processor and disk utilization is not constant as it might increase, decrease or
remain constant. The outcomes are examined on the basis of various graphs obtained from
the data. Hence, various analysis techniques can be used to further analyze the memory
dump acquired through these tools, which is out of the scope for this paper.

References
2020_Google_services_outages @ en.wikipedia.org. https://en.wikipedia.org/wiki/2020_Google_ser-
vices_outages#:~:text=On 14 December 2020%2C another, still accessible with private browsing.
The last access/edit date was 25 March, 2022.
Balogh, Š., & Mydlo, M. (2013). New possibilities for memory acquisition by enabling DMA using net-
work card. In 2013 IEEE 7th International Conference on Intelligent Data Acquisition and Advanced
Computing Systems (IDAACS) (Vol. 2, pp. 635–639).
Beebe, N. (2009). Digital forensic research: The good, the bad and the unaddressed. IFIP Advances in
Information and Communication Technology, 306, 17–36. https://doi.org/10.1007/978-3-642-04155-6_2
Carrier, B. D., & Grand, J. (2004). A hardware-based memory acquisition procedure for digital inves-
tigations. Digital Investigation, 1(1), 50–60. https://doi.org/10.1016/j.diin.2003.12.001
ftkimager3 @ marketing.accessdata.com. (n.d.). https://marketing.accessdata.com/ftkimager3.4.2
Garner, G. M. (2009). Forensic acquisition utilities. http://gmgsystemsinc. com/fau/
Harichandran, V. S., Breitinger, F., Baggili, I., & Marrington, A. (2016). A cyber forensics needs analy-
sis survey: Revisiting the domain’s needs a decade later. Computers and Security, 57(March),
1–13. https://doi.org/10.1016/j.cose.2015.10.007
Hilgers, C., Macht, H., Muller, T., & Spreitzenbarth, M. (2014). Post-mortem memory analysis of
cold-booted android devices. In Proceedings -8th International Conference on IT Security Incident
Management and IT Forensics, IMF 2014 (pp. 62–75). https://doi.org/10.1109/IMF.2014.8
Klein, T. (2006). Process dumper. http://www.trapkit.de/research/forensic/ pd/index.html
Latzo, T., Palutke, R., & Freiling, F. (2019). A universal taxonomy and survey of forensic memory
acquisition techniques. Digital Investigation, 28, 56–69.
McDown, R. J., Varol, C., Carvajal, L., & Chen, L. (2016). In depth analysis of computer memory
acquisition software for forensic purposes. Journal of Forensic Sciences, 61, S110–S116.
60 Mathematical Modeling for Intelligent Systems

Palutke, R., Block, F., Reichenberger, P., & Stripeika, D. (2020). Hiding process memory via anti-
forensic techniques. Forensic Science International: Digital Investigation, 33, 301012. https://doi.
org/10.1016/j.fsidi.2020.301012
Parekh, M., & Jani, S. (2018). Memory forensic: Acquisition and analysis of memory and its tools com-
parison. International Journal of Engineering Technologies and Management Research, 5(2), 90–95.
ram-capture-twitter @ go.magnetforensics.com. (n.d.). http://go.magnetforensics.com/ram-capture-twitter
ram-capturer @ belkasoft.com. (n.d.). https://belkasoft.com/ram-capturer
Rathnayaka, C., & Jamdagni, A. (2017). An efficient approach for advanced malware analysis
using memory forensic technique. In Proceedings -16th IEEE International Conference on Trust,
Security and Privacy in Computing and Communications, 11th IEEE International Conference on Big
Data Science and Engineering and 14th IEEE International Conference on Embedded Software and
Systems, Trustcom/BigDataSE/ICESS 2017 (pp. 1145–1150). https://doi.org/10.1109/Trustcom/
BigDataSE/ICESS.2017.365
Reust, J., & Friedburg, S. (2006). DFRWS 2005 Workshop Report. Online at http://www.dfrws.
org/2005/download. http://www.w1npp.org/ares/RFI/2005FI~1.PDF
Schatz, B. (2007). BodySnatcher: Towards reliable volatile memory acquisition by software. Digital
Investigation, 4(suppl.), 126–134. https://doi.org/10.1016/j.diin.2007.06.009
Shaw, A. L., Bordbar, B., Saxon, J., Harrison, K., & Dalton, C. I. (2014). Forensic virtual machines:
Dynamic defence in the cloud via introspection. In Proceedings -2014 IEEE International Conference
on Cloud Engineering, IC2E 2014 (pp. 303–310). https://doi.org/10.1109/IC2E.2014.59
Sokolov, A. N., Barinov, A. E., Antyasov, I. S., Skurlaev, S. V., Ufimtcev, M. S., & Luzhnov, V. S. (2018).
Hardware-based memory acquisition procedure for digital investigations of security incidents
in industrial control systems. In Proceedings -2018 Global Smart Industry Conference, GloSIC 2018
(pp. 1–7). https://doi.org/10.1109/GloSIC.2018.8570109
Stüttgen, J., Vömel, S., & Denzel, M. (2015). Acquisition and analysis of compromised firmware using
memory forensics. Digital Investigation, 12, S50–S60. https://doi.org/10.1016/j.diin.2015.01.010
Vasisht, V. R., & Lee, H. H. S. (2008). SHARK: Architectural support for autonomic protection against
stealth by rootkit exploits. In Proceedings of the Annual International Symposium on Microarchitecture,
MICRO, 2008 PROCEEDINGS (pp. 106–116). https://doi.org/10.1109/MICRO.2008.4771783
Vidas, T. (2007). The acquisition and analysis of random access memory. Journal of Digital Forensic
Practice, 1(4), 315–323. https://doi.org/10.1080/15567280701418171
Vömel, S., & Freiling, F. C. (2011). A survey of main memory acquisition and analysis techniques
for the windows operating system. Digital Investigation, 8(1), 3–22. https://doi.org/10.1016/j.
diin.2011.06.002
Wang, J., Zhang, F., Sun, K., & Stavrou, A. (2011). Firmware-assisted memory acquisition and analy-
sis tools for digital forensics. In 2011 Sixth IEEE International Workshop on Systematic Approaches
to Digital Forensic Engineering (pp. 1–5).
Witherden, F. (2010). Memory Forensics over the IEEE 1394 Interface (pp. 1–28). https://freddie.with-
erden.org/pages/ieee-1394-forensics.pdf
4
IoMT and Blockchain-Based Intelligent and
Secured System for Smart Healthcare

Hitesh Kumar Sharma, Ravi Tomar, and Preeti

University of Petroleum and Energy Studies

CONTENTS
4.1 Introduction .......................................................................................................................... 62
4.2 Application of IoT in Intelligent Systems ......................................................................... 62
4.2.1 Smart Homes ............................................................................................................ 62
4.2.2 Smart City ................................................................................................................. 62
4.2.3 Self-Driven Cars .......................................................................................................63
4.2.4 IoT in Healthcare ......................................................................................................63
4.3 Application of IoT in Healthcare........................................................................................63
4.3.1 Health Monitoring ...................................................................................................63
4.3.2 Remote Sensors ........................................................................................................64
4.3.3 Clinical Alert Systems .............................................................................................64
4.4 Advantages of IoMT ............................................................................................................64
4.5 Blockchain-Based Security of Intelligent IoMT-Based Smart Healthcare ...................65
4.6 Application of Blockchain in Intelligent Systems ........................................................... 66
4.6.1 Cash Transfer and Installment Handling ............................................................ 66
4.6.2 Supply Chains Checking ........................................................................................ 67
4.6.3 Tracking the Supplies .............................................................................................. 67
4.6.4 Verification of the Team .......................................................................................... 67
4.6.5 A Focus on Transactions ......................................................................................... 67
4.7 Advantages of Blockchain in Healthcare ......................................................................... 67
4.7.1 Transparency in the Supply Chain........................................................................ 68
4.7.2 Patient-Centric Electronic Health Records ........................................................... 68
4.7.3 IoT Security for Remote Monitoring ..................................................................... 69
4.8 Challenges for Blockchain in Healthcare ......................................................................... 69
4.8.1 Major Shift ................................................................................................................ 69
4.8.2 Refusal to Share ....................................................................................................... 69
4.8.3 Distributed Healthcare............................................................................................ 69
4.9 IoMT with Edge/Fog Computing ...................................................................................... 69
4.10 Conclusion ............................................................................................................................ 72
References....................................................................................................................................... 72

DOI: 10.1201/9781003291916-4 61
62 Mathematical Modeling for Intelligent Systems

4.1  Introduction
IoT market in the medical sector is the fastest growing market and is estimated to be
worth nearly $176 billion by 2026 [1]. With such vast applications, IoT is extensively used
in healthcare. IoMT is a special IoT-based platform where the patient can meet the doctor,
show their report in virtual manner, and take the proper guidance without physically
meeting them. Few years ago, diagnosis of diseases in the human body was possible only
when the patient physically met the doctor. Sometimes the patient even had to stay in
hospital for treatment, which increased the healthcare cost and also stained the health-
care facility in rural and remote locations. The technological advancement through these
years allowed us for diagnosis of various diseases and health monitoring [2]. In the last
decade, there were so many technologies created with the help of which now we can
monitor the patient blood pressure, level of glucose, oxygen level, and so on in an easier
way, and the best thing is that we do not need to physically meet with doctors to show
our reports.

4.2 Application of IoT in Intelligent Systems

4.2.1 Smart Homes
The Internet of things (IoT) is used in homes too to make our homes smart. It can be used
in various ways:

• Monitoring the activities.

• Automated lightings.
• Day night sensors.
• Advanced locking systems.

All these uses can be beneficial for everyone. This makes it easy for people by saving time.
And due to high technology being used, security concerns are solved. All these uses are
carried out after lots of testing to make things reliable [3]. Smart Homes are becoming
popular every day. Due to this, the IoT has gained the trust of many people. This technol-
ogy will keep growing because there’s no end to the needs and imagination of humans.
Hopefully, this can play a major role in sending people to places other than Earth [4].

4.2.2 Smart City
Every country wants to have many Smart Cities. But why? The answer is simple—because
of the ease of everything and time efficiency. A Smart City uses the IoT [5]. Few examples
are:

• Solar panels store energy from sun rays and later use it to light up the nights. This
method is being adopted by almost every city now.
• Day–night sensors.
Smart Healthcare 63

• Speed tracking machines.

• Fast tags at toll gates.

There are a few other uses like GPS tracking and internet connectivity that are available
everywhere [6]. Many cars use this feature to modernize their products. This makes smart
connectivity between everyone. India is also using the IoT to develop. The developments
allowed India to form a movement called Digital India [7].

4.2.3 Self-Driven Cars
We have seen Elon Musk use this technology to make an outstanding product, Tesla, where
cars can move from one destination to the other on their own. Many other brands have
also tried using this technology—Google tried to make its self-driven bicycles and bikes.
This can solve many transport issues as not only do they drive on their own, but they
also interact with other vehicles to ensure safety of the passenger. This can also overcome
human interaction as there wouldn’t be any confusion.

4.2.4 IoT in Healthcare

The IoT helps the health industry by its powerful sensory devices and data interactions.
Since doctors are mostly busy, self-medication applications are built and improved con-
stantly to ensure the health of an individual. Many sensory devices send real-time infor-
mation that are then diagnosed online and also provided with the treatments. Advantages
of IoT in healthcare:

• Minimal errors.
• Online consultancy which saves time.
• Decreases frequent visits.

Healthcare is using and promoting the IoT to decrease workloads and also ensure the
safety of a person.

4.3 Application of IoT in Healthcare

4.3.1 Health Monitoring
Nowadays, wellness groups, pulse-checking sleeves, circulatory strain estimat-
ing groups, glucometer, and further developed shrewd medical care gadgets furnish
patients with ongoing customized wellbeing status. They remind the patient to be
aware of their ordinary calorie utilization, further develop practice schedule, recognize
changing circulatory strain pattern, etc. For elderly patients, it is critical to follow their
medical issues consistently, and keen medical service gadgets make their way of life
simpler by telling them about any conceivable peril ahead of time! It alarms the patient
just as to the patient’s relatives to caution them about the startling variances in the
patient’s ailments [8].
64 Mathematical Modeling for Intelligent Systems

4.3.2 Remote Sensors
Remote sensors are used for automation in pathology labs to control the proper environ-
ment required for various kinds of lab samples taken from the patients’ body. Like proper
cooling of blood samples etc. [9].

4.3.3 Clinical Alert Systems

These IoT-based systems are used for alerting the emergency contacts in case of any medi-
cal emergency. The person wearing smart health band will send an alert message to his/her
emergency contacts saved in his/her mobile phone if any health related mishappen occur.

4.4 Advantages of IoMT

Internet of Medical Things (IoMT) refers to the physical objects or devices which are inter-
connected through the internet, all collectively taking in and sharing data without requiring
human touch or interaction. In the past couple of decades, IoT has spread in every house-
hold, connecting our phones, kitchen appliances, and other electrical devices (Figure 4.1).
Similar to our household, IoT also plays a huge part in healthcare. Cases for the same are
given below:

• In a hospital, various devices are used to track and monitor patients’ heartbeats;
their psychological condition, or their movements. All such instruments are con-
nected to each other and provide real-time information to the doctors or nurses.

FIGURE 4.1
Data flow in smart healthcare application using IoT.
Smart Healthcare 65

Such technology helps the staff to keep a check on all the patients at the same time
and even reduces hospital stay time [4].
• IoT also has a huge role in our daily lives. Everyday use devices, such as fitness
bands, oxymeters, sphygmomanometers, etc., can be tuned/programed to provide
daily reports on calories, blood pressure, oxygen levels, and even help in booking
appointments. IoT can also help in preventing emergency situations by sending
alert signals to close kin and concerned authorities when there is a disturbance in
daily patterns or reports of the person.
• Another huge part played by IoT in healthcare is in Healthcare Insurance compa-
nies. Such companies make use of large amounts of data and information collected
extensively from various healthcare facilities and health-monitoring devices. This
data helps them in understanding the most common ailments, injuries, and dis-
eases and coming up with the most helpful and profitable insurance policies for
their customers. It also aids in bringing transparency between people and them-
selves in terms of prices, terms and conditions, etc.

Some examples of IoT-based smart healthcare devices are:

• Remote patient monitoring.

• Glucose monitoring.
• Heart-rate monitoring.
• Depression and mood monitoring.
• Hand hygiene monitoring.

4.5 Blockchain-Based Security of Intelligent

IoMT-Based Smart Healthcare
Cyber-attack concerns are very common in the world and nowadays there are serious
attacks even in the healthcare sector. The current infrastructure is incapable of ensuring
protection against such data breaches, putting patients’ health information’s privacy and
security at jeopardy. The current smart healthcare record models create another door to
a problematic situation, namely, patients’ data being in the custody of healthcare organi-
zations, putting patients’ information at risk, and resulting in inefficient data delivery to
patients’ healthcare. For instance, if information concerning a patient’s health is not com-
municated from one service provider to the next on time, the patient’s treatment may be
delayed.
EHR has such limitations practically, which can be overcome by using Blockchain [10].
Blockchain allows data to be collected from a variety of sources and saved in a transaction
audit log, which helps maintain track of data accountability and transparency throughout
data exchange. As shown in Figure 4.2, Blockchain can provide more secure way of shar-
ing medical data between different users of smart healthcare systems.
Blockchain is an arising innovation stage for creating decentralized applications and
information stockpiling, over and past its job as the innovation basis for digital curren-
cies. The fundamental principle of this stage is that it permits one to make a dispersed
66 Mathematical Modeling for Intelligent Systems

FIGURE 4.2
Role of blockchain in smart healthcare.

and recreated record of occasions, exchanges, and information created through different
IT processes with solid cryptographic certifications of alter obstruction, changelessness,
and unquestionable status. Public blockchain stages permit us to ensure these proper-
ties with overpowering probabilities in any event, when untrusted clients are members
of dispersed applications with the capacity to execute on the stage [11]. Despite the fact
that blockchain innovation has become prominently known due to its utilization in
the execution of Cryptographic forms of money like Bit Coin, Ethereum, and so on, the
actual innovation holds significantly more guarantee in different regions, for example,
time-stepping, logging of basic occasions in a framework, recording of exchanges, reli-
able e-administration, and so on. Numerous specialists are chipping away at many such
use cases like decentralized public key foundation, self-sovereign personality the execu-
tives, library support, wellbeing record the board, decentralized verification, decentral-
ized DNS, and so on. Additionally, organizations, for example, IBM and Microsoft are
fostering their own applications in assorted fields like the IoT, and so forth, in any event,
empowering blockchain stages on the cloud. Thinking about the need to spread the aris-
ing ideas for understudies, we chose to set up another seminar on blockchain innovation
stages and applications [12].

4.6 Application of Blockchain in Intelligent Systems

4.6.1 Cash Transfer and Installment Handling
Conceivably the best and most reasonable utilization of blockchain innovation is utilizing
it as a way to speed up the exchange of assets starting with one party then onto the next.
Most exchanges extended by means of blockchain can be settled inside mere seconds,
while banks require 24 hours every day and surprisingly 7 days per week.
Smart Healthcare 67

4.6.2 Supply Chains Checking

Blockchain innovation is not difficult to apply with regard to checking supply chains. By
destroying paper-based preliminaries, ventures can spot shortcomings inside their inven-
tory chains quickly, just as to distinguish things progressively [13]. Blockchain likewise
empowers endeavors, and even shoppers, to see how items perform according to a quality-
control perspective as they move from their place of beginning to the retailer.

4.6.3 Tracking the Supplies

Besides privacy and security, Blockchain tech can also be used to keep an eye on the sup-
plies. For e.g., if we have ordered something from a medical provider, it can be tracked by
the customer to make him aware of the product. MediLedger is one of the examples that
uses Blockchain to increase the authenticity and transparency of the goods/supplies.

4.6.4 Verification of the Team

Blockchain can also be used to verify the credentials of the health support. This makes an
individual know more about his team. This can lead to better relationships with the team.
ProCredEX, an American company, has built an application that verifies the details of the
support.

4.6.5 A Focus on Transactions

The validity of transactions will be determined by consensus, and while transactions can
be interpreted widely, not all healthcare transactions will be appropriate for this tech-
nology. With blockchain, smart contracts are possible, which are downloadable computer
code that allows programs, functions, or transactions to be automated. In addition, the
technology provides a way to create value for transactions through tokens. Also, crypto-
currency settings have been implemented with tokenization.

4.7 Advantages of Blockchain in Healthcare

DLT can be applied in numerous medical services regions, yet every type of effort inside
medical services isn’t connected to exchanges. Be that as it may, public blockchains can’t
be utilized to store private data like recognizing wellbeing information, in light of the
fact that the information in them is broadly open. This straightforwardness commands
that suppliers consider security issues to guarantee ensured wellbeing data (PHI) [14].
Furthermore, blockchain innovation is powerless against certain sorts of assaults; how-
ever, it offers inbuilt security against others [15,16]. The blockchain code exposes it to mul-
tiday assaults and bugs, just as friendly designing. In this manner, data security should
be given concentrated consideration, particularly when utilized in medical services. DLT
can be applied in numerous medical services regions, yet the activity of any kind inside
medical care isn’t connected to exchanges. Be that as it may, public blockchains can’t be
utilized to store private data like recognizing wellbeing information, on the grounds that
the information in them is generally open. This straightforwardness commands that sup-
pliers consider security issues to guarantee ensured wellbeing data (PHI).
68 Mathematical Modeling for Intelligent Systems

Furthermore, blockchain innovation is defenseless against certain sorts of assaults; how-

ever, it offers inbuilt insurance against others. The blockchain code exposes it to multiday
assaults and bugs, just as friendly designing. Along these lines, data security should be
given escalated consideration, particularly when utilized in medical care.

4.7.1 Transparency in the Supply Chain

Monitoring and security in supply chain is one of the major concerns in supply chain
management. Storing the information for each movement of goods in the supply chain at
each level can be stored in blocks of a public/private blockchain which cannot be altered
and updated in real time.

4.7.2 Patient-Centric Electronic Health Records

Every country and area is struggling with the issue of data segregation, which means
that patients and their healthcare professionals have an incomplete picture of their medi-
cal history. One possible answer to this issue is to create a blockchain-based record that
can be then be joined with existing electronic medical records (EMRs) and serve as a
single file.
Any new record, such as a physician’s note, a prescription, or a test result, gets attached
to this existing blockchain. In this scenario, any change to a patient record, as well as the
patient’s agreement to disclose a portion of their medical information, is recorded as a
transaction on the blockchain. Medicalchain is a notable example of a firm that works with
healthcare providers to adopt blockchain-enabled EMRs. In Figure 4.3, we have shown the
benefits of Blockchain in smart healthcare systems.

FIGURE 4.3
Benefits of Blockchain in smart healthcare.
Smart Healthcare 69

4.7.3 IoT Security for Remote Monitoring

Blockchain technology has the potential to impact all recordkeeping processes, including
the way transactions are initiated, recorded, authorized, and reported. Blockchain could
be used to securely and efficiently transfer user data across platforms and systems All
important lost data and information of any such problems are overcome after introduction
of Blockchain technology, now the patient’s problem looks like it has moved very far away
from them as those important reports are stored over the Blockchain through which they
can easily access on time of use or in any emergency. Doctors have also got ease in check-
up of any patients with the help of Blockchain data as that very patient is either suffering
from earlier from any such problems that may lead to this new kind of problem to that
very patient.

4.8 C hallenges for Blockchain in Healthcare

4.8.1 Major Shift
Many physicians are still stuck on paper at the moment. As a result, convincing them
to switch from paper to electronic healthcare records (EHR) utilizing blockchain is a tall
order.

4.8.2 Refusal to Share

The active refusal of insurance payers and hospitals to disclose data is a famous illustra-
tion of this. Keeping expense data to themselves gives hospitals a competitive advantage.
They may receive different fees for different patients if they are required to share with
insurance companies.

4.8.3  Distributed Healthcare

In terms of how various organizations handle records, healthcare physician providers and
insurance payers, etc. all employ different methods. It would be extremely difficult to get
all of these groups together to utilize blockchain as a technology without a simplified
approach.

4.9 IoMT with Edge/Fog Computing

Edge/Fog Computing is concerned with high performance of intelligent systems. It is any-
thing other than a continuously virtualized paradigm in which cloud workers operate
with, accumulate, and create resources. In addition, the majority of information conveyed
by these IoMT contraptions should be updated and rapidly decommissioned in order to
increase the amplitude of IoMT requests. Fog Computing looks at, collects a smashing
point and interference network up to a specific association edge in order to deal with the
rapid issue of IoMT and complete IoMT warranty. Fog Processing provides a complete
70 Mathematical Modeling for Intelligent Systems

licensing arrangement to various applications and firms. In the middle of a few IoMT’s
that mention such as linked cars, the fog can pass persuasive short tranny. The analysis
of Fog Computing is the superior solution for applications with reduced time demands
such as increased realism, fluctuating press streaming and game playback, etc. The inte-
gration of IoMT with Fog Computing has a number of advantages to various IoT applica-
tions. Fog processing helps to reduce support times, particularly with IoT requests that
are time fragile, by short contact amid IoMT devices. Furthermore, the capacity to restore
sensor connections to a huge degree is often among the essential components connected
with fog computing. Fog Computing might give various IoMT applications several advan-
tages. This section highlights the distinctive research publications on the interconnection
between IoMT and Fog Computing. In this section, we assess the current work that has
been discussed in different applications regarding IoMT and Fog Computing joints. The
manufacturers have maintained an eye on the motive and characteristics of IoMT applica-
tions. It displays data that can be handled by fog effortlessly by IoMT devices. It also con-
veys the problems of holding up time and block, which can simply be driven by fog. Fog
Computing will also help in a wonderful way, regulate decentralized and fast-developing
IoMT plans, enhance new businesses inside the network and restrict the results of differ-
ent programs connected to network facilitator activity, as well as potential customers. The
builders began by contemplating many challenges in the construction of IoMT structures
and that this is extremely difficult in order to pick issues along with current versions of
the business and calculation frameworks. Then we talk about the advantages of the book
plan for the structuring of frames, how to cope with them, and the limit. Just how this
method may be applied to create amazing possibilities for businesses. In this way, the
unique framework analyzes the particular characteristics of Fog processing and helps to
reduce some IoMT issues. This pushed upon cellular IoT story plus explored the particular
key problems associated with conflict that may have the ability.
This gives the cloud a middleware that foreshadows fog that acts without reservation.
As a further alternative to fog cells, cloud-based middleware is investigated, tested, and
organised. Furthermore, the manufacturers suggested a structured fog framework. They
passed on software with respect to a breeze farm with a shrewd, light-fitting visitor to look
at the qualities of their style. Articles on the regulating causes of fog processing in research
content are seldom truly investigated. The particular process employs the specific task
and unambiguously determines resources for control sources and takes into consideration
everything. These people stated that they could better regulate their own job in order
to achieve much more specific research and development linked to one another so that
they might use Fog Computing plus IoMT to convert to grouped product sales associated
with the special impairment of professions. The makers have defined a big improvement
problem to prevent the reactivity of available nebulous care. They evaluated their specific
design, sometimes simply, by differentiating and active approaches, which results in a fall
in inactivity. The manufacturers have evaluated the best way of restoring IoMT devices,
which are generally resource-focused barriers. In addition, three motivating situations
with WSNs, clever automobiles, and clever organizations, will be presented, showing that
Fog’s activities may be used for a few useful purposes and organizations. The fog com-
pact utilizes a center area process to create goods possible integrated into a professional
slave interface, where expert centers use the planning resources to record information
from slave centers. Fog provides different advantages for using IoT, as they may collect
and work near the edge of the association. Fog movable rear load changes between the
procenters to correlate the slave center points with the stacking master, which updates the
Smart Healthcare 71

FIGURE 4.4
Edge/Fog computing services for IoMT-based intelligent systems.

whole system’s flexibility. The manufacturers have proposed a tonal decreasing approach
for Fog-related centers that often reduces IoT center inactivity companies. They proposed
a technique of just spreading the local store that would build up Fog to Fog to convey
corporate lethargy. With regard to unloading calculations, the specific approach obviously
represents simply not just the full line but also a few interest lessons, which otherwise
have a distinct estimate time (Figure 4.4).
The unique manufacturer has noticed the particular basic advances for the particular
viewpoint of the fog. In addition, a flexible phase for tasks in Fog Computing is provided
to start to finalize operations of fog about the requirement that the cycle be determined.
Furthermore, these folks presented a case for using IoMT environmental factors to relax
the prosperity problem by fog computing in confirmatory reversal data.
They also considered safety measures to be beneficial in ensuring that the IoT fog
emphasizes an interest in facilitating the provision of fog computing through the use
of diverse protective systems. There are scarcely any research publications considered
in automobiles in fog computing. A car strategy reported since the transmission with
handling business is known as automotive Fog Digesting. They are simply recom-
mended by physicians to increase the partnership with clients to finish the transmis-
sion and preparation, based on vehicle assets. They are optimistic in connection with
this building.
Unused approaches are used by fog vehicles to influence fog computing. In the same
vein, the manufacturers expect a cross-cutting design of the fog vehicles to depict the
dimensions of policy-making and, for instance, the fog centers in different organisations
(Figure 4.4).
72 Mathematical Modeling for Intelligent Systems

4.10 Conclusion
The development of the IoT will bring good fortune with it. Special IoT devices and
technologies associated with healthcare are called IoMT. It will greatly increase our
understanding and save our time, workload, and errors with reliability. We can expect a
good future for the IoT because it is being adopted by everyone. This technology is being
improved day by day, many tests are being held to make this technology handy. New ideas
and reviews by people also help in growing. Blockchain is a connection of data between
computers. It is used to save our data and protect it from others. It’s like a safe lock for
our data. This technology is majorly used in almost every field because everyone wants
their privacy. Storing and protecting data is not as easy as it looks. It requires lots of hard
work  to do these things because we are not dealing with our own but everyone’s data.
With the help of IoMT and Blockchain, an intelligent system can be developed for smart
healthcare.

References
1. Aazam, M., Zeadally, S., Harras, K.A., “Offloading in fog computing for IoT: Review, enabling
technologies, and research opportunities”, Future Generation Computer Systems 87, pp. 278–289
(2018). https://doi.org/10.1016/j.future.2018.04.057
2. https://labs.sogeti.com/iot-vs-edge-vs-fog-computing.
3. https://www.embeddedcomputing.com/technology/iot/wireless-sensor-networks/how-
fog-computing-can-solve-the-iot-challenges.
4. Atlam, H.F., Walters, R.J., Wills, G.B., “Fog computing and the internet of things: A review”, Big
Data and Cognitive Computing 2(2), p. 10 (2018). https://doi.org/10.3390/bdcc2020010
5. https://www.rtinsights.com/what-is-fog-computing-open-consortium.
6. https://www.powersystemsdesign.com/articles/five-reasons-why-your-iot-application-
needs-fog-computing/140/14857.
7. Alli, A.A, Alam, M.M. “The fog cloud of things: A survey on concepts, architecture, standards,
tools, and applications”, Internet Things 9, p. 100177 (2020).
8. https://labs.sogeti.com/iot-vs-edge-vs-fog-computing/
9. Patni, J.C., Ahlawat, P., Biswas, S.S., “Sensors based smart healthcare framework using inter-
net of things (IoT)”, International Journal of Scientific and Technology Research 9(2), pp. 1228–1234
(2020).
10. Taneja, S., Ahmed, E., Patni, J.C., “I-Doctor: An IoT based self patient’s health monitoring
system”, In 2019 International Conference on Innovative Sustainable Computational Technologies
(CISCT) (2019).
11. Khanchi, I., Agarwal, N., Seth, P., Ahlawat, P., “Real time activity logger: A user activity detec-
tion system”, International Journal of Engineering and Advanced Technology 9(1), pp. 1991–1994
(2019).
12. Tiwari, R., Sharma, H.K., Upadhyay, S., Sachan, S., Sharma, A., “Automated parking system-
cloud and IoT based technique”, International Journal of Engineering and Advanced Technology
8(4C), pp. 116–123 (2019).
13. Kshitiz, K., Shailendra, “NLP and machine learning techniques for detecting insulting com-
ments on social networking platforms”, In Proceedings on 2018 International Conference on
Advances in Computing and Communication Engineering (ICACCE), pp. 265–272 (2018).
Smart Healthcare 73

14. https://medium.com/yeello-digital-marketing-platform/what-is-fog-computing-why-fog-
computing-trending-now-7a6bdfd73ef.
15. Gupta, M., Benson, J., Patwa, F., Sandhu, R., “Dynamic groups and attribute-based access con-
trol for next-generation smart cars”, In Proceedings of the Ninth ACM Conference on Data and
Application Security and Privacy, pp. 61–72 (2019).
16. Sharma, A., Tomar, R., Chilamkurti, N., Kim, B.G. “Blockchain based smart contracts for inter-
net of medical things in e-healthcare”, Electronics 9, p. 1609 (2020). https://doi.org/10.3390/
electronics9101609
5
AI-Enabled Cloud-Based Intelligent
System for Telemedicine

Hitesh Kumar Sharma, Ravi Tomar, and Preeti

University of Petroleum and Energy Studies

Prashant Ahlawat
GL Bajaj Institute

CONTENTS
5.1 Introduction .......................................................................................................................... 75
5.2 Artificial Intelligence in Intelligent Systems.................................................................... 76
5.2.1 Application of Artificial Intelligence in Telemedicine Systems ........................ 76
5.2.1.1 AI and BCI..................................................................................................77
5.2.1.2 AI and Nanotechnology...........................................................................77
5.2.1.3 AI with Machine Learning in Research ................................................ 78
5.2.1.4 AI and Drug Creation............................................................................... 78
5.2.2 Challenges of AI in Telemedicine Systems .......................................................... 78
5.3 Cloud Computing in Intelligent Systems ......................................................................... 78
5.3.1 Application of Cloud Computing in Telemedicine Systems .............................80
5.4 Conclusion ............................................................................................................................83
References.......................................................................................................................................83

5.1 Introduction
In the current generation, Artificial Intelligence or AI is one of the most popular fields
in terms of research, applications, jobs, etc. It has a wide range of uses in every sector
and domain that also includes healthcare. Artificial Intelligence is the ability of the com-
puter to sense and analyze data provided to it. Healthcare big database [1] is useful for
AI to implement machine learning algorithms. It helps in finding out necessary solutions
where humans need much more effort to do so. AI can change the Health Infrastructure
by using the data provided by IoT devices over cloud and analyzing it in depth [2]. AI
helps in detecting diseases, administration of chronic situations, deliver health and secu-
rity services, and may help in inventing new drugs or discovering drugs by using the
existing database [2–4]. AI also can plan resources—depending upon the condition of
the patient, it can quickly allocate necessary resources (if required) for them and provide
the information to the doctor. In today’s world, AI can be used in mobile applications;
patient can use these apps to feed data regarding their mood swings, anxiety, and depres-
sion; weight, height, and other parameters, and AI could analyze this data and can provide
an approximate diagnostic for the symptoms [5]. More often, smart devices track such
DOI: 10.1201/9781003291916-5 75
76 Mathematical Modeling for Intelligent Systems

information and upload it over the AI database. In surgeries and operations, AI can use its
database to provide the exact location of a certain wound or internal bleeding. Surgeons
can use this information to carry out the operation/surgery with more accuracy. AI is also
being used to detect tumors that may cause cancer in the future. We can also disburden
Electronic Health Records (EHR) using AI. EHR developers now use AI to create better
processing power and limit the time spent by the user on maintaining EHR by automating
the functions.
Similarly, Cloud Computing is the delivery of different services through the internet,
including data storage, servers, databases, networking, software, and gaming. It makes it
possible to save files to a remote database and retrieve them on demand [6]. Cloud com-
puting in health care can be very beneficial—it will increase the efficiency in sharing of
database of every patient with different hospitals without using any extra efforts. It will
also need less money as hospitals depend on large storage devices to store patient data,
which cost lots of money and their energy consumption is also very high, resulting in high
expenses [7]. By using cloud computing, you can get a large amount of storage at very
less prices, and it will be more secure in emergency and its accessibility is way higher.
Institutes that provide insurances can also get a boost as they can also track the health
of their insurance holder as they can now get more accurate status of health and also can
amend new policies to increase their profits [8]. By uploading different medical conditions,
students all over the world can study different medical conditions and their cure without
much effort. It will also develop the healthcare industries.

5.2 A rtificial Intelligence in Intelligent Systems

AI is basically a machine mimicking basic human cognitive functions. In laymen’s terms, it
is a programming machine to think for themselves and make their own decisions without
the help of humans [9]. This, when applied in healthcare, can become a big boost to how
the whole sector in itself functions. The biggest advantage of using AI in the health sector
is its gathering and processing of data and giving a well-defined result to the user/patient.
These programs can be used for diagnosis, treatments, drug development and manufac-
ture, patient monitoring, and providing personal medication [10]. Basically, every task
done by a doctor can be programmed into or taught to an AI to perform it with perfection.

5.2.1 Application of Artificial Intelligence in Telemedicine Systems

In Figure 5.1, we have shown various roles played by AI in various fields of healthcare.
Another upcoming technology that uses Brain–Computer Interfaces (BCI) aims to help
patients with neurological diseases and trauma. Patients who are unable to communicate
with their bodies and have lost the ability to move or speak can be helped through this
technology. Using BCI and AI, neural activities associated with a simple body function
(such as waving one’s hand) can be analyzed, decoded, and relayed to output devices that
carry out the same desired functions [11]. Such technology can drastically improve many
people’s lives suffering from strokes, spinal cord injuries, or ALS.
Nanotechnology is already a huge part of the healthcare sector with scientists using this
technology to target tumors inside the human body. However, operating at such atomic
levels is very tricky and dangerous. This is where AI plays a very vital role. AI nanobots
AI-Enabled Cloud-Based Intelligent System 77

FIGURE 5.1 Role of AI in smart healthcare.

may currently just be a theory but their applications, when manufactured, are endless.
These tiny “living robots” could enter our bodies, carry out specific tasks, and even help
in curing cancer! AI and ML come together to form Artificial Neural Networks (ANN),
which forms a new branch, called “Deep Learning.” To give a basic idea, ANN uses the
processing of brain to model complex patterns. Through these patterns, it can then predict
unseen data. In healthcare, ANN is used to predict future diseases and inform about clini-
cal diagnosis [12].
Some of the popular application are defined in the following subsections.

5.2.1.1 AI and BCI

Another up-and-coming technology that uses BCI aims to help patients with neurological
diseases and trauma. Patients who are unable to communicate with their bodies and have
lost the ability to move or speak can be helped through this technology. Using BCI and AI,
neural activities associated with a simple body function (such as waving one’s hand) can
be analyzed, decoded, and relayed to output devices that carry out the same desired func-
tions. Such technology can drastically improve many people’s lives suffering from strokes,
spinal cord injuries, or ALS [13].

5.2.1.2 AI and Nanotechnology

Nanotechnology is already a huge part of the healthcare sector with scientists using this
technology to target tumors inside a human body. But operating at such atomic levels is
very tricky and dangerous. This is where AI plays a very vital role. AI nanobots may cur-
rently just be a theory but their applications, when manufactured, are endless. These tiny
“living robots” could enter our bodies and carry out specific tasks and even help in curing
cancer!
78 Mathematical Modeling for Intelligent Systems

5.2.1.3 AI with Machine Learning in Research

AI and ML come together to form ANN which forms a new branch called “Deep Learning.”
To give a basic idea, ANN uses the processing of brain to model complex patterns. Through
these patterns, it can then predict unseen data. In healthcare, ANN is used to predict
future diseases and inform about the clinical diagnosis. Many radiological departments
use physical tissue samples obtained via biopsies for research, but this carries considerable
risks for infection. Here, AI can be used to replace these tissues with a new generation of
radiology tools that are accurate and detailed. These images could also help to understand
cancer cells better and aim for more appropriate treatments [14].

5.2.1.4 AI and Drug Creation

With its own thinking capabilities, and the information from around the world at its disposal,
AI can create medicines, and it even has. DSP-1181, a drug made for OCD, was invented by
AI through joint efforts from some companies and has even been accepted for human trial.

5.2.2 Challenges of AI in Telemedicine Systems

AI in healthcare also faced some challenges due to the criticality of this system. It is directly
connected to human life, so high accuracy and precision are required for using AI-based
algorithms in smart healthcare systems.

5.3 C loud Computing in Intelligent Systems

The healthcare industry has historically been behind in the adoption of the bleeding
edge of technological developments. Given the gravity of the work this industry deals
AI-Enabled Cloud-Based Intelligent System 79

FIGURE 5.2 Cloud storage for medical data.

with, they have to be significantly more careful with any change made and have a smaller
margin of error [16]. If they unwittingly adopt a technology that was faulty in ways not
obvious in the early stages of widespread implementation of the same, then it can lead to
some disastrous situations. Despite this attitude, the healthcare industry has been one of
the fastest and most willing at adapting cloud computing technology. From West Monroe
Partner’s report, 35% of healthcare organizations surveyed held more than 50% of the data
or infrastructure in the “cloud,” or in such a network, as of 2018. Such monumental moves
to adopting this technology have not been seen in any other industry. In Figure 5.2, block
diagram is shown for data storage and usage through the cloud computing model.
Cloud computing, in layman speak is the ability to store data and use processing power
that isn’t exactly “yours,” i.e., physically a part of your computer. For example, it can allow
for barebones computers to execute highly demanding programs or tasks by simply
requesting another computer over some sort of network to do it in its stead, deflecting the
demands of processing, data, and such requirements to the other computer instead. This
strips the would-be stringent requirements of the computer hardware of the barebones
computer this request originated from, making it easier to perform a larger variety of tasks
from “anywhere” given a sufficiently well-made network. Cloud computing provides mul-
tiple models and multiple services to fulfill the computation and storage requirements as
per customers’ need [17–22].
The various types of cloud computing models are:

• Public cloud: Public clouds are those type of cloud computing models where it is
accessible by all. The resources in this model can be accessed by all people who
want those resources. This type of model is provided by Google Cloud, AWS,
Microsoft Azure. When privacy is not a concern, then an organization or indi-
vidual can go for this type of cloud computing model.
80 Mathematical Modeling for Intelligent Systems

• Private cloud: This type of cloud computing model is secured and has restricted
accessibility. The secret data resources are deployed on this kind of cloud com-
puting model. The charges for this type of cloud computing model are high in
comparison to public cloud.
• Hybrid clouds: It is built with both public and private clouds. Private clouds can
support a hybrid model by supplementing local infrastructure with computing
capacity from an external public cloud. An example of hybrid clouds is Research
Compute Clouds which was developed by IBM to connect the computing IT
resources at eight IT research centers. Hybrid cloud works in the middle of public
and private cloud with many compromises in terms of resource sharing.
• Multiclouds: Multiclouds are made from more than one cloud service and more
than one vendor that is public or private. Multicloud helps organizations to avail
benefits from more than one cloud service provider instead of being dependent on
only one service provider.

The various types of cloud computing services are the following:

• Infrastructure as a service (IaaS): The basic building blocks of cloud comput-

ing are contained in IaaS. It usually gives users access to networking capabilities,
computers, and data storage. IaaS allows us to have the most freedom and control
over our IT resources.
• Platform as a service (PaaS): This relieves us of the burden of managing the
underlying infrastructure, allowing us to concentrate on the deployment and
maintenance of our apps. This allows us to be more productive.
• Software as a service (SaaS): This gives us s fully functional product that is man-
aged and maintained by the service provider.

The characteristics of cloud computing services are the following:

• Agility: The cloud provides us quick access to a wide range of technologies, allow-
ing us to develop more quickly and construct almost anything we can dream.
• Elasticity: With cloud computing, we don’t need to over-provision resources to
accommodate future peal levels of the company activity; instead, we allocate the
exact quantity of resources that we require.
• Cost-savings: The cloud enables us to swap fixed costs for variable ones, allowing
us to pay just for IT. Furthermore, due to economies of scale, the variable expendi-
tures are far lower than what we would spend if we did it ourselves.
• Deploy globally in minutes: We may quickly grow to new geographic locations
and deploy internationally using the cloud. Taking the example of AWS, which
has infrastructure all over the world, allows us to deploy the app in different phys-
ical locations in a matter of seconds.

5.3.1 Application of Cloud Computing in Telemedicine Systems

Technology uses internet for storing the data and accessing it through the internet. They
are a cost-effective solution for enterprises. For example, Google cloud which includes app
engine, Google cloud storage, etc., quality has become an important feature of modern
AI-Enabled Cloud-Based Intelligent System 81

healthcare organizations. When it comes to healthcare, it comes to overcome two industry

challenges: increasing cost-effectiveness and building a self-sufficient health ecosystem.
The demand and supply fluctuations increased in the pandemic. The technology server is
a greater mission to establish a smart healthcare community. It’s a huge step toward treat-
ing lifestyle-related noncommunicable diseases.
The advantages of cloud computing platform are the following:

The disadvantages of adopting cloud computing services are the following:

Smart healthcare is a service system that uses technologies such as IoT, internet, etc., to
access people related to healthcare and then manages and responds to medical needs in
an intelligent manner. The adoption of cloud computing requires time and effort from
numerous industries.
As mentioned in the introduction, data storage is one of the primary methods that the
industry has put cloud computing to use. It might be slightly unintuitive, given the name,
but just this facet of cloud computing has tremendously helped the healthcare indus-
try. The first thing having such an efficient method of data storage means that everyone
involved will be able to retrieve it with ease and elegance. This means that people not
familiar with the practice, people like the patients, are easily able to see their own medical
reports. Things like mobile applications, more interactive and informative websites are
now available due to the ease of transfer and organization of data as a result of adopting
the cloud network in the medical industry. A mobile application can provide seamless and
regular updates to patients about their own health, while also reminding them of things
like when and what medicine to take. This has proven to increase the engagement that
patients have with their own health and also has had a positive reaction in general [2].
Another potential benefit to cloud technology is the ease at which an insurance company
or the workplace of the patient can have the information of the medical proceedings in
their hands, alleviating the need for the patient to painstakingly draft up a report. All of
82 Mathematical Modeling for Intelligent Systems

these are just the beginning of cloud computing, as its applications like EMRs, and other
such implementations are being made daily.
Finally, we come to the miscellaneous benefits of cloud computing in healthcare. One of
the most apparent benefits is the flexibility of this system. Most cloud services are handled
by third parties who can handle the large amounts of machinery involved. This allows a
very scalable program and no growth pains, unlike when the analog systems had to be
expanded. Other benefits like the maintenance being handled by the provider. The cloud
services are generally very affordable as compared to buying your own cloud system as
well. All in all, cloud computing can be said to be a unilaterally positive development for
the healthcare industry, and a massive plus for humanity in general.
The centralization of the data storage makes it much easier to manage, and even in such
a simple example in which many nuances are ignored, it is clear how streamlined the
data flow is under such an arrangement. Cloud computing is a useful tool, and the fun-
damental reason for its popularity is that businesses, educational institutions, and other
organizations can rent storage, processing capacity, and applications from a cloud ser-
vice provider. The infrastructure and system maintenance are taken care of by the Cloud
service provider. For businesses, this means less money spent on computers and fewer
employees needed to operate and maintain the system. There are three primary sorts
of clouds: public, private, and hybrid. Microsoft Azure, Amazon Web Services, Google
Cloud, and other public clouds are the most popular. The provider owns and manages
all of the infrastructure and apps in this scenario. The most significant benefit is that the
Cloud provider is responsible for all maintenance and management. Private Clouds, on the
other hand, are restricted to a single firm or organization. The infrastructure for the Cloud
is often managed by an organization’s IT department. The advantages of this strategy for
major organizations are that their data is secure, and they can deliver all of the benefits of
Cloud Computing services inside to their many departments, but at a cost. The best value
is available with Hybrid Cloud. The private cloud is used for very sensitive data, whereas
the public cloud is used for other services. This gives enterprises more control and security
over critical data while yet allowing them to use the public cloud when necessary. Nearly
87% of healthcare firms, for example, choose to employ a hybrid cloud architecture. The
private cloud is typically used to host systems that include sensitive patient data, such as
imaging and EHR. For storing patients’ health files in the cloud, a cloud-based EHR offers
a scalable, adaptable, intuitive, and cost-effective solution. We now have the ease of being
able to communicate with our doctor on the phone or computer from anywhere in the
world, with real-time access to information, anywhere, and compatibility across different
systems as a result of this development. Patients’ portals, telehealth services, healthcare
information exchanges, smart devices, remote monitoring services, and other components
of the new healthcare ecosystem are all powered by this. Patients and doctors do not have
to be physically present at the same area because of telemedicine and telehealth’s Virtual
Care platforms. Patients’ portals allow patients to access their medical records and contact
their doctors safely. Cloud computing is a type of internet-based computing that enables
virtual access to a variety of applications and services such as storage, servers, and net-
working. Cloud computing is a whole new virtualization technology for both individu-
als and businesses that departs from the old software business model. It is defined as
providing end users with remote dynamic access to services and computer resources via
the internet. Medical reports and personal information of the patient are available in the
online records and can be easily accessed by the physicians, pharmacies, and counsel-
ors without charging any extra cost over the diagnosis fees. The healthcare field has a
vast expanse and highly complicated ecosystem, including health insurance companies,
AI-Enabled Cloud-Based Intelligent System 83

hospitals, labs, pharmacies/medical stores, patients, and other bodies as well; hence, in
order to assure good, effective, and fast functionality of this ecosystem, it is very essen-
tial that data is transmitted quickly and securely between the different entities. Cloud
computing provides business models to various hospitals and clinic centers to store the
information of the patient and update it with his/her every visit. Cloud computing helps
in reviewing, exchanging, and sharing the images of MRI, CT-scan, and X-ray reports in
a faster and highly secured environment. Uninterrupted services can be provided to vari-
ous health organizations due to high availability of cloud computing services. The vari-
ous business models of cloud computing benefit the health organizations in many ways.
Cloud computing also provides benefits to the patients like improvement in the quality
of services and collaborations between various health organizations. Since human life is
priceless and precious and the medical facilities across the world are limited, therefore the
healthcare services provided in cloud offer cost-effective concepts so as to benefit both the
patients and the health organizations. Thus, we can see that cloud computing has a big
scope in terms of healthcare as well.

5.4 Conclusion
Artificial Intelligence and Cloud computing together can help the telemedicare services to
perform in a more effective manner. Close by 24*7 availability, clinical consideration pro-
viders certainly need to scale the data amassing and association essentials as per the help
demands. Cloud development can increase or decrease these limit as indicated by the need
of clinical consideration specialists. As cloud development can thoroughly upset your cus-
tomary methodology of data dealing with, clinical benefit providers need to design the
movement cycle well early. Cloud development method diminishes dangers just as it lim-
its the chances of get-away, thwarts information spills, further creates data managing, and
builds up security practices.
In this chapter, we have described the role of AI and Cloud computing in an advanced
Telemedicine System.

References
1. Kustwar, R. K., & Ray, S. (2020). eHealth and telemedicine in India: An overview on the health
care need of the people. Journal of Multidisciplinary Research in Healthcare, 6(2), 25–36. doi:
10.15415/jmrh.2020.62004.
2. Saha, A., Amin, R., Kunal, S., Vollala, S., & Dwivedi, S. K. (2019). Review on “Blockchain tech-
nology based medical healthcare system with privacy issues”. Security and Privacy, 2(5), e83.
doi: 10.1002/spy2.83.
3. Taneja, S., Ahmed, E., & Patni, J. C. (2019). I-Doctor: An IoT based self patient’s health monitor-
ing system. In 2019 International Conference on Innovative Sustainable Computational Technologies
(CISCT).
4. Sharma, A., Choudhury, T., & Kumar, P. (2018). Health monitoring & management using IoT
devices in a Cloud Based Framework. In 2018 International Conference on Advances in Computing
and Communication Engineering (ICACCE), pp. 219–224.
84 Mathematical Modeling for Intelligent Systems

5. Kamdar, N., & Jalilian, L. (2020). Telemedicine: A digital interface for perioperative anesthetic
care. Anesthesia & Analgesia, 130(2), 272–275. doi: 10.1213/ANE.0000000000004513.
6. Patni, J. C., Sharma, H. K., Sharma, S., Choudhury, T., Mor, A., Ahmed, M., & Ahlawat, P.,
(2022) COVID-19 pandemic diagnosis and analysis using clinical decision support systems.
In Tavares, J. M. R. S., Dutta, P., Dutta, S., Samanta, D. (eds) Cyber Intelligence and Information
Retrieval. Lecture Notes in Networks and Systems, vol. 291. Springer, Singapore.
7. Gupta, S., Arya, P., & Sharma, H. K. (2021). User anonymity based secure authentication protocol
for telemedical server systems. International Journal of Information and Computer Security 1(1), 1.
8. Shailender, C., & Sharma H. K. (2018). Digital cancer diagnosis with counts of adenoma and
luminal cells in plemorphic adenoma immunastained healthcare system. IJRAR 5(12), 869–874.
9. Joshi, A., Choudhury, T., Sai Sabitha, A., & Srujan Raju, K. (2020). Data mining in healthcare
and predicting obesity. In Proceedings of the Third International Conference on Computational
Intelligence and Informatics, pp. 877–888, Hyderabad, India.
10. Patni, J.C., Ahlawat, P., & Biswas, S.S. (2020). Sensors based smart healthcare framework using
internet of things (IoT). International Journal of Scientific and Technology Research, 9(2), 1228–1234.
11. Purri, S., Choudhury, T., Kashyap, N., & Kumar, P. (2017). Specialization of IoT applications in
health care industries. In 2017 International Conference on Big Data Analytics and Computational
Intelligence (ICBDAC), pp. 252–256.
12. Krishnan, D. S. R., Gupta, S. C., & Choudhury, T. (2018). An IoT based patient health moni-
toring system. In 2018 International Conference on Advances in Computing and Communication
Engineering (ICACCE), pp. 1–7.
13. Shamshad, S., Mahmood, K., Kumari, S., & Chen, C. M. (2020). A secure blockchain-based
e-health records storage and sharing scheme. Journal of Information Security and Applications 55,
102590.
14. Hussien, H. M., Yasin, S. M., Udzir, S. N. I., Zaidan, A. A., & Zaidan, B. B. (2019). A systematic
review for enabling of develop a blockchain technology in healthcare application: Taxonomy,
substantially analysis, motivations, challenges, recommendations and future direction. Journal
of Medical Systems, 43(10). doi: 10.1007/s10916-019-1445–8.
15. Negi, G., Kumar, A., Pant, S., & Ram, M. (2021). Optimization of complex system reliability
using hybrid grey wolf optimizer. Decision Making: Applications in Management and Engineering,
4(2), 241–256.
16. Koranga, P., Singh, G., Verma, D., Chaube, S., Kumar, A., & Pant, S. (2018). Image denoising based
on wavelet transform using visu thresholding technique. International Journal of Mathematical,
Engineering and Management Sciences, 3(4), 444–449.
17. Fore, V., Khanna, A., Tomar, R., & Mishra, A. (2016). Intelligent supply chain management sys-
tem. In 2016 International Conference on Advances in Computing and Communication Engineering
(ICACCE), pp. 296–302. doi: 10.1109/ICACCE.2016.8073764
18. Srivastava, R., Tomar, R., Gupta, M., Yadav, A.K., & Park, J. (2021). Image watermarking
approach using a hybrid domain based on performance parameter analysis. Information, 12,
310. doi: 10.3390/info12080310
19. Tomar, R., Prateek, M., & Sastry, H. G. (2017). Analysis of beaconing performance in IEEE 802.11p
on vehicular ad-hoc environment. In 2017 4th IEEE Uttar Pradesh Section International Conference
on Electrical, Computer and Electronics (UPCON), pp. 692–696. doi: 10.1109/UPCON.2017.8251133.
20. Gupta, M., Benson, J., Patwa, F., & Sandhu, R. (2019). Dynamic groups and attribute-based
access control for next-generation smart cars. In Proceedings of the Ninth ACM Conference on Data
and Application Security and Privacy, pp. 61–72.
21. Gupta, M., Patwa, F., & Sandhu, R. (2017). Object-tagged RBAC model for the Hadoop ecosys-
tem. In IFIP Annual Conference on Data and Applications Security and Privacy, pp. 63–81. Springer,
Cham.
22. Gupta, M., Patwa, F., & Sandhu, R. (2018). An attribute-based access control model for secure
big data processing in hadoop ecosystem. In Proceedings of the Third ACM Workshop on Attribute-
Based Access Control, pp. 13–24.
6
Fuzzy Heptagonal DEMATEL
Technique and Its Application

A. Felix and PP. Ajeesh

Vellore Institute of Technology

S. Karthik
Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology

R. Dinesh Jackson Samuel

Oxford Brookes University

CONTENTS
6.1 Introduction .......................................................................................................................... 85
6.2 Preliminaries ........................................................................................................................ 86
6.3 Fuzzy Heptagonal DEMATEL Method ............................................................................ 90
6.3.1 Numerical Illustration ............................................................................................. 93
6.4 Conclusion ............................................................................................................................ 96
References....................................................................................................................................... 97

6.1 Introduction
Multicriteria decision-making (MCDM) method is the distinguished division of decision-
making, which creates an interest to the researchers in evaluating, assessing, and ranking
alternatives. In the literature, there are about 20 MCDM techniques (Chen and Chen, 2010)
that have been designed to solve many real-world decision-making problems. Among
them, DEMATEL (decision-making trial and evaluation laboratory) is one of the powerful
tools. It was developed in Geneva Research Centre (Fontela and Gabus 1976; Gabus and
Fontela, 1972) by the Battelle Memorial Institute. This method is an efficient and practi-
cal tool for viewing any difficult structure with matrices/digraphs. The crisp DEMATEL
technique involves only the situation wherein the experts provide their opinion on the cor-
relation between the attributes using the crisp values {0,1}. However, in real-life situations,
correlations among the attributes involve a lot of uncertainty. Thus, the concept of fuzzy
set was introduced to handle the ambiguity and uncertainty of human thought (Zadeh,
1965). Bellmann and Zadeh (1970) introduced decision-making in a fuzzy environment
where the membership functions of the fuzzy decision attain maximum values. Zadeh’s
Fuzzy set theory has paved a way to bring out the fuzzy DEMATEL method (Lin and Wu
2004, 2008; Tseng, 2009a) under the uncertain environment.
When creating a structural model, crisp values are being used to represent human
decisions for determining the relationship among factors. However, in most of the

DOI: 10.1201/9781003291916-6 85
86 Mathematical Modeling for Intelligent Systems

circumstances, human’s decisions with favorites seem hard to evaluate by exact values.
Instead of assigning numerical values in the problem, linguistic variables can be assigned
to take a decision (Zadeh, 1975). In recent times, linguistic terms are widely adopted by
researchers in the decision-making problem. Input variables are represented by linguis-
tic terms in decision-making problems and assigned the values by the fuzzy numbers
(Karthik et al., 2019, 2020). Fuzzy DEMATEL was extended by Lin and Wu (2008) where the
linguistic variables are transformed into triangular fuzzy numbers. Devadoss and Felix
(2013a) used DEMATEL with triangular fuzzy numbers to model the cause and effect fac-
tor of youth violence. Suo et al. (2012) expanded the DEMATEL method in an uncertain
linguistic environment using trapezoidal fuzzy number. Devadoss and Felix (2013a, 2014)
extended DEMATEL technique for hexagonal and octagonal fuzzy numbers.
DEMATEL is one of the potent methods to analyze the cause/effect relationship among the
criteria (Wu and Lee, 2007). This method supports for analyzing the correlations among crite-
ria in the multifaceted systems (Hori and Shimizu, 1999). Currently, fuzzy DEMATEL method
has been used magnificently in various domains such as hospital service quality (Shieh et al.,
2010), hotel service quality (Tseng, 2009b), social problems on youth violence (Devadoss and
Felix, 2013a), human resource management (Abdullah and Zulkifli, 2015), portfolio of invest-
ment projects (Altuntas and Dereli, 2015), business process management (Bai and Sarkis,
2013), green supply chain management (Lin, 2013), global managers competencies (Wu and
Lee, 2007), hotel service quality perceptions (Tseng, 2009b), software system design (Hori and
Shimizu, 1999), extracting consumer’s uneasiness over foods (Tamura et al., 2006), solid waste
management (Tseng and Lin, 2009), emotive Music Composition Selection (Aseervatham
and Devadoss, 2015), climate change (Felix et al., 2017), solid waste management (Selvaraj
et al., 2018; Felix and Dash, 2021), IC fabrication process (Velmathi and Felix, 2019) and bipolar
environment (Deva and Felix, 2021). From this review, it is observed that DEMATEL method
can be extended using heptagonal fuzzy number under uncertain linguistic environment.
To illustrate and demonstrate the heptagonal DEMATEL techniques, the problem on youth
aggressive behavior is studied. The structure of this paper is as follows: the important defini-
tions of fuzzy number, linguistic terms, and modified heptagonal fuzzy numbers are given
in Section 6.2. In Section 6.3, fuzzy DEMATEL-heptagonal method and its application are
presented, and conclusion and direction for future are given in the last section.

6.2  Preliminaries
The important definitions (2.1–2.8) are reviewed and also the modified heptagonal fuzzy
number with respect to alpha cut is explored under an uncertain linguistic environment.
This heptagonal fuzzy number differs from ordinary heptagonal fuzzy numbers (Rathi
and Balamohan, 2014).

Definition 6.2.1

A fuzzy set A ⊆ X, in which each element is mapped to [0, 1] by a membership function µ  :

A
X → [ 0,1]. The membership grade µ A (τ ) represents the degree of truth that τ is a member
of the fuzzy set A . Fuzzy set can also be represented by ordered pair by the following
manner A = {(τ , µ A (τ ))|τ ∈ X }.

Fuzzy Heptagonal DEMATEL Technique 87

Definition 6.2.2

Fuzzy number can be thought as a subset of the real numbers , which is characterized
 : R → [0,1] holds the subsequent properties,
by the membership function, A

(i) A is convex.
µ A ( λτ 1 + (1 − λ )τ 2 ) ≥ min( µ A (τ 1 ), µ A (τ 2 )),

(ii) A is piecewise continuous.

(iii) A is normal, which means the maximum membership value is 1.

Definition 6.2.3
 of the universal set X is defined as the set of all elements
The α-cut of the fuzzy set A
whose membership grade is greater than α, and it is defined as A  α = {τ ∈ X / µ  (τ ) ≥ α },
A
where α ∈[ 0,1].

Definition 6.2.4

A Triangular Fuzzy Number Tr can be defined as a triplet (l, m, r) where l, m, and r are real
numbers such that l ≤ m ≤ r and the membership function µ N (τ ) is defined as

 0 τ <l

 τ
 −l
  l ≤τ ≤ m
 m − l
µTr (τ ) = 
  r −τ 
  m≤τ ≤ r
 r − m

 0 τ >r

where l, m, n are real numbers l ≤ m ≤ n.

Definition 6.2.5

A trapezoidal fuzzy number Tre can be defined as (l, m1 , m2 , n) and the membership func-
tion is defined as

  τ −l 
  m − l  l ≤ τ ≤ m1
 1
 1 m1 ≤ τ ≤ m2
µ Tre (τ ) = 
  n−τ 
  n − m  m2 ≤ τ ≤ n
2

 0 l ≤ 0& n ≥ 0

where l, m1 , m2 , n are real numbers l ≤ m1 ≤ m2 ≤ n.

 88 Mathematical Modeling for Intelligent Systems

Definition 6.2.6

A heptagonal fuzzy number H e can be defined as (a1 , a2 , a3 , a4 , a5 , a6 , a7 ), and the member-

ship function is defined as µ H e (τ )

 (τ − a1 ) , a1 ≤ τ ≤ a 2
 3(a2 − a1 )

 ( a3 − a2 ) + (τ − a2 )
 , a2 ≤ τ ≤ a3
 3(a3 − a2 )

 2(a4 − a3 ) + (τ − a3 ) , a3 ≤ τ ≤ a 4
 3( a4 − a3 )

µ He (τ ) = 
 3(a5 − a4 ) − (τ − a4 ) , a 4 ≤ τ ≤ a5
 3(a5 − a4 )

 2(a6 − a5 ) − (τ − a5 )
 , a 5 ≤ τ ≤ a6
 3(a6 − a5 )

 (a7 − τ ) , a6 ≤ τ ≤ a7
 3(a7 − a6 )

A Heptagonal fuzzy number H e can also be defined as µ H e (τ ) = ( P1 (t), Q1 (u), R1 (v), P2 (t), Q2 (u), R2 (v))
, R1 (v), P2 (t), Q2 (u), R2 (v)) for t ∈[0, 0.33), u ∈[0.33, 0.66), and v ∈[0.66,1], where

(τ − a1 ) ( a − a2 ) + (τ − a2 ) 2(a4 − a3 ) + (τ − a3 )
P1 (t) = , Q1 (u) = 3 , R1 ( v) =
3( a2 − a1 ) 3( a3 − a2 ) 3(a4 − a3 )

( a7 − τ ) 2(a6 − a5 ) − (τ − a5 ) 3(a5 − a4 ) − (τ − a4 )
P1 (w) = , Q2 (u) = , R2 ( v) =
3(a7 − a6 ) 3( a6 − a5 ) 3(a5 − a4 )

Here,

1. P1 (t), Q1 (u), and R1 ( v) is bounded and continuous increasing function over [0, 0.33),
[0.33, 0.66) and [0.66, 1] respectively.
2. P2 (t), Q2 (u) and R2 ( v) is bounded and continuous decreasing function over [0, 0.33),
[0.33, 0.66) and [0.66, 1] respectively.

Definition 6.2.7

The α-cut of the fuzzy set H e of the universe of discourse X is defined as

H e = {τ ∈ X / µH e (τ ) ≥ α } for α ∈[ 0,1]

[ P1 (α ), P2 (α )] for α ∈[0, 0.33)


= [Q1 (α ), Q2 (α )] for α ∈[0.33, 0.66)

[R1 (α ), S2 (α )] for α ∈[0.66,1]
Fuzzy Heptagonal DEMATEL Technique 89

Definition 6.2.8 (Generalized Heptagonal Fuzzy Number)

A generalized heptagonal fuzzy number is defined as GH = ( a1 , a2 , a3 , a4 , a5 , a6 , a7 ; m, n)

where m, n ∈ ( 0,1), and its membership function is defined as (Figure 6.1)

  τ − a1 
 m , a1 ≤ τ ≤ a2
  a2 − a1 

 m − ( m − n )  τ − a2  , a2 ≤ τ ≤ a3
  a − a 
3 2

  τ − a3 
 n − ( n − 1)  a − a  , a3 ≤ τ ≤ a 4
4 3

 1, x = a4
µGH ( x ) = 
 n − ( n − 1)  a5 − τ  , a 4 ≤ τ ≤ a5
  a − a 
5 4

  a6 − τ 
 m − ( m − n)  a − a  , a5 ≤ τ ≤ a6
 6 5

  a −τ 
 m 7 , a6 ≤ τ ≤ a7
  a7 − a6 
 0, τ ≤ a1 and τ ≥ a7


Definition 6.2.9
 = ( a1 , a2 , a3 , a4 , a5 , a6 , a7 ) and B = (b1 , b2 , b3 , b4 , b5 , b6 , b7 ) are the Heptagonal Fuzzy
If A
Numbers, then the addition, subtraction, multiplication, and division are defined as

 ⊕ B = ( a1 + b1 , a2 + b2 , , a7 + b7 )
• A

GH (τ )

0 a1 a2 a3 a4 a5 a6 a7 τ
FIGURE 6.1
Heptagonal fuzzy number.
90 Mathematical Modeling for Intelligent Systems

TABLE 6.1
Heptagonal Linguistic Values and Terms
Linguistic Terms Linguistic Values
No influence (NI) (0, 0, 0, 0, 0.05, 0.1, 0.15)
Very low influence (VL) (0, 0.05, 0.1, 0.15, 0.2, 0.25, 0.3)
Medium low influence (ML) (0.15, 0.2, 0.25, 0.3, 0.35, 0.4, 0.45)
Low influence (L) (0.3, 0.35, 0.4, 0.45 0.5, 0.55, 0.6)
High influence (H) (0.45 0.5, 0.55, 0.6, 0.65, 0.7, 0.75)
Medium high influence (MH) (0.6, 0.65, 0.7, 0.75, 0.8, 0.85, 0.9)
Very high influence (VH) 0.75, 0.85, 0.9, 0.95, 1, 1, 1)

 ΘB = ( a1 − b7 , a2 − b6 ,… , a7 − b1 )
• A

 ⊗ B = ( a1b1 , a2 b2 , a3 b3 , …, a7 b7 )
• A

 ÷ B =  a1 , a2 , a3 ,… , a7 
• A  b b b b1 
7 6 5

A ⊕ B , A
 ΘB , A
 ⊗ B , A
 ÷ B are also Heptagonal Fuzzy Numbers.
Next, the linguistic variables are represented by heptagonal fuzzy numbers in the fol-
lowing Table 6.1.

6.3 Fuzzy Heptagonal DEMATEL Method

In this section, the DEMATEL method is extended through heptagonal fuzzy numbers.

Step 1: Set up the initial linguistic uncertain direct-relation matrix Xˆ k = [ xˆ kij ]n× n ,
k = 1, 2, , n
Let C = {Ci : i = 1, 2, n} be set of attributes/factors. E = {Ek : k = 1, 2,..., m} be the
experts. Experts are asked to provide the relationship between the factors from
the set S = {NI, VL, ML, L, H, MH, VH}. Xˆ k = [ xˆ kij ]n× n given by expert Ek

C1 C2  Cn
C1 
− xˆ k 12  xˆ k 1n 
 
C2  xˆ −  xˆ k 2 n 
Xˆ k = [ xˆ kij ]n× n = 
k 21
, k = 1, 2,…, m
      
xˆ kn1 xˆ kn2  − 
Cn  

Step 2: Change uncertain linguistic matrix Xˆ k = [ xˆ kij ]n× n into heptagonal fuzzy matrix
X k = [x kij ]n×n .
Fuzzy Heptagonal DEMATEL Technique 91

The initial uncertain linguistic direct-relation matrices are transformed into

heptagonal fuzzy numbers. i.e.,
Xˆ k = [ xˆ kij ]n× n is transformed into X k = [ x kij ]n× n . x̂kij = (xk1ij , xkij
2 3
, xkij 4
, xkij 5
, xkij 6
, xkij 7
, xkij ),
k = 1, 2,…, m and i, j = 1, 2,…, n. Particularly, x̂kij = ' − ' is transformed into
x̂kij = (0, 0, 0, 0, 0, 0, 0).
Step 3: Design the group uncertain direct-relation matrix X k = [ x kij ]n× n .
All the uncertain direct-relation matrices X 1 , X 2 ,..., X m are aggregated into a
group uncertain direct-relation matrix X k = [ x kij ]n× n .
It is denoted as x kij = (xk1ij , xkij
2 3
, xkij 4
, xkij 5
, xkij 6
, xkij 7
, xkij ) where

m m m m
1
x =
ij
1
m ∑x
k =1
1
kij , 2
x =
ij
1
m ∑
k =1
2
x , x =
kij
1
m
3
ij ∑x k= 1
3
kij , 4
x =
ij
1
m ∑x
k =1
4
kij ,

m m m

xij5 =
1
m ∑
k= 1
5
xkij , xij6 =
1
m ∑
k =1
6
xkij , xij7 =
1
m ∑x
k=1
7
kij i, j = 1, 2,…, n,

Step 4: Frame normalized uncertain relation matrix Z k = [ z kij ]n× n .

X k = [ x kij ]n× n is changed into Z k = [ z kij ]n× n .
It is given as z ij = ( zij1 , zij2 , zij3 , zij4 , zij5 , zij6 , zij7 ),
where
 n   n   n 
x1
zij1 = ij max 

∑
1≤ i ≤ n  j=1
 x2
xij7  , zij2 = ij max 


1≤ i ≤ n  j=1
∑ 

x7
xij7  , .... , zij7 = ij max 

1≤ i ≤ n 
∑ 
xij7  ,

Here
   j=1

 n 
max 
1≤ i ≤ n

∑ 
xij7  ≠ 0 and 0 ≤ zij1 ≤ zij2 ≤ zij3 ≤ ... ≤ xij7 < 1.,
 j=1 
Z is factorized into seven crisp value matrices Z 1 , Z 2 , Z 3 ... , Z 7

 0 1
z12  z11n   0 2
z12  2
z1n 
   
1  z21
1
0  1
z2n   z21
2
0  2
z2n 
Z = 2
, Z =   , ... ,
           
 zn11 zn1 2 0    2
zn22 0 
   zn 1 

 0 7
z12  7
z1n 
 
 z7 0  7
z2n 
Z 7 =  21 
     
 zn71 zn7 2  0 


By the crisp DEMATEL method, the uncertain overall-relation matrix T is

(
defined as T = lim (Z )1 + (Z )2 + ... + (Z )τ
τ →∞
)
Step 5: Frame uncertain overall-relation matrix T = [tij ]n× n
 92 Mathematical Modeling for Intelligent Systems

 t11 t12  t1n 

 
  t21 t22  t21   1 2 3 4 5 6 7
Let T =   , where tij = (tij , tij , tij , tij , tij , tij , tij ) Then
     
 tn1 tn1  tnn 


[tij1 ]n× n = Z1 ( I − Z1 )−1 , i, j = 1, 2,..., n

[tij2 ]n× n = Z 2 ( I − Z 2 )−1 , i, j = 1, 2,..., n

...

[tij7 ]n× n = Z7 ( I − Z7 )−1 , i, j = 1, 2,..., n

Step 6: Derive the total intensities of influencing and influenced correlation of factors
Ci, ci and hi , i = 1, 2,, n,
Let ci be the overall intensity factor Ci influencing others. It is denoted by
(
ci = ci1 , ci2 , ci3 , ci4 , ci5 , ci6 , ci7 ,
n
) n n

where ci1 = ∑j=1

ti1j , i = 1, 2 ,..., n, ci2 = ∑
j=1
ti2j , i = 1, 2,..., n, ci3 = ∑ t , i = 1, 2,..., n,
j=1
3
ij

n n

ci4 = ∑ t , i = 1, 2,..., n,
j= 1
4
ij ..., ci7 = ∑ t , i = 1, 2,..., n
j=1
7
ij

Let h j be the overall intensity factor C j which is influenced by others. It is represen-

n n
(
ted as h j = h1j , h2j , h3j , h j4 , h5j , h6j , h7j , where h1j = 1
) 2
tij , j = 1, 2,..., n, h j = ∑
ti2j , j = 1, ∑
i= 1 i=1
n n n

2,..., n, h = 3
j ∑ t , j = 1, 2,..., n,
i=1
3
ij
4
h =
j ∑ t , j = 1, 2,..., n, ...,
i=1
4
ij
7
h =
j ∑ t , j = 1, 2,..., n,
i=1
7
ij

Step 7: Derive the uncertain prominence and relation of each factor p i and ri
Let p i be the uncertain prominence of factor Ci. It is denoted as
( )
p i = pi1 , pi2 , pi3 , pi4 , pi5 , pi6 , pi7 , where pi1 = ci1 + hi1 , pi2 = ci2 + hi2 , pi3 = ci3 + hi3 , ... , pi7 = ci7 + hi7 , i = 1, 2,
= ci3 + hi3 , ... , pi7 = ci7 + hi7 , i = 1, 2,..., n ,
Let ri be the uncertain relation of factor Ci. It is denoted as ri = ri1 , ri2 , ri3 , ri4 , ri5 , ri6 , ri7 , ( )
where ri1 = ci1 − hi1 , ri2 = ci2 − hi2 , ri3 = ci3 − hi3 , ... , ri7 = ci7 − hi7 , i = 1, 2,..., n
Step 8: Find the crisp prominence and relation between each factor pi and ri.
The crisp values of prominence (pi) and relation (ri) of factor Ci are determined as,

1 1
pi =
7
(
pi + pi2 + pi3 + pi4 + pi5 + pi6 + pi7 , i = 1, 2,..., n. )
1 1
ri =
7
(
ri + ri2 + ri3 + ri4 + ri5 + ri6 + ri7 , ) i = 1, 2,..., n

Step 9: Construct the causal diagram based on pi and ri.

The importance and classification of factors are determined through the value of
pi and ri. A causal diagram can be depicted based on pi and ri.
Fuzzy Heptagonal DEMATEL Technique 93

6.3.1 Numerical Illustration
Violence is a common event occurring everywhere, which disturbs our life in numerous
ways. Although human beings are taking a lot of effort to form civilized societies, we are
unable to stop the influence of violence and aggression. Therefore, it is our concern to
analyze the cause and effects of violence and the people who are responsible for it. It is
reported by WHO that in most cases of violence, the active participants are the youth. Also,
WHO stated that youth violence is the fourth leading cause of death among youngsters
worldwide, and due to this, two lakhs people die every year (WHO, 2016). Aggressiveness
among the youth is one of the unhealthy developmental problems in many families. The
anger of youth is an easy target and vulnerable section of the society to be misused by
anyone. The terrorist acts, for instance, are motivated by two factors, viz., social and politi-
cal injustice, and hence, violence as the only tool for change. Violence includes a range
of actions from bullying, physical fighting, severe sexual, physical assault to homicide,
and other atrocities. Youth engaging themselves in violence and aggressive behavior is
unpredictable because the factors that influence may vary and they belong to the category
of uncertain factors. It is extremely difficult to predict the behavior of the male youth in
comparison to their female counterpart. Therefore, the study aims to analyze the cause
and effect of youth aggressiveness.
We have collected the following attributes which are related to the youth aggressiveness
through the unsupervised method by interviewing the 100 youths from Chennai, Tamil
Nadu. From their statement, the key purposes for youth violence have been selected as the
attributes. C1—Poor monitoring and supervision of children by parents, C2—Academic
failure/dropping out of school, C3—Delinquent peers/Gang membership, C4—Addiction
to drugs and alcohol, C5—Poverty/unemployment, C6 —Involvement in other forms of
antisocial behavior, C7—Aggressive behavior, C8—Parental substance or criminality, C9—
Depression, C10 —Castisem/inequality. Then, three experts—a sociologist, a psychiatrist,
and victims—were interviewed to give opinions on the existence and intensities of the
correlation among the factors from S = {No Influence, Very Low, Medium Low, Low, High,
Medium High, Very High}.
To illustrate the proposed heptagonal DEMATEL technique, the proposed technique has
been applied to solve the problem of youth violence, and the computational procedure is
summarized as follows:

Step 1: At the initial stage, with the three expert’s view, the initial linguistic uncer-
tain direct-relation matrices Xˆ k , k = 1, 2, 3 are given below in Tables 6.2–6.4.
Step 2 and 3: After transforming initial linguistic uncertain direct-relation matri-
ces into heptagonal fuzzy numbers, the group uncertain direct-relation matrix
X k = [x kij ]n×n . is obtained by aggregating all three matrices, which is shown in
Table 6.5.
Step 4: The uncertain overall-relation matrix Z k is obtained from the group uncertain
direct-relation matrix X k .
Step 5: The uncertain overall-relation matrix T is obtained.
Step 6–8: The overall intensities of influencing and influenced correlation of factors
and prominence (p i) and relation (ri ) are obtained and the crisp prominence (pi)
and relation (ri ) are also obtained, which are shown in Table 6.6. To view the com-
plicated (10 by 10) matrices in a simple manner, the causal diagram is depicted
using pi and ri . The causal diagram divides the factors into causes and effect group.
94 Mathematical Modeling for Intelligent Systems

TABLE 6.2
Initial Uncertain Direct-Relation Matrix X̂1 Provided by E1
C1 C2 C3 C4 C5 C6 C7 C8 C9 C10
C1 – VH VH H H VH H NI L NI
C2 NI – VH H VH VH H NI L H
C3 H VH – L H VH H L H H
C4 L H H – VL H H VL L H
C5 VH H VH L – VL VH VL VH NI
C6 VH H VH VH NI – H H L H
C7 H L L H H VL – H VH VH
C8 NI VH VH H H VL L – VH NI
C9 H NI VL H H H VH H – VH
C10 NI H VH VH L H VH NI H –

TABLE 6.3
Initial Uncertain Direct-Relation Matrix X̂ 2 Provided by E2
C1 C2 C3 C4 C5 C6 C7 C8 C9 C10
C1 – H VH H H H L VL VL NI
C2 VL – VH H VH H L NI VL L
C3 L H – L VH VH H VL L VL
C4 L H H – VL H VH L H L
C5 H L VH H – H VH H VH L
C6 H H H VH L – H L H VH
C7 H L VH H VH H – H VH H
C8 NI VH H VH H VH H – VH H
C9 H L H H VH L VL H – H
C10 NI H VH H L H H NI VH –

TABLE 6.4
Initial Uncertain Direct-Relation Matrix X̂ 3 Provided by E3
C1 C2 C3 C4 C5 C6 C7 C8 C9 C10
C1 – VH VH H H L VL NI L NI
C2 VL – H VH L NI VL H L VL
C3 L H – H VH VH H VH NI L
C4 VL VH H – VL VH H L H VH
C5 VH L VH H – H VH H H VH
C6 H H VH VH VL – H L VL H
C7 H VH VH H H VL – H VH H
C8 NI H VH H VH VH H – VH NI
C9 VL H VL H VH H H H – H
C10 NI VL H H VH VH H NI H –
TABLE 6.5
Group Uncertain Direct-Relation Matrix X k
C1 C2 C3 C4

C1 (0, 0, 0, 0, 0, 0, 0) (0.6, 0.7, 0.7, 0.8, 0.8, 0.9, 0.9) (0.7, 0.8, 0.8, 0.9, 0.9, 1, 1) (0.4, 0.5, 0.5, 0.6, 0.6, 0.7, 0.7)
C2 (0.05, 0.08, 0.11, 0.2, 0.25, 0.3) (0, 0, 0, 0, 0, 0, 0) (0.6, 0.7, 0.7, 0.8, 0.8, 0.9, 0.9) (0.5, 0.6, 0.6, 0.7, 0.7, 0.8, 0.8)
C3 (0.3, 0.4, 0.4, 0.5, 0.5, 0.6, 0.6) (0.5, 0.6, 0.6, 0.7, 0.7, 0.8, 0.8) (0, 0, 0, 0, 0, 0, 0) (0.3, 0.4, 0.4, 0.5, 0.5, 0.6, 0.6)
C4 (0.2, 0.2, 0.3, 0.3, 0.4, 0.4, 0.5) (0.5, 0.6, 0.6, 0.7, 0.7, 0.8, 0.8) (0.4, 0.5, 0.5, 0.6, 0.6, 0.7, 0.7) (0, 0, 0, 0, 0, 0, 0)
C5 (0.6, 0.7, 0.7, 0.8, 0.8, 0.9, 0.9) (0.3, 0.4, 0.4, 0.5, 0.5, 0.6, 0.6) (0.7, 0.8, 0.8, 0.9, 0.9, 1, 1) (0.4, 0.4, 0.5, 0.5, 0.6, 0.6, 0.7)
C6 (0.5, 0.6, 0.6, 0.7, 0.7, 0.8, 0.8) (0.4, 0.5, 0.5, 0.6, 0.6, 0.7, 0.7) (0.6, 0.7, 0.7, 0.8, 0.8, 0.9, 0.9) (0.7, 0.8, 0.8, 0.9, 0.9, 1, 1)
C7 (0.4, 0.5, 0.5, 0.6, 0.6, 0.7, 0.7) (0.3, 0.4, 0.4, 0.5, 0.5, 0.6, 0.6) (0.6, 0.6, 0.7, 0.7, 0.8, 0.8, 0.8) (0.4, 0.5, 0.5, 0.6, 0.6, 0.7, 0.7)
C8 (, 0, 0, 0, 0, 0.0, 0.1, 0.1) (0.6, 0.7, 0.7, 0.8, 0.8, 0.9, 0.9) (0.6, 0.6, 0.7, 0.7, 0.8, 0.8, 0.8) (0.5, 0.6, 0.6, 0.7, 0.7, 0.8, 0.8)
C9 (0.3, 0.3, 0.4, 0.4, 0.5, 0.5, 0.6) (0.2, 0.2, 0.3, 0.3, 0.4, 0.4, 0.5) (0.2, 0.2, 0.3, 0.3, 0.4, 0.4, 0.5) (0.4, 0.5, 0.5, 0.6, 0.6, 0.7, 0.7)
C10 (0, 0, 0, 0, 0.05, 0.1, 0.15) (0.3, 0.3, 0.4, 0.4, 0.5, 0.5, 0.6) (0.6, 0.6, 0.7, 0.7, 0.8, 0.8, 0.8) (0.5, 0.6, 0.6, 0.7, 0.7, 0.8, 0.8)

C5 C6 C7 C8
Fuzzy Heptagonal DEMATEL Technique

C1 (0.4, 0.5, 0.5, 0.6, 0.6, 0.7, 0.7) (0.5, 0.5, 0.6, 0.6, 0.7, 0.7, 0.7) (0.2, 0.3, 0.3, 0.4, 0.4, 0.5, 0.5) (0, 0.0, 0.0, 0.1, 0.1, 0.2, 0.2)
C2 (0.6, 0.6, 0.7, 0.7, 0.8, 0.8, 0.8) (0.4, 0.4, 0.4, 0.5, 0.5, 0.6, 0.6) (0.2, 0.3, 0.3, 0.4, 0.4, 0.5, 0.5) (0.1, 0.1, 0.1, 0.2, 0.2, 0.3, 0.3)
C3 (0.6, 0.7, 0.7, 0.8, 0.8, 0.9, 0.9) (0.7, 0.8, 0.8, 0.9, 0.9, 1, 1) (0.4, 0.5, 0.5, 0.6, 0.6, 0.7, 0.7) (0.3, 0.4, 0.4, 0.5, 0.5, 0.6, 0.6)
C4 (0, 0.05, 0.1, 0.1, 0.2, 0.2, 0.3) (0.5, 0.6, 0.6, 0.7, 0.7, 0.8, 0.8) (0.5, 0.6, 0.6,0.7, 0.7, 0.8, 0.8) (0.2, 0.2, 0.3, 0.3, 0.4, 0.4, 0.5)
C5 (0, 0, 0, 0, 0, 0, 0) (0.3, 0.3, 0.4, 0.4, 0.5, 0.5, 0.6) (0.7, 0.8, 0.8, 0.9, 0.9, 1, 1) (0.3, 0.3, 0.4, 0.4, 0.5, 0.5, 0.6)
C6 (0.1, 0.1, 0.1, 0.2, 0.2, 0.3, 0.3) (0, 0, 0, 0, 0, 0, 0) (0.4, 0.5, 0.5, 0.6, 0.6, 0.7, 0.7) (0.3, 0.4, 0.4, 0.5, 0.5, 0.6, 0.6)
C7 (0.5, 0.6, 0.6, 0.7, 0.7, 0.8, 0.8) (0.1, 0.2, 0.2, 0.3, 0.3, 0.4, 0.4) (0, 0, 0, 0, 0, 0, 0) (0.4, 0.5, 0.5, 0.6, 0.6, 0.7, 0.7)
C8 (0.5, 0.6, 0.6, 0.7, 0.7, 0.8, 0.8) (0.5, 0.5, 0.6, 0.6, 0.7, 0.7, 0.7) (0.4, 0.4, 0.5, 0.5, 0.6, 0.6, 0.7) (0, 0, 0, 0, 0, 0, 0)
C9 (0.6, 0.7, 0.7, 0.8, 0.8, 0.9, 0.9) (0.4, 0.4, 0.5, 0.5, 0.6, 0.6, 0.7) (0.4, 0.4, 0.5, 0.5, 0.6, 0.6, 0.6) (0.4, 0.5, 0.5, 0.6, 0.6, 0.7, 0.7)
C10 (0.4, 0.5, 0.5, 0.6, 0.6, 0.7, 0.7) (0.5, 0.6, 0.6, 0.7, 0.7, 0.8, 0.8) (0.5, 0.6, 0.6, 0.7, 0.7, 0.8, 0.8) (0, 0, 0, 0, 0.0, 0.1, 0.1)

C9 C10

C1 (0.2, 0.2, 0.3, 0.3, 0.4, 0.4, 0.5) (0, 0, 0, 0, 0.0, 0.1, 0.1)
C2 (0.2, 0.2, 0.3, 0.3, 0.4, 0.4, 0.5) (0.2, 0.3, 0.3, 0.4, 0.4, 0.5, 0.5)
C3 (0.2, 0.2, 0.3, 0.3, 0.4, 0.4, 0.5) (0.2, 0.3, 0.3, 0.4, 0.4, 0.5, 0.5)
C4 (0.4, 0.4, 0.5, 0.5, 0.6, 0.6, 0.7) (0.5, 0.5, 0.6, 0.6, 0.7, 0.7, 0.7)
C5 (0.6, 0.7, 0.7, 0.8, 0.8, 0.9, 0.9) (0.3, 0.3, 0.4, 0.4, 0.5, 0.5, 0.5)
C6 (0.2, 0.3, 0.3, 0.4, 0.4, 0.5, 0.5) (0.5, 0.6, 0.6, 0.7, 0.7, 0.8, 0.8)
C7 (0.7, 0.8, 0.8, 0.9, 0.9, 1, 1) (0.5, 0.6, 0.6, 0.7, 0.7, 0.8, 0.8)
C8 (0.7, 0.8, 0.8, 0.9, 0.9, 1, 1) (0.1, 0.1, 0.1, 0.2, 0.2, 0.3, 0.3)
95

C9 (0, 0, 0, 0, 0, 0, 0) (0.5, 0.6, 0.6, 0.7, 0.7, 0.8, 0.8)

C10 (0.5, 0.6, 0.6, 0.7, 0.7, 0.8, 0.8) (0, 0, 0, 0, 0, 0, 0)
96 Mathematical Modeling for Intelligent Systems

TABLE 6.6
Computational Results
p i ri pi ri
C1 (1.5316, 1.9317, 2.4410, 3.1110, (0.1469, 0.1741, 0.2086, 0.2537, 3.8025 0.2736
4.1666, 5.7683, 7.6674) 0.3043, 0.3812, 0.4465)
C2 (1.8578, 2.3160, 2.8985, 3.6639, (−0.2290, −0.2666, −0.3143, −0.3766, 4.4040 −0.4153
4.8294, 6.5980, 8.6646) −0.4515, −0.5652, −0.7042)
C3 (2.3550, 2.8825, 3.5528, 4.4333, (−0.3063, −0.3394, −0.3816, −0.4371, 5.2418 −0.4811
5.7514, 7.7516, 9.9661) −0.5240, −0.6559, −0.7233)
C4 (2.0175, 2.5160, 3.1495, 3.9814, (−0.2688, −0.2986, −0.3366, −0.3866, 4.7462 −0.4320
5.2100, 7.0744, 9.2744) −0.4632, −0.5796, −0.6904)
C5 (2.1761, 2.6837, 3.3288, 4.1762, (0.1301, 0.1462, 0.1668, 0.1938, 4.9557 0.2140
5.4433, 7.3661, 9.5157) 0.2321, 0.2903,0.3385)
C6 (2.0976, 2.5951, 3.2273, 4.0577, (−0.0186, −0.0202, −0.0222, −0.0249, 4.8306 −0.0241
5.3014, 7.1886, 9.3467) −0.0296, −0.0369, −0.0163)
C7 (2.1532, 2.6674, 3.3206, 4.1786, (0.0529, 0.0583, 0.0652, 0.0744, 4.9613 0.0830
5.4460, 7.3693, 9.5942) 0.0892, 0.1117, 0.1291)
C8 (1.6839, 2.0998, 2.6292, 3.3254, (0.4588, 0.5150, 0.5864, 0.6802, 4.0403 0.7466
4.4234, 6.0897, 8.0307) 0.8151, 1.0200, 1.1509)
C9 (1.9758, 2.4555, 3.0652, 3.8661, (−0.0487, −0.0577, −0.0690, −0.0838, 4.6234 −0.0857
5.0718, 6.9013, 9.0279) −0.1007, −0.1264, −0.1137)
C10 (1.7698, 2.1933, 2.7323, 3.4414, (0.0827, 0.0889,0.0967, 0.1069, 4.1721 0.1210
4.5623, 6.2631, 8.2424) 0.1283, 0.1608, 0.1829)

Here, pi shows the degree of importance that factor Ci acts in the entire system. On
the contrary, the ri portrays the effect that factor Ci pays to the system. Precisely,
if ri is +ve, factor Ci is a cause group, while factor Ci is an effect group if ri is −ve.

To view the computational results in (Table 6.6), a causal diagram is provided in Figure 6.2.

From the causal diagram, it is observed that the evaluation criteria were virtually divided
in to the cause group, including C1—Poor monitoring and supervision of children by par-
ents, C5—Poverty/unemployment, C7—Aggressive behavior, C8—Parental substance or
criminality, C10 —Castisem/inequality, while the effect group was composed of criteria
C2—Academic failure/dropping out of school, C3—Delinquent peers/Gang membership,
C4—Addiction to drugs and alcohol, C6 —Involvement in other forms of antisocial behav-
ior, and C9—Depression. Moreover, by only considering the relation of factors, a ranking
order C8  C1  C5  C10  C7  C6  C9  C4  C2  C3 is obtained for making good deci-
sion. Consequently, C1—Poor monitoring and supervision of children by parents and C8—
Parental substance or criminality are the major causes of youths are involving in violence.
Therefore, parent’s role is very important in the behavior of the youth and their growth.

6.4 Conclusion
This present study has proposed an extension of DEMATEL method using heptagonal
fuzzy numbers for analyzing the relationship among factors. With the aid of this technique,
Fuzzy Heptagonal DEMATEL Technique 97

FIGURE 6.2
Causal diagram.

the complex interaction between factors and digraphs can be converted into an observ-
able model. The merit of this technique lies in catching the complexity of any problem
through which an insightful decision can be taken by the decision-maker. Compared with
the other existing DEMATEL methods, the unique characteristics of this new extended
DEMATEL method are the following: (a) appropriate to solve the problem of interrelated
factor analysis in fuzzy environment when the judgments and decisions are uncertain in
nature and (b) this technique acts better to reduce the uncertainty instead of using trian-
gular and trapezoidal fuzzy number. This proposed technique can also be applied in all
decision-making fields such as Engineering, Data Science, Medical, and Social Science.
Further research, linguistic variables may be represented by nonagonal and decagonal
fuzzy number to extend the DEMATEL method.

References
Abdullah, L., & Zulkifli, N. (2015). Integration of fuzzy AHP and interval type-2 fuzzy DEMATEL: An
application to human resource management. Expert Systems with Applications 42(9):4397–4409.
Altuntas, S., & Dereli, T. (2015). A novel approach based on DEMATEL method and patent cita-
tion analysis for prioritizing a portfolio of investment projects. Expert Systems with Applications
42(3):1003–1012.
Aseervatham, S., & Devadoss A. V. (2015). Analysis on criteria based emotive music composition
selection using a new trapezoidal fuzzy DEMATEL - TOPSIS hybrid technique. Journal of Fuzzy
Set Valued Analysis 2:122–133.
Bai, C., & Sarkis, J. (2013). A grey-based DEMATEL model for evaluating business process manage-
ment critical success factors. International Journal of Production Economics 146(1):281–292.
Bellmann, R. E., & Zadeh, L. A. (1970). Decision making in fuzzy environment. Management Science
17(4):141–164.
98 Mathematical Modeling for Intelligent Systems

Chen, J. K., & Chen, I. S. (2010). Using a novel conjunctive MCDM approach based on DEMATEL,
fuzzy ANP, and TOPSIS as an innovation support system for Taiwanese higher education.
Expert System with Application 37(3):1981–1990.
Deva, N., & Felix, A. (2021). Designing DEMATEL method under bipolar fuzzy environment. Journal
of Intelligent and Fuzzy Systems 41(6):7257–7273.
Devadoss, A. V., & Felix, A. (2013a). A Fuzzy DEMATEL approach to study cause and effect relation-
ship of youth violence. International Journal of Computing Algorithm 2:363–372.
Devadoss, A. V., & Felix, A. (2013b). A new Fuzzy DEMATEL method in an Uncertain Linguistic
Environment. Advances in Fuzzy Sets and Systems 16(2):93–123.
Devadoss, A. V., & Felix, A. (2014). An extension of fuzzy DEMATEL- hexagonal techniques. In
Proceedings of International Conference on Mathematical Sciences, Sathyabama University.
Felix, A., Christopher, S., & Mani, R. (2017). Fuzzy cognitive DEMATEL technique for modeling cause
and effect of climate change. International Journal of Pure and Applied Mathematics 117(4):655–661.
Felix, A., & Dash, S. K. (2021). Haar FCM with DEMATEL techniques to analyze the solid waste man-
agement. Advances in Smart Grid Technology, Lecture Notes in Electrical Engineering 688:393–402.
Fontela, E., & Gabus, A. (1976). The DEMATEL observer, DEMATEL 1976 report. Battelle Geneva
Research Centre, Geneva.
Gabus, A., & Fontela, E. (1972). World problems an invitation to further thought within the framework of
DEMATEL. Battelle Geneva Research Centre, Geneva.
Hori, S., & Shimizu, Y. (1999). Designing methods of human interface for supervisory control sys-
tems. Control Engineering Practice 7(11):1413–1419.
Karthik, S., Sarojkumar, D., & Punithavelan, N. (2019). Haar ranking of linear and non-linear hep-
tagonal fuzzy number and its application. International Journal of Innovative Technology and
Exploring Engineering 8(6):1212–1220.
Karthik, S., Sarojkumar, D., & Punithavelan, N. (2020). A fuzzy decision-making system for the
impact of pesticides applied in agricultural fields on human health. International Journal of
Fuzzy System Applications 9(3):42–62.
Lin, R. J. (2013). Using fuzzy DEMATEL to evaluate the green supply chain management practices.
Journal of Cleaner Production 40:32–39.
Lin, C. J., & Wu, W. W. (2004). A fuzzy extension of the DEMATEL method for group decision mak-
ing. European Journal of Operational Research 156(1):445-455.
Lin, C. J., & Wu, W. W. (2008). A causal analytical method for decision making under fuzzy environ-
ment. Expert System with Applications 34(1):205–213.
Rathi, K., & Balamohan, S. (2014). Representation and ranking of fuzzy numbers with heptago-
nal membership function using value and ambiguity index. Applied Mathematical Sciences
8(87):4309–4321.
Selvaraj, A., Dash, S. K., & Punithavelan, N. (2018). Hexagonal fuzzy DEMATEL approach to ana-
lyze the solid waste management. International Journal of Pure and Applied Mathematics 118(5):
475–492.
Shieh, J. I., Wu, H. H., & Huang, K. K. (2010). A DEMATEL method in identifying key success factors
of hospital service quality. Knowledge-Based System 23(3):277–282.
Suo, W. L., Feng, B., & Fan, Z. P. (2012). Extension of the DEMATEL method in an uncertain linguistic
environment. Soft Computing 16(3):471–483.
Tamura, H., Okanishi, H., & Akazawa, K. (2006). Decision support for extracting and dissolving
consumer’s uneasiness over foods using stochastic DEMATEL. Journal of Telecommunication
Information and Technology 4:91–95.
Tseng, M. L. (2009a). A causal and effect decision making model of service quality expectation using
grey-fuzzy DEMATEL approach. Expert System with Applications 36(4):7738–7748.
Tseng, M. L. (2009b). Using the extension of DEMATEL to integrate hotel service quality perceptions
into a cause-effect model in uncertainty. Expert Systems with Applications 36(5):9015–9023.
Tseng, M. L., & Lin, H. Y. (2009). Application of fuzzy DEMATEL to develop a cause and effect
model of municipal solid waste management in Metro Manila. Environmental and Monitoring
Assessment 158:519–533.
Fuzzy Heptagonal DEMATEL Technique 99

Velmathi, G., & Felix, A. (2019). IC fabrication process steps analysis using fuzzy DEMATEL method.
International Journal of Innovative Technology and Exploring Engineering 9(1):1968–1973.
WHO (2016). https://www.who.int/news-room/fact-sheets/detail/youth-violence
Wu, W. W., & Lee, Y. T. (2007). Developing global managers competencies using the fuzzy DEMATEL
method. Expert Systems with Applications 32:499–507.
Zadeh, L. A. (1965). Fuzzy Sets. Information and Control 8(2):338–353.
Zadeh, L. A. (1975). The concept of a linguistic variable and its application to approximate reasoning
(Part II). Information Science 8:301–357.
7
A Comparative Study of Intrapersonal and
Interpersonal Influencing Factors on the Academic
Performance of Technical and Nontechnical Students

Deepti Sharma, Rishi Asthana, and Vaishali Sharma

BML Munjal University

CONTENTS
7.1 Introduction ........................................................................................................................ 101
7.1.1 Intrapersonal Factors ............................................................................................. 102
7.1.2 Interpersonal Influencing Factors ....................................................................... 103
7.2 Aim of the Research .......................................................................................................... 105
7.3 Methodology ...................................................................................................................... 106
7.4 Data Analysis ..................................................................................................................... 107
7.5 Results and Discussions ................................................................................................... 108
7.6 Conclusion .......................................................................................................................... 114
References..................................................................................................................................... 115

7.1 Introduction
Students’ varying academic temperaments often surprise the academicians. Some stu-
dents’ enthusiasm to embrace academic challenges delights the academicians, whereas
others’ sheer reluctance and disengagement toward academics disheartens their endeav-
ors. Students’ diverse and miscellaneous academic behavior thus intrigues the researchers
to unravel the possible factors that help them understand their students’ academic perfor-
mance. Mere assessment of some assignments and examinations cannot give an elabora-
tive picture of students’ behavior and attitude toward their academic work and success in
their formal education
A lot of research has already been done to study the different factors which affect stu-
dents’ performance (Saa et al., 2019; Cavilla, 2017; Anderson & Good, 2016; Gibson, 1982)
but we have categorized the various factors into intrapersonal and interpersonal influenc-
ing factors to see how they impact students’ academic performance.
The cultural diversity of India makes it unique and remarkable in every sense. It is quite
challenging for universities to accept and embrace the multidimensional aspects of diver-
sity (Cooke & Saini, 2010). The students come from diverse backgrounds/cultures, which
to some extent propel their ways of formulating an opinion and also influence their intra-
personal and interpersonal influencing factors. Hence, the varied backgrounds of the stu-
dents fascinated us to study how different factors navigate the mindsets of the students
which affect their academic performance culture is “the way of life especially the general

DOI: 10.1201/9781003291916-7 101

102 Mathematical Modeling for Intelligent Systems

customs and beliefs, of a particular group of people at a particular time” (The Cambridge
English Dictionary, 2015).
Intrapersonal as well as interpersonal influencing factors consciously and sometimes
unconsciously affect an individual’s behavior. Kassarnig et al. (2018) state that several
behavioral patterns are closely attached to academic performance. They can influence
social ties and other activities in which students participate. These factors also shape the
individual’s mindset. Mindset is interpreted as a collection of our thoughts and beliefs,
which culminates our inner and outer world. Many researchers have shared their valu-
able insights to define Mindset. Dweck and Yeager (2019) state, Mindsets help to organize
multiple variables such as goals, attributions, helplessness, etc. to provide them a clear
meaning, which further adds a significant value of right efforts to achieve better results.
Various factors such as leadership (Deng et al., 2020), parenting style (Yang & Zhao,
2020) personality traits (Credê & Kuncel, 2008), or stress (Elias et al. 2011) collectively play
a significant role to affect the mindset of the students which consciously or unconsciously
impacts their academic performance. The cumulative effect of different factors paves a
route to consolidate different thoughts. People with fixed mindset believe that intelligence
is fixed and cannot be changed, whereas people with growth mindset feel that intelligence
can be developed and improved with the right efforts.
Various intrapersonal and interpersonal influencing factors pave our way of thinking
and sometimes motivate or demotivate us in different situations. It is hence difficult to
ascertain which factor has an edge over the other. Wentzel and Wigfield (1998) stated that
a comprehensive study of students’ academic behavior is possible when both social and
academic factors are sincerely taken into consideration.

7.1.1 Intrapersonal Factors
Intrapersonal factors are the individual’s beliefs such as a sense of self, awareness to
restrict behavior in extreme states, recognition of the power of expression, etc. Tatnell et al.
(2014) state that among intrapersonal factors, emotion regulation and self-esteem show the
strongest role in behavior. These factors have tremendous scope of bringing exponential
growth in life provided the individual is aware of his abilities. Rampullo et al. (2015) state
intrapersonal factors play a significant role; they can act as a source of support or a barrier.
They are the talents and abilities which reinforce the individual’s competence to face and
accomplish the challenges and goals. Flavian (2016) states that students’ self-fulfillment is
based on realistic self-awareness. Research shows that factors including students’ aware-
ness about themselves, their goals to achieve, and their expectations from themselves and
external world play a pivotal role in their academic performance and mental well-being
(Gu et al., 2015). Bandura, an eminent Social Cognitive Psychologist, stated that self-effi-
cacy or confidence is one of the pivotal factors which guides an individual to develop an
optimistic view about his goal. It is self-belief in one’s abilities that ascertains the success-
ful accomplishment of the task. They are the predictors of the guaranteed completion of
the task (Bandura, 1997; Bandura et al., 1999).
To hone intrapersonal skills, language determines a significant role. Language is the
most potent vehicle of expression. English is the most common language used in edu-
cation and professional world. Nonnative English learners may feel uncomfortable and
anxious while expressing their opinions. MacIntyre et al. (1998) describe foreign language
anxiety as “the worry and negative emotional reaction aroused when learning or using a
second language.” It is perceived that students suffer foreign language learning anxiety
and it reflects on their academic and peer performance: “Language anxiety can originate
Technical and Nontechnical Students 103

from learner’s own sense of self, their self-related cognitions, language learning difficul-
ties, differences in learners” (Hashemi, 2011). Students who suffer from language anxiety
usually avoid eye contact and feel uncomfortable in expressing their thoughts in the class-
room and gradually they start avoiding the class discussions.
Each student has a unique way to respond to setbacks and challenges. Some accept the
hurdles to improvise their performance and some may get intimidated and see them as
insurmountable blocks.
Anxiety, nervousness, and stress are the immediate responses, which occur when stu-
dents face some unexpected challenges. It is interesting to study to what extent anxiety
obstructs academic performance. Research shows different results, in a few cases, it was
found that anxiety impedes success (Seipp, 1991; Keogh et al. 2004; Dobson 2012; Macher
et al., 2012; Owens et al., 2012) On the contrary, in a few cases, anxiety helped the students
to perform better (Alkhalaf, 2018).
Joining university is a huge transition in students’ life. A major shift takes place in their
physical, mental, psychological, and social arena. The sudden shift may overwhelm some
students to adapt to the new changes. As mentioned earlier, individuals with growth
mindset will take this as a huge opportunity as they enter a new phase of life. On the
contrary, those who feel little intimidated by the change may get affected by some of their
preconceived notions or feedback they receive about the teachers. Students who frequently
miss the mark become fearful of new tasks and distraught by setbacks (Dweck, 2006).
Prejudices, preconceived notions, and biasedness can have a major influence on some-
body’s behavior if he lacks the vision to pierce through them to see the reality (Duckitt,
1992; Olson, 2009; Rattan & Dweck, 2010).
Some students are eager to take feedback from their seniors and peers about the uni-
versity, curriculum, faculty, other activities, etc. (Owen, 2016). They are not affected by
the feedback they receive as they wish to make their own opinions as their academic
journey progresses. They don’t allow their preconceived notions to affect their thinking.
Cognitive biases can impair the decision-making process and can disrupt collective out-
comes (Caviola et al. 2014).
Correct opinions are formed after an exhaustive analysis of the situation. Self-driven
individuals do not form opinions easily. They critically evaluate the situation, spend time
in thorough analysis, observe how others respond in the same situation (Moussaïd et al.,
2013), and then derive their conclusions. Conscious opinions help in taking wiser decisions
(Shahsavarani & Abadi, 2015).

7.1.2 Interpersonal Influencing Factors

Interpersonal influencing factors are the external factors that students face while coming
in contact with their external surroundings. In the book The Psychology of Interpersonal
Relationships, the authors quoted the work of social psychologists Baumeister and Leary
who stated that since evolution, humans showed a strong inclination toward interpersonal
attachments as they provided them a need to belong to a certain group (Berscheid & Rega,
2005).
Interpersonal influencing factors comprise different external entities which come into
our contact on daily basis. They consist of different social groups which are an integral
part of a personality development (Lamb & Bornstein, 2011).
To study the effect of interpersonal influencing factors on students we have extensively
concentrated on the teacher’s verbal and nonverbal skills, pedagogy, inclusion of tech-
nology, and socio-cultural diversity that enhance or affect student engagement. The ever
104 Mathematical Modeling for Intelligent Systems

engaging and active bond of the teacher and student always strengthens the overall learn-
ing which takes place in the class (Gablinske, 2014). The teacher not only provides academic
support but also looks after the emotional and social being of the students (Varga, 2017).
A well-informed and conscious teacher knows that his/her knowledge and skills have
a direct and indirect effect on the students. Many studies show that the teacher’s com-
munication style has an indelible effect on the students. Right paralinguistic features such
as voice modulation, pitch, rate of speech and tone, determine the effectiveness of com-
munication. Positive and efficient communication motivates the students significantly. On
the contrary, weak and fragile communication not only demotivates the students but also
pushes them to lose interest in a particular subject. It further affects students’ communica-
tion style due to which they might struggle to voice their opinion in public (Ataunal, 2003).
Effective teaching depends on the teacher’s effective communication. A teacher with
effective communication skills understands the needs of the students and uses their com-
munication skills to make learning an enjoyable and fruitful experience. Amadi and Paul
(2017) state that a teacher’s efficient communication skills have a direct positive effect on a
students’ academic performance.
Besides verbal, a teacher’s nonverbal communication also leaves a remarkable impression
on the students’ attentiveness. Nonverbal communication includes kinesics, proxemics,
chronemics, and paralinguistic features. Kinesics includes correct postures, appropriate
gesticulation, positive body language, right facial expressions, and correct eye contact.
Proxemics helps us to use the space while communicating with others. Bambaeeroo and
Shokrpour (2017) found that nonverbal cues should complement verbal communication,
and balanced communication enhances students learning and improves academic per-
formance. Sutiyatno (2018) also states that verbal and nonverbal communication has a
significant positive impact on a students’ learning process. Zeki (2009) indicated that the
teacher’s nonverbal communication creates a comfortable environment for learning and
motivates students to participate in group discussions.
A major transformation can be seen in teaching methodology due to the advancement of
technology. Technology has opened multiple platforms where students and teachers find
a constant inflow of knowledge (Bruenjes, 2002). Burton (2003) also supports that incor-
porating technology in the class also helps a teacher to build better rapport with the stu-
dents. In order to meet the pace, the teaching pedagogies are also evolving. Teachers are
experimenting with blended learning to improvise their teaching style. They are making
a fine balance by adopting multiple approaches to deal with different topics (Muema et al.,
2018). The conventional way of teaching is used, especially to explain theoretical concepts.
The interactive mode of teaching promotes an active and participative environment in the
class. The interactive mode of teaching enhances collaborative learning. An amalgama-
tion of novel techniques in teaching bridges the gap between the teacher and the student
(Jamian & Baharom, 2012). It fosters innovation and creativity, enhances the ability to ask
the right questions, gives an opportunity to everyone to participate in the tasks more effi-
ciently, and pushes the individuals to break their boundaries and open up to new chal-
lenges (Dexter et al, 1999; Loveless & Ellis, 2001).
University education is always progressive and encompasses diversity to spread its
arena. The inclusion of students from different backgrounds has always been a welcoming
aspect of university education. Cultural diversity opens new avenues for the students to
practically learn various things about different cultures. Each culture possesses a multi-
tude of different beliefs, values, assumptions, attitudes to respond in different situations.
It also affects our responses and decision-making and also influences the development of
personalities (Traindis & Suh, 2002; McCullers & Plant, 1964; Sears, 1948) Most of the time
Technical and Nontechnical Students 105

we judge people as per our own beliefs, which are to an extent influenced by our cultural
influences.
This diversity makes a university a vibrant place where people from diverse places
promote new language skills and novel ways of thinking by sharing their experiences.
Teaching ethnically diverse classrooms starts from embracing views and employing expe-
riential learning that can transmute the differences into positivity (Civitillo et al., 2018).
Students are often receptive to assimilate and accept each other without giving much
consideration to cultural differences; however, they may face some challenges in the begin-
ning, such as perceived discrimination, social isolation, and cultural adjustment (Wu et al.,
2015; Wright & Schartner, 2013; Russell et al. 2010). A study conducted by Newsome and
Cooper, in 2016, pointed the difficulties such as isolation, loneliness, racial discrimination,
and economic exploitation faced by international students in adjusting and accommodat-
ing to a new environment. Although these differences do not impact the students’ mindset
significantly, they may leave certain marks of their presence in the initial days. These bar-
riers can be overcome with the passage of time.

7.2 A im of the Research

This chapter is a comparative analysis of technical and nontechnical students to study and
discern the various factors which affect their academic performance. Human behavior
cannot be studied in isolation as both internal and external worlds and intra- and inter-
personal factors play a combined role to shape behavior. Through our topic we wanted
to study the different perspectives and responses of the intrapersonal and interpersonal
influencing factors on the academic performance of the technical and nontechnical stu-
dents. The authors have used the terms intrapersonal and interpersonal influencing fac-
tors to observe their effects on students’ academic performance. The term technical refers
to engineering undergraduates and nontechnical refers to Management and Law students.
To analyze intrapersonal factors, we focused on different behavioral patterns such as anxi-
ety, influence of preconceived notions, feedback from the seniors regarding the teachers,
and the relation of students’ language skills on their academic performance. For inter-
personal influencing factors, we have included the effects of teachers’ verbal and non-
verbal communication, teaching pedagogy: conventional and blended, social factors such
as cultural influences. Through our secondary research, we studied papers that either
focused on intrapersonal factors such as behavioral issues or interpersonal factors such as
a teacher’s influence on the students. Above all, we could not find any paper which made
a comparative study of technical and nontechnical students to observe the impact of the
mentioned factors. Our literature review research propelled us to take a new approach to
make a comprehensive and inclusive study of both the factors—intrapersonal and inter-
personal influencing factors to see their impact on academic performance. Hence, we took
both the factors and prepared broad questions which could possibly answer the important
issues—which we thought are very important and significantly impact the academic per-
formance. We also believe that the program structure of both technical and nontechnical
groups incites the students to develop a certain thought process and consequently they
behave and respond differently in different situations. Most of the technical courses are
experiment-based and challenge an inquiry-based learning; on the other hand, the subjec-
tive style of the nontechnical courses encourages a more detailed and subjective approach
106 Mathematical Modeling for Intelligent Systems

that provokes a discussion-based learning. An adequate understanding of the intraper-

sonal and interpersonal influencing factors will always support the facilitator to under-
stand and bridge the academic and nonacademic gaps that otherwise impede the teaching
learning process, if remain unaddressed.

7.3 Methodology
The population of the study consisted of undergraduate students of an Indian university,
located in Haryana. The sample population consisted of 310 undergraduate students, out
of which 258 students were in a technical group and 52 students were in a nontechnical
group. The term technical refers to engineering undergraduates and nontechnical refers
to management and law undergraduate students. The researchers prepared a question-
naire using a four-point Likert scale. The sample population of three schools of the same
university: School of Engineering and Technology, School of Management, and School of
Law filled the questionnaire shared through a Google form. As seen in Tables 7.1 and 7.2,
the Questionnaire had ten questions that were framed keeping the Intrapersonal and
Interpersonal influencing factors in mind.
As seen in Table 7.3, a four-point Likert scale has been used in this study ranging from
Always, Sometimes, Rarely, Never. We have transformed this to numeric values in the fol-
lowing ways: Always = 4, Sometimes = 3, Rarely = 2, Never = 1.

TABLE 7.1
Questionnaire—Intrapersonal Factors
S. No. Intrapersonal Factors
1 Which specific language skill do you feel is the cause of poor academic performance? (due to your
lack of understanding or efficiency of that skill)
2 Does lack of language proficiency impact the overall understanding and performance of core
engineering/management
3 To what extent/how frequently does stress or anxiety impact academic performance?
4 Do the preconceived notions, feedback from the seniors about the teacher affect the overall interest/
performance in any particular subject?

TABLE 7.2
Questionnaire -Interpersonal Influencing Factors
S. No. Interpersonal Influencing Factors
1 Does teacher’s accent or pronunciation affect your understanding of any subject?
2 Does cultural difference affect the academic performance?
3 To what extent does the usage of digital teaching aids like power point presentations, online
courses, models, videos etc., help in improving the understanding of the subject?
4 Does conventional teaching hinder the overall understanding of the subject?
5 Does the interactive mode of teaching help in better understanding of the subject?
6 Do nonverbal communication skills: Body language, postures, gestures, facial expressions and eye
contact of the teacher effect in generating interest in any particular subject?
Technical and Nontechnical Students 107

TABLE 7.3
Likert Scale and Weightage Used
Likert Scale Agreement Weightage
Always 4
Sometimes 3
Rarely 2
Never 1

The value of Cronbach’s alpha is 0.53. A possible reason for this Cronbach’s alpha value
could be the different types of questions that were not similar and consistent in nature.
Since this chapter aimed to see the impact of the intrapersonal and interpersonal factors, we
decided to take the first response of the students to avoid the biasedness of the responses.

7.4 Data Analysis
The first question of the questionnaire was the following: Which specific language skill do
you feel is the cause of poor academic performance? (Due to your lack of understanding
or efficiency of that skill). The students were asked to write any one skill from Listening,
Speaking, Reading and Writing. Many authors (Nan, 2018; Evans, 2018; Sadiku, 2015;
Wallace et al., 2004) have focused on LSRW—Listening, Speaking, Reading and Writing—
the four skills that are essentially required while learning a language and known as lan-
guage foundation skills. Since this question asked them to write only one skill and did not
provide options such as Always, Sometimes, Rarely and Never, the analysis of this ques-
tion has been done separately.
From Figures 7.1 and 7.2, we can see that both technical and nontechnical groups think
that their listening skill is the weakest among all the language skills. Although both the
groups acknowledge that listening is the weakest skill, still the percentage of listening as
the weakest skill is more in the technical group as compared with the nontechnical. There
could be a number of reasons for weak listening skills. The majority of the students feel
that listening comes naturally. They interpret listening and hearing as the same and don’t
put in the required efforts in listening.

FIGURE 7.1
Shows the language skills of the technical group.
108 Mathematical Modeling for Intelligent Systems

FIGURE 7.2
Shows the language skills of the nontechnical group.

The students get distracted easily due to their short span of concentration. Although
students give various reasons for their inept listening skills, at the same time they also
realize the barriers to listening can be overcome with their consistent practice. One of the
ways to improvise listening skills is active participation, which motivates the students to
stay focused while listening. The pie charts also present some interesting facts besides
the weakest skill. The technical group sees listening as the weakest skill and speaking as
the second weakest. The difference between the two skills is significant. Although in the
nontechnical group, the students feel listening is their weakest skill, they find writing and
speaking skills equally challenging after listening. This may happen because the nontech-
nical students often write subjective articles/essays and present them frequently. Due to
that they may feel their writing and speaking skills are constantly challenged.
Rest all questions were analyzed following the Likert scale. Each of the questions had
four options: Always, Sometimes, Rarely, Never. We have transformed this to numeric val-
ues (refer Table 7.3) in the following ways: Always = 4, Sometimes = 3, Rarely = 2, Never = 1.
We have taken the opinion of students from the following two groups: technical and non-
technical. There were 258 students in the technical group and 52 students in the non-
technical group. We have compared their average (mean) opinion on each question using
t-distribution for two samples of technical and nontechnical groups. In this case, we have
tested null hypothesis: mean of two populations are same vs alternative hypothesis: mean
of two populations are not same. We have also compared the variance of two groups using
F-distribution. In this case, we have tested null hypothesis: variances of two populations
are same vs. alternative hypothesis: Variances of two populations are not same. Here we
have discussed results related to each of these nine questions.

7.5 Results and Discussions

As shown in Table 7.4, we can see that there is no difference in average opinion of the popu-
lations of two groups (Technical and nontechnical groups) to measure/observe the impact
of lack of language proficiency on subject understanding and anxiety impact on the aca-
demic performance. The average opinion of both the samples for this intrapersonal factor
is close to 3. The mean results indicate that on average both groups think that sometimes
lack of language proficiency impacts the overall understanding and performance of core
Technical and Nontechnical Students 109

TABLE 7.4
Intrapersonal Factors—t Test
Preconceived Notions,
Lack of Language Feedback about the
Proficiency on Anxiety Impact on Teacher Affect the
Subject the Academic Interest/Performance in
Intrapersonal Factors Understanding Performance Subject
Mean of technical group 2.810 3.100 2.759
Mean of nontechnical group 2.846 2.923 2.5
Variance of technical group 0.714 0.581 0.665
Variance of nontechnical group 0.995 0.660 0.725
Hypothesized mean difference 0 0 0
t Stat −0.243 1.452 2.019
P(T≤t) two-tail 0.808 0.150 0.047
t critical two-tail 1.996 1.994 1.993
Conclusion Do not reject null Do not reject null Reject the null hypothesis.
hypothesis. hypothesis.

engineering/management subjects. Lack of language proficiency may hinder the students’

freedom of expressing their point of view, especially for those who come from semiurban
or remote areas where English is not used as the first language to teach different subjects.
This may affect their oral and written skills. Technical students may feel confident while
writing their core courses exams or assignments as their content is standardized around
the technical concepts, but they feel uneasy during presentations.
Nontechnical students generally deal with more subjective topics, and lack of language
proficiency impedes their written as well as oral communication and restricts their opin-
ions due to limited vocabulary.
The average opinion of both the samples for anxiety impact on the academic performance
is also close to 3. We can conclude that on average, both groups think that sometimes stress
or anxiety influences academic performance. Based on the results shown in Table 7.4, it can
be ascertained that anxiety to a certain degree affects academic performance. The reasons
for distress may vary in technical and nontechnical groups due to the differing nature and
assessment patterns of the technical and nontechnical courses. Technical courses are more
scientific and empirical in nature. The assessments are also designed to evaluate the scien-
tific and technical aptitude of the students. Besides theoretical, most of the assessments are
lab-based where the students check or prove the validity of the concepts, which demand
robust technical skill sets. The meticulous and ordered pattern of the course and assess-
ment may create anxiety in several students, especially in those who lack confidence in
performing such assessments. Nontechnical subjects are more theoretical and interactive
in nature. The assessments are also flexible as compared with the technical courses. Since
nontechnical courses have more presentations-based assessments, the students who have
low confidence and are conscious because of inept presentation skills may feel somewhat
anxious while expressing their opinions.
Means for two samples of technical and nontechnical groups are 2.759 and 2.5, respec-
tively, for interpersonal factor preconceived notions, feedback about the teacher affect
the interest/performance in subject. Since there is a difference in average opinion of two
populations, we can conclude that both groups think differently on how the preconceived
notions, for example, feedback from the seniors about the teacher, affect the overall interest/
110 Mathematical Modeling for Intelligent Systems

TABLE 7.5
Intrapersonal Factors-F Test
Lack of Language Anxiety Impact Preconceived Notions, Feedback
Intrapersonal Proficiency on Subject on the Academic about the Teacher Affect the
Factors Understanding Performance Interest/Performance in Subject
Variance of technical 0.714 0.581 0.802
group
Variance of 0.995 0.660 0.685
nontechnical group
Hypothesis Variance of two Variance of two Variance of two population are
populations are same populations are same
same
F 1.392 1.136 1.169
F critical 1.397 1.397 1.468
Conclusion Do not reject null Do not reject null Do not reject the null hypothesis
Hypothesis. Hypothesis.

performance in any subject. Majority of students in the technical group in comparison to

the nontechnical group believe that sometimes it affects the overall inderest/performance
in any subject. It is observed that Engineering students have very high expectations from
their teachers. They always feel that the course is challenging, and it requires tremendous
potential to perform well to attain high technical skills. Engineering students are more
conscious about their performance, so they may feel a little overwhelmed if they receive
any feedback about the teacher from their seniors. This can work in both ways: if they hear
positive feedback, then it may help them to increase their interest in that subject. On the
contrary, if they receive negative feedback, then it may increase their nervousness as they
are already a little daunted due to the complexity of the course. Nontechnical students may
also get affected by the feedback, but it does not affect their performance much as their
courses focus more on the soft skills where they are more interested to hone their interper-
sonal skills and hence are not much bothered by the feedback they receive from the seniors.
As shown in Table 7.5, we have compared the variability of the opinion of the two groups
(technical and nontechnical groups) for three interpersonal factors. Although there is a
difference in the variance of two groups for lack of language proficiency on subject under-
standing, we can conclude that there is not much variability in the opinion of the two
populations. Similarly, we can conclude for Anxiety Impact on the Academic Performance
that there is not much variability in the opinion of two populations. There is not much
difference in the sample variances of technical and nontechnical groups for preconceived
notions, feedback about the teacher affecting the interest/performance in a subject. As it
is evident from Table 7.5, we can conclude that there is no variability in the opinion of two
populations for this interpersonal factor.
As shown in Table 7.6, we have compared the average opinion of two groups (Technical
and nontechnical groups) for different interpersonal factors. We can see that there is no
difference in average opinion of the populations of two groups (technical and nontechni-
cal groups) for the factors: teachers’ accent/pronunciation affect subject understanding,
integrating digital teaching aids to improve the subject understanding, interactive mode
of teaching helps in better subject understanding, nonverbal communication of the teacher
affects in generating interest in the subject, cultural difference affects the academic per-
formance and there is a difference in average opinion of the two groups for the factor:
conventional teaching hinders the subject understanding.
Technical and Nontechnical Students 111

TABLE 7.6
Interpersonal Influencing Factors-t Test
Integrating Nonverbal
Teachers Digital Conventional Interactive Communication Cultural
Accent/ Teaching Aids Teaching Mode of of the Teacher Difference
Pronunciation Improving the Hinder the Teaching Help Affect in Affect the
Interpersonal Affect Subject Subject Subject in Better Subject Generating Academic
Factors Understanding Understanding Understanding understanding Interest in Subject Performance
Mean of 2.717 3.431 2.751 3.759 3.399 1.964
technical
group

Mean of 2.5 3.442 2.480 3.711 3.557 2.096

nontechnical
group
Variance of 0.966 0.535 0.802 0.237 0.684 0.947
technical
group
Variance of 1 0.486 0.685 0.483 0.526 0.951
nontechnical
group
Hypothesized 0 0 0 0 0 0
mean
difference
t Stat 1.431 −0.097 2.124 0.476 −1.402 −0.884
P(≤) two-tail 0.156 0.461 0.036 0.635 0.164 0.379
t critical two-tail 1.993 1.992 1.991 1.999 1.990 1.992
Conclusion Do not reject null Do not reject the Reject the null Do not reject the Do not reject the null Do not reject
hypothesis. null hypothesis. hypothesis. null hypothesis. hypothesis. the null
hypothesis.

Means for two samples of technical and nontechnical groups are 2.717 and 2.5, respec-
tively, for the factors: teachers’ accent/pronunciation affects subject understanding. Since
there is no difference in average opinion of two populations, we can conclude that both
groups think that the teacher’s accent or pronunciation rarely or sometimes affect their
understanding of any subject. Many of them believe it sometimes influences their under-
standing. The teacher’s paralinguistic skills play a vital role in overall students’ learning
process. For technical students, communication skills predominantly affect the students
in their skills and perspective courses, where they need good communication skills to
perform well. If some students struggle to understand the teacher’s accent, then they may
feel difficult to cope with the course.
On the other hand, if the students find the teacher’s accent unclear or faulty, they struggle
to understand the course content. Consequently, it may hinder their understanding of the
core engineering courses because these courses are full of technical jargon. Nontechnical
students don’t use too much jargon and can enjoy the flexibility of subjective terms and
expressions.
If some students are already apprehensive about their language skills and constantly
strive to understand things, then they may take time to understand the teacher’s accent.
On the other hand, if the students don’t appreciate the teacher’s accent but understand the
concept taught by him/her, then it may not have much impact on their academic perfor-
mance. The result also supports/highlights this fact.
Means for two samples of technical and nontechnical groups are 3.431 and 3.442, respec-
tively, for the factor: integrating digital teaching aids improve the subject understanding.
112 Mathematical Modeling for Intelligent Systems

Since there is no difference in average opinion of two populations (in this case sample
means are very close as well), we can conclude that both groups think that usage of
digital teaching aids like ppts, online courses, models, and videos, sometimes or always
help in improving the understanding of the subject. Majority of them believe that it
always helps in improving the understanding of the subject. For this question, average
opinions of two groups are very similar, and this fact is also verified with high p-value.
This clearly shows that the students in both groups feel that technology expands their
horizon of learning. Technology strengthens their problem-solving aptitude. It not only
helps them to identify the gaps of learning but also offers various platforms to bridge
those gaps. Through advanced technology, they can get access to various new courses.
Online learning is one such example. The result shows that both technical and nontech-
nical students welcome and embrace blended learning that amalgamates technology
with theoretical concepts.
Means for two samples of technical and nontechnical groups are 2.751 and 2.480, respec-
tively, for the factor: conventional teaching hinders the subject understanding. Since
there is a difference in average opinion of the two populations, we can conclude that both
groups think differently about conventional teaching. Many students of the technical
group in comparison to the nontechnical group believe that sometimes it hinders the over-
all understanding of the subject. Conventional teaching is interpreted as long descriptive
lectures with minimum or no usage of technology. Technical courses are more focused on
experiments. The technical students are always inquisitive to validate the concepts that
are taught in the class. It can be deduced from the results that technical students may
like to give more importance to task-oriented lectures that promote experiential learning
instead of listening to long descriptive lectures. That does not deny the value of theoretical
lectures but the majority of them welcome blended learning over conventional learning.
Nontechnical subjects are more subjective and descriptive such as subjects taught to law
students. The students understand the nature of the course and hence are prepared to
listen to the lectures. This could be one of the reasons that the majority of nontechnical
students are comfortable with conventional methods of teaching.
Means for two samples of technical and nontechnical group are 3.759 and 3.711, respec-
tively, for the factor: interactive mode of teaching help in better subject understanding.
Since there is no difference in average opinion of two populations (in this case sample
means are very close as well), we can conclude that both groups think that the interac-
tive mode of teaching always helps in better understanding of the subject as two samples
means are very close to four. For this question, the average opinions of the two groups are
very similar and this fact is also verified with high p-value. Through interactive teaching, a
teacher actively involves the students in their learning process by way of regular teacher–
student interaction, student–student interaction, use of audio–visuals, and hands-on dem-
onstrations. The students act as active contributors and learn significantly through such
interactions. It is clearly demonstrated in the result that both groups hail and acknowledge
the advantages of interactive mode of teaching.
Means for two samples of technical and nontechnical groups are 3.399 and 3.557, respec-
tively, for the factor: nonverbal communication of the teacher affects in generating interest
in the subject. Since there is no difference in average opinion of two populations, we can
conclude that both groups think that nonverbal communication skills that include body
language, postures, gestures, facial expressions, and eye contact of the teacher sometimes
or always affect in generating interest in any subject. The majority of them believe that
nonverbal communication skills: body language, postures, gestures, facial expressions,
and eye contact of the teacher always promote interest in any subject. Both the groups feel
Technical and Nontechnical Students 113

that nonverbal communication improves a person’s ability to relate, engage, and establish
meaningful interactions in everyday life and leads to developing a good rapport with the
teacher, which helps them to develop more interest and participation in the subject.
The means for two samples of the technical and nontechnical group are 1.964980545 and
2.096153846, respectively, for the factor: cultural difference affects academic performance.
Since there is no difference in average opinion of two populations (in this case sample
means are very close as well), we can conclude that both groups think that cultural dif-
ferences rarely affect academic performance as 2 samples means are very close to 2. For
this question, the average opinions of the two groups are very similar and this fact is also
verified with high p-value. The students in both groups feel that cultural diversity does
not have any tangible effect on their performance.
As shown in Table 7.7, we have compared the variability of the opinion of two groups
(technical and nontechnical groups) for the questions addressing different interpersonal
influencing factors.
As it is evident from Table 7.7, there is not much difference in the sample variances of
technical and nontechnical groups for the interpersonal influencing factors discussed in
the table. We can conclude that there is no variability in the opinion of two populations
for the factors: teacher’s accent/pronunciation affects subject understanding, integrating
digital teaching improves the subject understanding, conventional teaching hinders the
subject understanding, teacher’s nonverbal communication yields in generating interest
in the subject and, cultural differences do affect the academic performance. There is sig-
nificant variability in the opinions of two populations for the factor: interactive mode of
teaching promotes better subject understanding, although there is no difference in the
average opinion of two populations for this factor. In this case, we can conclude that there
is variability in opinion of two groups that the interactive mode of teaching helps in better
understanding of the subject. This variation is more for nontechnical group in comparison
to the technical group as evident from sample variances of technical and nontechnical
groups.

TABLE 7.7
Interpersonal Influencing Factors-F Test
Teachers Integrating Conventional Interactive NVC of the Cultural
Accent/ Digital Teaching Teaching Mode of Teaching Teacher Affect Difference
Pronunciation Aids Improving Hinder the Help in Better in Generating Affect the
Interpersonal Affect Subject the Subject Subject Subject Interest in Academic
Factors Understanding Understanding Understanding Understanding Subject Performance
Variance of 0.966 0.535 0.665 0.237 0.684 0.947
technical
group
Variance of 1 0.486 0.725 0.483 0.526 0.951
nontechnical
group
Hypothesis Variance of two Variance of two Variance of two Variance of two Variance of two Variance of
population are population are population are population are population are two
same same same same same population
are same
F 1.034 1.098 1.089 2.034 1.301 1.003
F critical 1.397 1.468 1.397 1.397 1.468 1.397
Conclusion Do not reject the Do not reject the Do not reject the reject the null Do not reject the Do not reject
null hypothesis null hypothesis null hypothesis hypothesis null hypothesis the null
hypothesis
114 Mathematical Modeling for Intelligent Systems

7.6 Conclusion
Coping with students’ academic challenges is a complex and rigorous experience for aca-
demicians. Motivating students to channelize their intellect and emotions in the right
direction requires robust and consistent efforts. The present study presents the impact of
intrapersonal and interpersonal influencing factors on the academic performance of techni-
cal and nontechnical students and highlights how different factors differently affect each
group’s academic performance. The responses to the questionnaire helped us to under-
stand how these factors play their roles in two different groups. We could deduce some
interesting results through the varied responses of each group. For a few factors, the groups
showed similar opinions: as the fondness of digital learning in both groups, the impact of
lack of proficiency in language skills on their academic performance, nonverbal commu-
nication of the teachers and their effect on their interest in that subject, the impact of anxi-
ety on academic performance, least effect of cultural diversity on academic performance.
Although the average opinion of the two groups toward an interactive mode of teaching
is similar, there is variability in the opinion of the two groups. Besides similarities, we
also found some differences in the opinions of the technical and nontechnical groups. The
technical group believes that conventional teaching may hinder their performance but the
nontechnical group shows no such indifference toward conventional teaching, the techni-
cal group believes that the teacher’s accent and paralinguistic skills can have an effect on
their understanding of the subject, but the nontechnical group seems less affected from
this. The technical group indicates the possibility of the effect of the preconceived notions
or the feedback they receive about the teacher from their seniors on their academic perfor-
mance, but the nontechnical group does not agree to this thought much. It is an enriching,
scholastic, and edifying experience to see the differences in the responses of the technical
and nontechnical students. The study of the intrapersonal and interpersonal influencing
factors on two different groups has broadened our view as teachers to understand the effect
of different factors on the students’ academic performance. The reasons for these differ-
ences are explained in the result analysis section, and we feel that one of the major reasons
for the difference of opinions in the two groups could be the nature of the courses that the
technical and nontechnical students pursue. We are very hopeful that the results of this
study will be promising for the academicians to fathom the academic challenges of the
technical and nontechnical students and provide them their timely assistance to boost their
academic performance. A future study on overcoming the factors can be pursued to see the
direct advantages on academic performance. Factors can be undertaken individually to see
their vast implication in the academic realm. These factors can also be studied to see their
effects on postgraduate students. Following are the limitations of this research:

• We conducted the survey only once to avoid the biasedness of the responses. Had
the survey been conducted more than once, we could have applied test–retest
reliability.
• We could only collect the sample from one university from different courses that
the university offers. To obtain better results, the sample size could have been
varied and vast.
• We had decided the influencing factors—intrapersonal and interpersonal—before
drafting the questionnaire and adhered to those factors only. In the future, we
would like to explore factors by conducting factor analysis.
Technical and Nontechnical Students 115

References
Alkhalaf AM (2018) Positive and negative affect, anxiety, and academic achievement among medical
students in Saudi Arabia. Inter Journal of Emergency Mental Health and Human Resilience 20(2):
397.
Amadi G, Paul AK (2017) Influence of student-teacher communication on students’ academic
achievement for effective teaching and learning. American Journal of Educational Research 5(10):
1102–1107.
Anderson AS, Good, DJ (2016) Increased body weight affects academic performance in university
students. Preventive Medicine Reports 5: 220–223.
Ataunal A (2003) Niçin ve nasıl bir öğretmen? (Why and what kind of teacher?). Ankara: Milli Eğitim
Vakfı Yayınları.
Bambaeeroo F, Shokrpour N (2017) The impact of the teachers’ non-verbal communication on suc-
cess in teaching. Journal of Advances in Medical Education and Professionalism 5(2): 51–59.
Bandura A (1997) Self-efficacy: The exercise of control. New York: W H Freeman/Times Books/Henry
Holt & Co.; MacMillan Publisher.
Bandura A, Freeman WH, Lightsey R (1999) Self-efficacy: The exercise of control. Journal of Cognitive
Psychotherapy 13(2): 158–166.
Berscheid ES, Rega PC (2005) The psychology of interpersonal relationships. Upper Saddle River, NJ:
Pearson Prentice Hall.
Bruenjes LS (2002) A multi-case study investigating the disposition of faculty use of technology as a
teaching and learning tool in the higher education classroom. Doctoral dissertation University
of Massachusetts, Lowell. UMI 3041377.
Burton DB (2003) Technology professional development: A case study, Academic Exchange Quarterly
7(2): 2378–2381.
Cambridge English Dictionary (2015) Meaning of culture- Archived from the original on August 15,
2015. Retrieved July 26, 2015.
Cavilla D (2017) The effects of student reflection on academic performance and motivation. Sage
Open 7(3), 2158244017733790.
Caviola L, Mannino A, Savulescu J, Faulmüller N (2014) Cognitive biases can affect moral intuitions
about cognitive enhancement. Frontiers in Systems Neuroscience. https://www.frontiersin.org/
articles/10.3389/fnsys.2014.00195/full
Civitillo S, Juanga LP, Schachner MK (2018) Challenging beliefs about cultural diversity in educa-
tion: A synthesis and critical review of trainings with pre-service teachers. Educational Research
Review 24: 67–83.
Cooke FL, Saini DS (2010) Diversity management in India: A study of organizations in different
ownership forms and industrial sectors. HR Science Forum 40(3): 477–500.
Credê M, Kuncel NR (2008) Study habits, skills, and attitudes: The third pillar supporting collegiate
academic performance. Perspectives on Psychological Science 3(6): 425–453.
Deng W, Li X, Wu H, Xu G (2020) Student leadership and academic performance. China Economic
Review 60
Dexter SL, Anderson RE, Becker HJ (1999) Teachers’ views of computers as catalysts for changes in
their teaching practice. Journal of Research on Computing in Education 31(3): 221–239.
Dobson C (2012) Academic anxiety and coping with anxiety effects of academic anxiety on the
performance of students with and without learning disabilities and how students can cope
with anxiety at school, Master Thesis, Northern Michigan University, https://www.nmu.edu/
education/sites/DrupalEducation/files/UserFiles/Dobson_Cassie_MP.pdf
Duckitt J (1992) Prejudice and behavior: A review. Current Psychology 11(4): 291–307.
Dweck CS (2006) Mindset: The new psychology of success. New York, NY: Random House.
Dweck CS, Yeager DS (2019) Mindsets: A view from two eras. Perspectives on Psychological Science
14(3), 481–496.
116 Mathematical Modeling for Intelligent Systems

Elias H, Ping WS, Abdullah MC (2011) Stress and academic achievement among undergraduate stu-
dents in Universiti Putra Malaysia. Procedia-Social and Behavioral Sciences 29: 646–655.
Evans MB (2018) The Integration of Reading, Writing, Speaking, and Listening Skills in the Middle
School Social Studies Classroom in the Middle School Social Studies Classroom. Utah State
University Utah State University, All Graduate Theses and Dissertations. 7157. https://digitalc-
ommons.usu.edu/cgi/viewcontent..cgi?article=8259&context=etd
Flavian H (2016) Towards teaching and beyond: Strengthening education by understanding stu-
dents’ self-awareness development. Power and Education 8(1), 88–100.
Gablinske PB (2014) A case study of student and teacher relationships and the effect on student learn-
ing and the effect on student learning, Doctorate Thesis, University of Rhode Island University.
https://digitalcommons.uri.edu/cgi/viewcontent..cgi?article=1284&context=oa_diss
Gibson MA (1982) Reputation and respectability: How competing cultural systems affect students’
performance in school, Anthropology & Education Quarterly. https://anthrosource.onlinelibrary.
wiley.com/doi/abs/10.1525/aeq.1982.13.1.04x0462t
Gu J, Strauss C, Bond R, Cavanagh K (2015) How do mindfulness-based cognitive therapy and
mindfulness-based stress reduction improve mental health and wellbeing? A systematic
review and meta-analysis of mediation studies. Clinical Psychology Review 37: 1–12.
Hashemi M (2011) Language stress and anxiety among the English language learners. Procedia-
Social and Behavioral Sciences 30: 1811–1816.
Jamian AR, Baharom R (2012) The application of teaching aids and school supportive factors in
learning reading skill among the remedial students in under enrolment schools. Procedia-
Social and Behavioral Sciences 35:187–194
Kassarnig V, Mones E, Nielsen AB, Sapiezynski P, Lassen DD, Lehmann S (2018) Academic perfor-
mance and behavioral patterns. EPJ Data Science 7: 1–16.
Keogh E, Bond FW, French CC, Richards A, Davis RE (2004) Test anxiety, Susceptibility to distrac-
tion and examination performance. J of Anxiety, Stress and Coping 17(3): 241–252.
Lamb ME, Bornstein MH (2011) Social and Personality Development: An Advanced Textbook, Psychology
Press Taylor & Francis Group, New York.
Loveless A, Ellis V (2001) ICT, Pedagogy and the Curriculum. Taylor and Francis Group. https://
teknologipendidikankritis.files.wordpress.com/2011/11/loveless-ellis_ict-pedagogy-and-cur-
riculum.pdf
Macher D, Paechter M, Papousek I, Ruggeri K (2012) Statistics anxiety, trait anxiety, learning behav-
ior, and academic performance. European Journal of Psychology of Education 27: 483–498.
MacIntyre PD, Dörnyei Z, Clément R, Noels KA (1998) Conceptualizing willingness to communicate
in a L2: A situational model of L2 confidence and affiliation, Journal of The Modern Language
82(4): 545–562.
McCullers JC, Plant WT (1964) Review of educational research growth. Development, and Learning
34(5): 599–610.
Moussaïd M, Kämmer JE, Analytis PP, Neth H (2013) Social influence and the collective dynamics of
opinion formation. PLoS One, 8(11): e78433.
Muema JS, Mulwa DM, Mailu, SN (2018) Relationship between teaching method and students. per-
formance in mathematics in public secondary schools in Dadaab Sub County, Garissa County;
Kenya. IOSR Journal of Research & Method in Education 8(5): 59–63.
Nan C (2018) Implications of Interrelationship among four language skills for high school english
teaching, Journal of Language Teaching and Research 9(2) 418–423.
Newsome LK, Cooper P (2016) International students’ cultural and social experiences in a British
University: “Such a hard life [it] is here”. Journal of International Students 6(1): 195–215.
Olson MA (2009) Measures of prejudice. In: Nelson TD, editor. Handbook of prejudice, stereotyping, and
discrimination. New York: Psychology Press 367–386.
Owen L (2016) The impact of feedback as formative assessment on student performance. Inter J of
Teaching and Learning in Higher Education 28(2): 168–175.
Technical and Nontechnical Students 117

Owens M, Stevenson J, Hadwin JA, Norgate R (2012) Anxiety and depression in academic per-
formance: An exploration of the mediating factors of worry and working memory. School
Psychology International 33: 433–449.
Rampullo A, Licciardello O, Castiglione C (2015) Intrapersonal factors effects on professional orien-
tation and environmental representations. Procedia-Social and Behavioral Sciences 205(9):422–428.
Rattan A, Dweck CS (2010) Who confronts prejudice? The role of implicit theories in the motivation
to confront prejudice. Psychological Science 21(7): 952–959.
Russell J, Rosenthal D, Thomson G (2010) The international student experience: Three styles of adap-
tation. Higher Education, 60(2): 235–249.
Saa AA, Emran MA, Shaalan K (2019) Factors affecting students’ performance in higher education:
a systematic review of predictive data mining techniques. Technology, Knowledge and Learning
24: 567–598.
Sadiku LM (2015) The importance of four skills reading, speaking, writing, listening in a lesson
hour. European Journal of Language and Literature Studies 1(1): 29–31.
Sears RR (1948) Personality development in contemporary culture. Proceedings of the American
Philosophical Society 92(5): 363–370.
Seipp B (1991) Anxiety and academic performance: A meta-analysis of findings” Journal of Anxiety
Research 4(1): 27–41.
Shahsavarani AM, Abadi EAM (2015) The bases, principles, and methods of decision-making: A
review of literature. International Journal of Medical Reviews 2(1): 214–225.
Sutiyatno S (2018) The effect of teacher’s verbal communication and non-verbal communication on
students english achievement. Journal of Language Teaching and Research 9(2): 430–437.
Tatnell R, Kelada L, Hasking P, Martin G (2014) Longitudinal analysis of adolescent NSSI: The role
of intrapersonal and interpersonal factors. Journal of Abnormal Child Psychology 42(6): 885–896.
Traindis H, Suh EM (2002) Cultural influences on personality. Annual Review of Psychology 5(1):
133–160.
Varga M (2017) The effect of teacher-student relationships on the academic engagement of students.
Master Thesis, Graduate Programs in Education, Goucher College.
Wallace T, Stariha E, Walberg, HJ (2004) Teaching, speaking, listening and writing. Educational practices
series-14. International Academy of Education, International Bureau of Education. https://files.
eric.ed.gov/fulltext/ED495377.pdf
Wentzel KR, Wigfield A (1998) Academic and Social Motivational Influences on Students’ Academic
Performance. Educational Psychology Review 10:155–175.
Wright C, Schartner A (2013) “I can’t...I won’t?” International students at the threshold of social inter-
action. Journal of Research in International Education 12(2):113–128.
Wu HP, Garza E, Guzman N (2015) International student’s challenge and adjustment to col-
lege. Education Research International 2015: 202753. https://www.hindawi.com/journals/
edri/2015/202753/
Yang J, Zhao X (2020) Parenting styles and children’s academic performance: Evidence from middle
schools in China. Children and Youth Services Review 113: 105017.
Zeki CP (2009) The importance of non-verbal communication in classroom management. Procedia-
Social and Behavioral Sciences 1(1): 1443–1449.
8
PIN Solar Cell Characteristics:
Fundamental Physics

Aditya N. Roy Choudhury

Techno India University

CONTENTS
8.1 Introduction ........................................................................................................................ 119
8.2 Working of the PIN Solar Cell ......................................................................................... 120
8.3 Dark Current....................................................................................................................... 123
8.4 Photocurrent ....................................................................................................................... 127
8.5 Conclusion .......................................................................................................................... 131
References..................................................................................................................................... 131

8.1 Introduction
Solar cells are prevalent in our lives today. Roadside lamp posts use solar cells, solar vehi-
cles are roaming in the city, and space stations up in the sky are using solar panels for pow-
ering up. Some natural questions arise: what are solar cells? How do they function? Well,
a short answer is that solar cells are power sources: they convert sunlight into electricity.
They act like batteries when they are illuminated. That is, when there is light, a solar cell
produces a voltage.
Like many other devices, solar cells have two terminals. Under illumination, if the two
terminals are shorted, a short-circuit current flows. To understand solar cell action, further
a current–voltage study is needed; questions need to be addressed like, under illumination,
and subject to a voltage, what happens to a solar cell’s current? Under a known intensity of
light, if a certain voltage is applied against a solar cell, can we know how much current will
flow in the circuit? This chapter ventures out to answer this question, i.e., it works toward
figuring out the current–voltage–light phase space of a solar cell. For that, we would first
go through the operations of a solar cell in dark. We would figure out the solar cell’s dark
current–voltage characteristics. Then we would illuminate the solar cell and would derive
the current–voltage characteristics under illumination. Details of the dark characteristics
can be found in Refs. [1–4]. The illuminated characteristics were discussed in Ref. [5].
From various types of solar cells that are getting researched at present, we would pick
the PIN type. The PIN solar cell has three layers of semiconductors—a P type semiconduc-
tor, an I type (intrinsic) semiconductor, and an N type semiconductor. The P layer is doped
with negatively charged acceptor ions and has more holes than electrons. Similarly, the N
layer is doped with positively charged donor ions and has more electrons than holes. The I
layer is intrinsic; it has an equal number of holes and electrons. The I layer can also instead
be lightly doped (P or N type) but the doping density should ideally be less than that of
the P and N layers.
DOI: 10.1201/9781003291916-8 119
120 Mathematical Modeling for Intelligent Systems

8.2 Working of the PIN Solar Cell

Under illumination, the I layer acts as an absorber of light. Light is absorbed and exci-
tons are produced inside the semiconductor. These excitons are then dissociated into free
electrons and holes. As the P layer is negatively charged (as a result of the PIN structure
formation, the P layer has more holes than I layer, in the dark, and so some holes diffuse
out of the P layer into the I layer) the positively charged holes in the I layer, under illumina-
tion, feel an attraction toward the P layer and move to it. Similarly, the negatively charged
electrons feel an attraction toward the positively charged N layer and move to it. These
movements of electrons and holes from the absorber to the P and N layers cause an electric
current inside the solar cell. If one electrically connects the P and the N layers with a wire,
one should be able to measure a current under illumination. This is the simplest under-
standing of a PIN solar cell. Figure 8.1 shows the above operations schematically. The
short-circuit current density flowing through a solar cell is symbolized as JSC. Similarly, if
the P and the N sides of the illuminated solar cell in Figure 8.1b is not connected (directly
or through a resistor), but kept open, an open-circuit voltage develops across the cell and
the cell behaves, in principle, like a battery. This voltage is symbolized as Voc.
For deriving the current–voltage characteristics, an understanding of the electron energy
band diagram is required. A semiconductor is characterized by its energy band diagram.
It is a potential energy landscape for electrons: if the band tilts or bends down electrons
find it easy to move downhill. In the other case if the band bends up electrons find it dif-
ficult to climb that energy barrier. In the diagram, the vertical axis is potential energy, and
the horizontal axis is space. A straight line in the energy band diagram, parallel to the
X-axis, refers to a potential energy level. For the entire semiconductor, as there are many
atoms and, therefore, many levels, energy levels merge to form a band. An electron, there-
fore, can be anywhere within a band. Besides this, a semiconductor is always characterized
by a bandgap. The bandgap is a forbidden zone for electrons, no electron can be inside the
gap unless, of course, there are trapping levels induced by defects. The bandgap has its
upper level as the conduction band edge EC. For an electron at any energy E>EC, i.e., above
EC, horizontal movement across the band (without any voltage or force from outside) is
decided by the slope of EC. The down level of the gap is the valence band edge EV. Thus,
electrons can be anywhere above EC or below EV. Also, in the valence band holes reside.
Once a bound electron is pumped up, from the valence band to the conduction band, to
become free, it leaves behind a free hole in the valence band. The Fermi level EF decides
the total number of electrons and holes present in a semiconductor. Closer the Fermi level
is to the conduction band edge, more electrons (than holes) the semiconductor has (i.e., it
is n type), and closer the EF is to the valence band, likewise, more holes than electrons the
semiconductor has (it is p type). The electron and the hole concentrations are expressed

FIGURE 8.1
(a) A PIN solar cell showing the three layers—the P, the I, and the N. (b) Under illumination the I layer absorbs
light and produces free electron and hole pairs. These opposite charge carriers travel in opposite directions
under the effect of the built-in electric field present in the PIN structure. This causes an electric current from the
P side to the N side, or an open-circuit voltage when the cell is left open.
PIN Solar Cell Characteristics 121

by equations (8.1) and (8.2). NC and NV are the effective conduction and the valence band
density of states, kB is the Boltzmann constant, and T is temperature.

 E − EC 
n =  N C  exp  F  (8.1)
 k BT 

 E − EF 
p =  NV   exp  V  (8.2)
 k BT 

The electron-hole concentration product, i.e., np product is always equal to the semiconduc-
tor’s intrinsic carrier density squared. The intrinsic carrier concentration ni is the (equal)
number of electrons and holes present in the undoped semiconductor due to thermal acti-
vation. ni is expressed in equation (8.3). EG is the bandgap.

 −E 
ni2 = np =   N C NV   exp  G  (8.3)
 k BT 

The PIN energy band diagram is drawn keeping in mind what would happen to the P, I,
and N layers due to each other’s presence. The P layer has more holes than the I layer. This
would make holes diffuse out from the P layer into the I layer when the P and the I layer
are brought in close proximity. Similarly, the N layer will lose electrons to the I layer. As a
result, the I layer’s carrier concentration profiles become nontrivial: toward the P side, the I
layer will have more holes, and toward the N side, it will have more electrons. This makes
the I layer tilt (with respect to the Fermi level) as shown in Figure 8.2a. The P layer and the
N layer will not get affected by this tilt. They will just become devoid of respective carriers
up to a width (xp –wp) and (wn –xn).
Electric field inside a semiconductor can be calculated using Poisson’s Equation: at any
point in space, the gradient of the electric field is given by the total charge density stored
at that point divided by the electric permittivity at that point. Applying Poisson’s Equation
to the PIN structure is done below. NA and ND are the doping densities of the P and the N
layers (i.e., the total hole and electron concentrations), εr is the dielectric constant, and ε 0 is
the permittivity of the vacuum.
dε −qN A
=  for w p < x < x p . (8.4a)
dx ε r ,P   ε 0

FIGURE 8.2
(a) The PIN energy band diagram. The P semiconductor is xp units wide. The I layer (absorber of the solar cell) is
(xn –xp) units wide. The N layer’s width is (d–xn). xp –wp is the width of the affected region in the P layer due to the
presence of the neighboring I layer, and similarly (wn –xn) is the width of the affected portion of the N layer. d is
the total width of the solar cell. (b) Electric field in the P, I, and N layers. The absolute value of the field is plotted,
i.e., the field ℇ is actually negative. The field is constant across the I layer and drops to zero in the regions of the
P and N layers that are unaffected by the I layer.
122 Mathematical Modeling for Intelligent Systems

dε p ( x ) − n ( x )

dx
= 
ε r ,I   ε 0
( )
for x p < x < xn − x p 2 (8.4b)

dε n( x) − p ( x)

dx
= −
ε r ,I   ε 0
( )
for xn − x p 2 < x < xn (8.4c)

dε qN D
=  for xn < x < wn (8.4d)
dx ε r ,N   ε 0

Solutions of the above Poisson Equations lead to the electric field distribution as shown in
Figure 8.2b. Figure 8.2b gives the absolute value of the solved electric field. The real elec-
tric field is negative which is easy to find from Figure 8.2a. Down the slope of the tilted I
layer, an electron travels from left to right which means the electric field under which the
electron is in is negative (the current is negative, and from Ohm’s Law, so is the field). This
electric field is sufficient to make the PIN structure gain solar cell functionalities. Electrons
and holes can follow this field and can travel toward the N and the P side when the I layer
is illuminated. This gives rise to a Voc. Thus Figure 8.2 shows the PIN band diagram in dark
and gives a rough idea of what can happen when the solar cell is illuminated.
There is a second variant of the PIN band diagram. In reality, having a perfectly intrinsic
semiconductor is not possible. So the I layer always remains very lightly doped. In addition,
if the I layer is also sufficiently wide, the free carriers inside the I layer can rearrange them-
selves in a way so as to cancel the electric field in the bulk. The electric field arises due to the
immobile charges in the P and N layers’ depletion regions, i.e., in the portions (xp –wp) and
(wn–xn). These regions are devoid of carriers and so are negatively and positively charged
due to the ionized acceptor and donor atoms. After the field is canceled at the I layer’s bound-
aries, it drops to zero in the I layer’s bulk. As a result, the bands in the bulk are flat and don’t
tilt or bend. The energy band diagram of this type of PIN structure is shown in Figure 8.3.
A natural question, therefore, arises—if the electric field in the I layer’s bulk is zero what
aids the solar cell function when the device is illuminated? The answer is—diffusion. After
the I layer is illuminated, it gets filled with free electrons and holes. Both of these entities
tend to diffuse out of the I layer because they are more in number inside the I layer. Due to
the band bending present on either sides of the I layer (Figure 8.3a), electrons travel to the
right, i.e., toward the N layer (down the band diagram) and holes travel to the P layer (up
the band diagram). This ensures solar cell functions when the device is illuminated. Just as
it is easy for electrons to move down an electron energy band diagram, it is easy for holes
to move up. That is how the band diagram is.

FIGURE 8.3
(a) The energy band diagram of the PIN structure where the I layer is intentionally/unintentionally doped and
is also wider. The electric field arising from the P and N layers is canceled at the I layer boundaries making the
field in the I layer interior zero. (b) The electric field in this type of PIN structure.
PIN Solar Cell Characteristics 123

8.3 Dark Current
After getting the fundamentals of the energy band diagram right, one should ponder how
will a solar cell’s current–voltage curve look in the dark. That is what we will discuss next.
Across all the three solar cell layers—P, I, and N—the current–voltage relation will be cal-
culated. In principle, the total current should be independent of space.
There are three kinds of currents involved in semiconductors—drift current, diffusion
current, and generation–recombination (RG) current.

J drift =  σε = cqµε (8.5a)

dc
J diff = qD (8.5b)
dx
x
J RG = 
∫ qMdx
0
(8.5c)

The drift current density is given by equation (8.5a). Here, by c we mean either the hole
concentration p or the electron concentration n. μ is mobility. By drift, we mean Ohm’s Law.
Likewise, the diffusion current density is given by Fick’s Law in equation (8.5b). D is diffu-
sion coefficient or diffusivity. When a semiconductor is illuminated, after an electron and
hole pair is created, a current can flow known as the generation (G) current. Similarly an
electron and hole pair can combine back and vanish (i.e., recombine) to give rise to another
current component called the recombination (R) current. The generation or recombination
current density should go so as to make 1/q (dJRG/dx) = M, where M stands for G or R. Thus,
the G or R current density is given by equation (8.5c).
While deriving currents inside semiconductors, one comes across the Continuity
Equations which states that a carrier just cannot appear or disappear at a given point of
space. In other words, if there are some electrons, getting created or destroyed per sec-
ond, for example, inside the P layer (as minority carriers), then those electrons must have
arrived at that point of space following the physical processes of either drift, or diffusion,
or recombination, or generation. The Continuity Equations are given in equations (8.6a)
and (8.6b).

 dp   dp   dp   dp   dp 
=  +  +  +  (8.6a)
 dt  total  dt  drift  dt  diff  dt  G  dt  R

 dn   dn   dn   dn   dn 
  =    +     +    +     (8.6b)
dt total  dt  drift  dt  diff  dt  G  dt  R

Further, because there are only two types of currents that are physically possible—the cur-
rents due to drift and diffusion—one can write

 dp   dp  1 dJ p
+  =  (8.7a)
 dt  drift  dt  diff q dx

 dn   dn  1 dJ n
  +    =   (8.7b)
dt drift  dt  diff q dx
124 Mathematical Modeling for Intelligent Systems

In our derivation, we would assume that there is no recombination present in the I

layer. As, in addition, it is dark, there is no generation present also. This would mean,
in the I layer, G = R = 0, and, therefore, JRG = 0 (from equation (8.5c)). Thus, by applying the
Continuity Equation in the I layer, one sees that the carriers arrive by drift and leave by
diffusion. Therefore,

 dp   dp 
+  =0 (8.8a)
 dt  drift  dt  diff

 dn   dn 
  +    = 0 (8.8b)
dt drift  dt  diff

From equations (8.7) and (8.8) one sees that, in the I layer, both the electron and the hole
currents, individually, are constant spatially. We call these values Jsn and Jsp.
Interestingly, just as drift balances diffusion in the I layer, for the P and N layers, in the
quasi-neutral regions (i.e., for x < wp, and for x > wn), it’s recombination that balances diffu-
sion. In other words, diffusion brings the carriers and recombination subtracts them. There
is no generation or drift present in the quasi-neutral regions. Therefore, the Continuity
Equations in the P and N layers go as

 dp   dp 
  +    = 0 (8.8c)
dt diff dt R

 dn   dn 
  +    = 0 (8.8d)
dt diff dt R

The recombination here follows the monomolecular Shockley Read Hall (SRH) model. In
other words, the Continuity Equation terms for recombination becomes

 dp  p − p0 ∆p
  = =  (8.9a)
dt R τp τp

 dn  n − n0 ∆n
  = =  (8.9b)
dt R τn τn

where p0 and n0 are the carrier concentrations that are present in the quasi-neutral regions at
zero applied voltage V, i.e., before the applied bias brings in more minority carriers in the P
and the N layers. τ is the minority carrier lifetime for holes (τp) or electrons (τn) accordingly.
This means, in the dark and under zero applied voltage, the hole concentration in the N layer
for x > wn is p0. Likewise, the electron concentration in the P layer for x < wp is n0. It is easy to see
that p0 = ni2/ND, and n0 = ni2/NA. As voltage is applied to the PIN structure, with the positive
terminal at P and the negative terminal at N, the minority carrier concentrations (i.e., the con-
centration of holes in the N layer and that of the electrons in the P layer) increase. The recom-
bination is proportional to this increase and inversely proportional to the carrier lifetime τ.
At an applied voltage V the minority carrier concentration, say, increases to p1 and n1 in
the quasi-neutral regions’ edges, i.e., at wn and wp.

ni2  qV 
p1 =    exp  (8.10a)
ND  kBT 
PIN Solar Cell Characteristics 125

ni2  qV 
n1 =    exp  (8.10b)
NA  kBT 

From the fact that ℇ = 0 in the quasi-neutral regions, and from equations (8.5b), (8.7), (8.8c),
(8.8d), and (9), we have

d 2 ( ∆p ) ( ∆p )
Dp =  for x ≥ wn (8.11)
dx 2
τp

Redefining the spatial variable to x′ = (x–wn) we have

d 2 ( ∆p ) ( ∆p )
Dp =  for x′ ≥ 0 (8.12)
dx '2
τp

with boundary conditions

∆p =   p1 −   p0 at x ′ = 0 (8.13a)

∆p = 0 at x ′ → ∞ (8.13b)

Solving equation (8.12) with the above boundary conditions one arrives at the solution

 x′ 
∆p  ( x ′ ) =  ( p1 − p0 )  exp  −  (8.14)
 Lp 

where L is called the minority carrier diffusion length. Simplifying the expression one
arrives at –

ni2   qV    x′ 
∆p  ( x ′ )  =     exp   − 1  exp  −  (8.15)
ND  k BT    Lp 

i.e.,

ni2   qV    x′  n2
p ( x ′ )  =      exp   −   1 exp  −  +  i (8.16)
ND  k BT    Lp  N D

The electron minority carrier concentration will have a similar expression.

After deriving the carrier concentrations, deriving the carrier current densities become
trivial. From equations (8.5b) and (8.16) one has

qni2 Dp   qV    x′ 
J p ( x ′ ) =     exp   − 1 exp  −  (8.17)
N D Lp  k BT    Lp 
126 Mathematical Modeling for Intelligent Systems

One must balance for the minus sign as the hole minority carriers travel from left to right,
and, therefore, the current must be positive, and, on the other hand, the slope of p(x’), for
x′ > 0, is negative.
The electron current will have a similar expression.
From equation (8.17), at x′ = 0, one can get the expression of the hole current density, i.e.,

qni2 Dp   qV  
J sp =      exp   − 1 (8.18)
N D Lp  k B T 

Similarly, for the electron current density

qni2 Dn   qV  
J sn =     exp  − 1 (8.19)
N A Ln  kBT  

Keeping in mind that in the I layer the current densities do not change spatially, for the
PIN solar cell in dark, the current density–voltage characteristics go as

J =   J sp +   J sn (8.20a)

  qV  
J =  JS      exp   −   1 (8.20b)
  k BT  

 Dp D 
JS =   qni2     +  n  (8.20c)
 N D Lp N A Ln 

Strictly speaking equation (8.20b) is valid for a PN junction diode, but it can also hold for
a PIN structure when recombination in the I layer is considered negligible. JS is called the
saturation current density. When recombination in the I layer is significant an ideality fac-
tor nID enters in equation (8.20b), but that we do not discuss here. Equation (8.20b) is also
called the Ideal Diode Equation or the Shockley Equation. Figures 8.4 and 8.5 show the
dark carrier concentrations and the dark current densities of the PIN solar cell. Figure 8.6,
finally, shows the current–voltage characteristics that is given by equation (8.20b).

FIGURE 8.4
(a) Hole concentration in the solar cell under dark and at an applied voltage V. The dashed profile is at zero
applied voltage. At the N layer edge (more precisely, at the depletion layer edge of the N layer) the hole concen-
tration at V = 0 is p0 and becomes p1 at applied voltage V. (b) The hole current density for the PIN solar cell under
dark at applied bias V.
PIN Solar Cell Characteristics 127

FIGURE 8.5
(a) Likewise, electron concentration in the solar cell under dark and at an applied voltage V. The dashed profile
signifies the pattern at zero applied voltage. At the depletion layer edge of the P layer the electron concentration
at V = 0 is n0 and becomes n1 at applied bias V. (b) The electron current density for the PIN solar cell under dark
and at applied voltage V.

FIGURE 8.6
The dark current–voltage characteristics of the solar cell showed in normal (a) and in natural logarithm (b) scale.

8.4 Photocurrent
After deriving the dark current for our PIN solar cell, the next aim will be to derive the
current bias dependence under illumination. For this, we would follow a method similar
to what we did for the dark current. We will start with the Continuity Equation, equate the
generation current to the current caused by drift and diffusion, and will derive the spatial
profiles of carrier concentrations and current densities. The total current density will be
independent of space as before. An important point on deriving the photocurrent is that all
calculations are done for the I layer, i.e., the absorber. Just as the P and the N layers formed
the center of attraction while deriving the dark current, this time, the absorber will take
the front seat for getting into photocurrent. In what follows the spatial variable x″ is con-
sidered to be 0 at xp, and to be di at xn.
Following equation (8.5c), the generation term can be written, for both electrons and holes, as

1 dJ p
= G (8.21a)
q dx′′

1 dJ n
=   −G (8.21b)
q dx′′

The electron and hole drift and diffusion current densities can be expressed as

J p =   J p ,drift +   J p ,diff   (8.22a)

128 Mathematical Modeling for Intelligent Systems

qµ pV ′
J p ,drift =   p. (8.22b)
di

dp
J p ,diff =   − qDp (8.22c)
dx′′

J n =   J n,drift +   J n,diff (8.23a)

qµnV ′
J n,drift =  n (8.23b)
di

dn
J n,diff =  qDn (8.23c)
dx′′

Here, the width of the I layer is considered to be di, i.e., (xn –xp) = di. The total bias is consid-
ered as V′. It must be borne in mind that at zero applied bias the photogenerated charge
carriers will be under the built-in electric field ℇ, and, therefore, under the built-in voltage
Vbi, where Vbi = ℇ/di. Any externally applied voltage V will add to this built-in voltage and
therefore the total bias V′ = V–Vbi. Meanwhile, solving equation (8.21) we get

J p = qGx′′ + Ap (8.24a)

J n = − qGx′′ + An (8.24b)

By using equations (8.22) and (8.24a) we can solve for the hole spatial profile p(x″)

  µ pV′  
Bp   exp  x ′′  +   q 2GDp   
qGx ′′ +  Ap   Dp di  
p =  +  (8.25a)
qµ pV ′  qµ pV ′ 
2

di  di 

Similarly, by using equations (8.23) and (8.24b), we can solve for the electron spatial profile
n(x″)

  µnV′  
Bn   exp  − x ′′  −   q 2GDn   
−qGx ′′ +  An   Dn di  
n =  +  (8.25b)
qµ pV ′
( )
2
qµnV ′
di di

The two unknowns A and B in each of equation (8.25) can be solved by using boundary
conditions. Under high-level injection, i.e., when the concentration of photogenerated car-
riers is higher than the doping densities, it is easy to see that p = n = 0 at x″ = 0 (i.e., at x = xp)
and at x″ = di (i.e., at x = xn). In other words, the electron and hole densities in these points
will be negligibly small compared to that in the I layer. Under low-level injection, i.e., when
the concentration of photogenerated carriers are higher than the minority carrier concen-
trations in the P and the N layers but lower than the doping densities, these boundary
PIN Solar Cell Characteristics 129

conditions change. This makes the p and n spatial dependencies different, but in either
cases the current density graphs and expressions stay approximately the same, and the
final bias-dependent photocurrent expression is that which is obtained using the p = n = 0
boundary conditions at the boundaries of the I layer. The solutions are given as

   qV′ x′′  
  exp   −  1  
Gdi2  x′′   kBT di   
p =  −   (8.26a)
µ pV′  di   qV′   
 exp   − 1 
   k BT   

   qV′ x′′  
 exp  −  −  1 
2
Gd  x′′ 
i  k B T di  
n = − −  (8.26b)
µnV′  di   qV′   
 exp  −  − 1 
   k BT   

Following equations (8.22a) and (8.23a) the photocurrent densities can be deduced

 
 
 x′′ kBT 1 
J p =  J L −  +  (8.27a)
 di qV′   qV ′  
 exp   − 1 
   k BT   

 
 
 x′′ kBT 1 
Jn =   −   JL +  +  (8.27b)
 di qV′   −qV′  
 exp   − 1 
   k BT   

J L = qGdi (8.27c)

  qV′   2kBT  
J =   J p +   J n =  J L  coth  −   (8.27d)
  2kBT   qV′  

Thus the final bias-dependent photocurrent expression is given by equations (8.27c) and
(8.27d).
The photocurrent densities in the P or N layers should be independent of space because
both the generation and recombination there are zero. The generation is zero because
in our derivation we have considered illumination only in the I layer, i.e., the absorber.
The recombination is also zero because for majority carriers in the P and N quasi-neutral
regions there is no recombination present.
Figure 8.7 shows the carrier concentration profiles for V′ close to zero (but negative). It
assumes high-level injection. Figures 8.8 and 8.9 show the carrier profiles under low-level
injection, and also the current density profiles. The photocurrents are always negative,
however, in the diagrams only the magnitudes are shown.
130 Mathematical Modeling for Intelligent Systems

FIGURE 8.7
The solved spatial profiles of holes (a) and electrons (b) in the I layer under the assumption of high-level injection.

FIGURE 8.8
(a) The hole spatial profile under low-level injection. (b) The hole current density profile. JLP = JLn = JL. The mag-
nitude of the hole current is drawn.

FIGURE 8.9
(a) The electron spatial profile under low-level injection. (b) The electron current density profile. JLP = JLn = JL. The
magnitude of the electron current is drawn.

FIGURE 8.10
(a) The bias-dependent photocurrent of the solar cell. (b) The total current–voltage characteristics of the solar cell.
PIN Solar Cell Characteristics 131

In summary, the bias-dependent photocurrent is shown in Figure 8.10a. It is also often

referred to as the S curve. When this photocurrent adds to the solar cell’s dark current
(shown in Figure 8.6), the total solar cell current looks like what is shown in Figure 8.10b.
This is the final current–voltage curve for an illuminated solar cell. The current at zero
voltage is called the short-circuit current (Jsc) and the voltage at zero current is called the
open-circuit voltage (Voc).

8.5 Conclusion
Thus, in summary, in this chapter, we have derived the PIN current–voltage character-
istics in dark and under illumination. We have started with the energy band diagram,
have introduced the Continuity Equations, and from the Drift Diffusion Equations, we
arrived at the fact that for the P and N layers the diffusion will be equal to the recombina-
tion and thus the exponential current–voltage curve and the Ideal Diode Equation were
derived. Next we assumed high-level injection in the absorber and with those boundary
conditions, we equated drift and diffusion to generation in the I layer and arrived at the
bias-dependent photocurrent S curve. When both of these characteristics are added, the
total current–voltage characteristics of the solar cell are obtained under illumination. With
more illumination G increases, and, from Equation 27(c), JL increases, and thus the current–
voltage–light phase space of the solar cell can be chalked out.

References
[1] R. F. Pierret, Semiconductor Device Fundamentals, Addison-Wesley Publishing Company,
New York (1996).
[2] D. A. Kleinman, “The forward characteristic of the PIN diode”, Bell Syst. Tech. J. 35 (1956)
685–706.
[3] A. Herlet, “The forward characteristic of Silicon power rectifiers at high current densities”,
Solid State Electron. 11 (1968) 717–742.
[4] F. Berz, “A simplified theory of the PIN diode”, Solid State Electron. 20 (1977) 709–714.
[5] R. Sokel and R. C. Hughes, “Numerical analysis of transient photoconductivity in insulators”,
J. Appl. Phys. 53 (1982) 7414.
9
A New Approximation for Conformable Time-
Fractional Nonlinear Delayed Differential
Equations via Two Efficient Methods

Brajesh Kumar Singh and Saloni Agrawal

Babasaheb Bhimrao Ambedkar University

CONTENTS
9.1 Introduction ........................................................................................................................ 133
9.2 Basic Literature ................................................................................................................... 134
9.3 Description of IDTM....................................................................................................... 137
9.4 Description of O HAITM ................................................................................................. 138
9.4.1 Error Estimation of O HAITM ............................................................................ 140
9.5 Test Examples ..................................................................................................................... 141
9.6 Conclusion .......................................................................................................................... 150
9.7 Result and Discussion ....................................................................................................... 150
Acknowledgments ...................................................................................................................... 157
References..................................................................................................................................... 157

9.1 Introduction
Delayed differential equations are concerned to be very beneficial in biological and physi-
cal sciences, in which the evaluation of the structure confides on both present and former
positions. In the literature, vigorous physical/biological models arise in terms of frac-
tional-order systems that could not be handled by any integer order system, for instance,
Robertson diffusion model, signal processing, electrical network system, space–time delay
models, and so forth [1–4]. The investigation of Burgers’ equation assumes an imperative
role in understanding distinctive numerical and analytical methods. These models give a
premise to the hypothesis of shock waves and turbulence flow, refer [5–7] for more details.
Integro-delay differential equations are generally named pantograph-type delay differ-
ential equations or generalized delay equations, which are frequently utilized to demon-
strate certain issues with delayed consequence in mechanics and the related logical fields.
Numerous models, for example, stress–strain conditions of materials, models of polymer
crystallization, movement of rigid bodies are described in Kolmanovskii and Myshkis’s
monograph [8] and the references therein. In addition, we have considered the time-
fractional pantograph differential equation (pantograph: z-formed mechanical connection
appended to the top of an electric train/transport to acquire power flexibly by an over-
head electric wire) that has a wide range of application in several systems of mechanical/
electro-dynamic framework [9].
DOI: 10.1201/9781003291916-9 133
134 Mathematical Modeling for Intelligent Systems

As we know, it is quite challenging to calculate exact results of fractional differential

equations. In order to overcome this problem, an easy and demanding definition of new
type of fractional-order derivative “conformable fractional derivative” was first suggested
by Khalil et al. [10]. Latterly, the concept of conformable fractional calculus has been applied
to compute the approximate solutions of linear and nonlinear partial differential equa-
tions (PDEs). Lio [11,12] developed a homotopy analysis method to determine the solution
of various nonlinear classical order (PDEs). Thereafter, this technique has been extended
to analyze many fractional-order PDEs, see [13–18]. Moreover, q-homotopy analysis
method ( qHAM ) was used to estimate the solution of conformable space–time–fractional
Robertson equation with 1-D diffusion [13], generalized Hirota–Satsuma coupled KdV sys-
tem [17], ERCDTM was implemented to figure out the solution of the nonlinear PDEs with
delay argument [19] and so forth.
To the best of our knowledge, no effort has been made in the computation of approxi-
mate solution of conformable time-fractional PDEs (CTFPDEs) with proportional delay
using IDTM and O HAITM. In this chapter, we have implemented two reliable tech-
niques for the analytical study of following the autonomous initial valued structure of
nonlinear CTFPDEs with proportional delay.

N [t T α φ ( x , t)] − g( x, t) = 0 φ ( x , 0) = f ( x), 0 ≤ α < 1, xi , t ∈[0,1] (9.1)

where

( ( m)
N [t T α φ ( x , t)] = t T α φ ( x , t) − F x , φ ( a0 x , b0 t), φ x ( a1 x , b1t),…, φ x ( am x , bm t) ) (9.2)

 be symbolize the nonlinear operator, t  α symbolize the conformable derivative, g( x ,t)

be a known function, f be a smooth function, C be a constant and φ ( x , t) is the unknown
function to be computed.

9.2 Basic Literature
There are numerous definitions of fractional derivatives proposed in history. Researchers
have found some weaknesses in various types of definitions by which few models do not
approach toward reality. For instance, Caputo fractional derivative [2] is noncommutative
[20]. Keeping this in mind, a new sort of meaning of fractional derivative, i.e., Conformable
Fractional Derivative as mentioned above is defined as f.

Definition 1 [Conformable Time-fractional derivative (CTFD) [10]]

The CTFD of a function φ ( x , t) :  × (0, ∞) →  is described as

φ ( x , t + ht 1−α ) − φ (x , t)
t  α φ ( x , t) = lim , α ∈ (0,1].
h→0 h
where α is the fractional order of derivative.
Nonlinear Delayed Differential Equations 135

Further, the CTFD (left) of fractional order α of a function φ ( x , t) :  × ( a, ∞) →  , a ≥ 0 is

characterized as below:
φ ( x , t + h(t − a)1−α ) − φ ( x , t)
t  α φ ( x , t) = lim , α ∈ (0,1].
h→0 h
The fundamental conditions of CTFD are given as:

Theorem 9.1 [20]

Suppose α ∈(0,1],C ,  D are arbitrary constants and φ ( x , t) and ψ ( x , t) are fractional-order

differentiable functions at the given point (x , t) ∈  × ( a, ∞), a ≥ 0. Then

a. t  α (C  φ + D ψ ) = C    t  α φ + D   t  αψ ,
b. t  α (C) = 0,
c. t  α (φ  ψ ) = φ    t  αψ + ψ    t  α φ ,
ψ    t  α φ − ψ t  α  φ
d. t  α (φ /ψ ) = ,
ψ2
e. For every r ∈ R , i)   T α (t r ) = r   t r −α ,    ii)   t T aα (t − a)r = r(t − a)r −α .
Appropriately, t  aα (t − a)α = α .
f. If φ , ψ are differentiable functions w.r.t. t, then

∂φ ∂ψ
i)   t  aα φ = t 1−α ,    ii)   t  aαψ = (t − a)(1−α ) ,
∂t ∂t
 λ  (t− a )α +x   ( t − a )α 
λ +x 
g. t a
α e  α 
= λe  α 
.
 
 
∂ mψ  ∂ mψ 
h. Take m < α ≤ m + 1, β = α − m, if exists, t  aαψ = t  aβ  m  .
∂t m
 ∂t 
i. Consider α ∈(0,1] and φ ( x , t) be k − times differentiable function at (x , t0 ) ∈  × (0, ∞)
∂i φ
s.t. |( x ,t0 ) = 0, ∀i ∈ {1, 2… k − 1}, therefore, the k − times CTFD φ ( x , t) is
∂t i
 ∂k φ 
t  t0kα φ = (t − t0 )k −kα  k  , where t  t0kα ≡ t  t0kα …t  t0kα .
 ∂t  t=t0   
k-times

The conformable integration of fractional order α as defined in Ref. [21] is given below:

Definition 9.2 [21]

Consider φ :  →  is a real-valued function. Therefore α th order (0 < α ≤ 1) “conformable

integration” of φ is characterized as follows

φ ( x)
t
(tα φ )(t) =
∫
0 x
1−α
dx , t > 0 (9.3)

Moreover if α ∈ (n, n + 1], β = α − n,

136 Mathematical Modeling for Intelligent Systems

if the nth derivative φ n (t) of φ :  →  exists, therefore the conformable integration of φ

for order α is
φ n ( x)
t
(tα φ )(t) = ( tβ φ n )(t) =
∫ 0 x
1− β
dx. (9.4)

Definition 9.3 [New integral transform, [22–24]]

The NIT of φ (τ ) ∈ is defined by

1 ∞
 −tα 
Φ( x , µ ) = α {φ ( x , t)} :=
µ ∫ 0
exp  2  φ ( x , t) tα −1 dt,
 αµ 
(9.5)

Here  be the class of real-valued exponential order functions:

 |t| 
 = φ ( x , t) : ∃r1 , r2 > 0, 0 < K < ∞ ;|φ ( x ,t)|≤ K exp  2  ,   if   t ∈ (−1) × [0, ∞)  ,
  r  
where for given φ ∈ , the constant K must be finite although r1 , r2 arbitrary.
Some properties of (NIT) required to complete understanding of the present work are
reported below:

Theorem 9.2 [Properties of NIT, [22–24]]

∂κ g( x ,t)
a. The NIT of ∈ is defined by
∂tκ
k−1
 ∂ k g( x ,t)  G( x , µ ) ∂  g( x , 0 + )
Kα 
 ∂t
k =
 µ2k
− ∑µ =0
1
2 ( k−  )− 1
∂t 
, k ≥ 1.

b. The NIT of CTFD t  α g( x ,t) read as follows:

i. Kα {T t
−α
}
g( x ,t) = µ 2α G ( x , µ ),
k −1
G( x , µ ) ∂l g( x , 0)
ii. Kα { α
t T g( x , t) = }µ 2α
− ∑µ
=0
1
2 (α −  )− 1
∂t l
, k − 1 < α ≤ k ∈,

Some conformable NITs of certain functions are listed below:

i. α {1} = µ
λ 2λ
λ
{ }
+1
ii. α t λ = α α µ α Γ(1 + ), κ = 0,1, 2,…, n
α
  tα   µ
iii. α exp    =
  α   1 − µ2
  tα   µ3
iv. α sin    = ;
  α   1 + µ
4

  tα   µ
v. α cos    =
  α   1 + µ4
Nonlinear Delayed Differential Equations 137

9.3 Description of IDTM
This section represents the IDTM [25] for the system of CTFPDEs. Consider a nonlinear
PDEs with conformable derivative of the type equation (9.1) whose standard form is given as

t T α (φ ( x , t)) + L[φ ( x , t)] + N [φ ( x , t)] = h( x, t);  φ ( x , 0) = g( x) (9.6)

where the  and  are linear and nonlinear operators, respectively; h( x ,t) is the known
source term, t  α be the conformable fractional operator of order α .
Applying conformable NIT to both sides of equation (6)

Kα {Tt
α
}
(ϕ ( x , t)) + Kα {L[ϕ ( x , t)] + N [ϕ ( x , t)] − h( x, t)} = 0. (9.7)

kα
utilizing the property of conformable NIT operator for t  (φ ( x , t)) with initial condition,
we obtain

Kα {φ ( x , t)}
k−1
∂ φ ( x , 0)
µ 2α
− ∑µ
=0
1
2 (α −  )− 1
  
∂t 
= −Kα {L[φ ( x , t)] + N [φ ( x , t)] − h( x ,t)} . (9.8)

Now implementing inverse NIT on both sides of equation (9.8), we get

φ ( x , t) = G ( x , t) − Kα−1  µ 2α Kα {L[φ (x , t)] + N [φ( x , t)] − h( x ,t)}  (9.9)

where ( x , t) represents the source term with initial condition. Formulate linear and non-
linear terms as
∞ ∞

L[φ ( x , t)] = ∑k=0

L[φk ( x , t)]; N [φ ( x , t)] = ∑φ (x, t).
k= 0
k (9.10)

where
   ∞

(
Pk φk ( x , t) =
1 ∂k
)
k ! ∂ℵk
N 
  ∑
k=0
ℵkφ ( x , t,ℵ)  .
 ℵ=0

Substituting equation (9.10) in equation (9.9), we obtain

  
    ∞ 
∞ ∞

∑ φk ( x , t) = G (x , t) − Kα−1  µ 2α Kα
 ∑
L  φk ( x , t)  + N ∑
 φk ( x , t) − h( x ,t)  
 
(9.11)
k =0    k = 0   k= 0 

Comparing the coefficient, we calculate the following terms:

φ0 ( x , t) = G ( x , t) + Kα−1  µ 2α Kα h( x ,t)  ;φ1 ( x , t) = − Kα−1  µ 2α Kα {L [φ0 ( x , t)] + N [φ0 ( x , t)]} 

In general, for k ≥ 1:

φk +1 ( x , t) = −Kα−1  µ 2α Kα {L [φk ( x , t)] + N [φk ( x , t)]}  (9.12)

138 Mathematical Modeling for Intelligent Systems

9.4 Description of O HAITM
This section elaborates the O HAITM [13] for the system of CTFPDEs. Consider nonlin-
ear PDEs with conformable derivative of the type equation (9.1) whose standard form is
given as

t T α (φ ( x , t)) + L[φ ( x , t)] + N [φ ( x , t)] − h( x, t) = 0 (9.13)

where the  and  are linear and nonlinear operators, respectively; h( x ,t) is the known
source term, and t  kα be the conformable fractional operator of order kα , k be a positive
integer.
Applying conformable NIT on both sides of equation (9.13)

Kα {T t
α
}
(ϕ ( x , t)) + Kα {L[ϕ ( x , t)] + N [ϕ ( x , t)] − h( x, t)} = 0. (9.14)

kα
and utilizing the property of conformable NIT operator for t  (φ ( x , t)) :

k−1
∂ φ ( x , 0)
Kα { T {φ(x, t)}} = µ1
t
α
2α
Φα ( x , µ ) − ∑µ
=0
1
2 (α −  )− 1
  
∂t 
.

we get,

k −1
∂ φ ( x , 0)
Φα ( x , µ ) − ∑µ
=0
1
−2 − 1
  
∂t 
+ µ 2α Kα {L[φ (x , t)] + N [φ ( x , t)] − h( x ,t)} = 0.

Formulate nonlinear operator as follows:

[ϕ ( x , t ;ℵ)]
k−1

= Kα {ϕ ( x , t;ℵ)} − ∑µ
 =0
1
2 − 1
  
∂ φ ( x , 0)
∂t 
+ µ 2α Kα {L[ϕ (x , t ;ℵ)] + N [ϕ (x , t ;ℵ)] − h(x , t)} .

where ϕ ( x , t;ℵ) is real-valued function of ℵ, x , t; ℵ ∈[0,1] is the standard embedded param-

eter, α stands for conformable NIT operator.
The deformation expression of zeroth-order as in [11] is constructed as

(1 −ℵ)t T α [ϕ ( x , t;ℵ) − φ0 ( x , t)] = ℵH ( x, t) [ϕ ( x , t ;ℵ)], (9.15)

where  ≠ 0, H ( x ,t) be the auxiliary parameter & auxiliary function, respectively; φ0 ( x , t) be

the initial value of the function φ ( x , t). It is worth mentioning that O HAITM suggests a
huge opportunity for selection of the auxiliary things in proceedings. It is easy to observe
that ϕ ( x , t;0) = φ ( x , 0) and ϕ ( x , t ;1) = φ ( x , t) truly hold, respectively, for ℵ = 0,1.
This signifies that as ℵ grows from 0 to 1, the solution ϕ ( x , t;ℵ) grows simultaneously
from the initial guess: φ0 ( x , t) to the exact solution: φ ( x , t).
Nonlinear Delayed Differential Equations 139

The function ϕ ( x , t;ℵ) can be expanded utilizing Taylor’s formula in the form of ℵ as
follows:

ϕ ( x , t;ℵ) = φ0 ( x , t) + ∑φ (x, t)ℵ ,

k=1
k
k
(9.16)

where the parameter  regulates the convergence area of the solution of equation (9.16). The
convergence of the solution equation (9.16) at ℵ = 1 can be established by suitable selection
of the auxiliary parameters: , H ( x, t) ≠ 0 and the initial guess. Thus,

φ ( x , t) = φ0 ( x , t) + ∑φ (x, t).
k=1
k (9.17)

1 ∂k ϕ
where φk ( x , t) = .
k ! ∂ℵk ℵ=0

φk ( x , t) = (φ0 ( x , t), φ1 ( x , t), φ2 ( x , t), …, φk ( x , t)) .

Subsequently, the deformation equation of order k is evaluated through differentiating

k-times to the deformation equation of zeroth-order with respect to the embedded param-
eter ℵatℵ = 0, thereafter dividing the resulting expression by k !.
In this consequence, the deformation equation of order k is obtained as follows:


Kα [φk ( x , t) − χ kφk −1 ( x , t)] = ℵH ( x ,t)Pk (φk −1 ( x ,t)), (9.18)

where χ k = 0 if k ≤ 1 and 1 otherwise. On employing inverse conformable Laplace trans-

form to the above equation (9.19) and setting ℵ = 1, H ( x ,t) = 1, we get

φk ( x , t) = χ kφk −1 ( x , t) + Kα−1  Pk (φk−1 ( x , t))  (9.19)

where

 1 ∂ k−1 T [ϕ ( x , t ,ℵ)]
Pk ( ϕ k −1 ( x , t)) = .
( k − 1)! ∂ℵk−1 ℵ=0

After computing φk ( x , t) for k ≥ 1, Mth-order approximate series solution of equation (9.13)

is calculated as follows:

SM ( x ,t) = ∑φ (x, t),

k=0
k (9.20)

which converges to the exact solution φ ( x , t) of the equation (9.14) accurately as M → ∞.

140 Mathematical Modeling for Intelligent Systems

Theorem 9.3

As far as the series approximation equation (9.17) converges, where φk ( x , t) are computed
from equation (9.19), then the series approximation equation (9.17) must be the exact solu-
tion to the conformable differential equation (9.13).
Let the series approximation of equation (9.13) converges. If we set
∞

Ξ( x , t) = ∑φ (x, t),
k=0
k

then following condition (*) holds true.

(*)lim φκ ( x , t) = 0.
κ →∞

Since, (**)φκ ( x , t) = ∑[φ (x, t) − χ φ

λ =1
k k k−1 ( x , t)].

Utilize the condition (*) and (**) in equation (9.20) with property ℵ ≠ 0, we get

κ ∞
 
lim
κ →∞ ∑
k=1
k (φk −1 ( x , t)) = ∑  (φ
k=1
k k−1 ( x , t)) = 0.

Therefore,
∞

∑  (ϕ

k k−1 ( x , t))
k=1

∞   k −1
∂ φ ( x , 0)  
= ∑k =1
 Kα [φk −1 ( x , t)] − (1 − χ k ) 
 
∑µ
= 0
2 + 1
  
∂t 
+ µ 2α h( x , t) + µ 2α Kα {L[φk −1 ( x , t)] + N [φk −1 ( x , t)]} 
 

= µ 2α Kα  t T α (φ (x , t)) + L[φ ( x , t)] + N [φ (x , t)] − h( x, t)  = 0,

⇒t T α (φ ( x , t)) + L[φ (σ , t)] + N [φ ( x , t)] = h(x , t)

which confirms that the series solution φ ( x , t) in equation (9.18) is the exact solution of the
differential equation equation (9.14).

9.4.1 Error Estimation of O HAITM

The efficiency and accurateness of the present method can be calculated using root mean
square error norm L 2 and maximum error norm (L∞ )
For the problems considered in the previous section. The square residual error [14] in the
kth order approximate solution is

2
  k

∑
1 1
∆ k () =
∫∫ 0

0 
  i=0
φi ( x , t)  dxdt.

(9.21)
Nonlinear Delayed Differential Equations 141

Here S (φ ) = N [t T α φ ( x , t)], is the residual error. The controlling parameter  appeared in

equation (9.17) plays a significant character in governing the convergence area and rate of
the convergence for the proposed results. As mentioned in equation (9.16) of Section 9.4,
q-HAM yields a family of solutions in view of , and so, ∆ k () treated as the function of .
More rapidly ∆ k () decreases to zero, the corresponding q-HAM solution converges faster
to the exact solution (i.e., accelerate the convergence of the solution). More precisely, the
acceptable interval of  is the line segment in -curve approximately parallel to the hori-
zontal axis. Thus, -curve helps us in an easy form to control & fix the convergence area of
the approximate solution equation (9.16) of CTFPDEs with proportional delay. The optimal
value of  is defined by the values of  within the -region for which ∆ k () is minimum,
d∆ k ()
i.e., = 0.
d
As mentioned in [11,14], it is seen that exact square residual error ∆ k () as defined in
equation (9.21) takes too much CPU-time in their computation even for small order of
approximation, it cause the computation of ∆ k () through equation (9.21) is often not wor-
thy to use. To overcome such type of disadvantage, the authors of Ref. [11] introduced a
more efficient formula to compute the residual error, which is known as averaged residual
error, and is given in the following
2
k1 k2   k

∆ k () =
1
k1   k 2 ∑∑ j=0 l=0
N
 
∑
i=0
φi ( jdx , ldt)  .

(9.22)

1 1
where δ x = and δ t = , k1 = k2 = 10.
k1 k2
The bound on the maximum absolute truncated error of the O HAITM, as proved in
[14], is reported in the following

Theorem 9.4 [14]

k

Let 0 < ρ < 1. If φi+1 ( x , t) ≤ ρ φi ( x , t) for each i. In addition, if the truncated series
represents an approximation for the solution of φ ( x , t).
∑φ (x, t)
i=0
i

Then the maximum truncated absolute error can obtain by the following
k

∑φ (x, t) ≤ 1ρ− ρ φ (x, t) .

k+1
φ ( x , t) − i 0 (9.23)
i=0

9.5 Test Examples
Example 9.1
Firstly, assume nonlinear nonhomogeneous time-fractional pantograph differential
equation [26]

 t
t  α φ (t) + 2φ 2   = 1 φ (0) = 0, t ∈[0, 1], 0 < α ≤ 1. (9.24)
 2

In particular case when α = 1 the exact solution of this equation is φ (t) = sin t.
142 Mathematical Modeling for Intelligent Systems

Case 1. IDTM: On implementing IDTM as described in equation (9.12) on the system

of equation (9.24), we get the following recursive equation

φ0 (t) = α−1  µ 2α α {1} 

∞   k 
∑ φk +1 (t) = −α−1  µ 2α α


2

∑φ  2t  φ
j k− j
 t 
    .
2 
(9.25)
k =0  j=0 

By solving the recurrence relation equation (9.25), the first three iterations can be com-
puted by using Mathematica software as

tα 2 1− 2α t 3α 81− 2α t 5α
φ0 (t) = ; φ1 (t) = − ; φ2 (t) = ;
α 3α 3
15α 5
2 3 − 10α t 7α 2 5 − 12α t 7α
φ3 (t) = − − . (9.26)
63α 7 105α 7

In this consequence, the kth iterative solutions φk (t) for k ≥ 4 can be evaluated from equa-
tion (9.26). Therefore, sixth-order approximate solution for α = 1 is obtained as

∑φ (t) = t − t3! + t5! − 7!t + t9! − 11!

3 5 7 9 11 13
t t
S6 (t) = k + (9.27)
k =0
13!

which is the nearest form of exact result φ (t) = sin t at α = 1.

Case 2. O HAITM : On implementing α operator on the system of equation (9.24) and

using the properties of NIT, we get

  t 
α {φ (t)} − µφ (0) + µ 2α α 2φ 2   − 1 = 0
  2  

Construct nonlinear operator as

 t  
Nϕ (t ,ℵ) = Kα {ϕ (t ;ℵ)} + µ 2α Kα 2ϕ 2  ;ℵ − 1 . (9.28)
  2  

Utilizing equation (9.28) in equation (9.19), we get the recursive formula to evaluate the
term φk (t)
for each k ≥ 1 :


φk (t) = χ kφk −1 (t) + Kα−1  Pk (φk −1 (t))  (9.29)

where

   k −1 
Pk (φk − 1 (t)) = Kα {φk − 1 (t)} − (1 − χ k )µ 2 + µ 2α Kα
 ∑
 2 φk − 1− j  t  φ j  t   
 2  2 
 
  j=0
Nonlinear Delayed Differential Equations 143

With the aid of Mathematica software, solve recurrence relation equation (9.30) for φk (t)
for k ≥ 1 as follows, on solving recurrence equation (9.30), the first three iterations are
obtained as:

tα  2 tα tα 2 1− 2α  3 t 3α  3 tα 2  2 tα tα

φ1 (t) = − ; φ2 (t) = − − ; φ3 (t) = − − −
α α α 3α 3 α α α

In this consequence, the kth iterative solutions φk (t) for k ≥ 4 can be evaluated from equa-
tion (9.29). Thus, sixth-order approximate solution at  = −1 is obtained as:

6 α 1− 2α 3α
81− 2α t 5α
S6 (t) = ∑φ (t) = tα − 2 3α t
k =0
k 3 +
15α 5

which is nearest form of exact solution φ (t) = sin t at α = 1.

Example 9.2
Let us consider the following first-order nonlinear nonhomogeneous pantograph-type
integro-differential equation [26]

t
t   s
t  α φ (t) +  − 2  φ (t) − 2
2  ∫φ 0
2
  ds = 1, 0 < α ≤ 1, φ (0) = 0, t ∈[0, 1].
2
(9.30)

In special case when α = 1 the exact result is φ (t) = te t .

Case 1. IDTM : On implementing IDTM as mentioned in equation (9.12) on the sys-

tem of equation (9.30), we obtain the following recursive equation

φ0 (t) = α−1  µ 2α α {1} 

∞   t k 
∑ 
∫∑
 s
t
  s
φk +1 (t) = −α−1  µ 2α α  − 2  φ j (t) − 2 φ j   φk − j     (9.31)
  2  2 
k =0   2 0
j=0  

By solving the relation equation (9.31), the first two iterations are obtained as:

tα
φ0 (t) = ;
α
 1  
−1/α 1/α
 1  1   tα    1 
 4 −α   Γ  2 +     4Γ  3 +  tα − 4α α 2 Γ  4 +   
 α   α  α
    α   α  
t 2α  2α 2 +
  1  1 
 Γ3 +  Γ4 +  
 α   α 
 
φ1 (t) = ;
2α 4

In this consequence, the kth iterative solutions φk (t) for k ≥ 2 can be evaluated from equa-
tion (9.31). Thus sixth-order approximate solution for α = 1 is obtained as:
144 Mathematical Modeling for Intelligent Systems

S6 (t) = ∑φ (t) = t + t
k =0
k
2
+
t3 t4 t5
+ + +
t6
+
t7
−
t8
+
97t 9
2 6 24 120 720 336 40320 44800
−
39t 10

11311t 11 2861t 12 20265209t 13 2448137t 14

+ − + −
63866880 132710400 12752938598400 35267385753600
49241431t 15 40692071863t 16
+ −
27700928446464000 1404104661094367232000
33521726461t 17 2460828499t 18
+ −
152766587127067154841600 2444265394033074477465600
77094233t 19 (9.32)
+
35762086192911720697036800

which is nearest form of exact solution φ (t) = te t at α = 1.

Case 2. O HAITM : On implementing α operator on the system of equation (9.30) and

using the properties of NIT, we get

 t  t
 s 
α {φ (t)} − µ 2φ (0) + µ 2α α  − 2  φ (t) − 2
 2  ∫φ 0
2
  ds − 1 = 0
2 

Construct nonlinear operator as:

 t  t
s  
Nϕ (t ,ℵ) = Kα {ϕ (t ;ℵ)} + µ 2α Kα  − 2  ϕ (t;ℵ) − 2
 2  ∫ϕ 0
2
 ;ℵ ds − 1 .
2 
(9.33)

Utilizing equation (9.33) in equation (9.19), to evaluate the term φk (t) for each k ≥ 1 :


φk (t) = χ kφk −1 (t) + Kα−1  Pk (φk −1 (t))  (9.34)

where

Pk (φk − 1 (t)) = Kα {φk − 1 (t)} − (1 − χ k )µ 2

 k −1 
 t
∑∫
  s  
t
  s
+ µ 2α Kα   − 2  φ (t) − 2  φ j   φk − 1− j   ds   
  2   2  2  
j=0   
0


With the aid of Mathematica software, solve recurrence relation equation (34) for φk (t)
for k ≥ 1. On solving recurrence equation (9.34), the first three iterations are obtained as
follows:

  1
−1/α
 1  α  tα 
1/α

   Γ  2 +  t 
tα  2 tα  α α α   α  α
 − t ;
φ1 (t) = − ; φ2 (t) = 2  −2α + 2t −
α 2α  1  α
 Γ3 +  
  α 
Nonlinear Delayed Differential Equations 145

  1
−1/α
 1  α  tα 
1/α

   Γ  2 +  t 
 2 tα  α
α  α   α  α 2 2α 2 α 3 α
 − t +  t −  t −  t
φ3 (t) = α
2  −2 + 2t −
2α  1  α α 2
α α
 Γ3 +  
  α 

1
−1/α +2
 1  1   tα  α
2
(2 + 1) Γ  2 +   
2 3 t 3α 2  3 t 2α  α   α α 
− + −
3α 3 α2  1
2Γ  3 + 
 α
1 2
−1/α +3 −2/α +3
 1  1   tα  α  1  2   tα  α
(4α + 1)   3
 Γ2 +      α Γ 3 +   
3
α  α α  α  α α 
+ −
 1  2
2αΓ  4 +  4(2α + 1)Γ  4 + 
 α  α
1
−1/α +3
 1  1   tα  α
2 
α
3
 Γ2 +   
 α α 
( sinh(α log 4) − cosh(α log 4))
+
 1
αΓ  4 + 
 α

In this consequence, the kth iterative solutions φk (t) for k ≥ 4 can be evaluated from equa-
tion (9.34). Thus sixth-order approximate solution at α = 1 and  = −1 is obtained as:

S6 (t) = ∑φ (t) = t + t
k =0
k
2
+
t3 t4 t5
+ + +
t6
−
19t 7
2 6 24 120 1680 13440
+
61t 8

10189t 9 4769t 10 t 11
− + − (9.35)
11612160 58060800 332640

which is nearest form of exact solution φ (t) = te t at α = 1.

Example 9.3
Take the following initial valued problem of T GBE PD :

∂2 ∂   t   x t  1
 α φ ( x , t) = [φ ( x , t)] + φ  x,  φ  ,  + φ( x , t),
∂x   2    2 2  2
t
∂x 2 (9.36)
φ (x , 0) = x , x , t ∈[0, 1], 0 < α ≤ 1.

Case 1. IDTM : On implementing IDTM as mentioned in equation (9.12) on the sys-

tem of equation (9.36), we obtain the following recursive equation:

∑φ (x, t)
k =0
k

 
  ∂2 k −1
 
 t   x t  1
= φ ( x , 0) − α−1  µ 2α α


 2 [φk − 1 ( x , t)] +
 ∂ x
∑  ∂x∂ φ k − 1− j
2  2 2  2

 x ,   φ j  ,   + φk −1 ( x , t)  
 
(9.37)
 j=0
146 Mathematical Modeling for Intelligent Systems

By solving the recurrence relation equation (9.37), the first three iterative terms are
obtained as:

 − α 1  2α
2 +  xt
xtα  2
φ0 ( x , t) = x ; φ1 ( x , t) = ; φ2 ( x , t) = ;
α 2α 2

φ3 ( x , t) =
(
2 −3(α + 1) 3 2α + 1 + 2 2α + 1 + 8α + 4 xt 3α )
3α 3

In this consequence, the kth iterative solutions φk (t) for k ≥ 4 can be evaluated from equa-
tion (9.37). Thus sixth-order approximate solution for α = 1 is obtained as:

6
 
∑φ (x, t) = x  1 + t + t2! + t3! + t4! + t5! + 6!t 
2 3 4 5 6
S6 ( x , t) = k (9.38)
k=0

which is nearest form of exact solution φ ( x , t) = xe t at α = 1.

Case 2. O HAITM : On implementing α operator on the system of equation (9.36) and

using the properties of NIT, we get

 ∂ 2 ∂   t   x t  1 
α {φ ( x , t)} − µφ ( x , 0) − µ 2α α  2 [φ (x , t)] + φ  x,   φ  ,  + φ (x , t)  = 0
 ∂x ∂x   
2  2 2  2 

Construct nonlinear operator as:

Nϕ (t ,ℵ) = Kα {ϕ ( x , t ;ℵ)} − xµ − µ 2α

 ∂ 2 ∂   t   x t  1 
× Kα  2 [ϕ ( x , t ;ℵ)] + ϕ  x, ;ℵ  ϕ  , ;ℵ + ϕ ( x , t;ℵ)  . (9.39)
 ∂x ∂x   2  
 2 2  2 

Utilizing equation (9.39) in equation (9.19), we get the recursive formula to evaluate the
term φk ( x , t) for each k ≥ 1 :


φk ( x , t) = χ kφk − 1 ( x , t) + Kα−1  Pk (φk −1 ( x , t))  (9.40)

where

Pk (φk − 1 (t)) = Kα {φk − 1 (t)} − (1 − χ k )µφk −1 ( x, 0)

 2 k −1
 
 t   x t  1
 ∂
+ µ 2α Kα   2 [φk − 1 ( x , t)] +
  ∂x ∑  ∂∂x φ k − 1− j
2  2 2  2

 x ,   φ j  ,   + φk − 1 ( x , t)  
 
 j=0

With the aid of Mathematica software, solve recurrence relation equation (9.40) for
φk ( x , t) for k ≥ 1. On solving the recurrence equation (9.40), the first three iterations are
obtained as follows:
Nonlinear Delayed Differential Equations 147

φ1 ( x , t) = −
xtα
; φ2 ( x , t) =  
( )
 2 −α − 2 2α + 2 xt 2α xtα  xtα
− − ;
α  α2 α  α

2 −3α − 3  3 xt 3α  4 + 6 sinh(α log 2) + 2 sinh(α log 4) + sinh(α log 8) + 6 cosh(α log 2) 

φ3 ( x , t) =  
3α 3
+2 cosh(α log 4)cosh(α log 8) 

2 −3α − 2  2 xt 2α 3α 2α +1
+ 2
 +2  + 2( + 1)sinh(α log 4) + 2( + 1)cosh(α log 4)
α2
+ ( + 1)sinh(α log 4) ( sinh(α log 2)) + cosh(α log 2)

+ ( + 1)cosh(α log 4) ( cosh(α log 2) + sinh(α log 2)) 

+
( 3 (− x) −  2 x)tα xtα
− + 
( −
)
 2 −α − 2 2α + 2 xt 2α xtα 

α α  α2 α 

In this consequence, the kth iterative solutions φk (t) for k ≥ 4 can be evaluated from equa-
tion (9.40). Thus sixth-order approximate solution at  = −1 is obtained as:

2 −2α − 5 xt 4α 3 2 −3α − 5 xt 4α 2 −5α − 4 xt 4α 2 −α − 5 xt 4α

S6 ( x , t) = + + +
α4 α4 α4 3α 4
7 2 −4α − 4 xt 4α 2 −6α − 3 xt 4α xt 4α 2 −2α − 2 xt 3α 2 −α −2 xt 3α
+ + + + +
3α 4
3α 4
192α 4
α3 3α 3
2 −3α − 1 xt 3α xt 3α 2 −α −1 xt 2α xt 2α xtα
+ + + + + + x +…
3α 3
24α 3
α2 4α 2 α

which is nearest form of exact solution.

Example 9.4
In the last example consider the CTFPDEs as follows:

∂2  x t  ∂  x t  1 ∂
t  α φ ( x , t) = φ ,  φ ,  − φ ( x , t) − φ (x , t)
∂x 2  2 2  ∂ x  2 2  8 ∂x (9.41)
φ ( x , 0) = x , x , t ∈[0, 1], 0 < α ≤ 1.
2

Case 1. IDTM : On implementing IDTM as mentioned in equation (9.12) on the sys-

tem of equation (9.41), we obtain the following recursive equation

∞   k ∂2 
∑ φk ( x , t) = φ ( x, 0) − α−1  µ 2α α


∑ ∂x
φ
x t ∂
2 j
, φ
x t 1 ∂
, − φ
 2 2  ∂ x k − j  2 2  8 ∂x k
( x , t) − φ k ( x , t )



k =0   j =0  
(9.42)
148 Mathematical Modeling for Intelligent Systems

By solving the relation equations (9.42), the first three iterations are obtained as

φ0 ( x , t) = x 2 ; φ1 ( x , t) = −
x 2 tα
; φ 2 ( x , t) =
(
2 −α − 3 xt 2α 2α + 2α + 2 x − 2
;
)
α α2

φ 3 ( x , t) = −
( (
8−α − 2 t 3α 8α − 2α + 1 − 2 2α + 1 + 2 3α + 5 x 2 + 2α + 4 −2α + 4α − 2 x + 4 ) )
3α 3

In this consequence, the kth iterative solutions φk (t) for k ≥ 4 can be evaluated from equa-
tion (9.42). Thus sixth-order approximate solution for α = 1 is obtained as

6
 
∑φ (x, t) = x  1 − t + t2! − t3! + t4! − t5! + 6!t 
2 3 4 5 6
S6 ( x , t) = k
2

k =0

Which is nearest form of exact solution φ ( x , t) = x 2 e −t at α = 1.

Case 2. O HAITM : On implementing α operator on the system of equation (9.41) and

using the properties of NIT, we get

 ∂2  x t  ∂  x t  1 ∂ 
α {φ ( x , t)} − µφ ( x, 0) − µ 2α α  2 φ  ,  φ ,  − φ ( x , t) − φ ( x , t)  = 0
 ∂x   
2 2 ∂x 2 2  8 ∂x 

Construct nonlinear operator as:

Nϕ (t ,ℵ) = Kα {ϕ ( x , t ;ℵ)} − x 2 µ − µ 2α Kα

 ∂2  x t  ∂  x t  1 ∂ 
 2 ϕ  , ;ℵ ϕ  , ;ℵ − ϕ ( x , t;ℵ) − ϕ ( x, t;ℵ)  . (9.43)
 ∂x  2 2  ∂x  2 2  8 ∂x 

Utilizing equation (9.43) in equation (9.19), we get the reclusive formula to evaluate the
term φk ( x , t) for each k ≥ 1 :


φk ( x , t) = χ kφk −1 ( x , t) + Kα−1  Pk (φk −1 ( x , t))  (9.44)

where


Pk (φk − 1 (t)) = Kα {φk − 1 (t)} − (1 − χ k )x 2 µ + µ 2α

  k −1 
 ∂2  x t  ∂  x t  1 ∂


× Kα   ∑
 ∂x 2 φ k − 1− j 
 2
, 
2  ∂x
φj  ,   −
 2 2   8 ∂ x

φk − 1 ( x , t) − φk − 1 ( x , t)  
 
  j =0

With the aid of Mathematica software, solve recurrence relation equation (9.44) for
φk ( x , t) for k ≥ 1. On solving recurrence equation (9.44), the first three iterations are
obtained as follows:
Nonlinear Delayed Differential Equations 149

hx 2 tα h2 x 2 t 2α 2 −α − 2 h2 xt 2α h2 xt 2α h2 x 2 tα hx 2 tα
φ1 ( x , t) = ; φ2 ( x , t) = − + + + ;
α 2α 2 α2 8α 2 α α
 3 x 2 tα 2  2 x 2 tα x 2 tα
φ3 ( x , t) = + +
α α α
2 −3α − 3  2 ( + 1)x cosh(α log 4) ( cosh(α log 2) + sinh(α log 2)) t 2α
+
α2
2 −3α −1  2 ( + 1)x 2 ( cosh(α log 8) + sinh(α log 8)) t 2α
+
α2
2 −3α − 3  2 ( + 1)x sinh(α log 4) ( cosh(α log 2) + sinh(α log 2)) t 2α
+
α2

2 −3α − 4 h3 t 3α 2 h ( cosh(α log 2) + sinh(α log 2)) t 3α

−3α − 5 3
+ −
3α 3
3α 3
2 −α − 2  2 ( + 1)xt 2α  2 ( + 1)x 2 t 2α  2 ( + 1)xt 2α
− − +
α2 2α 2 8α 2

2 −3α − 2  2 ( + 1)x ( cosh(α log 4) + sinh(α log 4)) t 2α

+
α2

+
(
2 −3α − 2 h3 x cosh 3 (α log 2) + sinh 3 (α log 2) t 3α )
3α 3

2 −3α − 2 h3 x ( −1 / 3 + cosh(α log 2)) sinh 2 (α log 2)t 3α

+
α3
2 −3α − 6 h3 ( cosh(α log 8) + sinh(α log 8)) t 3α
+
3α 3

2 −3α − 1 h3 x 2 ( cosh(α log 8) + sinh(α log 8)) t 3α

+
3α 3
2 −3α − 2 h3 x cosh 2 (α log 2) ( −1 / 3 + sinh(α log 2)) t 3α
+
α3
2 −3α − 1 h3 x ( cosh(α log 2) − sinh(α log 2) − cosh(α log 2)sinh(α log 2)) t 3α
−
3α 3
2 −3α − 5 h3 ( cosh(α log 4) + sinh(α log 4)) t 3α
−
3α 3

In this consequence, the kth iterative solutions φk (t) for k ≥ 4 can be evaluated from equa-
tion (9.44). Thus sixth-order approximate solution at and α = 1 reduces to

6
 
∑φ (x, t) = x  1 − t + t2! − t3! + t4! − t5! + 6!t 
2 3 4 5 6
S6 ( x , t) = k
2

k =0

Which is nearest form of exact solution.

150 Mathematical Modeling for Intelligent Systems

9.6 Conclusion
In this chapter, a comparative study for CTFPDEs with proportional delay has been made
by adopting two efficient techniques: New integral decomposition transform method
(IDTM) and optimal homotopy analysis NIT method (O HAITM). Specially, conform-
able time-fractional nonlinear nonhomogeneous pantograph-type integro-differential
equation and Burger equation with proportional delay along with two more problems are
effectively solved and analyzed via these two techniques. The findings from the graph-
ics and tables represent that the approximate solutions from both techniques agreed well
with the exact solution and the error decreases with increasing the order of iterations.
Moreover, (IDTM) converges faster than (O HAITM) but computational cost of (IDTM)
is higher than (O HAITM) for Examples 9.1 and 9.2. But for Examples 9.3 and 9.4 the solu-
tions obtained from IDTM and (O HAITM) for  = −1 coincides. Although (O HAITM)
produces more accurate results in comparison to IDTM and existing results [19] at opti-
mal value of  with higher computational cost.

9.7 Result and Discussion

Tables 9.1 and 9.2 report approximate solutions obtained from O HAITM at optimal value
 = −1 for different values of α = 0.8, 0.9,1 and exact solution at α = 1.
In addition, absolute error in sixth-order results computed from IDTM and O HAITM
at different time levels are also reported in the table. Figures 9.1a and 9.2a depicts the behav-
ior of sixth-order results for different α = 0.8, 0.9,1; Figures 9.1b and 9.2b show the two-
dimensional plots of the absolute error for α = 1 in kth order results (k = 4, 5, 6); Figures 9.1c

TABLE 9.1
Approximate Solutions in k = 6th Order Results of Example 9.1
Absolute Error

t α = 0.8 α = 0.9 α =1 Exact  = −1 IDTM

0.1 0.687114809 0.338579687 0.099833417 0.099833417 1.98387E-11 1.38778E-17
0.2 1.060792659 0.561720142 0.198669333 0.198669331 2.53827E-09 2.77556E-17
0.3 1.338783663 0.743043822 0.295520250 0.295520207 4.33387E-08 0.00000E+00
0.4 1.555763829 0.895494803 0.389418667 0.389418342 3.24358E-07 0.00000E+00
0.5 1.726874559 1.024671393 0.479427083 0.479425539 1.54473E-06 0.00000E+00
0.6 1.860708146 1.133664777 0.564648000 0.564642473 5.52660E-06 3.33067E-16
0.7 1.962954201 1.224464662 0.644233917 0.644217687 1.62294E-05 3.55271E-15
0.8 2.037818345 1.298519863 0.717397333 0.717356091 4.12424E-05 2.68674E-14
0.9 2.088683584 1.357005549 0.783420750 0.783326910 9.38404E-05 1.56986E-13
1 2.118457180 1.400966770 0.841666667 0.841470985 1.95682E-04 7.61946E-13
CPU Time (In sec) 144.187 79.8

For α = 0.8, 0.9 & 1 compared with exact result and absolute error in 6th order results obtained from O HAITM for
optimal value of  (  = −1) and IDTM at distinct time levels 0 < t ≤ 1.
Nonlinear Delayed Differential Equations 151

TABLE 9.2
Approximate Solutions in k = 6th Order Results of Example 9.2
Absolute Error

t α = 0.8 α = 0.9 α=1 Exact  = −1  = −1.1240 NIDTM

0.1 0.4836016 0.2316341 0.1105170 0.1105170 1.2273E-09 1.2908E-09 2.9451E-11
0.2 1.0779989 0.5375138 0.2442804 0.2442805 1.5188E-07 9.6011E-08 6.9938E-09
0.3 1.7403048 0.9009098 0.4049551 0.4049576 2.5097E-06 6.6505E-07 1.6628E-07
0.4 2.4548178 1.3149513 0.5967116 0.5967298 1.8189E-05 1.4747E-08 1.5409E-06
0.5 3.2151700 1.7772602 0.8242766 0.8243606 8.3938E-05 4.3351E-06 8.5221E-06
0.6 4.0193992 2.2879798 1.0929800 1.0932712 2.9120E-04 7.3893E-06 3.4011E-05
0.7 4.8681144 2.8489681 1.4087971 1.4096268 8.2979E-04 6.2423E-06 1.0839E-04
0.8 5.7636440 3.4634006 1.7783850 1.7804327 2.0477E-03 5.5071E-05 2.9302E-04
0.9 6.7095869 4.1355520 2.2091146 2.2136428 4.5282E-03 1.3852E-04 6.9885E-04
1 7.7105552 4.8706663 2.7090975 2.7182818 9.1843E-03 1.9646E-04 1.5102E-03
CPU (In sec) 95.657 154.031

For α = 0.8, 0.9 &1 compared with exact result and absolute error in 6th order results obtained from O HAITM for
 = −1 and optimal value of ) and IDTM at distinct time levels 0 < t ≤ 1.

and 9.2c display the comparison of the absolute error in k = 6th order solution obtained from
IDTM and O HAITM. The findings show that the computed result obtained from both
techniques agreed well with the exact solution and the error decreases with increasing the
iterations. It can also be depicted from table that IDTM converges faster than O HAITM
but the computational cost of IDTM is much higher than O HAITM for Examples 9.1
and 9.2.
Tables 9.3 and 9.4 report approximate solutions obtained from O HAITM at  = −1 and
at the optimal value  = −1.1240 for different values of α = 0.8, 0.9,1 and exact solution at
α = 1. In addition, absolute error in sixth-order results computed from IDTM and
O HAITM & comparison with [19] at different time levels also reported in table.
Figures 9.3a and 9.4a depict the behavior of sixth-order results for different α = 0.8, 0.9,1;
Figures 9.3b and 9.4b show 3D plots of the absolute error for α = 1 in kth order results (k = 4,
5, 6); Figures 9.3c and 9.4c show the two-dimensional plots of the absolute error for α = 1 in
kth order results (k = 4, 5, 6); Figures 9.3d and 9.4d display the comparison of absolute error
in k = 6th order solution obtained from IDTM and O HAITM.The finding shows that the
computed result obtained from both techniques agreed well with the exact solution and
the error decreases with increasing the iterations. It can also be depicted from table that
the solutions obtained from IDTM and O HAITM for -coincides but at optimal value of
 absolute error in k = 6th order solution decrease more rapidly than IDTM and ERCDTM
[19]. Also, computational cost of O HAITM is higher than IDTM for Examples 9.3 and 9.4.
152 Mathematical Modeling for Intelligent Systems

(a)

(b)

(c)

FIGURE 9.1
(a) 2D plots of the behavior k = 6th order results for α = 0.8, 0.9, 1 and exact result in t ∈(0, 1) for Example 9.1.
(b) Absolute error plot for kth iterative (k = 4, 5, 6) results for Example 9.1. (c) Comparison in absolute error for
k = 6th order iterative results obtained from both techniques in Example 9.1.
Nonlinear Delayed Differential Equations 153

(a)

(b)

(c)

FIGURE 9.2
(a) 2D plots of the behavior k = 6th order results for α = 0.8, 0.9, 1 and exact result in t ∈(0, 1) for Example 9.2.
(b) Absolute error plot for kth iterative (k = 4, 5, 6) results for Example 9.2. (c) Comparison in absolute error for
k = 6th order iterative results obtained from both techniques in Example 9.2.
154 Mathematical Modeling for Intelligent Systems

TABLE 9.3
Approximate Solution in k = 6th Order Results of Example 9.3
Absolute Error

(x, t) α = 0.8 α = 0.9 α =1 Exact [19] IDTM  = −1

(0.25,0.25) 0.5044843 0.4333137 0.3210064 0.3210064 3.12484E-09 3.12484E-09 3.58611E-10
(0.25,0.5) 0.6419288 0.5778248 0.4121799 0.4121803 4.13161E-07 4.13161E-07 1.44368E-08
(0.25,0.75) 0.7627260 0.7251787 0.5292427 0.5292500 7.29785E-06 7.29785E-06 7.95825E-08
(0.25,1) 0.8857146 0.8857146 0.6795139 0.6795705 5.65682E-05 5.65682E-05 8.30344E-07
(0.5,0.25) 1.0089687 0.8666274 0.6420127 0.6420127 6.24969E-09 6.24969E-09 7.17221E-10
(0.5,0.5) 1.2838577 1.1556495 0.8243598 0.8243606 8.26322E-07 8.26322E-07 2.88735E-08
(0.5,0.75) 1.5254521 1.4503575 1.0584854 1.0585000 1.45957E-06 1.45957E-06 1.59165E-07
(0.5,1) 1.7714292 1.7714292 1.3590278 1.3591409 1.13136E-05 1.13136E-05 1.66069E-06
(0.75,0.25) 1.5134530 1.2999410 0.9630191 0.9630191 9.37453E-09 9.37453E-09 1.07583E-09
(0.75,0.5) 1.9257865 1.7334743 1.2365397 1.2365410 1.23948E-06 1.23948E-06 4.33103E-08
(0.75,0.75) 2.2881781 2.1755362 1.5877281 1.5877500 2.18936E-05 2.18936E-05 2.38748E-07
(0.75,1) 2.6571439 2.6571439 2.0385417 2.0387114 1.69705E-04 1.69705E-04 2.49103E-06
CPU (In sec) 1.280 75.155 230.531

For α = 0.8, 0.9 & 1 and comparison of absolute error from IDTM and O HAITM in M = 6th order results with
ERC-DTM [19] at distinct time levels 0 < x ≤ 1 and 0 < t ≤ 1.

TABLE 9.4
Approximate Solution in k = 6th order Results of Example 9.4
Absolute Error

(x, t) α = 0.8 α = 0.9 α =1 Exact [19] IDTM  = −1.06462

(0.25,0.25) 0.1459650 0.1083284 0.0802516 0.0802516 7.81211E-10 7.81211E-10 1.09656E-10
(0.25,0.5) 0.1968759 0.1444562 0.1030451 0.1030451 1.03290E-07 1.03290E-07 3.66342E-09
(0.25,0.75) 0.2416755 0.1812947 0.1323125 0.1323125 1.82446E-06 1.82446E-06 2.25854E-08
(0.25,1) 0.2844234 0.2214287 0.1698928 0.1698926 1.41421E-05 1.41421E-05 1.84690E-07
(0.5,0.25) 0.5838600 0.4333137 0.3210064 0.3210064 3.12484E-09 3.12484E-09 4.38623E-10
(0.5,0.5) 0.7875038 0.5778248 0.4121803 0.4121803 4.13161E-07 4.13161E-07 1.46537E-08
(0.5,0.75) 0.9667019 0.7251787 0.5292501 0.5292500 7.29785E-06 7.29785E-06 9.03416E-08
(0.5,1) 1.1376934 0.8857146 0.6795712 0.6795705 5.65682E-05 5.65682E-05 7.38759E-07
(0.75,0.25) 1.3136851 0.9749558 0.7222643 0.7222643 7.03090E-09 7.03090E-09 9.86902E-10
(0.75,0.5) 1.7718835 1.3001057 0.9274057 0.9274057 9.29613E-07 9.29613E-07 3.29708E-08
(0.75,0.75) 2.1750793 1.6316522 1.1908127 1.1908125 1.64202E-05 1.64202E-05 2.03268E-07
(0.75,1) 2.5598102 1.9928579 1.5290352 1.5290335 1.27279E-04 1.27279E-04 1.66221E-06
CPU (In sec) 2.406 80.094 379.531

For α = 0.8, 0.9 & 1 and comparison of absolute error obtained from IDTM and O HAITM in k = 6th order results
with [19] at distinct time levels 0 < x ≤ 1 and 0 < t ≤ 1.
Nonlinear Delayed Differential Equations 155

(a)

(b)

(c)

(d)

FIGURE 9.3
(a) 2D plots of the behavior k = 6th order results for α = 0.8, 0.9, 1 and exact result in t ∈(0, 1) for Example 9.3.
(b) 3D plots of the behavior k = 6th order results for α = 0.8, 0.9, 1 and exact result in x , t ∈ (0, 1) for Example 9.3.
(c) Absolute error plot for kth iterative (k = 4, 5, 6) results for Example 9.3. (d) Comparison in absolute error for
k = 6th order iterative results obtained from both techniques for x = 0.5 and t ∈(0, 1) in Example 9.3.
156 Mathematical Modeling for Intelligent Systems

(a)

(b)

(c)

(d)

FIGURE 9.4
(a) 2D plots of the behavior k = 6th order results for α = 0.8, 0.9, 1 and exact result in t ∈(0, 1) for Example 9.4.
(b) 3D plots of the behavior k = 6th order results for α = 0.8, 0.9, 1 and exact result in x , t ∈ (0, 1) for Example 9.4.
(c) Absolute error plot for kth iterative (M = 4, 5, 6) results for Example 9.4. (d) Comparison in absolute error for
k = 6th order iterative results obtained from both techniques for x = 0.5 and t ∈(0, 1) in Example 9.4.
Nonlinear Delayed Differential Equations 157

Acknowledgments
The authors are obliged to the editor and reviewers for their advice & instructions. S.
Agrawal thanks to BBA University, Lucknow, India for their economic support to perform
her research article.

References
1. Miller, K. S., Ross, B., An Introduction to the Fractional Calculus and Fractional Differential Equations,
Wiley, New York (1993).
2. Podlubny, I., Fractional Differential Equations, Academic Press, San Diego (1999).
3. Caputo, M., Mainardi, F., Linear models of dissipation in anelastic solids, Rivista del Nuovo
Cimento, 1 (1971) 161–98.
4. He, J.H., Nonlinear oscillation with fractional derivative and its applications. In International
Conference on Vibrating Engineering (1998) 288–291.
5. Burgers, J.M., A mathematical model illustrating the theory of turbulence, Adv. Appl. Mech. 1
(1948) 171–199.
6. Duan, Y., Tian, P., Zhang, S., Oscillation and stability of nonlinear neutral impulsive delay dif-
ferential equations, J. Appl. Math. Comput. 11 (2003) 243–253.
7. Jordan, P.M., A note on Burgers’ equation with time delay: Instability via finite-time blow-up,
Phys. Lett. A 372 (2008) 6363–6367.
8. Kolmanovskii, V., Myshkis, A., Introduction to the theory and applications of functional dif-
ferential equations, 1999.
9. Ockendon, J.R., Tayler, A.B., The dynamics of a current collection system for an electric loco-
motive, Proc. R. Soc. Lond. Ser. A. 322(1971) 447–468.
10. Khalil, R., Al Horani, M., Yousef, A., Sababheh, M., A new definition of fractional derivative, J.
Comput Appl Math, 264 (2014) 65–70. doi: 10.1016/j.cam.2014.01.002
11. Lio, S., The proposed homotopy analysis techniques for the solution of nonlinear problems
(Ph.D. Thesis), Shanghai Jiao Tong University, Shanghai (1992).
12. Liao, S., Homotopy Analysis Method in Nonlinear Differential Equations, Springer, Heidelberg,
Dordrecht, London (2012).
13. Iyiola, O. S., Tasbozan, O., Kurt, A., Çenesiz, Y., On the analytical solutions of the system of con-
formable time-fractional Robertson equations with 1-D diffusion, Chaos, Solitons and Fractals 94
(2017), 1–7. doi: 10.1016/j.chaos.2016.11.003.
14. Kumar, S., Kumar, A., Baleanu, D., Two analytical methods for time fractional nonlinear cou-
pled Boussinesq Burger’s equations arise in propagation of shallow water waves, Nonlinear
Dyn. 85 (2) (2016) 699–715.
15. Zurigat, M., Momani, S., Odibat, Z., Alawneh, A., The homotopy analysis method for handling
systems of fractional differential equations, Appl. Math. Model. 34 (2010) 24–35.
16. Ray, S.S., Patra, A., Application of homotopy analysis method and adomian decomposition
method for the solution of neutron diffusion equation in the hemisphere and cylindrical reac-
tors, J. Nucl. Eng. Technol. 1 (2011) 1–12.
17. Cenesiz, Y., Kurt, A., Tasbozan, O., On the new solutions of the conformable time fractional
generalized hirota-satsuma coupled KdV system, Ann. West Univ. Timisoara-Math. Comput. Sci.
1 (2017) 37–49. doi: 10.1515/awutm–2017–0003
18. Saad, K.M., Al-Shareef, E.H., Mohamed, M.S., Yang, X.-J., Optimal q-homotopy analysis
method for time-space fractional gas dynamics equation, Eur. Phys. J. Plus 132 (2017) 23.
158 Mathematical Modeling for Intelligent Systems

19. Singh, B.K., Agrawal, S., A new approximation of conformable time fractional partial differ-
ential equations with proportional delay, App. Numer. Math. 157(2020) 419–433. doi: 10.1016/j.
apnum.2020.07.001
20. Thabet, H., Kendre, S., Analytical solutions for conformable space-time fractional partial dif-
ferential equations via fractional differential transform, Chaos Solitons Fractals 109 (2018) 238–
245. doi: 10.1016/j.chaos.2018.03.001
21. Abdeljawad, T., On conformable fractional calculus, J. Comput. Appl. Math. 279 (2015) 57–66.
doi: 10.1016/j.cam.2014.10.016.
22. Kashuri, A., Fundo, A., A new integral transform, Adv. Theor. Appl. Math. 8(1) (2013) 27–43.
23. Singh, B.K., Homotopy perturbation new integral transform method for numeric study of
spaceand time- fractional (n+1)-dimensional heatand wave-like equations, Waves Wavelets
Fractals 4 (2018) 19–36. doi: 10.1515/wwfaa–2018–0003
24. Shah, K., Twinkle, S., Kilicman, A., Combination of integral and projected differential trans-
form methods for time-fractional gas dynamics equations, Ain Shams Eng. J. 9(4) (2017) 1683–
1688. doi: 10.1016/j.asej.2016.09.012
25. Khan, H., Khan, A., Chen, W., Shah, K., Stability analysis and a numerical scheme for fractional
Klein-Gordon equations, Math Methods Appl. Sci. 42 (2019) 723–732. doi: 10.1002/mma.5375
26. Smardal, Z., Diblikl, J., Khan, Y., Extension of the differential transformation method to non-
linear differential and integro-differential equations with proportional delays, Adv. Differ.
Equ. 2013 (2013) 1–12.
10
Numerical Treatment on the Convective
Instability in a Jeffrey Fluid Soaked
Permeable Layer with Through-Flow

Dhananjay Yadav
University of Nizwa

Mukesh Kumar Awasthi

Babasaheb Bhimrao Ambedkar University

U. S. Mahabaleshwar
Davangere University

Krishnendu Bhattacharyya
Banaras Hindu University

CONTENTS
10.1 Introduction ........................................................................................................................ 159
10.2 Mathematical Model.......................................................................................................... 160
10.2.1 Basic State ................................................................................................................ 162
10.3 Stability Analysis ............................................................................................................... 162
10.4 Procedure of Solution ........................................................................................................ 163
10.5 Outcomes and Discussion ................................................................................................ 163
10.6 Conclusions......................................................................................................................... 166
Acknowledgments ...................................................................................................................... 166
References..................................................................................................................................... 167

10.1 Introduction
Convective instability is a phenomenon that frequently occurs in a horizontal layer of fluid
warmed from the bottom. Due to heat from the bottom, the density of the bottom fluid
decreases, so buoyancy force drives the less-dense fluid upward in the direction of the
colder end of the plate [1–4]. In the meantime, the colder fluid at the top is denser, so it goes
down and substitutes the hot fluid. This convective instability problem in a layer of liquid-
soaked permeable media has been extensively explored by numerous investigators in the
past as current because of its significant applications in various areas, like geosciences
(geothermal basins, geological carbon repository), astrophysics (pore-liquid convective

DOI: 10.1201/9781003291916-10 159

160 Mathematical Modeling for Intelligent Systems

motion within carbonaceous chondrite parent bodies), manufacturing and industrialized

procedure (water management practice, nuclear dissipate removal) [5–10].
In many such applications, the control of the instability of fluid flow is a very important and
fascinating task. Throughflow is one of the effectual ways that manages the convective fluid
movement in permeable media and it is studied by many researchers. Throughflow amends
the basic shape for the temperature field from linear to nonlinear with a depth of the layer,
which influences the steadiness of the scheme considerably. Higgins [11] and Sherwood and
Homsy [12] analyzed the impact of throughflow on convective instability for the situ dealing
of power resources in coal gasification. The limit for the occurrence of convective progress in
permeable media with throughflow was studied by Nield [13]. Chen [14] inspected the weight
of upright throughflow on the convective activity in permeable layers and verified the limit
presented by Nield [13] numerically. An analytical examination of the power of perpendicu-
lar throughflow on the start of convection in a layered permeable matrix was made by Nield
and Kuznetsov [15]. They established that the throughflow has a stabilizing influence whose
amount may be augmented or diminished with the heterogeneity. Later, Hill [16], Hill et al. [17],
Yadav [18–21], Shankar and Shivakumara [22–24], Brevdo and Ruderman [25], and Kiran [26]
inspected the power of throughflow on the convective instability under various conditions.
The convective flow for non-Newtonian fluids involving the Jeffrey model is an area of
research undergoing rapid growth in the past few years due to the best fit model for biologi-
cal fluid (spit, synovia, blood, and chyme) and thin polymer solution [27–39]. Nazeer et al. [40]
investigated the multiphase motion of Jeffrey liquid with warmth transport throughout a flat
channel. They observed that the drive of multiphase movement improves due to shear thinning
influences sourced by Jeffrey parameter. Impacts of radiation and magnetic force on compress-
ible Jeffrey fluid were explored by Khan and Rafaqat [41]. Mehboob et al. [42] deliberated the
transportation of a Jeffrey liquid throughout a permeable slit of microchannel. They found that
the flood rate at the middle line of slit and shear stress on the boundaries of slit decompose
due to the occurrence of permeable medium and Jeffrey parameter. Ali et al. [43] investigated
the magnetic force impact on the nonisothermal movement of an incompressible Jeffrey liquid.
They detected that the Jeffrey parameter plays an important role to manage the velocity, flood
rate, and web coating width. The force of electric power on the initiate of stationary Jeffrey-
nanofluid convection in a permeable matrix was scrutinized by Gautam et al. [44]. They created
that the Jeffrey parameter progress in the launch of marginal convection. Yadav [45–47] studied
the effects of warmth nonequilibrium, anisotropy, and inner heating on the arrival of Jeffrey
fluid convection and found that the steadiness of the arrangement reduced with the expansion
of the Jeffrey parameter. The magnetic power influence on the convective development in a
permeable medium layer soaked by Jeffrey liquid was inspected by Yadav et al. [48]. They found
that the extent of convective cells reduces with Jeffrey and magnetic force parameters.
The literature evaluation confirms that till now no exertion has been taken to explore the
impact of throughflow on the Jeffrey fluid convection, while it is important in many appli-
cations as cited above. Therefore, the chief aspiration of the present work is to study the
impact of throughflow on the convective instability in a Jeffrey fluid-soaked porous layer.

10.2 Mathematical Model
The physical setup engages a horizontal Jeffrey fluid flooded permeable layer bounded by
restrictions at z = 0 and z = d as exposed in Figure 10.1. The base boundary is mentioned
at an identical temperature T0 + ∆T, while the top boundary is kept at a temperature T0.
Numerical Treatment on the Convective Instability 161

FIGURE 10.1
Graphic representation of the considered problem.

The layer is activated by a uniform upward throughflow q0 and the gravity which pro-
ceeds in the reverse z-pathway. The Darcy law is expanded to comprise the viscoelasticity
of the Jeffrey fluid and the Boussinesq finding is employed for density digressions. The
pertinent governing equations are [5,44,46]:

∇ ⋅ q = 0, (10.1)
µK
∇P + q = ρ0 1 − βT (T − T0 )  g , (10.2)
(1 + η )
∂T
γ + (q ⋅ ∇ ) T = α∇ 2T , (10.3)
∂t
w = q0 , T = T0 +∆T , at z = 0 and w = q0 , T = T0 , at z = d. (10.4)

where q ( u, v , w ) means the velocity vector, t corresponds to the time, ρ0 represents the
reference density at T = T0 , K and α are the permeability and the thermal diffusivity of
the permeable medium, correspondingly, βT illustrates the heat expansion coefficient, P
designates the pressure, γ characterizes the heat capacity fraction, η symbolizes the Jeffrey
parameter and described by the fraction of relaxation to retardation instants, and µ denotes
the viscosity of the Jeffrey liquid.
The following scaling is used to get the nondimensional formulation of the problem:

x  tα qd  PK  T − T0
x → , t → 2 , q = , P= ,T= , (10.5)
d γd α µα ∆T

By using equation (10.5) in to equations (10.1)–(10.4), we have

 ⋅ q = 0,
∇ (10.6)
q
(1 + η )
= −∇ ( 0 )
 P + zR ρ eˆ z + R T T eˆ z , (10.7)

∂T
∂t
(
+ q ⋅ ∇ )  2T ,
 T = ∇ (10.8)

w = Pe at z = 0,1, T = 1 at z = 0 and T = 0 at z = 1. (10.9)

162 Mathematical Modeling for Intelligent Systems

ρ0 g Kd ρ0 βT g∆TKd
Here, R ρ0 = (density Rayleigh–Darcy number), R T = (Rayleigh–Darcy
µα µα
q0 d
number), and Pe = (Péclet number).
α

10.2.1 Basic State
It is believed that the basic status of affairs to be steady and fully developed in the upright
z—direction as q b = ( 0, 0, Pe ) and Tb = Tb ( z ) . Then, from the equations (10.8) and (10.9), the
basic temperature field is found as:

e Pe − e Pez
Tb = Pe . (10.10)
e −1

10.3 Stability Analysis
For stability analysis, we disturb the basic state by very small amplitude disturbances as

P = Pb + φ P ′, q = q b + φq ′ , T = Tb + φT ′ (10.11)

Here prime denotes the perturbed nondimensional variables and φ << 1. Replacing equa-
tion (10.11) into equations.(10.6)–(10.9) with the basic state results and eliminating the pres-
sure term from equation (10.7), we have the stability equations as:

 2 w ′
∇  2 T ′,
=R ∇ (10.12)
(1 + η ) T H
∂T ′
∂t
(
+ q ′ ⋅ ∇ ) (  T ′ = ∇
 Tb + q b ⋅ ∇
)  2T ′. (10.13)

w ′ = 0, T ′ = 0, at z = 0,1. (10.14)
2 2
Here ∇ 2H = ∂ + ∂ . By means of the normal modes [1, 49–58], the perturbed variables
∂x 2 ∂ y 2
can be taken as:
 w ′  W ( z ) 
 =   exp  iλ x x + iλ y y + ω t. (10.15)
T ′  Θ ( z ) 
   

where λ x and λ y are the wave numbers in the x and y directions, respectively and ω is the
growth rate of instability. On using equation (10.15) into equations (10.12)–(10.14), we have:

(D2 − λ 2 )W + (1 + η )λ 2 RT Θ = 0, (10.16)
dT
− b W +  D2 − λ 2 − ω − PeDΘ
 = 0, (10.17)
dz
Numerical Treatment on the Convective Instability 163

W = Θ = 0, at z = 0,1. (10.18)

d
where ≡ D and λ = λ x 2 + λ y 2 is the resulting dimensionless wave number.
dz

10.4 Procedure of Solution
The resulting system of equations (10.16)–(10.18) are solved with high numerical correct-
ness by utilizing the higher-order Galerkin process. Thus, we assumed as

n N

W= ∑
p=1
Ep sin pπ z and Θ = ∑ F sin pπ z (10.19)
p=1
p

where Ep and Fp are undisclosed coefficients. Exploiting equation (10.19) into equations
(10.16) and (10.17) and using the orthogonal properties, we have

G jp Ep + H jp Fp = 0,
(10.20)
I jp Ep + J jp Fp = ω K jp Fp .

Here G jp to K jp are G jp = DWj DWp − λ 2WjWp , H js = λ 2 RT Wj Θ p ( 1 + η ) , I jp = −Θ jWp DTb ,

1

J jp = DΘ j DΘ p − λ Θ j Θ p − PeΘ j DΘ p , K jp = Θ j Θ p , where AB =
2
∫ ABdz.
0
The above arrangement of equation (10.20) is a generalized eigenvalue condition and
resolved in MATLAB® exploiting QZ algorithm, EIG function and Newton’s routine. The
pattern of the convective activity is stationary for an investigated problem.

10.5 Outcomes and Discussion

The criterion for the start of convective progress in a flat Jeffrey fluid-soaked permeable
layer with a continuous upward throughflow is analyzed numerically applying the higher-
order Galerkin scheme.
The essential dimensionless parameters that direct the movement are the Rayleigh–
Darcy number RT , the Péclet number Pe, and the Jeffrey parameter η . The outcomes are
given in terms of computed critical estimates of RT with respect to the wavenumber λ for
diverse estimates of Péclet number Pe and the Jeffrey parameter η .
The rightness of our numerical results has been confirmed by contrasting the results
obtained under the subcase of this analysis with the results obtained by Yadav [20]. The
comparisons are also listed in Table 10.1. From this comparison, it is found that the out-
comes calculated from the current code are in outstanding accord with the available ones.
164 Mathematical Modeling for Intelligent Systems

TABLE 10.1
Contrast of RT , c and λc with Pe in the Nonexistence of η
Present Study Yadav [20]

Pe RT , c λc RT , c λc
0 39.478 3.14 39.478 3.142
0.5 39.826 3.15 39.827 3.151
1.0 40.873 3.18 40.873 3.179
1.5 42.621 3.23 42.621 3.225
2.0 45.071 3.29 45.071 3.292

FIGURE 10.2
The impact of Péclet number Pe on neutral stability curves at η = 0.4.

FIGURE 10.3
The effect of Jeffrey parameter η on neutral stability diagrams at Pe = 0.5.

Figures 10.2 and 10.3 show the neutral stability diagrams in the ( RT , λ )-plane for diverse
estimates of Péclet number Pe and the Jeffrey parameter η , respectively, using two-term
Galerkin method. The neutral stability graphs provide the threshold among the stability
and instability. The smallest of the neutral stability diagrams gives the critical assessments
of RT and λ which are indicated by RT ,c and λc, correspondingly. The physical significance
of the critical assessments is that the system will be stable for RT < RT ,c .
Numerical Treatment on the Convective Instability 165

FIGURE 10.4
Digression of R T , c with Pe for diverse estimates of η .

TABLE 10.2
Estimation of the RT , c and λc for Diverse Estimates of Pe and η
Pe η RD , c ac Pe η RD , c ac
0 0 39.478 3.14 1.5 0 42.621 3.23
0.2 32.899 3.14 0.2 35.518 3.23
0.4 28.199 3.14 0.4 30.444 3.23
0.6 24.674 3.14 0.6 26.638 3.23
0.8 21.932 3.14 0.8 23.678 3.23
1 19.739 3.14 1 21.311 3.23
0.5 0 39.827 3.15 2.0 0 45.071 3.29
0.2 33.189 3.15 0.2 37.559 3.29
0.4 28.448 3.15 0.4 32.194 3.29
0.6 24.892 3.15 0.6 28.169 3.29
0.8 22.126 3.15 0.8 25.039 3.29
1 19.913 3.15 1 22.536 3.29
1.0 0 40.873 3.18 2.5 0 48.220 3.38
0.2 34.061 3.18 0.2 40.183 3.38
0.4 29.195 3.18 0.4 34.443 3.38
0.6 25.546 3.18 0.6 30.137 3.38
0.8 22.707 3.18 0.8 26.789 3.38
1 20.437 3.18 1 24.110 3.38

Figure 10.4 shows the disparity of the critical Rayleigh–Darcy number R T ,c with a varia-
tion of Péclet number Pe for assorted estimates of the Jeffrey parameter η . The outcomes
are also listed in Table 10.2. The corresponding critical wave number λc is presented in
Figure 10.5. From Figure 10.4, it is established that on increasing the worth of Pe, the
assessment of R T ,c boosts. Thus, the effect of Pe is to postponement the start of convective
movement. This came about because the throughflow transports the imperative warmth
166 Mathematical Modeling for Intelligent Systems

FIGURE 10.5
Digression of λc with Pe for diverse values of η .

gradient into a thermal boundary layer at the boundary toward which it is implemented.
From Figure 10.4, we can also find that the stability of the system decreases with raising
the Jeffery parameter λ . This is because increasing value of Jeffery parameter η reduces the
retardation time of Jeffery fluid, thus making the system more unstable.
From Figure 10.5, it is found that the worth of λc enlarges with an increase in the values
of Pe. Thus, Pe reduces the magnitude of convective cells, whereas the Jeffery parameter η
has no major control on λc.

10.6 Conclusions
The influence of throughflow is analyzed on the convective movement in a Jeffrey fluid-
saturated permeable layer numerically. The following conclusions are made:
Increasing Péclet number Pe increases the critical Rayleigh–Darcy number R T ,c. Thus, it
postpones the onset of convective instability by increasing R T ,c.
The arrangement becomes more unstable with escalating Jeffrey parameter η .
The size of convective cell shrinks with Pe, while the Jeffrey parameter η has no effect
on it.

Acknowledgments
This study was supported by the University of Nizwa Research Grant (Grant No.: A/2021-
2022-UoN/3/CAS/IF), the Sultanate of Oman.
Numerical Treatment on the Convective Instability 167

References
1. Chandrasekhar S., Hydrodynamic and hydromagnetic stability. Dover Publication, New York, (2013).
2. Banzhaf, W., Self-organizing systems, in: R.A. Meyers (Ed.) Encyclopedia of physical science and
technology (Third Edition), Academic Press, New York (2003) 589–598.
3. Chandrasekhar, S., Thermal convection, Daedalus, 86(4) (1957) 323–339.
4. Kadanoff, L. P., Turbulent heat flow: structures and scaling, Physics Today, 54(8) (2001) 34–39.
5. Nield, D.A., Bejan, A., Convection in porous media, Springer (2006).
6. Capone, F., Gentile, M., Massa, G., The onset of thermal convection in anisotropic and rotat-
ing bidisperse porous media, Zeitschrift für angewandte Mathematik und Physik, New York, 72(4)
(2021) 1–16.
7. Xu, X., Chen, S., Zhang, D., Convective stability analysis of the long-term storage of carbon
dioxide in deep saline aquifers, Advances in Water Resources, 29(3) (2006) 397–407.
8. Pozzoni, P., Di Filippo, S., Manzoni, C., Locatelli, F., The relevance of convection in clinical
practice: a critical review of the literature, Hemodialysis International, 10 (2006) S33–S38.
9. Zhang, K., Liao, X., Schubert, G., Pore water convection within carbonaceous chondrite par-
ent bodies: Temperature-dependent viscosity and flow structure, Physics of Fluids, 17(8) (2005)
086602.
10. Kim, M. C., Yadav, D., Linear and nonlinear analyses of the onset of buoyancy-induced insta-
bility in an unbounded porous medium saturated by miscible fluids, Transport in Porous Media,
104(2) (2014) 407–433.
11. G. Higgins, A new concept for in situ coal gasification: Livermore, Ca, Univ. Ca., Lawrence
Livermore Lab., Rept. UCRL-51217, (1972).
12. Sherwood, A., Homsy, G. M., Convective instability during in situ coal gasification, Lawrence
Livermore Laboratory (1975).
13. Nield, D., Convective instability in porous media with throughflow, AIChE Journal, California,
33(7) (1987) 1222–1224.
14. Chen, F., Throughflow effects on convective instability in superposed fluid and porous layers,
Journal of Fluid Mechanics, 231 (1991) 113–133.
15. Nield, D., Kuznetsov, A., The onset of convection in a layered porous medium with vertical
throughflow, Transport in Porous Media, 98(2) (2013) 363–376.
16. Hill, A. A., Unconditional nonlinear stability for convection in a porous medium with vertical
throughflow, Acta Mechanica, 193(3) (2007) 197–206.
17. Hill, A. A., Rionero, S., Straughan, B., Global stability for penetrative convection with through-
flow in a porous material, IMA Journal of Applied Mathematics, 72(5) (2007) 635–643.
18. Yadav, D. The influence of pulsating throughflow on the onset of electro-thermo-convection
in a horizontal porous medium saturated by a dielectric nanofluid, Journal of Applied Fluid
Mechanics, 11(6) (2018) 1679–1689.
19. Yadav, D., Throughflow and magnetic field effects on the onset of convection in a Hele Shaw
cell, Revista Cubana de Física, 35(2) (2018) 108–114.
20. Yadav, D., Numerical investigation of the combined impact of variable gravity field and
throughflow on the onset of convective motion in a porous medium layer, International
Communications in Heat and Mass Transfer, 108 (2019) 104274.
21. Yadav, D., The effect of pulsating throughflow on the onset of magneto convection in a layer
of nanofluid confined within a Hele-Shaw cell, Proceedings of the Institution of Mechanical
Engineers, Part E: Journal of Process Mechanical Engineering, 233(5) (2019) 1–12.
22. Shankar, B., Shivakumara, I., Stability of porous-Poiseuille flow with uniform vertical through-
flow: High accurate solution, Physics of Fluids, 32(4) (2020) 044101.
23. Shankar, B., Shivakumara, I., Changes in the hydrodynamic stability of plane porous-Couette
flow due to vertical throughflow, Physics of Fluids, 33(7) (2021) 074103.
168 Mathematical Modeling for Intelligent Systems

24. Shankar, B., Shivakumara, I., Instability of natural convection in a vertical fluid layer with net
horizontal throughflow, Zeitschrift für angewandte Mathematik und Physik, 72(3) (2021) 1–17.
25. Brevdo, L., M.S. Ruderman, On the convection in a porous medium with inclined temperature
gradient and vertical throughflow. Part I. Normal modes, Transport in Porous Media, 80(1) (2009)
137–151.
26. Kiran, P., Nonlinear throughflow and internal heating effects on vibrating porous medium,
Alexandria Engineering Journal, 55(2) (2016) 757–767.
27. Ahmad, K., Hanouf, Z., Ishak, A., Mixed convection Jeffrey fluid flow over an exponentially
stretching sheet with magnetohydrodynamic effect, AIP Advances, 6(3) (2016) 035024.
28. Bhatti, M., Abbas, A., Simultaneous effects of slip and MHD on peristaltic blood flow of Jeffrey
fluid model through a porous medium, Alexandria Engineering Journal, 55(2) (2016) 1017–1023.
29. Hayat, T., Qayyum, S., Imtiaz, M., Alsaedi, A., Impact of Cattaneo-Christov heat flux in Jeffrey
fluid flow with homogeneous-heterogeneous reactions, PLoS One, 11(2) (2016) e0148662.
30. Javed, M., Farooq, M., Anjum, A., Ahmad, S., Insight of thermally stratified Jeffrey fluid flow
inside porous medium subject to chemical species and melting heat transfer, Advances in
Mechanical Engineering, 11(9) (2019) 1687814019876187.
31. Kahshan, M., Lu, D., Siddiqui, A., A Jeffrey fluid model for a porous-walled channel: Application
to flat plate dialyzer, Scientific Reports, 9(1) (2019) 1–18.
32. Krishna, M. V., Bharathi, K., Chamkha, A. J., Hall effects on MHD peristaltic flow of Jeffrey
fluid through porous medium in a vertical stratum, Interfacial Phenomena and Heat Transfer, 6(3)
(2018) 253–268.
33. Naganthran, K., Nazar, R., Pop, I., Effects of heat generation/absorption in the Jeffrey fluid past
a permeable stretching/shrinking disc, Journal of the Brazilian Society of Mechanical Sciences and
Engineering, 41(10) (2019) 1–12.
34. Nallapu, S., Radhakrishnamacharya, G., Jeffrey fluid flow through porous medium in the
presence of magnetic field in narrow tubes, International Journal of Engineering Mathematics,
2014 (2014) 713831.
35. Narayana, P. S., Babu, D. H., Numerical study of MHD heat and mass transfer of a Jeffrey fluid
over a stretching sheet with chemical reaction and thermal radiation, Journal of the Taiwan
Institute of Chemical Engineers, 59 (2016) 18–25.
36. Noor, N A., Shafie, M. S., Admon, M. A., Unsteady MHD squeezing flow of Jeffrey fluid in a
porous medium with thermal radiation, heat generation/absorption and chemical reaction,
Physica Scripta, 95(10) (2020) 105213.
37. Ojjela, O., Raju, A., Kumar, N. N., Influence of induced magnetic field and radiation on free
convective Jeffrey fluid flow between two parallel porous plates with Soret and Dufour effects,
Journal of Mechanics, 35(5) (2019) 657–675.
38. Santhosh, N., Radhakrishnamacharya, G., Chamkha, A. J., Flow of a Jeffrey fluid through a
porous medium in narrow tubes, J. Porous Media, 18(1) (2015) 71–78.
39. Vajravelu, K., Sreenadh, S., Lakshminarayana, P., The influence of heat transfer on peristaltic
transport of a Jeffrey fluid in a vertical porous stratum, Communications in Nonlinear Science and
Numerical Simulation, 16(8) (2011) 3107–3125.
40. Nazeer, M., Hussain, F., Ahmad, M. O., Saeed, S., Khan, M. I., Kadry, S., Chu, Y.-M., Multi-phase
flow of Jeffrey Fluid bounded within magnetized horizontal surface, Surfaces and Interfaces, 22
(2021) 100846.
41. Khan, A., Rafaqat, A. R., Effects of radiation and MHD on compressible Jeffrey fluid with peri-
stalsis, Journal of Thermal Analysis and Calorimetry, 143(3) (2021) 2775–2787.
42. Mehboob, H., Maqbool, K., Ellahi, R., Sait, S. M., Transport of Jeffrey fluid in a rectangular
slit of the microchannel under the effect of uniform reabsorption and a porous medium,
Communications in Theoretical Physics, 73(11) (2021) 115003.
43. Ali, F., Hou, Y., Zahid, M., Rana, M. A., Usman, M., Influence of magnetohydrodynamics and
heat transfer on the reverse roll coating of a Jeffrey fluid: A theoretical study, Journal of Plastic
Film & Sheeting, 38(1) (2022) 72–104.
Numerical Treatment on the Convective Instability 169

44. Gautam, P. K., Rana, G. C., Saxena, H., Stationary convection in the electrohydrodynamic ther-
mal instability of Jeffrey nanofluid layer saturating a porous medium: free-free, rigid-free, and
rigid-rigid boundary conditions, Journal of Porous Media, 23(11) (2020) 1043–1063.
45. Yadav, D., Thermal non-equilibrium effects on the instability mechanism in a non-Newtonian
Jeffery fluid saturated porous layer, Journal of Porous Media, 25(2) (2022) 1–12.
46. Yadav, D., Influence of anisotropy on the Jeffrey fluid convection in a horizontal rotary porous
layer, Heat Transfer, 50(5) (2021) 4595–4606.
47. Yadav, D., Electric field effect on the onset of Jeffery fluid convection in a heat generating
porous medium layer, Pramana - Journal of Physics, 96, 19, (2022).
48. Yadav, D., Mohamad, A. A., Awasthi, M. K., The Horton–Rogers–Lapwood problem in a Jeffrey
fluid influenced by a vertical magnetic field, Proceedings of the Institution of Mechanical Engineers,
Part E: Journal of Process Mechanical Engineering, 235(6) (2021) 2119–2128.
49. Akbarzadeh, P., Mahian, O., The onset of nanofluid natural convection inside a porous layer
with rough boundaries, Journal of Molecular Liquids, 272 (2018) 344–352.
50. Alex, S. M., Patil, P. R., Effect of variable gravity field on Soret driven thermosolutal convec-
tion in a porous medium, International Communications in Heat and Mass Transfer, 28(4) (2001)
509–518.
51. Awad, F., Sibanda, P., Motsa, S. S., On the linear stability analysis of a Maxwell fluid with
double-diffusive convection, Applied Mathematical Modelling, 34(11) (2010) 3509–3517.
52. Banu, N., Rees, D., Onset of Darcy–Benard convection using a thermal non-equilibrium model,
International Journal of Heat and Mass Transfer, 45(11) (2002) 2221–2228.
53. Capone, F., Gentile, M., Hill, A. A., Penetrative convection via internal heating in anisotropic
porous media, Mechanics Research Communications, 37(5) (2010) 441–444.
54. Yadav, D., Mohamad, A., Rana, G., Effect of throughflow on the convective instabilities in an
anisotropic porous medium layer with inconstant gravity, Journal of Applied and Computational
Mechanics, 7(4) (2020) 1964–1972.
55. Yadav, D., Effects of rotation and varying gravity on the onset of convection in a porous
medium layer: A numerical study, World Journal of Engineering, 17(6) (2020) 785–793.
56. Yadav, D., Maqhusi, M., Influence of temperature dependent viscosity and internal heating
on the onset of convection in porous enclosures saturated with viscoelastic fluid, Asia‐Pacific
Journal of Chemical Engineering, 15(6) (2020) e2514.
57. Yadav, D., Numerical solution of the onset of Buoyancy‐driven nanofluid convective motion
in an anisotropic porous medium layer with variable gravity and internal heating, Heat
Transfer—Asian Research, 49(3) (2020) 1170–1191.
58. Yadav, D., The density-driven nanofluid convection in an anisotropic porous medium layer
with rotation and variable gravity field: A numerical investigation, Journal of Applied and
Computational Mechanics, 6(3) (2020) 699–712.
11
Computational Modeling of Nonlinear
Reaction-Diffusion Fisher–KPP Equation
with Mixed Modal Discontinuous Galerkin Scheme

Satyvir Singh
Nanyang Technological University

CONTENTS
11.1 Introduction ........................................................................................................................ 171
11.2 Mixed Modal Discontinuous Galerkin Scheme ............................................................ 173
11.3 Numerical Results.............................................................................................................. 176
11.4 Concluding Remarks ......................................................................................................... 181
Acknowledgements .................................................................................................................... 182
References..................................................................................................................................... 182

11.1 Introduction
It is constantly difficult for researchers to simulate nonlinear partial differential equations
(PDEs) since these nonlinear PDEs are utilized to solve several challenging problems that
occurred in science, nature, engineering, and other fields. For example, Cahn–Hilliard PDE
in material sciences; nonlinear Schrödinger PDE sinquantum mechanics; nonlinear reac-
tion–diffusion PDEs in biology; Euler and Navier–Stokes PDEs in fluid dynamics (Zheng,
2004), Schrödinger/Gross–Pitaevskii and Klein–Gordon–Zakharov PDEs (Dehghan and
Abbaszadeh, 2018), generalized Zakharov and Gross–Pitaevskii PDEs (Dehghan and
Abbaszadeh, 2017) are several illustrious examples of the nonlinear PDEs. Many scholars
have spent decades trying to find effective numerical methods for solving nonlinear PDEs.
The purpose of this chapter is to describe an efficient numerical technique based on modal
discontinuous Galerkin (DG) method for simulating nonlinear reaction–diffusion PDEs in
two dimensions.
A nonlinear reaction–diffusion model was introduced by Fisher (1937) and Kolmogorov
(1937) autonomously to delineate the propagation of a mutant gene across a population.
This model is commonly referred to as the Fisher or Fisher–Kolmogorov–Petrovsky–
Piscounov (Fisher–KPP) equation in the literature. This Fisher–KPP equation has since
been applied in a variety of fields of research, most notably biological applications. The
Fisher–KPP equation, for example, can be used to represent the logistic growth of coral

DOI: 10.1201/9781003291916-11 171

172 Mathematical Modeling for Intelligent Systems

reefs (Roessler and Hüssner, 1997) and the stirrings of neutrons in a nuclear reactor (José,
1969). The mathematical expression of the Fisher–KPP equation is given by

∂w
= ∇ ⋅ ( λ∇w ) + µ f ( w ) , (11.1)
∂T

where w ≡ w ( X , Y ,T ) denotes the two-dimensional unknown population density, λ is the

population diffusivity, µ is the population proliferation rate and ∇ 2 ≡ ∂2 ∂X 2 + ∂2 ∂Y 2
presents the two-dimensional gradient operator. The function f ( w ) = w ( 1 − w ) is a non-
linear term that represents the effects of reaction or multiplication. Assuming λ1 and λ2 as
the paramount values of λ with ( X , Y ) coordinates along the principal axes, the Fisher–KPP
equation (11.1) is expressed as follows:

∂w
= λ1 wXX + λ2 wYY + µ w ( 1 − w ) . (11.2)
∂T

On rescaling X , Y and T as

µ µ
t = µT , x= X , and y = Y, (11.3)
λ1 λ2

Equation (11.2) becomes

∂w
= λ1 wxx + λ2 wyy + µ w ( 1 − w ) , (11.4)
∂t

which is a general form of two-dimensional Fisher–KPP equation.

In literature, several analytical and numerical approaches exist for solving the Fisher–
KPP equation. In 1974, the Fisher–KPP equation was examined numerically for the first
time using a pseudo-spectral scheme (Gazdag and Canosa, 1974) after which it was inves-
tigated by a variety of additional numerical schemes, including Petrov–Galerkin finite ele-
ment method (Tang and Weber, 1991), tanh method (Wazwaz, 2004), Homotopy analysis
method (Tan et al., 2007), alternating group explicit iterative method (Evans and Sahimi,
1989), hybrid approach of homotopic method and theory of traveling wave transform
(Mo et al., 2007), centered finite-difference method (Hagstrom and Keller, 1986), implicit
and explicit finite-differences methods(Parekh and Puri, 1990), wavelet Galerkin method
(Mittal and Kumar, 2006), collocation of cubic B-splines basis (Mittal and Arora, 2010), dif-
ferential quadrature method (Arora and Joshi, 2021).
In this study, a mixed modal DG method is proposed for simulating the two-dimensional
Fisher–KPP equation. The DG approaches for solving nonlinear PDEs have been increas-
ingly popular in recent decades. Initially, Reed and Hill (1973) proposed the DG method
to simulate the unsteady neutron transport equation. Afterward, several DG algorithms
were proposed and developed numerically for solving nonlinear hyperbolic systems by
Cockburn and Shu (1989a, 1989b, 1998, 2001). The DG approaches integrate the advantages
of modern CFD methods, including Finite Element (FE), and Finite Volume (FV) methods
and have been effectively utilized to a broad spectrum of scientific problems, including
computational fluid dynamics, plasma physics, quantum physics, biological sciences, and
many others (Le et al. 2014; Raj et al., 2017; Singh and Myong, 2017; Singh, 2018; Chourushi
et al., 2020; Singh, 2020; Singh and Battiato, 2020a; Singh and Battiato, 2020b; Singh and
Computational Modeling 173

Battiato, 2021a; Singh and Battiato, 2021b; Singh et al., 2021; Singh, 2021a; Singh, 2021b;
Singh, 2021c). The DG approaches have several key characteristics that make them interest-
ing for usage in applications. These characteristics include their capacity to easily address
complex geometry and boundary conditions, their flexibility for easy hp-adaptivity, their
ability for nonconforming elements having hanging nodes, and efficient parallel imple-
mentation along with time-stepping algorithms.
This study presents a two-dimensional mixed modal DG scheme different from previous
studies for solving the nonlinear Fisher–KPP equations. The third-order scaled Legendre
polynomials are adopted for DG spatial discretization, while a third-order Strong Stability
Preserving Runge–Kutta (SSP-RK33) scheme is employed for a temporal marching algo-
rithm is utilized to temporally discretize the emerging semi-discrete differential equation.
The accuracy and efficiency of the proposed algorithm are evaluated by solving four well-
known Fisher–KPP test problems.

11.2 Mixed Modal Discontinuous Galerkin Scheme

Here, a mixed modal DG scheme is described for solving the nonlinear Fisher–KPP equa-
tion. Thus, in order to employ this numerical scheme, equation (11.4) can be reformulated as

∂w
+ ∇F ( ∇w ) = S ( w ) , (11.5)
∂t

where F ( ∇w ) = λ1 wx + λ2 wy , and S ( w ) = µ w ( 1 − w ). For the spatial discretization of such

type of nonlinear equation that contains second-order derivatives in diffusion part, a
mixed DG formulation was developed previously in literature. An auxiliary variable Θ is
provided in this formulation to address the higher-order derivatives which are considered
as the derivative of solution variable. Thus, the Fisher–KPP equation (11.5) is rewritten as a
coupled system of w and Θ for the mixed DG construction,

w − ∇Θ = 0,

∂w (11.6)
+ ∇F ( Θ ) = S ( w ) .
∂t

In this work, the computational simulation of Fisher–KPP equation is considered in one

and two-dimensional spaces. Recently, a mixed modal DG scheme for solving one-dimen-
sional nonlinear Fisher–KPP equation has been developed (Singh, 2021c; Singh, 2021d). A
two-dimensional Fisher–KPP equation is here computed by an explicit mixed modal DG
approach based on rectangular elements. A rectangular element based on the tensor prod-
uct of mesh points in x- and y-directions has been used to discretize the coupled system of
equation (11.6). Let [ ℑ] be a family of partitions of the physical domain Ω = Ω x × Ω y which
is divided into N x × N y regular rectangular elements as

xL = x1/2 < x3/2 < ⋅ ⋅ ⋅ ⋅ ⋅ < xi−1/2 < xi+1/2 < ⋅ ⋅ ⋅ ⋅ ⋅ < xN +1/2 = xR ,
(11.7)
y B = y1/2 < y 3/2 < ⋅ ⋅ ⋅ ⋅ ⋅ < y j−1/2 < y j+1/2 < ⋅ ⋅ ⋅ ⋅ ⋅ < y N +1/2 = yU .
174 Mathematical Modeling for Intelligent Systems

A two-dimensional Cartesian mesh ℑ is defined as

{
ℑ := Tij = I i × J j , 1 ≤ i ≤ N x , 1 ≤ j ≤ N y , }
I i := [ xi−1/2 , xi+1/2 ] , ∀i = 1 ⋅ ⋅ ⋅ ⋅ ⋅ N x ; (11.8)

J j :=  y j−1/2 , y j+1/2  , ∀j = 1 ⋅ ⋅ ⋅ ⋅ ⋅ N y .

( ) ( ( ) )
where xi , y j = ( xi−1/2 + xi+1/2 ) 2 , y j−1/2 + y j+1/2 2 denotes the cell-center points of the
elements. In the case of Ω x − domain, the piecewise polynomial space of the functions
ξ : Ω x → ℜ is defined as

{
Zhk = ξ : ξ |Ωx ∈  k ( I i ) , i = 1 ⋅ ⋅ ⋅ ⋅ ⋅ N x , } (11.9)

where  k ( I i ) is the space of polynomial functions of degree at most k on I i . For the domain
Ω = Ω x × Ω y , it is defined ϕ : Ω → ℜ as

{ ( )
Vhk = ϕ : ϕ |Ω ∈  k Tij , 1 ≤ i ≤ N x , 1 ≤ j ≤ N y , } (11.10)

( ) ( )
where F k Tij = S k ( I i ) ⊗ S k J j denotes the space of polynomials of degree at most k on Tij .
( )
The number of degrees of freedom of  k Tij can be calculated by N k = ( k + 1) ( k + 2 ) / 2.
After then, the exact solutions of w and Θ can be approximated by the DG polynomial
approximations of wh ∈ Vhk and Θ h ∈ Vhk as

Nk

Θh ( x, y ) = ∑ Θ (t ) b ( x, y ) ,
i=0
i
h i

(11.11)
Nk

wh ( x, y , t ) = ∑ w (t ) b ( x, y ) ,
i=0
i
h i ∀ ( x , y ) ∈Tij ,

where whi denotes the unknown coefficients for w to be corrected with time, and bi ( x , y )
represents the polynomial (basis) function. The two-dimensional scaled Legendre polyno-
mial functions are formulated here as a tensor product of the so-called principal functions,
which are defined as (Singh, 2018)

bk (ξ , η ) = ψ i (ξ ) ⊗ ψ j (η ) , (11.12)

with

2 i ( i !) 0,0
2
ψ i (ξ ) = P (ξ ) , − 1 ≤ ξ ≤ 1,
( 2 i )!
2 i ( j !)
2

ψ j (η ) = P 0,0 (η ) , − 1 ≤ η ≤ 1,
( 2 j )!
Computational Modeling 175

FIGURE 11.1
Elemental transformation from physical element to reference element in a 2D computational domain.

where P 0,0 (ξ ) is the Legendre polynomial function. A standard rectangular element Ωest
is defined using a local Cartesian coordinate system (ξ , η ) ∈[ −1,1], which is illustrated in
Figure 11.1. The standard element can be mapped from the computational space (ξ , η ) to an
arbitrary rectangular element in the physical space ( x , y ) under the linear transformation
T : Ωest → Ωe defined by

1
x= ( 1 − ξ ) x1 + ( 1 + ξ ) x2  ,
2
(11.13)
1
y = ( 1 − η ) y1 + ( 1 + η ) y 2  ,
2

where ( xi , y i : i = 1, 4 ) denotes the physical coordinates of the vertices of Ωe .

The DG discretization of the coupled system of Fisher–KPP equation (11.6) can be
obtained by replacing the exact solutions with the corresponding approximations defined
in equation (11.11), and multiplying by a polynomial function bh ( x , y ) and then integrated
by parts over the element Tij ∈ℑ. Taking bh ∈ Vhk ( ℑ), wh ∈ Vhk ( ℑ), and Θ h ∈Vhk ( ℑ) , we obtain

∫ Θ b dV + ∫ ∇b ⋅ w dV − ∫ b w
Ωe
h h
Ωe
h h
∂Ωe
h h ⋅ ndΓ = 0,

(11.14)
∂
∂t ∫ w b dV − ∫ ∇b ⋅ F (Θ ) dV + ∫ b F (Θ ) ⋅ ndΓ = ∫ b S( w ) dV.
Ωe
h h
Ωe
h h
∂Ωe
h h
Ωe
h h

where n denotes the outward unit normal vector; Γ and V are the boundary and volume of
the element Ωe, respectively. The interface fluxes are not uniquely defined because of the
discontinuity in the solution wh and Θ h at the elemental interfaces. The functions F ( Θ h ) ⋅ n
and wh ⋅ n emerging in equation (11.14) can be substituted by the numerical fluxes at the
elemental interfaces denoted by H BR1 and H aux , respectively. Here, the central flux or Bassi
Rebay (BR1) scheme is employed for the viscous and auxiliary numerical fluxes to calcu-
late the flux at elemental interfaces.
176 Mathematical Modeling for Intelligent Systems

1 −
(
wh ⋅ n ≡ H aux wh− , wh+ = )  wh + wh+  ,
2
(11.15)
1
(
F ( Θ h ) ⋅ n ≡ H BR1 Θ −h ,Θ +h = ) 2
( ) ( )
F Θ −h + F Θ +h  .

Here, the superscripts ( − ) and ( + ) are the left and right states of the elemental interface. As
a result, the DG weak formulation is obtained as

∫ Θ b dV + ∫ ∇b ⋅ w dV − ∫ b H
Ωe
h h
Ωe
h h
∂Ωe
h
aux
d Γ = 0,

(11.16)
∂
∂t ∫ w b dV − ∫ ∇b ⋅ F (Θ ) dV + ∫ b H
Ωe
h h
Ωe
h h
∂Ωe
h
BR1
dΓ =
∫ b S( w ) dV.
Ωe
h h

In the expression equation (11.16), the emerging volume and surface integrals are approxi-
mated by the Gaussian–Legendre quadrature rule within the elements to ensure the
high-order accuracy (Cockburn and Shu, 2001). Furthermore, a nonlinear TVB limiter
(Cockburn and Shu, 1998) is utilized to remove the artificial oscillations arising in the
numerical solution.
Finally, the DG spatial discretization equation (11.16) can be expressed in semi-discrete
ordinary differential equations form as

dwh
M = R ( wh ) , (11.17)
dt

where M and R ( wh ) are the orthogonal mass matrix and the residual function, respec-
tively. Here, an explicit form of Strongly Stability Preserving Runge–Kutta method with
third-order accuracy (Shu and Osher, 1988) is adopted as the time marching scheme.

h = w h + ∆t M R ( w h ) ,
−1
w(1) n

w(2)
h =
3 n 1 (1) 1
4
wh + wh + ∆t M −1 R w(1)
4 4
h , ( ) (11.18)

whn+1 =
1 n 2 (2) 2
3
wh + wh + ∆t M −1 R w(h2 ) ,
3 3
( )
( )
where R whn is the residual approximation at time tn , and ∆t is the suitable time step value.

11.3 Numerical Results
In this section, four test problems of Fisher–KPP equations are examined by simulating
numerical results with the proposed mixed modal DG scheme.
Computational Modeling 177

PROBLEM 11.1
For testing the accuracy of the present DG scheme, we consider a one-dimensional Fisher’s
KPP problem (Wazwaz and Gorguis, 2004)

∂w ∂2 w
= λ1 2 + µ w ( 1 − w ) , (11.19)
∂t ∂x

subject to the initial condition and boundary conditions

1
w ( x, 0 ) = ,
(1 + e )x 2
(11.20)
lim w ( x , t ) = 1.0, lim w ( x , t ) = 0.
x→−∞ x→∞

The analytical solution of the standard form of Fisher–KPP equation is given by

−2
 µ 5
x − µt 
w ( x, t ) = 1 + e 6 6
 . (11.21)
 

The numerical solution of one-dimensional Fisher–KPP problem equation (11.19) has been
obtained for λ1 = 1.0 with N = 200 elements and ∆t = 0.0001. Figure 11.2 illustrates the
numerical and exact solutions at different time instants t = 0, 0.2, 0.4, 0.6, 0.8 and 1. The
computed results are depicted graphically in such a way that the numerical result can be
compared with the analytical solutions. The provided solutions are found in good match
with the analytical solutions. In addition, Table 11.1 shows the comparison between com-
puted DG and analytical solutions, and the maximum error (L∞ ) at two different time
instants t = 0.5, and 1.

FIGURE 11.2
Profiles and space-time plots for one-dimensional Fisher–KPP Problem 1 at early time instants. Arrow symbol
represents the direction of traveling wave solutions.
178 Mathematical Modeling for Intelligent Systems

TABLE 11.1
Comparison of Exact and Present Numerical Solutions of Problem 1 at Two Different Time
Level st = 0.5 and 1.0.
t = 0.5 t = 1.0
x Present Exact Error Present Exact Error
-10 0.10000E+1 0.99999E+0 0.42469E-4 0.10001E+1 0.99999E+0 0.49076E-4
−8 0.99999E+0 0.99994E+0 0.47085E-4 0.10001E+1 0.99999E+0 0.49529E-4
−6 0.99962E+0 0.99957E+0 0.46834E-4 0.10000E+1 0.99996E+0 0.49533E-4
−4 0.99689E+0 0.99681E+0 0.44718E-4 0.99979E+0 0.99974E+0 0.49400E-4
−2 0.97708E+0 0.97704E+0 0.36362E-4 0.99814E+0 0.99809E+0 0.49064E-4
0 0.85748E+0 0.85744E+0 0.47593E-4 0.98604E+0 0.98599E+0 0.50665E-4
2 0.40232E+0 0.40213E+0 0.18926E-3 0.91172E+0 0.91165E+0 0.70946E-4
4 0.36165E-1 0.36101E-1 0.64248E-4 0.54906E+0 0.54889E+0 0.17092E-3
6 0.94896E-3 0.94672E-3 0.22402E-5 0.77927E-1 0.77833E-1 0.93739E-4
8 0.18347E-4 0.18301E-4 0.46535E-7 0.24806E-2 0.24761E-2 0.45552E-5
10 0.54037E-6 0.27650E-6 0.26387E-7 0.80924E-4 0.40598E-4 0.40325E-4

PROBLEM 11.2
Consider a one-dimensional Fisher’s KPP equation (Tang and Weber, 1991)

∂w ∂2 w
= λ1 2 + µ w ( 1 − w ) , (11.22)
∂t ∂x

with the parameters λ1 = 1, µ = 1 and the initial conditions are provided by

w ( x , 0 ) = e 10( x+1) , if x < −1

w ( x , 0 ) = 1, if − 1 ≤ x ≤ −1 (11.23)

w ( x , 0 ) = e −10( x−1) , if x > 1.

The parameters are selected as ∆t = 0.01 and ∆x = 0.01. Figures 11.3–11.5 illustrate the
numerical solutions at various time instants. In Figure 11.3, the numerical solutions are
drawn for Ω x = [ −3, 3 ] at t = 0.1, 0.2, 0.3, 0.4 and 0.5, while in Figure 11.4, is shown for
Ω x = [ −6,6 ] at time instants t = 1, 2, 3, 4 and 5. Although, in general, the influence of reac-
tion–diffusion is minor early in the simulation, reaction is more dominant than diffusion
because there is a change from sharpness to smoothness near x = ±1 in the solutions. In
Figure 11.5, solution profiles across the interval Ω x = [ −30, 30 ] at time t = 0 − 35 are illus-
trated, where the numerical solution grows flatter because of diffusion impacts.

PROBLEM 11.3
Next, a two-dimensional Fisher’s KPP based test problem is considered.

∂w ∂2 w ∂2 w
= λ1 2 + λ 2 + µw (1 − w ) , (11.24)
∂t ∂x ∂y 2
Computational Modeling 179

FIGURE 11.3
Profiles and space-time plots for one-dimensional Fisher–KPP Problem 2 at early time instants. Arrow symbol
represents the direction of traveling wave solutions.

FIGURE 11.4
Profiles and space-time plots for one-dimensional Fisher–KPP Problem 2 at middle time instants. Arrow sym-
bol represents the direction of traveling wave solutions.

FIGURE 11.5
Profiles and space-time plots for one-dimensional Fisher–KPP Problem 2 at longer time instants. Arrow symbol
represents the direction of traveling wave solutions.
180 Mathematical Modeling for Intelligent Systems

FIGURE 11.6
Surface plots of numerical solutions for two-dimensional Fisher–KPP Problem 3 at various time instants t = 0,
1, 2, 3, 4 and 5.

with the parameters λ1 = 0.1, λ2 = 0.01, µ = 0.1, Ω = [ −2 , 2 ] × [ −1,1] and the initial condi-
tions are as follows:

w ( x , y , 0 ) = 1, if x 2 + 4 y 2 ≤ 0.25
(11.25)
w ( x, y , 0) = e
(
−10 x 2 + 4 y 2 − 0.25 ) , otherwise.

Figure 11.6 shows the surface plots of the numerical solution at time 0, 1, 2, 3, 4, and 5. This
example demonstrates that as time passes, the spreading disturbance causes the top to fall
first, although the initial disturbance has a flat top in the middle with zero diffusion and a
significant diffusion at the edge.

PROBLEM 11.4
Finally, a two-dimensional Fisher’s KPP equation is considered as (Tang et al. 1993)

∂w ∂2 w ∂2 w
= λ1 2 + λ 2 + µw (1 − w ) , (11.26)
∂t ∂x ∂y 2

with the parameters λ1 = 0.02, λ1 = 0.01, µ = 1, Ω = [ −10,10 ] × [ −8, 8 ] and the initial condi-
tions are as following:

w ( x, y , 0) = x 2 + 4y 2 , if x 2 + 4 y 2 ≤ 1 and x ≥ 0

w ( x , y , 0 ) = 1, if 1 ≤ x 2 + 4y 2 ≤ 2.25 and x ≥ 0
(11.27)
w ( x, y , 0) = e
(
−10 x 2 + 4 y 2 − 2.25 ) , if x 2 + 4y 2 ≥ 2.25 and x ≥ 0

w ( x , y , 0 ) = 0, if x > 0.
Computational Modeling 181

FIGURE 11.7
Surface plots of numerical solutions for two-dimensional Fisher–KPP Problem 4 at various time instants t = 0,
2, 5, 10, 20 and 30.

The parameters are selected to depict the concave and convex distributions across a vast
area. Figure 11.7 illustrates the surface plots of the numerical solutions at several time
instants t = 0, 2, 5, 10, 20 and 30. It can be observed that as time passes, the concave part
of w and its isopleths will become convex. The two-dimensional quasi-traveling waves are
obtained as shown and discussed by Tang and Weber (1991).

11.4 Concluding Remarks
The aim of this chapter is to find approximate solutions to the nonlinear Fisher–KPP reac-
tion-diffusion equation emerging in biology. A mixed modal DG scheme is developed to
simulate the two-dimensional Fisher–KPP equation. This numerical approach is based
on the concept of addressing an additional auxiliary unknown in the high-order deriv-
ative diffusion term. The scaled Legendre polynomials with third-order accuracy are
used for spatial discretization, while, a third-order Strongly-Stability-Preserving Runge-
Kutta (SSP-RK33) scheme is used for temporal discretization. This numerical approach
is widely applicable for several nonlinear reaction–diffusion problems. To verify the
accuracy and reliability of the DG scheme, four well-known one-dimensional and two-
dimensional numerical problems in literature are solved. The derived numerical solu-
tions and errors show that the results are in good match with the analytical solutions. In
the case of two-dimensional work, surface plots are also shown, which is identical to the
results reported graphically in the literature. This proposed DG approach demonstrates
that it is an efficient scheme to find numerical solutions in a variety of linear and non-
linear physical models. It may thus be extended to tackle higher-dimensional problems
efficiently.
182 Mathematical Modeling for Intelligent Systems

Acknowledgements
The author gratefully acknowledges the financial support provided by the Nanyang
Technological University, Singapore through the NAP-SUG grant program.

References
Arora, G. & Joshi, V. (2021). A computational approach for one and two dimensional Fisher’s equa-
tion using quadrature technique. American Journal of Mathematical and Management Sciences,
40(2), 145–162.
Chourushi, T., Rahimi, A., Singh, S., & Myong, R.S. (2020). Computational simulation of near con-
tinuum external gas flows using the Navier-Stokes-Fourier equations with slip/jump condi-
tions based on a modal discontinuous Galerkin method, Advances in Aerodynamics, 2(8), 1–37.
Cockburn, B., & Shu, C.-W. (1989a). TVB Runge-Kutta local projection discontinuous Galerkin finite
element method for conservation laws II, General framework. Mathematics of Computation, 52,
411–435.
Cockburn, B., & Shu, C.-W. (1989b). TVB Runge-Kutta local projection discontinuous Galerkin finite
element method for conservation laws III: one-dimensional systems. Journal of Computational
Physics, 84, 90–113
Cockburn, B., & Shu, C.-W. (1998). The Runge-Kutta discontinuous Galerkin method for conserva-
tion laws V: Multidimensional systems. Journal of Computational Physics, 141(2), 199–224.
Cockburn, B., & Shu, C.-W. (2001). Runge–Kutta discontinuous Galerkin methods for convection-
dominated problems. Journal of Scientific Computing, 16(3), 173–261.
Dehghan, M. & Abbaszadeh, M. (2017). Numerical investigation based on direct meshless local Petrov
Galerkin (direct MLPG) method for solving generalized Zakharov system in one and two
dimensions and generalized Gross-Pitaevskii equation. Engineering Computations, 34(4), 983–996.
Dehghan, M. & Abbaszadeh, M. (2018). Solution of multi-dimensional Klein-Gordon-Zakharov and
Schrödinger/Gross-Pitaevskii equations via local radial basis functions-differential quadra-
ture (RBF-DQ) technique on non-rectangular computational domains. Engineering Analysis
with Bundary Elements, 92, 156–170.
Evans, D.J., & Sahimi, M.S. (1989). The alternating group explicit iterative method to solve parabolic
and hyperbolic partial differential equations. Annual Review of Numerical FluidMechanics and
Heat Transfer, 2, 283–389.
Fisher, R.A. (1937). The wave of advance of advantageous genes. Annals of Eugenics, 7(4), 355–369.
Gazdag, J., & Canosa, J. (1974). Numerical solution of Fisher’s equation. Journal of Applied Probability,
11(3), 445–457.
Hagstrom, T., & Keller, H.B. (1986). The numerical calculation of travelling wave solutions of non-
linear parabolic equations. SIAM Journal on Scientific and Statistical Computing, 7(3), 978–988.
José, C. (1969). Diffusion in nonlinear multiplicative media. Journal of Mathematical Physics, 10(10),
1862–1868.
Kolmogorov, A.N. (1937). Étude de l‘équation de la diffusion avec croissance de la quantité de
matière et son application à un problème biologique. Moskow University Bulletin, International
Series, Section A, 1, 1–25.
Le, N.T.P., Xiao, H., & Myong, R.S. (2014). A triangular discontinuous Galerkin method for non-New-
tonian implicit constitutive models of rarefied and microscale gases. Journal of Computational
Physics, 273, 160–184.
Mittal, R.C. & Kumar, S. (2006). Numerical study of Fisher’s equation by wavelet Galerkin method.
International Journal of Computer Mathematics 83(3), 287–298.
Computational Modeling 183

Mittal, R.C., & Arora, G. (2010). Efficient numerical solution of Fisher’s equation by using B-spline
method. International Journal of Computer Mathematics, 87(13), 3039–3051.
Mo, J. Q., Zhang, W. J., & He, M. (2007). Asymptotic method of travelling wave solutions for aclass of
nonlinear reaction diffusion equation. Acta Mathematica Scientia, 27(4), 777–780.
Parekh, N., & Puri, S. (1990). A new numerical scheme for the Fisher’s equation. Journal of Physics A:
Mathematical and General, 23, 1085–1091.
Raj, L.P., Singh, S., Karchani, A., & Myong, R.S. (2017). A super-parallel mixed explicit discontinuous
Galerkin method for the second-order Boltzmann-based constitutive models of rarefied and
microscale gases. Computers & Fluids, 157, 146–163.
Reed, W.H., & Hill, T.R. (1973). Triangular mesh methods for the neutron transport equation. Los
Alamos Scientific Laboratory Report LA-UR, 13(7), 73–79
Roessler, J. & Hüssner, H. (1997).Numerical solution of the 1+ 2 dimensional Fisher’s equation by
finite elements and the Galerkin method. Mathematical and Computer Modelling, 25(3), 57–67.
Shu, C.-W., & Osher, S. (1988). Efficient implementation of essentially non-oscillatory shock-captur-
ing schemes. Journal of Computational Physics, 77(2), 439–471.
Singh, S. (2018). Development of a 3D discontinuous Galerkin method for the second-order
Boltzmann-Curtiss based hydrodynamic models of diatomic and polyatomic gases. (Doctoral
dissertation, Gyeongsang National University South Korea. Department of Mechanical and
Aerospace Engineering.)
Singh, S. (2020). Role of Atwood number on flow morphology of a planar shock-accelerated square
bubble: A numerical study. Physics of Fluids, 32, 126112.
Singh, S. (2021a). Numerical investigation of thermal non-equilibrium effects of diatomic and poly-
atomic gases on the shock-accelerated square light bubble using a mixed-type modal discon-
tinuous Galerkin method. International Journal of Heat and Mass Transfer, 169, 121708.
Singh, S. (2021b). Contribution of Mach number on the evolution of Richtmyer-Meshkov instability
induced by a shock-accelerated square light bubble. Physical Review Fluids 6(10), 104001.
Singh, S. (2021c). Mixed-type discontinuous Galerkin approach for solving the generalized
FitzHugh-Nagumo reaction-diffusion model. International Journal of Applied and Computational
Mathematics, 7, 207.
Singh, S. (2021d). A mixed-type modal discontinuous Galerkin approach for solving nonlinear reac-
tion diffusion equations. AIP Conference Proceedings (accepted).
Singh, S., & Battiato, M. (2020a). Effect of strong electric fields on material responses: The Bloch oscil-
lation resonance in high field conductivities. Materials, 13, 1070.
Singh, S., & Battiato, M. (2020b). Strongly out-of-equilibrium simulations for electron Boltzmann
transport equation using explicit modal discontinuous Galerkin method. International Journal
of Applied and Computational Mathematics, 6,133.
Singh, S., & Battiato, M. (2021a). An explicit modal discontinuous Galerkin method for Boltzmann
transport equation under electronic nonequilibrium conditions. Computers & Fluids, 224,
104972.
Singh, S., & Battiato, M. (2021b). Behavior of a shock-accelerated heavy cylindrical bubble under
nonequilibrium conditions of diatomic and polyatomic gases. Physical Review Fluids, 6,
044001.
Singh, S., & Myong, R.S. (2017). A computational study of bulk viscosity effects on shock-vortex
interaction using discontinuous Galerkin method. Journal of Computational Fluids Engineering,
22, 86–95.
Singh, S., Battiato, M. & Myong, R.S. (2021). Impact of bulk viscosity on flow morphology of shock-
accelerated cylindrical light bubble in diatomic and polyatomic gases. Physics of Fluids, 33,
066103.
Tan, Y., Xu, H., & Liao, S.J. (2007). Explicit series solution of travelling waves with a front of Fisher
equation. Chaos, Solitons and Fractals, 31(2), 462–472.
Tang, S., Qin, S. & Weber, R.O. (1993). Numerical studies on 2-dimensional reaction-diffusion equa-
tions. The Journal of the Australian Mathematical Society. Series B. Applied Mathematics, 35(2),
223–243.
184 Mathematical Modeling for Intelligent Systems

Tang, S. & Weber, R.O. (1991). Numerical study of Fisher’s equation by a Petrov–Galerkin finite ele-
ment method. The Journal of the Australian Mathematical Society. Series B. Applied Mathematics,
33(1), 27–38.
Wazwaz, A.M. (2004). The tanh method for traveling wave solutions of nonlinear equations. Applied
Mathematics and Computation, 154(3), 713–723.
Wazwaz, A.-M. & Gorguis, A. (2004). An analytic study of Fisher's equation by using Adomian
decomposition method. Applied Mathematics and Computation, 154(3), 609–620.
Zheng, S. (2004). Nonlinear evolution equations. Monographs and surveys in pure and applied
Mathematics. CRC Press, Boca Raton.
12
A Numerical Approach on Unsteady Mixed
Convection Flow with Temperature-Dependent
Variable Prandtl Number and Viscosity

Govindaraj N and Iyyappan G

Hindustan Institute of Technology and Science

A. K. Singh
VIT University

S. Roy
Indian Institute of Technology Madras

P. Shukla
VIT University

CONTENTS
12.1 Introduction ........................................................................................................................ 185
12.2 Formation of Governing Equations ................................................................................ 186
12.3 Numerical Computation with Quasilinearization Technique .................................... 189
12.4 Results and Discussion ..................................................................................................... 190
12.5 Conclusions......................................................................................................................... 194
References..................................................................................................................................... 195

12.1 Introduction
The unsteady heat transfer fluid flow is very important for many real-life applications where
fluid has direct contact with solid such as plastic sheet extrusion, wire and fiber coating,
metal and glass spinning, polymer sheets, the boundary layer along with material handlings
conveyers, etc. The boundary layer fluid flow is very thin and sometimes is visible to the
naked eye. For example, near the ship, there is a narrow band of water and the relative veloc-
ity of the ship is less than that of water. Water is used for working in many industrial and
engineering fields and hence the current study is based on water. The study on convection
pattern flow over various surfaces was performed with constant physical parameters of flu-
ids. Cortell [1] presented the numerical study of the fluid over moving surface and discussed
the direction of wall shear and temperature variations. Fang [2] discussed the unsteady flow
over moving plate, and the characteristics of the heat transfer and fluid motion are found.
Patil and Roy [3] numerically analyzed fluid flow over moving vertical plates along with par-
allel to the free stream. They discussed heat absorption or generation within the boundary
DOI: 10.1201/9781003291916-12 185
186 Mathematical Modeling for Intelligent Systems

layer. Bhattacharyya et al. [4] analyzed the slip impact on the flow over a flat plate. They have
found that velocity and temperature overshoot acting on near the plate. The fluid flow over
moving a plate is theoretically investigated with viscous dissipation term by Bachok et al.
[5]. The results indicated that the solution’s nature of the similarity equations depends on
the velocity ratio parameter, mixed convection parameter and the Eckert number. Aman and
Ishak [6] discussed the mixed convection flow toward a plate. The results indicate that dual
and unique solutions are based on opposing and assisting flow.
In general, various studies of boundary layer flow are based on constant Prandtl number
and variable viscous fluid. But in practical, liquids say water, are particularly sensitive to tem-
perature changes as for as Prandtl number and viscosity are concerned. It is assumed that the
Prandtl and viscosity are varied on inverse of linear function of temperature [7,8]. Pop et al. [9]
examined the impact on variable viscosity of the fluid flow over moving flat plate. The non-
uniform slot suction can be used for delayed separation, but the opposite effect is observed
in the case of nonuniform slot injection. Saikrishnan and Roy [10] presented the forced con-
vection flow impact on Prandtl number and viscosity with temperature. They discussed the
effects of skin friction and heat transfer rate on viscous dissipation. Roy et al. [11] studied the
laminar flow and heat transfer over a sphere and analyzed the impact of injection and suc-
tion. They have found that nonuniform slot suction can be used for delaying the separation
but the opposite effect is observed in the case of nonuniform slot injection. Mureithi et al. [12]
performed the numerical study of fluid flow on moving flat plates with the streaming flow
with temperature-dependent viscosity. Das et al. [13] discussed unsteady fluid flow over a
parallel plate with different temperature and mass diffusion in the presence of thermal radia-
tion. Govindaraj et al. [14,15] study the convection fluid flow over a stretching exponential
surface with a porous plate. Maitil and Mandal [16] have investigated impacts of thermal
radiation and buoyancy force over an infinite vertical flat plate with slip conditions. Talha
Anwar et al. [17] discussed magneto hydrodynamics effects on natural convection nanofluid
flow over a porous media with time-dependent and heat source/Sink. The effects of thermal
radiation heat absorption over an accelerated infinite vertical plate and ramped plate of con-
vective flow of nanofluids with time-dependent are discussed in Refs. [18] and [19].
The objective of the current work is to focus on the unsteady flow over a vertical plate
with temperature-dependent and physical parameters of fluids. This investigation of the
current problem is considered mainly focused on the time-dependent, various values
of Reynolds number, Prandtl number, and viscosity of fluid flow moving on a flat plate.
The nonlinear governing equations have been converted into linear equation using the
Quasi-linearization technique and further obtained the system of equations based on the
finite difference method. The effects of physical parameters on velocity, temperature, heat
transfer, and skin friction have been investigated. The unsteadiness in the flow and tem-
perature fields is induced by the free stream and moving plate velocity. The validation of
current results has been done by comparing them with previously published results that
are shown in Table 12.2.

12.2 Formation of Governing Equations

Consider a vertical plate with an unsteady laminar water-based mixed convection flow
with variable viscosity and temperature. In the range of 0°C–45°C, Prandtl number ( Pr )
and dynamic viscosity μ are considered too different as an inverse linear function of
Numerical Approach on Unsteady Mixed Convection Flow 187

FIGURE 12.1
Physical coordination of model figure.

temperature. At constant strain, the variance of both specific heat Cp and density (ρ) is ( )
assumed to be less than 1%, so they are treated as constants. Reference [7] provided the
values for thermophysical properties of water at various temperatures. Buoyancy force rise
to the fluid properties to relate the density changes to temperature, and the thermophysi-
cal properties of water with various temperature has been shown in [20]. In an incompress-
ible fluid, the contribution of heat due to the contraction of the fluid is very small and it is
ignored. The fluid at the surface is kept at a constant temperature Tw , and the boundary
layer’s edge is kept at the same temperature T∞ .
The variable Prandtl number and variable viscosity are defined as
1 1
Pr = and  µ=
c1 + c 2 T b1 + b2T
where b1, b2 , c1 and c2 are the best approximation of water and the numerical data presented
[20] (Figure 12.1).
The governing equations of the current problem are given as
∇.V = 0 (12.1)
1
(∇u)t + u ( ∇u)i + v ( ∇u) j = (∇U e )t + ( ∇ ) j  µ (∇u) j  + g  β (Tw − T∞ ) (12.2)
ρ
1 µ
(∇T )t + u (∇T )i + v (∇T ) j = ( ∇ ) j  ( ∇T ) j  (12.3)
ρ  Pr 
The boundary constraints:

( )
y = 0 :  u ( t , x , y ) = U w ( t ) = U 0 wφ t * ,  T ( t , x , y ) = Tw ,   v ( t , x , y ) = 0,
(12.4)
y → ∞   :  u ( t , x , y ) → U ( t ) → U φ ( t ) ,  T ( t , x , ∞ ) → T
e ∞
*
∞
188 Mathematical Modeling for Intelligent Systems

To determine the velocity field and temperature profiles within the boundary layer, the
problem is well posted with boundary condition equations (12.1)–(12.4) grouping into non-
dimensional fluid parameters with the following similarity transformation:
1/2
x U∞ U
ξ=
L
( )
,  ψ ( x , y ) = φ t * (Uxυ ) f (ξ , η ) ,   ε =
1/2

U∞ + U 0w
,  η = 
 xυ 
y;

(Uυ x )1/2
v=−
2x
{ f + 2ξ fη − η F} ,   fη = F ,  t* = Ux t; 
gβT (Tw − T∞ ) L3
GrL =
υ2
,  U = U 0 w + U∞   ,  T − T∞ = (Tw − T∞ ) θ t * , ξ , η ( )
GrL ∂ψ UL ∂ψ
λ=
ReL2
,  u =
∂y
,    ReL =
ν
,  v = −
∂x
,  u = Uφ t * F ( ) (12.5)

The transformed governing equations (12.2) and (12.3) are as follows:

( NFη )η + φ (t* )  2f  Fη + ReLξφ (t* )

( ) ( − F ) − Re ξ dF + λθξφ
dφ t *
(t )
−1 * −1
* L *
dt dt

( )
= ξφ t * ( FFξ − fξ Fη ) (12.6)

 f dθ
( NPr −1
θη )
η
( )  2 dt
( )
+ θηφ t *   − ReLξ * = ξφ t * ( Fθξ − θη fξ ) (12.7)

The transformed boundary constrains are

F = 1 − ε ,  θ = 1,   at  η = 0

(12.8)
F = ε ,  θ = 0,        at   η = η∞

η
where f =
∫ Fd
0
η + fw  ,   fw = 0 (Table 12.1)

TABLE 12.1
The Values of Thermo-Physical Properties at Different Temperature. Values of Water as Given
Below [7]
Temperature Density Specific Heat Thermal Conductivity Viscosity Prandtl
(T) (C o ) (g /m 3) (J.107/kg K) (erg.105)/cm.s.K) (g.10 −2-cm.s) Number
0 1.00228 4.2176 0.5610 1.7930 13.48
10 0.99970 4.1921 0.5800 1.3070 9.45
20 0.99821 4.1818 0.5984 1.0060 7.03
30 0.99565 4.1784 0.6154 0.7977 5.12
40 0.99222 4.1785 0.6305 0.6532 4.32
50 0.98803 4.1806 0.6435 0.5470 3.55
Numerical Approach on Unsteady Mixed Convection Flow 189

*
( )
i.e., fw = 0 is impermeable plate. Noted that φ t * = 1 + α t * ; α > 0 or α < 0, where the time
t = 0 is steady-state flow are obtained from equations (12.6) and (12.7) by substituting
( )
d∅ t * dF
dt *
dt
( )
= * = 0,∅ t * = 1    and if t * > 0 become unsteady flow with free stream velocity

 ( )
U e ( t ) = U∞ φ t *  and velocity at wall U w ( t ) = U 0 wφ t *  .
   ( )
( )
The skin friction coefficient C fx as below:

 ∂u 
2µ  
 ∂y  y=0
C fx =
ρU 2
w
( )
= 2φ t * Fη (ξ , 0 )( ReLξ )
−1/2
(12.9)

(
i.e.,\dsC fx ( ReLξ )
1/2
) = 2φ (t ) F (ξ , 0)
*
η

The Nusselt number ( Nux ) as below:

 ∂T 
−x 
 ∂y  y=0
= −θη (ξ , 0 ) ( ReL ~ ξ )
1/2
Nux = (12.10)
(Tw − T∞ )

(
i.e.,\dsNux ( ReLξ )
−1/2
) = −θ (ξ , 0).
η

12.3 Numerical Computation with Quasilinearization Technique

In the process of numerical computation, the set of nonlinear governing equations are
playing a vital role given its efficiency methods and accuracy in yielding the solutions. The
nonlinear equations (12.6) and (12.7) with boundary conditions (12.8) have been linearized
numerically by using quasi-linearization technique, to obtain the following coupled PDEs.

s+ 1
X 1s Fηη + X 2s Fηs+ 1 + X 3s F s+ 1 + X 4s Fξs+ 1 + X 5s Fts* + 1  + X 6sθηs+ 1 + X 7sθ s+1 = X 8s (12.11)

Y1sθηη
s+ 1
+ Y2sθηs+1 + Y3sθ s+1 + Y4sθηs+1 + Y5sθ ts*+1 + Y6s F s+1 = Y7s (12.12)

The coefficient functions are determined by an iterative method, subject to constraints,

F s+1 = 1 − ε ,  θ s+1 = 1        at    η = 0

F s+1 = ,  θ s+1 = 0                     at   η = η∞ (12.13)

The coefficients in the equations (12.11) and (12.12) are given by

X 1s = N
190 Mathematical Modeling for Intelligent Systems

X 2s = − a1 N 2θη + φ t * ( ) 2f + φ (t )ξ f
*
ξ

( ) − ξφ
dφ t *
( ) (t ) F
−1
X 3s = −ξ Reφ t * \ ds *
ξ
dt *
( )
X 4s = −ξφ t * F

X 5s = −ξ Re

X 6s = − a1 N 2 Fη

( )
−1
X 7s = − a1 N 2 Fηη + 2 a12 N 3 Fηθη + λξφ t *

( ) − ξφ t FF
dφ t *
( ) ( )
−1\
X 8s = − a1 N 2 Fηθη − a1 N 2 Fηηθ + 2 a12 N 3 Fηθηθ − ξ Reφ t * ds *
*
ξ
dt
Y1s = NPr −1

Y2s = −2 a1 N 2 Pr −1θη + 2 a3θη N + φ t * ( ) 2f + φ (t )ξ f*

ξ

Y3s = a3 Nθηη − a1 N 2 Pr −1θηη − 2 a1 a3 N 2θη2 + 2 a12 N 3 Pr −1θη2

( )
Y4s = −ξφ t * F

Y5s = −ξ Re

( )
Y6s = −ξφ t * θξ

Y7s = a3θη2 N − a1θη2 Pr −1 N 2 + a3θηηθ N − a1θηηθ Pr −1 N 2

− 2 a1 a3θη2θ N 2 + 2 Pr −1 a12θη2θ N 3 − ξφ t * Fθξ ( )

12.4 Results and Discussion

To obtain the nonsimilar solutions, the numerical procedure in the above section is
implemented. Unsteady mixed convection flow with variable viscous fluid and variable

TABLE 12.2
Comparison of the Steady State Results
Pr 2 5 7 10 100
Ali [21] – – – 1.6713 –
Chen [22] 0.68324 – 1.38619 1.68008 5.54450
A.K. Singh [23] 0.6830 1.151 1.38600 1.6801 5.5450
Present work 0.6832 1.149 1.385 1.6800 5.5449
Numerical Approach on Unsteady Mixed Convection Flow 191

Prandtl number is considered. Numerical computation process has been carried out for
different parameters values of α ( 0 ≤ α ≤ 2.5 ),  ( 0.1 ≤  ≤ 0.9 ), ξ ( 0 ≤ ξ ≤ 1), λ ( 0 ≤ λ ≤ 4 ), and
(
t* 0 ≤ t* ≤ 2 .)
Comparison of the steady state results −θη (ξ , 0 ) for λ = 0, ξ = 0, ε = 0 with those of Ali and
Yousef [21], Chen [22] and A.K. Singh et al. [23].
( )
Figure 12.2 displays the influence of accelerating fluid flow φ t * , time parameter t * , ( )
and buoyancy force ( λ ) on the velocity profile. Noted that, near the wall high overshoot
is presents in the velocity profile as increase λ with decrease t *. Also, the velocity profile
enhanced near the wall due to the increase of buoyancy force ( λ ). In case of time parameter
increase, the steepness of velocity profile is almost absent, which means that the intensity
of the fluid flow goes down. For example, when λ increases from 2 to 4 the fluid accelerated
near the wall and increased the velocity profile for  = 0.4, α = 0.5, ξ = 0.5, t * = 1 and Re = 1.0.
( )
Figure 12.3 shows the impact of  t *, ε and accelerating flow ∅ t *  on θ   (temperature
profile). It is noted that intensity of temperature profile θ increases with increase of ε and
its followed decreasing trends with the increase of time t * . However, due increase of ε the
fluid gets accelerated and the magnitude of temperature is increased. As ∅ t *  increase( )
the thickness of the thermal boundary layer is close to the lower segment of the plate. The
temperature gradient increased when t * decreases from 1 to 0 and hence thermal thickness
are much smaller for t = 1 as compared to t = 0. For example, the intensity of θ increases by
36% as ε varies from 0.1 to 0.7 at t * = 0, ξ = 0, λ = 2, α = 0.5 and Re = 1.
Figure 12.4 depicts the impact of the accelerating flow α and t * on the velocity profile.
The impacts of velocity profile indicate that it is strongly dependent on α and t * . In F,
the decreasing trend is observed for all values of α . Physically, furthermore accelerating
parameter is directly related to α and more significant effects are observed in velocity pro-
files. However, velocity profiles are enhanced for low values of α (α = 0.5) and time t * = 1.
In particular, for ε = 0.2, η = 1, ξ = 0, λ = 2.0 and Re = 100, the magnitude of F (η, ξ ) increases
by 20% with decreases of α from 2.5 to 0.5 at t * = 1.0.

FIGURE 12.2
( )
Variations of t * and α on F for φ t * when Re = 100,  = 0.2, ξ = 0, and λ = 2.0.
192 Mathematical Modeling for Intelligent Systems

FIGURE 12.3
( )
Variations of t * and α on F for φ t * when Re = 100,  = 0.2, ξ = 0, and λ = 2.0.

FIGURE 12.4
*
( )
Variations of t and α on F for φ t * when Re = 100,  = 0.2, ξ = 0, and λ = 2.0.

Figure 12.5 presents the effects of the accelerating flow α and t * on the temperature
profile. The temperature profile gradually decreases with the increase of α and time (t * ).
The temperature profiles are directly related to the energy equation of the right-hand side,
which gives the high friction near the wall due to increase of α . In particular, for ε = 0.2,
ε = 0.2, ξ = 0, λ = 2.0 and Re = 100 the magnitude of temperature profile increases by 26% as
α decreases 2.3 to 0.4 at t * = 1.5.
Numerical Approach on Unsteady Mixed Convection Flow 193

FIGURE 12.5
( )
Variations of α and t * on θ for φ t * when  = 0.2, Re = 100, ξ = 0 and λ = 2.0.

FIGURE 12.6
(
Variations of α and λ on C fx ( ReL ξ ) )
1/2
( )
for φ t * when  = 0.2, ξ = 0.5 and Re = 1.0.

Figure 12.6 depicts the impacts of λ and accelerating flow α on the skin-friction. For
the higher values of (λ >0), the fluid moves faster near the wall and increase the thermal
and acceleration of the fluid. Due to the fluid acceleration, the fluid gets hot and moves
towards the momentum boundary layers. In fact, the shear stress increased near the
wall and the magnitude of the C fx ( ReLξ )
−1/2
is increased. Initially (t * ≤ 0.4), the magni-
tude of C fx ( ReLξ )
−1/2
is very high for higher values of α = 1.0, whereas, the magnitude
194 Mathematical Modeling for Intelligent Systems

FIGURE 12.7
(
Variations of ∆T and  on C fx ( ReL ξ ) )
1/2
( )
for φ t * , α = 0.5  when ξ = 0, Re = 100, and λ = 1.0.

of C fx ( ReLξ )
−1/2
is decreasing trend observed for lesser value of α = 0.5 when t * ≥ 0.5. In
particular, for ε = 0.2, Re = 1.0, ξ = 0.5, and t * = 0.2, the skin-friction increases by 179% as λ
increase from 2 to 3 at α = 0.5.
Figure 12.7 represents the effects of ε and temperature dependent (∆Tw) on C fx ( ReLξ ) .
1/2

( )
The accelerating flow parameter ∅ t * , which gives the acceleration to the fluid due to
this the increasing trends observed with increase of velocity ratio (ε ). In case of lesser
values of (ε ) the decreasing trends is observed with increase of ∆Tw. In particular, for ξ =0,
α = 0.5, ξ = 0, λ = 1.0 and Re = 100 the magnitude of the skin-friction increases about 195%
when ε increases from 0.4 to 0.6 at ∆Tw = 15. Figure 12.8 shows the effects of ε and tem-
perature dependent (∆Tw) on Nux ( ReLξ ) . The heat-transfer coefficient increases with
−1/2

the increase of ε , but reverse trend is observed when ∆Tw decreases. The velocity ratio
(ε ) increase from 0.1 to 0.5, which produces the higher heat generation. In particular, for
α = 0.5, t * = 0.3, λ = 1.0, ξ = 0 and Re = 100 the magnitude of the heat transfer increases about
93%, when ε increases from 0.4 to 0.6 at ∆Tw = 15.

12.5 Conclusions
This work investigated numerically an unsteady laminar fluid flow of stretching sheets
the following results are summarized.

• The velocity profile overshoots with an increase of buoyancy parameters (λ) and
undershoot velocity profile with an increase of t * . In particular, for ε = 0.4, α = 0.5,
Re = 1.0, t * = 1 and ξ = 0.5, the magnitude of velocity profile maximum overshoot
by 21% between λ = 2 and λ = 4.
Numerical Approach on Unsteady Mixed Convection Flow 195

FIGURE 12.8
(
Variations ∆T and  on Nux ( ReL ξ ) )
−1/2
( )
for φ t * , α = 0.5, when Re = 100, ξ = 0 and λ = 1.0.

• The temperature distribution increases with an increase of composite velocity (ε )

and decreases with an increase of t * . For example, the temperature (θ ) increases by
10% as t * decreases from 1 to 0 for ε = 0.2, ξ = 0 and Re=1.0.
• The α acts as suction/injection factor on the velocity and temperature profiles. In
particular, for ε = 0.2, Re = 100, t * = 1 and ξ = 0, the velocity (F) and temperature
(θ ) decreased by 23% and 90%, respectively as α increases from α = 0.5 to
α = 2.5 at η = 1.
• The magnitude of skin friction (C fx ( ReLξ ) ) and heat transfer. (Nux ( ReLξ ) ) are
−1/2 −1/2

increased with an increase of ε . In particular t * = 1 the skin friction coefficient by

129% and heat transfer rate is increased by 117%.

References
1. R. Cortell (2007), Flow and heat transfer in moving fluid over a moving flat surface. Theor
Comput Fluid Dyn 21, 435–446.
2. T. Fang (2008), A note on the unsteady boundary layer over a flat plate, I. J. Non-linear Mech 43,
1007–1011.
3. P. M. Patil and S. Roy (2010), Unsteady mixed convection flow from a vertical plate in paral-
lel free stream: influence of heat generation or absorption, Int J Heat Mass Transf 53(21–22),
4749–4756.
4. K. Bhattacharyya, S. Mukhopadhyay, and G. C. Layek (2013), Similarity solution of mixed con-
vective boundary layer slip flow over a vertical plate, Ain Shams Eng J 4, 299–305.
196 Mathematical Modeling for Intelligent Systems

5. N. Bachok, A. Ishak, and I. Pop (2013), Mixed convection boundary layer flow over a moving
vertical flat plate in an external fluid flow with viscous dissipation effect, PLoS One 8(4), e60766.
6. F. Aman and A. Ishak (2012), Mixed convection boundary layer flow towards a vertical plate
with a convective surface boundary condition mathematical problems in engineering, Math
Probl Eng 2012, 453457.
7. I. Govindhasamy and A. K. Singh (2020), Boundary layer flow and stability analysis of forced
convection over a diverging channel with variable properties of fluids, Heat Transf 49(8),
5050–5065.
8. D. R. Lide (Ed.) (1990), CRC Handbook of Chemistry and Physics, 71st ed., CRC Press, BocaRaton,
FL.
9. H. Schlichting (2000), Boundary Layer Theory, McGraw Hill, New York.
10. I. Pop, R. S. R. Gorla, and M. Rashidi (1992), The effect of variable viscosity on flow and heat
transfer to a continuous moving flat plate, Int J Eng Science 30, 1–6.
11. P. Saikrishnan and S. Roy (2002), Steady nonsimilar axisymmetric water boundary layer with
variable and Prandtl number, Acta Mech 157 187–199.
12. S. Roy, P. Saikrishnan, and B. D. Pandey (2009), Influence of double slot suction (injection) into
water boundary layer flows over sphere, Int Commun Heat Mass Transf 36, 646–650.
13. E. M. Mureithi, J. J. Mwaonanji, and O. D. Makinde (2013), On boundary layer flow over a mov-
ing surface in a fluid with temperature-dependent viscosity, Open J Fluid Dyn 3, 135–140.
14. S. Das, R. N. Jana, and A. Chamkha (2015), Unsteady free convection flow between two vertical
plates with variable temperature and mass diffusion, J Heat Mass Transf Res 2 49–58.
15. N. Govindaraj, A. K. Singh, S. Roy, and P. Shukla (2019), Analysis of a boundary layer flow over
moving an exponentially stretching surface with variable viscosity and Prandtl number, Heat
Transf 48(7), 2736–2751.
16. N. Govindaraj, A. K. Singh, and S. Roy (2018), Water boundary layer flow over an exponen-
tially permeable stretching sheet with variable viscosity and Prandtl number, In International
Conference on Mathematical Modelling and Scientific Computation (pp. 207–217). Springer,
Singapore.
17. D. K. Maiti and H. Mandal (2019), Unsteady slip flow past an infinite vertical plate with ramped
plate temperature and concentration in the presence of thermal radiation and boyancy, J Eng
Thermophys 28(3), 431–452.
18. T. Anwar, P. Kumam, Z. Shah, W. Watthayu, and P. Thounthong (2020), Unsteady radiative
natural convective MHD nanofluid flow past a porous moving vertical plate with heat source/
sink, Molecules 25, 854.
19. M. V. Krishna, N. A. Ahamad, and A. J. Chamkha (2021), Radiation absorption on MHD con-
vective flow of nanofluids through vertically travelling absorbent plate, Ain Shams Eng J 12,
3043–3056.
20. Y. D. Reddy, B. S. Goud, and M. A. Kumar (2021), Radiation and heat absorption effects on an
unsteady MHD boundary layer flow along an accelerated infinite vertical plate with ramped
plate temperature in the existence of slip condition, Partial Diff Equ Appl Math 4, 100166.
21. M. Ali and F. Al-Yousef (1998), Laminar mixed convection from a moving vertical surface with
suction or injection, Heat Mass Transf 33(4), 301–306.
22. C. H. Chen (1998), Laminar mixed convection adjacent to vertical, continuously stretching
sheets, Heat Mass Transf 33, 471–476.
23. A. K. Singh, A. K. Singh, and S. Roy (2019), Analysis of mixed convection in water boundary
layer flows over a moving vertical plate with variable viscosity and Prandtl number, Int. J.
Numer. Methods Heat Fluid Flow 29(2), 602–616.
13
Study of Incompressible Viscous Flow
Due to a Stretchable Rotating Disk
Through Finite Element Procedure

Anupam Bhandari
University of Petroleum & Energy Studies (UPES)

CONTENTS
13.1 Introduction ........................................................................................................................ 197
13.2 Mathematical Formulation ............................................................................................... 198
13.3 Finite Element Procedure for Numerical Solution ........................................................ 200
13.4 Results and Discussion ..................................................................................................... 203
13.5 Conclusion .......................................................................................................................... 211
References..................................................................................................................................... 212

13.1 Introduction
The study of three-dimensional incompressible flow in a rotating system has been
researched owing to its significant role in rotating equipment. Such types of interest-
ing problems have been solved using the similarity transformation. It reduces the three-
dimensional Navier–Stokes equations into nondimensional nonlinear ordinary differential
equations. Then the converted nondimensional equations represent the nondimensional
velocity distribution for the nondimensional similarity variable of the flow.
Cochran implemented the power series approximations method for the solution of non-
dimensional differential equations (Cochran, 1934). In 1966, Benton tried to obtain a more
accurate solution than Cochran (Benton, 1966). In magnetohydrodynamic flow, Hassan
and Attia used a finite difference scheme for solving problems of rotational flow (Aboul-
Hassan & Attia, 1997). Turkyilmazoglu used the fourth-order Runge–Kutta scheme for
solving the nonlinear coupled differential equations for magnetohydrodynamic stagna-
tion revolving flow with the help of MATHEMATICA Software (Turkyilmazoglu, 2012).
Similarity solutions have been obtained through the homotopy analysis method for dif-
ferent types of steady flow in the rotating system considering the influence of variable
thickness and slip effects (Rashidi et al., 2011, 2012; Sravanthi, 2019).
Due to the importance of the finite element method over other numerical methods, this
method is being commercialized (Reddy & Gartling, 2010). Gupta et al. developed the
finite element procedure to investigate the micropolar flow over a shrinking sheet (Gupta
et al., 2018). Anjos et al. implemented the finite element method procedure for the flow
over a rotating plate (Anjos et al., 2014). Takhar et al. presented the finite element solu-
tion for micropolar fluid flow from an enclosed rotating plate (Takhar et al., 2001). Yoseph

DOI: 10.1201/9781003291916-13 197

198 Mathematical Modeling for Intelligent Systems

and Olek described the solution of compressible flow over a rotating porous disk using
asymptotic and finite element approximation (Bar-Yoseph & Olek, 1984). Sheikholeslami
and Ganji used the control volume finite element technique to investigate ferrofluid flow
and heat transfer in a semiannulus enclosure (Sheikholeslami & Ganji, 2014). Vasu et al.
demonstrated the magnetohydrodynamic flow through a stenosed coronary artery using
finite element analysis (Vasu et al., 2020). Feistauer et al. demonstrated the finite element
solution for nonviscous subsonic irrotational flow (Feistauer et al., 1992).
This work investigates the steady, incompressible, and axisymmetric flow of viscous
fluid due to rotating and stretching disks. The influence of rotation and stretching of
the disk on the velocity profiles is graphically presented. The finite element procedure is
explained for solving nonlinear coupled differential equations of the flow due to a rotating
and stretching disk. The problem formulation is presented over a typical element of the
solution domain. In the present analysis, the solution of nonlinear differential equations is
demonstrated through finite element procedure in COMSOL Multiphysics Software and
is compared with the previously reported solutions. The present solution finite element
solution is corrected up to ten decimal places.

13.2 Mathematical Formulation
Figure 13.1 demonstrates the flow configuration of viscous over a rotating and stretching
plate. Let ( u, v , w ) be the velocity components in the cylindrical coordinates ( r , ϕ , z ) direc-
tion, respectively. It is assumed that the disk uniformly rotates with the angular velocity ω
about z-axis. Along with the rotation, the disk is stretching at a uniform rate s in the direc-
tion of. The governing equations for the steady and axisymmetric flow due to rotating disk
are as follows (Benton, 1966; Cochran, 1934; Turkyilmazoglu, 2014):

∂u u ∂w
+ + =0 (13.1)
∂r r ∂z

 ∂u ∂ u v2  ∂p ∂2u 1 ∂u u ∂2u 
ρ u +w − =− +µ 2 + − + (13.2)
 ∂r ∂z r  ∂r ∂z r ∂ r r 2 ∂ r 2 

FIGURE 13.1
Flow configuration of viscous fluid due to rotating and stretching disk.
Study of Incompressible Viscous Flow 199

 ∂v ∂ v uv  ∂2v 1 ∂ v ∂2v v 
ρ u +w + = µ 2 + + − (13.3)
 ∂r ∂z r  
∂z r ∂ r ∂ r 2 r 2 

 ∂w ∂w  ∂p  ∂2 w 1 ∂w ∂2 w 
ρ u +w =− +µ 2 + + (13.4)
 ∂r 
∂z  ∂z  ∂z r ∂r ∂r 2 

With boundary conditions:

z = 0; u = rs, v = rω , w = 0 ; z → ∞; u → 0, v → 0 (13.5)

The following similarity transformation reduces equations (13.1)–(13.5) into nondimen-

sional nonlinear differential equations (Bhandari, 2020; Cochran, 1934; Ram & Bhandari,
2013; Benton, 1966):

ω
u = rω F (η ) , v = rω G (η ) , w = υω H (η ) , p − p∞ = − ρυω P (η ) , η = z (13.6)
υ

The reduced nondimensional equations are as follows:

dH
+ 2F = 0 (13.7)
dη

d2 F dF
−H − F 2 + G2 = 0 (13.8)
dη 2
dη

d 2G dG
−H − 2 FG = 0 (13.9)
dη 2 dη

With boundary conditions:

F ( 0 ) = C , G ( 0 ) = 1, H ( 0 ) = 0, F ( ∞ ) = 0, G ( ∞ ) = 0 (13.10)

s
where C = denotes the stretching strength parameter.
ω
A case C = 0 demonstrates that there is no stretching in the disk and the problem reduces
into the flow due to rotating disk only. However, the present problem can also be inves-
tigated for different speeds of rotation of the disk. In this case, the following similarity
transformation can be used:

s
u = rsF (η ) , v = rsG (η ) , w = υω H (η ) , p − p∞ = − ρυω P (η ) , η = z (13.11)
υ
In this case, the boundary conditions can be written as

F ( 0 ) = 1, G ( 0 ) = Ω, H ( 0 ) = 0, F ( ∞ ) = 0, G ( ∞ ) = 0 (13.12)
200 Mathematical Modeling for Intelligent Systems

ω
where Ω = represents the rotation strength parameter.
s
A case Ω = 0 represents there is no rotation of the disk and the flow of viscous fluid is
appeared due to only stretching. If only stretching of the plate is present, then the three-
dimensional flow reduces into the two-dimensional flow.

13.3 Finite Element Procedure for Numerical Solution

Figure 13.2(a) demonstrates the entire domain of the solution and, Figure 13.2(b) repre-
sents the element-wise discretization of the domain. The weighted integral form of the
( )
above differential equation over a typical element Ωe = ηae , ηbe is (Reddy et al., 2020;
Reddy & Gartling, 2010):

ηbe
 dH 
∫ W  dη + 2F dη = 0
ηae
(13.13)

ηbe
 d2 F dF 
∫
ηe
a
W 2 − H
 dη dη
− F 2 + G 2 dη = 0

(13.14)

ηbe
 d 2G dG 
∫
ηe
a
W 2 − H
 dη dη
− 2 FG dη = 0

(13.15)

(a)

(b)

FIGURE 13.2
(a) Domain of the solution and (b) finite element mesh of the domain
Study of Incompressible Viscous Flow 201

Equations (13.14) and (13.15) can be written as

ηbe ηe
 dW dF dF   dF  b

∫
ηae
 dη dη + W H
dη
+ W FF − WGG  dη =  W dη  (13.16)
ηae

ηbe ηe
 dW dG dG   dG  b

∫
ηae
 dη dη + W H
dη
− 2W FG  d η =  W dη  (13.17)
ηae

The above equations, F, G and H are nonlinear and considered the known quantity using
Lagrange’s interpolation formula. Therefore, the system of equations becomes linear to
obtain the unknown dependent variables.
( )
For a typical element Ωe = ηae , ηbe , we consider the quadratic Lagrange’s interpolation
functions (J. N. Reddy, 2019; Reddy & Gartling, 2010)
3

F (η ) ≈ Fhe (η ) = ∑F ψ
j=1
j
e e
j (η ) (13.18)

G (η ) ≈ Ghe (η ) = ∑G ψ
j=1
e
j
e
j (η ) (13.19)

H (η ) ≈ H he (η ) = ∑H ψ
j=1
e
j
e
j (η ) (13.20)

where ψ ej (η ) are the Lagrange’s interpolation functions. In the present analysis, quadratic
interpolation functions are used, which are defined as follows (J. N. Reddy, 2019):

 η  2η  4η  η η 2η 
ψ 1e (η ) =  1 −   1 −  , ψ 2e (η ) =  1 −  , ψ 3e (η ) =  1 − (13.21)
 he   he  he  he  he  he 

where η denotes the local coordinates. Therefore, the local coordinate η and global coordi-
nate η are related to each other with the linear relation η = η − ηae.
Consider the weight function W = ψ ie (η ), equations (13.13), (13.16), and (13.17) can be
written as

ηbe


∫ ψ e
1
d
dη
( ) (
H 1eψ 1e + H 2eψ 2e + H 3eψ 3e + 2ψ 1e F1eψ 1e + F2eψ 2e + F3eψ 3e )dη = 0 (13.22)
ηae


ηbe
 d 

∫ ψ e
2
dη
( ) ( )
H 1eψ 1e + H 2eψ 2e + H 3eψ 3e + 2ψ 2e F1eψ 1e + F2eψ 2e + F3eψ 3e dη = 0 (13.23)
ηae


ηbe
 d 

∫ ψ e
3
dη
( ) ( )
H 1eψ 1e + H 2eψ 2e + H 3eψ 3e + 2ψ 3e F1eψ 1e + F2eψ 2e + F3eψ 3e dη = 0 (13.24)
ηae

202 Mathematical Modeling for Intelligent Systems

ηbe
 dψ e d  3
 d 
3
  3
  3
 
∫ 
 d η
1

dη  ∑ Fieψ ie  + ψ 1e H


dη  ∑ Fieψ ie  + ψ 1e F 
 
∑ Fieψ ie  − ψ 1e G 
 
∑ Gieψ ie  dη
 
ηae  i=1 i=1 i= 1 i=1

( )
= ψ 1e ηae Q1e + ψ 1e ηbe Q2e( ) (13.25)

ηbe
 dψ 2e d   d       
3 3 3 3

∫ 
e  dη dη 
 ∑
i= 1
Fieψ ie  + ψ 2e H


dη  ∑i=1
Fieψ ie  + ψ 2e F 
 
∑i= 1
Fieψ ie  − ψ 2e G 
 
∑
i=1
Gieψ ie  dη
 
ηa

( ) ( )
= ψ 2e ηae Q1e + ψ 2e ηbe Q2e (13.26)

ηbe
 dψ 3e d   d       
3 3 3 3

∫ 
 dη dη 
 ∑ Fieψ ie  + ψ 3e H


dη  ∑ Fieψ ie  + ψ 3e F 
 
∑ Fieψ ie  − ψ 3e G 
 
∑ Gieψ ie  dη
 
ηae  i= 1 i=1 i= 1 i=1

( )
= ψ 3e ηae Q1e + ψ 3e ηbe Q2e( ) (13.27)

ηbe
 dψ e d  3
 d 
3
  3

∫ 
 d η
1

dη  ∑ G ψ  + ψ 1e H
e
i

e
i 
dη  ∑ G ψ  − 2ψ 1e F 
e
i

e
i

∑ G ψ  dη
e
i
e
i

ηae  i= 1 i=1 i=1 

( )
= ψ 1e ηae Q3e + ψ 1e ηbe Q4e ( ) (13.28)

ηbe
 dψ e d  3
 d 
3
  3

∫ 
dη
2

dη  ∑ G ψ  + ψ 2e H
e
i

e
i 
dη  ∑ G ψ  − 2ψ 2e F 
e
i

e
i

∑ G ψ  dη
e
i
e
i

ηae 
 i=1 i=1 i= 1 

( )
= ψ 2e ηae Q3e + ψ 2e ηbe Q4e ( ) (13.29)

ηbe
 dψ e d  3
 d 
3
  3

∫ 
 d η
3

dη  ∑ G ψ  + ψ 3e H
e
i

e
i 
dη  ∑ G ψ  − 2ψ 3e F 
e
i

e
i

∑ G ψ  dη
e
i
e
i

ηae  i=1 i=1 i=1 

( )
= ψ 3e ηae Q3e + ψ 3e ηbe Q4e ( ) (13.30)

In the above equations Q1e, Q2e, Q3e and Q4e represents the secondary variables. These are
defined as follows:

 dF   dF   dG   dG 
Q1e = −   , Q2e = −   , Q3e = −   , Q4e = −  (13.31)
 dη  ηae  dη  η e  dη  ηae  dη  ηae
b
Study of Incompressible Viscous Flow 203

The matrix form of the algebraic equations can be written as:

 A 
 ij   Bij  Cij    F
e e e
   0 
  
  D 
e
ij  E 
e
ij  Fije    G 
=  Q1 + Q2e
e
 (13.32)
   H   
Q3e + Q4e
 M   N  Oij    
e e e
 ij ij

ηbe ηbe ηbe

dψ ej  dψ ie dψ ej e dψ j
e

e
ij
e
∫
A = 2 ψ ψ dη , C = ψ
ηa
e
i
e
j
e
d η
dη , Dije =
e
ij
∫
ηa
e
i
∫
ηae
 dη dη + Hψ i
dη
+ Fψ iψ j  dη ,


ηbe ηbe
 dψ ie dψ ej e dψ j
e

e
ij
∫
E = − Gψ iψ j dη , N =
ηae
e
ij
∫
ηae
 dη dη + Hψ i
dη
− 2 Fψ iψ j  dη

(13.33)

Bije = 0, Fije = 0, Mije = 0, Oije = 0,

After the assembly of element equations, a set of nonlinear equations can be obtained.
We assume the functions for F and H at a lower iteration level to obtain a system of linear
equations. The present domain of the solution is considered (0,10) to obtain the asymp-
totic solution. If we consider the element length 0.001, then 10,000 element equations
need to assemble through the finite element method. This attempt is used in COMSOL
Multiphysics Software and then automatic Newton is used to obtain the numerical results.

13.4 Results and Discussion

Current results describe finite element solution of similarity equations for flow over a
rotating and stretching plate. In the previous section, the mathematical procedure of the
finite element procedure is presented. The finite element equations are obtained here for
a typical element Ω e . The size and geometry of each element same, therefore equations of
each element can be assembled in a matrix form using quadratic Lagrange’s interpolation.
The graphical representation of the present model is obtained using COMSOL Multiphysics
Software and is compared with the previously published solutions.
Figures 13.3–13.5 represent the radial, tangential, and axial velocity profiles with the
variation of the parameter C. Cochran (1934) solved the similarity equations for C = 0 and
Ω = 1 using power series approximation and obtained the solution up to three decimal
places. In the literate (Benton, 1966; Cochran, 1934), the radial, tangential, and axial veloci-
ties are demonstrated in the domain ( 0, 4.4 ). However, in the present solution, the domain
of the similarity solution is ( 0, 10 ) for the variation of stretching and rotation parameters.
The present solution represents that the velocities reach asymptotically zero and this trend
of the graph was absent in the previous study due to its small domain. Escalating the
strength of the stretching parameter enhances the radial and axial velocities. Stretching of
the disk favors the two-dimensional flow in the radial and axial directions. Therefore, the
204 Mathematical Modeling for Intelligent Systems

radially stretch of the disk diminishes the azimuthal velocity profile. Figures 13.6 and 13.7
represents the radial and axial velocity profiles for different values of stretching strength
parameter. In this case, the rotation of the disk is considered zero. The stretching of the
disk without rotation produces only a two-dimensional flow. The comparative study of
rotation and stretching indicates that the stretching parameter influences the radial and
axial velocities more than the rotation parameter. Stretching of the disk increases the axial
velocity more than the rotation of the disk as shown in Figures 13.5, 13.7, and 13.10. It rep-
resents the dominance of stretching effects on the velocity.
Figures 13.8–13.10 represent the radial, tangential, and axial velocity profiles, respectively
for variation of the strength of rotation parameter along with constant radial stretching.
Increasing the rotation parameter enlarges the momentum boundary layer. At Ω = 0, the
tangential velocity becomes zero which reduces the three-dimensional boundary layer flow
into the flow in the radial and axial directions only.
Table 13.1 represents the comparison of the finite element method with previous results
(Bachok et al., 2011; Benton, 1966; Cochran, 1934; Kelson & Desseaux, 2000; Turkyilmazoglu,
2014). In references (Benton, 1966; Cochran, 1934), the similarity equations were solved
using power series approximation and recurrence relation method and the solution was
not achieved to a higher degree of accuracy due to the unavailability of a high computa-
tional computer. However, using computational software, the similarity equations were
numerically solved using the finite difference scheme, shooting method, and Runge–Kutta
fourth-order method (Bachok et al., 2011; Kelson & Desseaux, 2000; Turkyilmazoglu, 2014).
Using the finite element method, we obtained the solution with better accuracy in com-
parison to the previous results. The derivative of the radial and tangential velocity profiles
are shown in Figures 13.11 and 13.12.
Tables 13.2 and 13.3 show the values of velocities and derivatives of radial and tangen-
tial velocities. The values of Tables 13.2 and 13.3 can be compared with the finite element

FIGURE 13.3
Influence of stretching on the radial velocity profile.
Study of Incompressible Viscous Flow 205

FIGURE 13.4
Influence of stretching on the tangential velocity profile.

FIGURE 13.5
Influence of stretching on the axial velocity profile.

solution presented in Tables 13.4 and 13.5. The finite element method improves the previ-
ous solution of incompressible viscous flow due to the rotating disk. The degree of accu-
racy is higher as compared to previous methods. The present solution represents better
asymptotic behaviors since the domain of the solution is extended.
206 Mathematical Modeling for Intelligent Systems

FIGURE 13.6
Influence of stretching in the absence of the rotation of the disk on the radial velocity profile.

FIGURE 13.7
Influence of stretching in the absence of the rotation of the disk on the axial velocity profile.
Study of Incompressible Viscous Flow 207

FIGURE 13.8
Influence of rotation of the disk in the presence of constant stretching on the radial velocity profile.

FIGURE 13.9
Influence of rotation of the disk in the presence of constant stretching on the tangential velocity profile.
208 Mathematical Modeling for Intelligent Systems

FIGURE 13.10
Influence of rotation of the disk in the presence of constant stretching on the axial velocity profile.

FIGURE 13.11
Derivative of the radial velocity profile at Ω = 1 and C = 0.
Study of Incompressible Viscous Flow 209

FIGURE 13.12
Derivative of the tangential velocity profile at Ω = 1 and C = 0.

TABLE 13.1
Comparison of the Finite Element Solution with Previous Results
 dF   dG  H∞
   
 dη η = 0  dη η = 0

Cochran (1934) 0.510 −0.616 −0.844

Benton (1966) 0.5102 −0.6159 −0.8845
Kelson & Desseaux(2000) 0.510233 −0.65922 −0.884474
Bachok et al. (2011) 0.5102 −0.6159 –
Turkyilmazoglu (2014) 0.51023262 −0.61592201 −0.88447411
Present Result 0.5102134911 −0.6159097466 −0.8822962810

TABLE 13.2
Steady-State Radial, Tangential and Axial Velocities in Previous Studies
η Benton (1966) Cochran (1934) G Benton (1966) G Cochran (1934) Benton (1966) −H Cochran (1934)
0.50 0.1536 0.154 0.7076 0.708 0.0919 0.092
1.0 0.1802 0–180 0.4766 0.468 0.2655 0.266
1.5 0.1559 0–156 0.3132 0.313 0.4357 0.435
2 0.1189 0–118 0.2033 0.203 0.5732 0.572
2.5 0.0848 0–084 01313 0.131 0.6745 0.674
3 0.0581 0–058 0.0845 0.083 0.7452 0.746
3.5 0.0389 – 0.0544 – 0.7851 –
4 0.0257 0–026 0.0349 0.035 0.8251 0.826
210 Mathematical Modeling for Intelligent Systems

TABLE 13.3
Derivative of Radial and Tangential Velocities in Previous Investigations
η dF Benton (1966) dF Cochran (1934) dG Benton (1966) dG Cochran (1934)
− −
dη dη dη dη

0.50 0.1467 0.147 0.5321 0.532

1.0 −0.0157 −0.016 0.3911 0.391
1.5 −0.0693 −0.070 0.2677 0.268
2 −0.0739 −0.074 0.1771 0.177
2.5 −0.0643 −0.061 0.1153 0.116
3 −0.0455 −0.046 0.0745 0.075
3.5 −0.0319 – 0.0480 –
4 −0.0216 −0.022 0.0309 0.031

TABLE 13.4
Steady-State Radial, Tangential, and Axial Velocities Obtained Through Finite Element Procedure
η F G H
0.50 0.1536131403 0.7075852568 −0.0918753802
1.0 0.1801341021 0.4766341550 −0.2654525478
1.5 0.1558260932 0.3129478876 −0.3189416011
2 0.1187269872 0.2031634678 −0.5733460188
2.5 0.0846300712 0.1311306716 −0.6744852792
3 0.0579977303 0.0844028956 −0.7451407153
3.5 0.0387932911 0.0542459886 −0.7929703003
4 0.0255481732 0.0348269501 −0.8247142674
4.5 0.0166514271 0.0223339711 −0.8455125851
5 0.0107715388 0.0142996297 −0.8590187475
5.5 0.0069307523 0.00913324178 −0.8677299880
6 0.0044236713 0.00581107831 −0.8733141462
6.5 0.0028043468 0.0036747368 −0.8768694802
7 0.0017585143 0.0023008742 −0.8791128449
7.5 0.0010841529 0.0014173033 −0.8805095641
8 6.497563E-4 8.4901977E-4 −0.8813606353
8.5 3.701088E-4 4.8349759E-4 −0.8818603319
9 1.901497E-4 2.483802E-4 −0.8821338701
9.5 7.436672E-5 9.71374E-5 −0.8822618862
10 0 0 −0.8822962810
Study of Incompressible Viscous Flow 211

TABLE 13.5
Derivative of Radial and Tangential Velocities Through Finite Element Procedure
η dF dG
−
dη
dη
0.50 0.1467227786 0.5321338435
1.0 −0.0157293502 0.3911387287
1.5 −0.0693942529 0.2675140128
2 −0.0739028582 0.1770084427
2.5 −0.0611822145 0.1152706230
3 −0.0454564158 0.0745204320
3.5 −0.0318893113 0.0480213081
4 −0.0216399180 0.0309029877
4.5 −0.0143921656 0.0198760889
5 −0.0094532799 0.0127814181
5.5 −0.0061608755 0.0082188276
6 −0.0039953788 0.0052850668
6.5 −0.0025829536 0.0033986920
7 −0.0016665482 0.0021857371
7.5 −0.001073940 0.0014057558
8 −6.915250E-4 9.041639E-4
8.5 −4.450734E-4 5.815783E-4
9 −2.863757E-4 3.741021E-4
9.5 −1.842369E-4 2.406515E-4
10 −1.1862427E-4 1.549458E-4

13.5 Conclusion
The present investigation describes the flow of viscous fluid over a rotating and stretch-
ing plate. Without rotation of the plate, the three-dimensional flow reduces into a two-
dimensional flow case. However, in the absence of the stretching, the flow remains
three-dimensional. The present results show that the influence of stretching on the radial
and axial velocity profiles is more dominant than the rotation of the disk. The results in the
present solution are improved to a higher degree of accuracy as compared to the previous
studies.
212 Mathematical Modeling for Intelligent Systems

References
Aboul-Hassan, A. L., & Attia, H. A. (1997). Flow due to a rotating disk with Hall effect. Physics Letters,
Section A: General, Atomic and Solid State Physics, 228(4–5), 286–290. https://doi.org/10.1016/
S0375–9601(97)00086–8
Anjos, G. R., Mangiavacchi, N., & Pontes, J. (2014). Three-dimensional finite element method for
rotating disk flows. Journal of the Brazilian Society of Mechanical Sciences and Engineering, 36(4),
709–724. https://doi.org/10.1007/s40430-013-0120–0
Bachok, N., Ishak, A., & Pop, I. (2011). Flow and heat transfer over a rotating porous disk in a
nanofluid. Physica B: Physics of Condensed Matter, 406(9), 1767–1772. https://doi.org/10.1016/j.
physb.2011.02.024
Bar-Yoseph, P., & Olek, S. (1984). Asymptotic and finite element approximations for heat transfer in
rotating compressible flow over an infinite porous disk. Computers and Fluids, 12(3), 177–197.
https://doi.org/10.1016/0045–7930(84)90003–3
Benton, E. R. (1966). On the flow due to a rotating disk. Journal of Fluid Mechanics, 24(4), 781–800.
https://doi.org/10.1017/S0022112066001009
Bhandari, A. (2020). Effect of magnetic field dependent viscosity on the unsteady ferrofluid flow due
to a rotating disk. International Journal of Applied Mechanics and Engineering, 25(2), 22–39. https://
doi.org/10.2478/ijame–2020–0018
Cochran, W. G. (1934). The flow due to a rotating disc. Mathematical Proceedings of the Cambridge
Philosophical Society, 30(3), 365–375. https://doi.org/10.1017/S0305004100012561
Feistauer, M., Felcman, J., Rokyta, M., & Vlášek, Z. (1992). Finite-element solution of flow prob-
lems with trailing conditions. Journal of Computational and Applied Mathematics, 44(2), 131–166.
https://doi.org/10.1016/0377–0427(92)90008–L
Gupta, D., Kumar, L., Anwar Bég, O., & Singh, B. (2018). Finite element analysis of MHD flow of
micropolar fluid over a shrinking sheet with a convective surface boundary condition. Journal
of Engineering Thermophysics, 27(2), 202–220. https://doi.org/10.1134/S1810232818020078
Kelson, N., & Desseaux, A. (2000). Note on porous rotating disk flow. ANZIAM Journal, 42, 837.
https://doi.org/10.21914/anziamj.v42i0.624
Ram, P., & Bhandari, A. (2013). Negative viscosity effects on ferrofluid flow due to a rotating disk.
International Journal of Applied Electromagnetics and Mechanics, 41(4), 467–478. https://doi.
org/10.3233/JAE–121637
Rashidi, M. M., Hayat, T., Erfani, E., Mohimanian Pour, S. A., & Hendi, A. A. (2011). Simultaneous
effects of partial slip and thermal-diffusion and diffusion-thermo on steady MHD convective
flow due to a rotating disk. Communications in Nonlinear Science and Numerical Simulation, 16(11),
4303–4317. https://doi.org/10.1016/j.cnsns.2011.03.015
Rashidi, M. M., Mohimanian Pour, S. A., Hayat, T., & Obaidat, S. (2012). Analytic approximate solu-
tions for steady flow over a rotating disk in porous medium with heat transfer by homotopy
analysis method. Computers and Fluids, 54(1), 1–9. https://doi.org/10.1016/j.compfluid.2011.08.001
Reddy, J. N. (2019). Introduction to the Finite Element Method, Fourth Edition. https://www.accessengi-
neeringlibrary.com/content/book/9781259861901
Reddy, J. N., & Gartling, D. K. (2010). The finite element method in heat transfer and fluid dynam-
ics. In The Finite Element Method in Heat Transfer and Fluid Dynamics, Third Edition. https://
books.google.co.in/books?hl=en&lr=&id=sv0VKLL5lWUC&oi=fnd&pg=PP1&dq=Finite+el
ement+method+in+Fluid+Mechanics+JN+Reddy&ots=r-MY7AjsSs&sig=OrMQdhVV45vXpS
oYkzaOa_z-qIU&redir_esc=y#v=onepage&q=Finite element method in Fluid Mechanics JN
Reddy&f=false
Reddy, J. N., Nampally, P., & Srinivasa, A. R. (2020). Nonlinear analysis of functionally graded beams
using the dual mesh finite domain method and the finite element method. International Journal
of Non-Linear Mechanics, 127. https://doi.org/10.1016/j.ijnonlinmec.2020.103575
Study of Incompressible Viscous Flow 213

Sheikholeslami, M., & Ganji, D. D. (2014). Ferrohydrodynamic and magnetohydrodynamic effects

on ferrofluid flow and convective heat transfer. Energy, 75, 400–410. https://doi.org/10.1016/j.
energy.2014.07.089
Sravanthi, C. S. (2019). Effect of nonlinear thermal radiation on silver and copper water nanofluid
flow due to a rotating disk with variable thickness in the presence of nonuniform heat source/
sink using the homotopy analysis method. Heat Transfer - Asian Research, 48(8), 4033–4048.
https://doi.org/10.1002/htj.21581
Takhar, H. S., Bhargava, R., & Agarwal, R. S. (2001). Finite element solution of micropolar fluid flow
from an enclosed rotating disc with suction and injection. International Journal of Engineering
Science, 39(8), 913–927. https://doi.org/10.1016/S0020–7225(00)00075–6
Turkyilmazoglu, M. (2012). Three dimensional MHD stagnation flow due to a stretchable rotat-
ing disk. International Journal of Heat and Mass Transfer, 55(23–24), 6959–6965. https://doi.
org/10.1016/j.ijheatmasstransfer.2012.05.089
Turkyilmazoglu, M. (2014). Nanofluid flow and heat transfer due to a rotating disk. Computers and
Fluids, 94, 139–146. https://doi.org/10.1016/j.compfluid.2014.02.009
Vasu, B., Dubey, A., & Bég, O. A. (2020). Finite element analysis of non-Newtonian magnetohemody-
namic flow conveying nanoparticles through a stenosed coronary artery. Heat Transfer - Asian
Research, 49(1), 33–66. https://doi.org/10.1002/htj.21598
14
Instability of a Viscoelastic Cylindrical Jet:
The VCVPF Theory

Mukesh Kumar Awasthi

Babasaheb Bhimrao Ambedkar University

Sudhir Kumar Pundir

S. D. (P.G.) College

Manu Devi
Motherhood University

Dhananjay Yadav
University of Nizwa

Vivek Kumar
Shri Guru Ram Rai (PG) College

A. K. Singh
VIT University

CONTENTS
14.1 Introduction ........................................................................................................................ 215
14.2 Formulation of the Problem ............................................................................................. 217
14.3 Viscoelastic Potential Flow (VPF) Analysis ................................................................... 218
14.4 Viscoelastic Contribution for the Viscoelastic Potential Flow (VCVPF) Analysis ... 219
14.5 Dissipation Calculation for the Cylindrical Jet of Viscoelastic Fluid .........................222
14.6 Dimensionless Form of the Dispersion Relation ........................................................... 224
14.7 Results and Discussions ................................................................................................... 227
14.8 Conclusions......................................................................................................................... 229
References..................................................................................................................................... 229

14.1 Introduction
It is necessary to know the breakup behavior of liquid jet of non-Newtonian fluids because
this phenomenon has many engineering applications, including fertilizers, inkjets, paint
leveling, and roll coating, etc. The cylindrical free surface containing jets of viscoelastic
fluids can be applied in many engineering processes such as oil-drilling fiber spinning,
bottle-filling, etc.

DOI: 10.1201/9781003291916-14 215

216 Mathematical Modeling for Intelligent Systems

Viscoelastic fluids are non-Newtonian fluids that can display a response that resem-
bles that of a viscous liquid under some circumstances or the response of an elastic solid
under other circumstances. The most common type of fluids with visco-elasticity is poly-
meric liquids: melts and solutions of polymers. Middleman [1] considered the stability of
a jet in the cylindrical flow of three-constant linearized Oldroyd fluid for axisymmetric
disturbances in an inviscid medium. Goldin et al. [2] extended the study of Middleman
[1] for the generalized viscoelastic material. Bousfield et al. [3] considered the combined
effect of nonlinearity and surface tension on the breakup of viscoelastic filament. Later,
Chang et al. [4] extended the outcomes of Bousfield et al. [3] by performing an extended
wave investigation of the stretching dynamics of bead string filament for FENE as well as
Oldroyd-B fluids. Brenn et al. [5] analyzed the stability of non-Newtonian liquid linearly
taking the effects of aerodynamic forces as well as surface tension. El-Dib and Moatimid
[6] investigated the instability of a cylindrical interface of viscoelastic fluids which sup-
ports free-surface currents. Moatimid [7] also considered the nonlinear instability of two
dielectric viscoelastic fluids. The linear stability investigation of a viscoelastic liquid jet
has been studied by Liu and Liu [8,9] taking both symmetrical as well as unsymmetrical
disturbances.
In 1994, Joseph and Liao [10] have given the idea of viscoelastic flow theory. In the visco-
elastic flow concept, tangential elements of stresses which are irrotational assumed to be
zero and normal stresses balance has to include viscosity. This concept does not require
a no-slip condition at rigid boundaries. Several vorticity and theorems of circulation of
inviscid potential flow are also valid in viscous potential flow theory. At some finite value
of Reynolds number, viscous potential flow theory produces excellent physical results
in many cases. A theory, in which both pressure and motion are irrotational and nor-
mal stresses balance equation was used to include viscosity at the interface, is termed as
viscoelastic potential flow (VPF). The stability of viscous liquid jets into incompressible
and viscous gases and liquids using viscous flow theory has been measured by Funada
et al. [11]. They showed that instability may be motivated by Kelvin–Helmholtz instability
(KHI) because of the difference in the velocity at the interface and a neck-down because of
capillary instability.
At the high Weber number, the Rayleigh–Taylor stability of viscoelastic fluids was exam-
ined by Joseph et al. [12]. They found that the majority of unstable waves, are the functions
of the retardation time and it fits into experimental data excellently if the ratio of the relax-
ation time to that of retardation time is of order 103. Funada and Joseph [13] study capillary
instability of viscoelastic fluid of Maxwell-type and observe that the growth is bigger for
the viscoelastic fluid as compared to the similar Newtonian fluids. The stability of a cylin-
drical jet of viscoelastic fluid has been considered by Awasthi et al. [14] using potential
theory and observed that the elastic property enhances the instability.
If the analysis is based on VPF theory; normal stresses balance has to include viscosity
and there is no contribution of the tangential element of viscous stresses. An additional
way to analyze the stability of viscous fluids by using potential flow is the dissipation
approximation. The dissipation approximation is based on an assessment of the equa-
tions that govern the progress of energy in fields. The coefficient of the dissipation integral
includes viscosity and estimates, certain viscous stress at the boundaries. This technique
has been used by Lamb [15] and by Funada et al. [16] to study the stability of cylindrical
incompressible jet into incompressible fluids, and also by Wang et al. [17,18] to analyze the
capillary instability of two liquids.
These two theories, VPF and DM, are not the same and, therefore, it is important to
study the difference between the VPF and dissipation approximation. It was established
Instability of a Viscoelastic Cylindrical Jet 217

by Joseph et al. [19] that the above two theories could provide identical observations if
extra viscous stress in terms of a viscoelastic modification of the extra irrotational pres-
sure, which was computed through a technique based on the concept that this extra con-
tribution is developed in a very thin boundary layer. The theory based on added extra
pressure, which gives almost identical results as given by dissipation approximation, is
termed VCVPF (viscoelastic contribution of VPF). Therefore, VCVPF is the VPF adding
extra pressure. In this paper, we use VCVPF to study the stability of a cylindrical jet of
viscoelastic fluid moving in a viscous medium.

14.2 Formulation of the Problem

A liquid cylindrical jet of viscoelastic fluid of density ρ1, viscosity µ1, and of signifying
radius R moves with an unvarying axial velocity U in an infinite viscous fluid of density
ρ2, viscosity µ2 with a polar cylindrical reference frame (r , θ , z). The viscoelastic fluid cyl-
inder is lying in the region 0 ≤ r < R + η and −∞ < z < ∞ where η = η(θ , z, t) represents the
interfacial displacement. It is assumed that σ denotes surface tension at the interface of
two fluids.
If we impose small asymmetric perturbations to the balanced state, the equation of the
interface can be denoted as

f (r , θ , z, t) = r − R − η (θ , z, t ) = 0. (14.1)

the outward unit normal at the interface is given by

−1/2
∂η    1 ∂η   ∂η  
2 2
 1 ∂η
n =  er − eθ − e z  1 +   +    (14.2)
 r ∂θ ∂ z    r ∂θ  ∂ z 

where ez , er and eθ are unit vectors along the z, r, and θ directions, respectively.
In the present analysis, the velocity can be denoted as the gradient of some potential
function and the potential functions always satisfy the Laplace equation because the fluids
are incompressible, i.e.,

∇ 2φ j = 0, ( j = 1, 2) (14.3)

∂2 1 ∂ 1 ∂2 ∂2
where ∇ 2 = + + 2 + 2.
∂r 2
r ∂r r ∂θ 2
∂z
We assume that every particle, which is on the interface currently, will remain on the
interface throughout the analysis, we derive the following kinematic conditions:

∂η ∂η ∂φ1
+U = at r = R (14.4)
∂t ∂z ∂r

∂η ∂φ2
= at r = R (14.5)
∂t ∂r
218 Mathematical Modeling for Intelligent Systems

Using normal mode technique and taking the help of kinematics boundary conditions
equations (14.4) and (14.5), the solution of equation (14.3) can be expressed as

η = C e( ikz+ inθ − iω t ) + c.c. (14.6)

φ (1) = (ikU − iω )CE1 ( kr ) e( ikz+ inθ − iω t ) + c.c. (14.7)

φ (2) = −iω CE2 ( kr ) e( ikz+ inθ − iω t ) + c.c. (14.8)

where C denotes complex constant, m is the positive integer, k represents real wave num-
I ( kr ) K ( kr )
ber. ω represents growth rate and E1 ( kr ) = n , E2 ( kr ) = n . I n ( kr ) and K n ( kr ) are
I n′ ( kR) K n′ ( kR)
the modified Bessel’s function of first and second kind of order n, respectively, and c.c.
denotes the complex conjugate of the previous term.
Here, we consider the fluid is of the Oldroyd-B type [13] because it has a feature that
combines the effects of relaxation and nonlinearity, with a relative case of execution, better
than any other viscoelastic fluid. The governing equation of the linear viscoelastic fluid of
Oldroyd-B type is written as

 ∂  ∂
1 + λ1 ∂t  τ = 2 µ1 1 + λ2 ∂t  γ , (14.9)

where τ denotes viscous stress tensor, µ1 represents viscosity, γ denotes strain tensor, and
λ2 and λ1 are the retardation and relaxation time, respectively.

14.3 Viscoelastic Potential Flow (VPF) Analysis

As per dynamical condition, we know that the normal components of stresses must be
uninterrupted at the interface, i.e.,

∂ 2 φ1 ∂ 2 φ2  ∂2 η 1 ∂2 η η 
p1 − p2 − 2 µ1 + 2 µ 2 = −σ  ∂z 2 + r 2 ∂θ 2 + R 2  (14.10)
∂r 2 ∂r 2

where p j ( j = 1, 2) represent the pressure for the inner and outer fluid, respectively. If we
use the equation given by Bernoulli for irrotational flow and linearize it, we have

∂φ1 ∂φ ∂φ ∂2 φ ∂ 2 φ2
− ρ1 ( + U 1 ) + ρ2 ( 2 ) − 2 µ1 21 + 2 µ2
∂t ∂z ∂t ∂r ∂r 2
(14.11)
 ∂2 η 1 ∂2 η η 
= −σ  2 + 2 +
 ∂z r ∂θ 2 R 2 

Putting the values of η , φ1 and φ2 in equation (14.11), we get the expression

 n2 − 1 
ρ1 (ω − kU )2 E1 ( kR) − ω 2 ρ2 E2 ( kR) − 2 iµ1 k 2 ( kU − ω )F1 ( kR) − 2 iµ2 k 2ω F2 ( kR) = σ  k 2 +
 R 2 
(14.12)
Instability of a Viscoelastic Cylindrical Jet 219

I n′′( kR)  n2  1
where F1 ( kR) = =  1 + 2 2  E1 ( kR) −
I n′ ( kR)  k R  kR

K n′′( kR)  n2  1
F2 ( kR) = =  1 + 2 2  E2 ( kR) −
K n′ ( kR)  k R  kR

Since the inside fluid is viscoelastic fluid of Oldroyd-B type, viscosity µ1 is modified as
1 − iλ2ω
µ1 .
1 − iλ1ω
Using the expression for µ 1 in equation (14.12), the dispersion relation for the VPF analy-
sis can be written as

a3ω 3 + ( a2 + ib2 )ω 2 + ( a1 + ib1 )ω + ( a0 + ib0 ) = 0 (14.13)

where

a3 = λ1 ( ρ1E1 ( kR) − ρ2 E2 ( kR))

a2 = −2 kλ1Uρ1E1 ( kR)

b2 = ( ρ1E1 ( kR) − ρ2 E2 ( kR)) + 2 k 2 ( λ2 µ1 F1 ( kR) − λ1 µ2 F2 ( kR))

 n2 − 1 
a1 = λ1  k 2U 2 ρ1E1 ( kR) − σ k( k 2 + ) − 2 k 2 ( µ1 F1 ( kR) − µ2 F2 ( kR))
 R 2 

b1 = −2 kUρ1E1 ( kR) − 2 k 3Uλ2 µ1 F1 ( kR)

a0 = 2 k 3Uµ1 F1 ( kR)

n2 − 1
b0 = k 2U 2 ρ1E1 ( kR) − σ k(k 2 + )
R2

14.4 Viscoelastic Contribution for the Viscoelastic

Potential Flow (VCVPF) Analysis
Taking viscoelastic pressure as an extra pressure into the normal stress balance equation,
we can fluidly take account of the effect of shearing components of stresses in the VPF
analysis of cylindrical jet.
In this study, we are not considering the small deformation η found in the linear analy-
sis. We assume that n1 = er denotes the outward unit normal for inner fluid at the interface
and, therefore, n2 = −n1 represents the outward unit normal for outer fluid in the jet, and
also t = ez is the unit tangent vector at the interface. Here, the superscripts “i” shows irro-
tational and “v” represents viscous. The tangential and normal components of the viscous
stresses are denoted by τ s and τ n , respectively.
220 Mathematical Modeling for Intelligent Systems

The mechanical energy balance equations for outer and inner fluids are as follows;

d ρ2
∫2 ∫ ( u ⋅ T ⋅ n )dA − ∫ 2µ D
2
u2 dV = 2 2 2 2 : D2 dV
dt
V2 A V2
(14.14)
=−
∫(
A
u2 ⋅ n1 (− p2i + τ 2n ) + u2 ⋅ t τ 2s dV − ) ∫
V2
2 µ2 D2 : D2 dV

d ρ1
∫2 ∫ ( u ⋅ T ⋅ n )dA − ∫ 2µ D : D
2
u1 dV = 1 1 1 1 1 dV
dt
V1 A V1
(14.15)
=
∫ ( u ⋅ n (− p
A
1 1
i
1 + τ ) + u1 ⋅ t τ
n
1
s
1 ) dV − ∫ 2 µ D
V1
1 1 : D1 dV

where D j ( j = 1, 2) represents the symmetric component of the rate of strain tensor for inner
and outer fluids, respectively.
As the normal components of velocities are continuous at the interface, we have

u2 ⋅ n1 = u1 ⋅ n1 = un,

Adding equations (14.14) and (14.15), we have

d ρ2 d ρ1
∫2 ∫2 ∫ ( u (− p )
2 2
u2 dV + u1 dV = n
i
1 + τ 1n + p2i − τ 2n ) + u2 ⋅ t τ 2s − u1 ⋅ t τ 1s dA
dt dt
V2 V1 A
(14.16)
−
∫ 2µ D
V2
2 2 : D2 dV −
∫ 2µ D : D dV
V1
1 1 1

If we introduce two viscoelastic pressure terms, namely, p1v and p2v for the inner and outer
flow, the horizontal velocity and shear stress will be continuous at the interface, and,
therefore,

u2 ⋅ t = u1 ⋅ t = us and τ1s = τ s2 = τ s

Here, we have taken an assumption that the boundary layer estimate has an insignificant
outcome on the flow in the bulk liquid, but it disturbs the continuity and pressure condi-
tions at the interface. Therefore, the equation (14.16) can be written as

d ρ2 d ρ1
∫2 ∫2 ∫ ( u (− p )
2 2
u2 dV + u1 dV = n
i
1 − p1v + τ 1n + p2i + p2v − τ 2n ) dA
dt dt
V2 V1 A
(14.17)
−
∫
V2
2 µ2 D2 : D2 dV −
∫
V1
2 µ1 D1 : D1 dV

Now we can obtain a relation between irrotational shear stresses and viscoelastic pressure
from equations (14.16) and (14.17)
Instability of a Viscoelastic Cylindrical Jet 221

We consider here that an extra pressure termed as viscoelastic pressure will be present in
the normal stress balance equation with irrotational pressures and, therefore, the normal
stress balance equation can be written as

∂ 2 φ1 ∂ 2 φ2  ∂2 η 1 ∂2 η η 
p1i + p1v − p2i − p2v − 2 µ1 + 2 µ = −σ + + (14.23)
 ∂z 2 r 2 ∂θ 2 R 2 
2
∂r 2 ∂r 2

where p ij denote irrotational pressure, p vj representing viscous pressure for inner and outer
fluid ( j = 1, 2), respectively.
To find irrotational pressure, we use Bernoulli’s equation and then linearizing it, we
have

 ∂φ ∂φ   ∂φ  ∂2 φ ∂ 2 φ2  ∂2 η 1 ∂2 η η 
− ρ1  1 + U 1  + p1v + ρ2  2  − p2v − 2 µ1 21 + 2 µ2 = −σ  2 + 2 +
 ∂t ∂z   ∂t  ∂r ∂r 2
 ∂z r ∂θ 2 R 2 
(14.24)

On solving equation (14.22) along with equation (14.18) we get

[Ck In (kR) − Dk Kn (kR)] = 2 kC  µ1 ( ikU − iω ) E1 (kR) − µ2 iω E2 ( kR) (14.25)

Putting the values of η , φ1 ,φ2 and contribution of viscous pressure in equation (14.24) we
get the expression

ρ1 (ω − kU )2 E1 ( kR) − ω 2 ρ2 E2 ( kR) − 2 iµ1 k 2 ( kU − ω ) ×

 n2 − 1  (14.26)
( F1 (kR) + E1 (kR)) − 2iµ2 k 2ω ( F2 (kR) + E2 (kR)) = σ  k 2 +
 R 2 
222 Mathematical Modeling for Intelligent Systems

Since the inside fluid is viscoelastic fluid of Oldroyd-B type so viscosity µ1 is modified as
1 − iλ2ω
µ1 .
1 − iλ1ω
Using the expression for µ1 in equation (14.26), the dispersion relation for VCVPF solu-
tion is expressed as:

a3ω 3 + ( a2 + ib2 )ω 2 + ( a1 + ib1 )ω + ( a0 + ib0 ) = 0 (14.27)

where

a3 = λ1 ( ρ1E1 ( kR) − ρ2 E2 ( kR))

a2 = −2 kλ1Uρ1E1 ( kR)

b2 = ( ρ1E1 ( kR) − ρ2 E2 ( kR))

+2 k 2 ( λ2 µ1 ( F1 ( kR) + E1 ( kR)) − λ1 µ2 ( F2 ( kR) + E2 ( kR)) )

 n2 − 1 
a1 = λ1  k 2U 2 ρ1E1 ( kR) − σ k( k 2 + )
 R 2 

−2 k 2 ( µ1 ( F1 ( kR) + E1 ( kR)) − µ2 ( F2 ( kR) + E2 ( kR)) )

b1 = −2 kUρ1E1 ( kR) − 2 k 3Uλ2 µ1 ( F1 ( kR) + E1 ( kR))

a0 = 2 k 3Uµ1 ( F1 ( kR) + E1 ( kR))

n2 − 1
b0 = k 2U 2 ρ1E1 ( kR) − σ k(k 2 + )
R2

14.5 Dissipation Calculation for the Cylindrical Jet of Viscoelastic fluid

Here, we are going to derive dispersion relation for liquid cylindrical jet using dissipa-
tion approximations. As we know that dissipation approximation is a technique to take
account of viscous effects into the stability of fluid layer taking flow as irrotational.
The addition of the mechanical energy balance equations for inner and outer fluids can
be expressed as

d ρ2 d ρ1
∫2 ∫2 ∫ u (− p
2 2
u2 dV + u1 dV = n
i
1 + τ 1n + p2i − τ 2n ) + u2 ⋅ t τ 2s − u1 ⋅ t τ 1s  dA
dt dt
V2 V1 A
(14.28)
−
∫ 2µ D
V2
2 2 : D2 dV −
∫ 2µ D : D dV
V1
1 1 1
Instability of a Viscoelastic Cylindrical Jet 223

Here, we consider the annular cylinder together with the surface r = R; the length
of the cylinder is taken as λ . Therefore, the volume of the inner cylinder is written as
2π R z+λ
V1 =
0 0
∫∫ ∫
z
r dθ dr dz.

2π ∞ z+λ
Similarly, the volume of the outer cylinder is given by V2 =
0 R z
r dθ dr dz
∫∫ ∫
The conservation equation of momentum in linear form is expressed as

 ∂2 η 1 ∂2 η η 
p2 − τ 2n − p1 + τ 2n = σ  2 + 2 + (14.29)
 ∂z R ∂θ 2 R 2 

and also we know that horizontal velocity and stresses are continuous at the interface, i.e.,

u2 ⋅ t = u1 ⋅ t = us , τ 2s = τ 1s = τ s (14.30)

Here, the potential flow theory approximates the flows of the fluids, and therefore, we can
use Gauss’s divergence theorem, i.e.,

∫ 2 µ D : D dV = ∫ n ⋅ 2 µ D ⋅ u dA (14.31)
V A

where V is the volume bounded by the surface A and n represents outward unit normal
vector. Putting equations (14.29)–(14.31) into equation (14.28), we obtain

d ρ2 d ρ1 2   ∂2 η 1 ∂2 η η  
∫ ∫ ∫
2
u2 dV + u1 dV =  unσ  2 + 2 +  dA
dt 2 dt 2   ∂z R ∂θ 2 R 2  
V2 V1 A
(14.32)
+
∫ n ⋅ 2µ D ⋅ u dA − ∫ n ⋅ 2µ D ⋅ u dA
A
1 2 2 2
A
1 1 1 1

The integrals in equation (14.32) are solved as

2π z+λ ∞
d ρ2 ρ2 ∂
∫ ∫0 ∫z R∫ 2 ∂t u
2 2
u2 dV = 2 r dθ dzdr
dt 2
V2
(14.33)
R α  α  2
= 2 iπωρ2 λ  − iω   + iω  C E2 ( kR)exp  i (ω − ω ) t 
k  ρ2   ρ2 

2π z+λ R
d ρ1 2 ρ1  ∂ ∂
∫ ∫0 ∫z ∫0 2  ∂t + U ∂z  u
2
u1 dV = 1 r dθ dzdr
dt 2
V1
(14.34)
R α  α 
= πρ1 λ ( 2ikU − 2iω )  + ikU − iω   − ikU + iω  C 2 E1 (kR) exp  i (ω − ω ) t 
k  ρ1   ρ1 
224 Mathematical Modeling for Intelligent Systems

  ∂2 η 1 ∂2 η η  
∫  unσ  2 + 2
A 
 ∂z
+  dA
R ∂θ 2 R 2  

2π z+ λ
 n2 1  α  2
=
∫0 dθ
∫z σ  − k 2 − 2 + 2   − ikU + iω  C E1′ ( kR) exp  i (ω − ω ) t  rdz
 R R   ρ1 
(14.35)

 n2 1  α  2
= 2πλ Rσ  − k 2 − 2 + 2   − ikU + iω  C exp  i (ω − ω ) t 
 R R   ρ1 

∫ n1 ⋅ 2 µ2 D2 ⋅ u2 dA
A
(14.36)
α  α  2
= 4πλµ2 kR  − iω   + iω  C ( F2 ( kR) + E2 ( kR)) exp  i (ω − ω ) t 
ρ  ρ
2  2

∫ n1 ⋅ 2 µ1 D1 ⋅ u1 dA
A
(14.37)
α  α  2
= 4πλµ1 kR  + ikU − iω   − ikU + iω  C ( F1 ( kR) + E1 ( kR)) exp  i (ω − ω ) t
ρ 1  ρ 1

Putting the values from equations (14.33)–(14.37) into equation (14.32) and using the con-
cept that normal velocity is continuous at the interface, we have

α  α   n2 1 
ρ1  + ikU − iω  ( ikU − iω ) E1 ( kR) − ρ2  − iω  × ( − iω ) E2 ( kR) = kσ  − k 2 − 2 + 2 
 ρ1   ρ2   R R 

α  α 
+2 µ2 k 2  − iω  × ( F2 ( kR) + E2 ( kR)) − 2 µ1 k 2  + ikU − iω  ( F1 ( kR) + E1 ( kR)) (14.38)
 ρ2   ρ1 

1 − iλ2ω
By putting µ1 at the place of µ1, we are getting the same dispersion relation as we
1 − iλ1ω
obtain in VCVPF analysis.

14.6 Dimensionless Form of the Dispersion Relation

If we consider the non-dimensional variables;

λ1V λ2 V ωD ρ (2) µ (2)

D = 2R, Λ1 = , Λ2 = , ω̂ = , ρ̂ = , µ̂ =
D D V ρ (1) µ (1)

R 1 ˆ ρ (1)VD ρ (1) DV 2 U
Rˆ = = , k = kD, Re = , W= , Uˆ =
D 2 µ (1) σ V
Instability of a Viscoelastic Cylindrical Jet 225

where Re denotes the Reynolds number that is defined as the ratio of inertia forces to vis-
cous forces. The ratio of the inertia force to the surface tension force is known as Weber
number and it is denoted by W . Here, Λ 1 denotes The Deborah number and is defined as
the ratio of relaxation time to the characteristic time scale. ρ̂ and µ̂ represents the density
ratio and viscosity ratio, respectively.
The dimensionless form of the dispersion relation for the VPF analysis is written as

( ) ( ) (
aˆ 3ωˆ 3 + aˆ 2 + ibˆ2 ωˆ 2 + aˆ 1 + ibˆ1 ω̂ + aˆ 0 + ibˆ0 = 0 ) (14.39)

where

( ˆ
aˆ 3 = Λ 1 E1 ( k̂/2) − ρˆ E2 ( k/2) )
aˆ 2 = −2 kˆUˆ Λ 1E1 (k̂/2)

(
bˆ2 = E1 ( kˆ/2) − ρˆ E2 ( kˆ/2) +
2k̂ 2
Re
) (
Λ 2 F1 ( kˆ/2) − Λ 1 µˆ F2 ( kˆ/2) )
2kˆ 2
(
aˆ 3 = Λ 1 kˆ 2Uˆ 2 E1 (kˆ/2) − W −1 kˆ( kˆ 2 + 4(n2 − 1)) −
Re
) (
F1 ( kˆ/2) − µ̂ F2 ( kˆ/2) )
3
ˆ ˆ ( kˆ/2) − Λ 2k̂ UF
bˆ1 = −2kUE ˆ
ˆ 1 (k/2)
1 2
Re

2k 3 ˆ ˆ
â0 = UF1 (k/2)
Re

b̂0 = k̂ 2Uˆ 2 E1 ( kˆ/2) − W −1 kˆ(kˆ 2 + 4(n2 − 1))

The dimensionless dispersion relation for VCVPF analysis can be written as

( ) ( ) (
aˆ 3ωˆ 3 + aˆ 2 + ibˆ2 ωˆ 2 + aˆ 1 + ibˆ1 ω̂ + aˆ 0 + ibˆ0 = 0 ) (14.40)

( ˆ
where aˆ 3 = Λ 1 E1 ( k̂/2) − ρˆ E2 ( k/2) )
aˆ 2 = −2 kˆUˆ Λ 1E1 ( k/2)
ˆ

(
b̂2 = E1 ( kˆ/2) − ρˆ E2 ( kˆ/2) )
+
2kˆ 2
Re ( ( ) (
Λ 2 F1 ( kˆ/2) + E1 ( kˆ/2) − Λ 1 µˆ F2 ( kˆ/2) + E2 ( k/2)
ˆ
))
226 Mathematical Modeling for Intelligent Systems

(
aˆ 1 = Λ 1 kˆ 2Uˆ 2 E1 (kˆ/2) − W −1 kˆ( kˆ 2 + 4(n2 − 1)) )
−
2kˆ 2
Re (( F (kˆ/2) + E (k/2)
1
ˆ
1 ) − µ̂ ( F (kˆ/2) + E (k/2)
2
ˆ
))
2

ˆ3
1 2
Re
1 1 (
ˆ ˆ ( kˆ/2) − Λ 2k Uˆ F ( kˆ/2) + E (k/2)
bˆ1 = −2kUE ˆ
)

â0 =
2k 3 ˆ
Re
(
U F1 (kˆ/2) + E1 (k/2)
ˆ
)
b̂0 = kˆ 2Uˆ 2 E1 ( kˆ/2) − W −1 kˆ(kˆ 2 + 4(n2 − 1))

In the case of inviscid viscoelastic fluid (Re → ∞), the VPF and VCVPF solution gives the
same dispersion relation. From the above two expressions, it can also be observed that the
VCVPF and VPF give the same solution at the higher values of W −1, while the difference
between the two theories increases at the smaller values of W −1.
For axisymmetric disturbances, i.e., for n = 0, the dimensionless dispersion relation
becomes

  2kˆ 2 
 1 (α 1 + ρ̂α 2 )  ωˆ + ωˆ  −2kUΛ 1α 1 + i  (α 1 + ρˆ α 2 ) + Re (Λ 1 µˆ ( β 2 + α 2 ) + Λ 2 ( β1 + α 1 )) 
Λ 3 2 ˆˆ
   

ˆ3
ˆ ˆ α − i 2k Λ Uˆ ( β + α )  ωˆ

( )
2
ˆ ˆ 2 − 4)) − 2 k (( β + α ) + µˆ ( β + α )) − 2 ikU
+  Λ 1 kˆ 2Uˆ 2α 1 − W −1 k(k 1 1 2 2 1 2 1 1
 Re Re 

 2k 3 ˆ
+
 Re
( ˆ ˆ 2 − 4)  = 0
U ( β1 + α 1 ) + i kˆ 2Uˆ 2α l − W −1 k(k 

) (14.41)

If we put Deborah number Λ 1 = Λ 2 and Û = 1 in the equation (14.41), the dispersion relation
becomes

 2 kˆ 2 
(α l + ρˆ α a )ωˆ 2 +  −2kˆα l + i (( β1 + α 1 ) + µ̂ ( β 2 + α 2 ))  ω̂
 Re 
(14.42)
 2k 3

+  kˆ 2α l − i ( β1 + α 1 ) − W −1 kˆ(kˆ 2 − 4) = 0
 Re 

It is the same expression as obtained by Funada et al. [16].

Instability of a Viscoelastic Cylindrical Jet 227

14.7 Results and Discussions

The previous study of the stability of a viscoelastic cylindrical jet has been made by
Awasthi et al. [14]. They used VPF theory without taking an additional viscoelastic pres-
sure (VCVPF). In the literature, we have seen that the VCVPF solution is more stable than
the VPF solution, and therefore, we recommend VCVPF theory to study the stability prob-
lem studied here.
As we have shown that the discrepancy between VCVPF and VPF vanishes in the
inviscid limiting case. In the case of inviscid viscoelastic fluid (Re → ∞), The VPF and
VCVPF solution gives the same dispersion relation. The instability is the liquid jet is
driven by capillary instability because of surface tension and KHI because there is a
velocity difference at the interface. For Weber number, W → ∞ the effect of surface ten-
sion vanishes, so the instability in the liquid jet becomes pure KHI. If Weber number
W → 0, the instability in the liquid jet is driven by pure capillary instability. It has also
been shown that the discrepancy between the above two theories vanishes when capil-
lary instability dominates (small value of W ) and is reasonable when KHI dominates
(large value of W ).
The dispersion relation for viscoelastic cylindrical jets for both VCVPF and VPF solu-
tion is cubic in terms of growth rate parameter and instability arises due to a positive
imaginary part of the growth rate parameter (i.e., ω I > 0). The viscoelastic fluid parameters
and their properties used here are the same as taken by Awasthi et al. [14] and given in
Table 14.1.
First, we compare the growth rate curves for axisymmetric disturbance. In Figure 14.1,
the highest growth rate curves versus Weber number obtained for the VPF solution
and VCVPF solution for the cylindrical jet of viscoelastic fluid have been compared.

TABLE 14.1
Viscoelastic Fluid Parameters and Their Properties
2% PAA 2% PO

ρ1 (g/cm )
3 0.99 0.99
µ1 ( P ) 96.0 350.0
ρ2 (g/cm 3 ) 1.947 × 10−3 1.776 × 10−3
µ2 ( P) 1.8 × 10 −4 1.8 × 10−4
σ (dyn cm )−1 45.0 63.0
λ1 (s) 0.039 0.21
λ2 (s) 0.0 0.0
V(cm/s) 0.4688 0.18
Re 206.8466 1964.10
W 206.8246 1964.10
ρ̂ 1.967 × 10−3 1.794 × 10−3
µ̂ 1.875 × 10−6 5.143 × 10−7
Λ1 0.01828 0.0378
Λ2 0.0 0.0
228 Mathematical Modeling for Intelligent Systems

In Figure 14.1(a), the viscoelastic fluid (2% PAA) is taken while (2% PO) is considered in
Figure 14.1(b). It has been found that the highest growth rates for the VCVPF solution are
almost identical to the VPF solution for lower values of Weber number. At the larger values
of Weber number, the maximum growth rate obtained in the VCVPF solution is lower as
compared to the maximum growth rate obtained in the VPF solution. This concludes that
at low Weber number, the VPF and VCVPF solutions are the same but as Weber number
increases, the VCVPF solution most stable waves. The VPF solution only contains the effect
of normal stresses, while in the VCVPF solution, the effect of both normal and tangen-
tial stresses is included. Therefore, it is concluded that the effect of horizontal irrotational
stresses resists the breakup of the viscoelastic liquid jet.
Figure 14.2 compares the highest growth rate curves for the VCVPF and VPF analysis
for the asymmetric disturbances. Figure 14.2(a) is drawn for (2% PAA) viscoelastic fluid
and (2% PO) is considered in Figure 14.2(b). Here, we also obtain that the highest growth
rate obtained in the VCVPF solution is nearly the same as obtained for the VPF solution;
however, at larger values of Weber number, the VPF solution is less stable.

FIGURE 14.1
Comparison of highest growth rate curves versus Weber number for VCVPF solution and VPF solution for
axisymmetric disturbances (a) viscoelastic jet (2% PAA) and (b) viscoelastic jet (2% PO).

FIGURE 14.2
Comparison of highest growth rate curves versus Weber number for VCVPF solution and VPF solution for
asymmetric disturbances (a) viscoelastic jet (2% PAA) and (b) viscoelastic jet (2% PO).
Instability of a Viscoelastic Cylindrical Jet 229

14.8 Conclusions
The temporal stability behavior of viscoelastic cylindrical jet with both asymmetric and
axisymmetric perturbations in an infinite viscous fluid has been studied by taking the
assumption that the motion in the fluids is irrotational. In the earlier theory (VPF), the
stability of the viscoelastic cylindrical jet was investigated taking the assumption that both
pressure and motion are irrotational and normal stresses balance equation was used to
include viscosity at the interface (Awasthi et al. [14]). Here, a different irrotational theory
identified as the VCVPF which incorporates the differences in the shearing stresses and
horizontal velocities at the plane interface is used to study the instability of viscoelastic
cylindrical jet. The key results are given as follows:

1. The dispersion relation found in the dissipation approximation is the one that
comes in the VCVPF solution.
2. In both the theories, the breakup is dominated by capillary instability if Weber’s
number is low and by KHI if the Weber number is very large.
3. The discrepancy between VCVPF and VPF is significant when the Weber number
is large and the Reynolds number is small.
4. The difference between the two theories is irrelevant if capillary instability is
dominant and is reasonable when KHI dominates.

References
1. Middleman, S., Stability of a viscoelastic jet, Chem. Eng. Sci., 20 (1965) 1037–1040.
2. Goldin, M., Yerushalmi J., Pfeffer, R. and Shinaar, R., Breakup of laminar capillary jet of visco-
elastic fluid, J. Fluid Mech. 38 (1969) 689–711.
3. Bousfield, D.W., Keunings, R., Marruci, G. and Denn, M. M., Nonlinear analysis of surface
driven breakup of viscoelastic filaments, J. Non-Newtonian Fluid Mech. 21(1986) 79–97.
4. Chang, H. C., Demekhin, E. A. and Kalaidin, E., Interated stretching of viscoelastic jets, Phy.
Fluids. 11(7) (1986) 1717–1737.
5. Brenn, G., Liu, Z. and Durst, F., Linear analysis of temporal instability of axisymmetrical non-
Newtonian liquid jets, Int. J. Multiph. Flow 26 (2000) 1621–1644.
6. El-Dib, Y. O. and Moatimid, G.M., The instability of a viscoelastic conducting cylindrical inter-
faces supporting free-surface currents, Z. Naturforsch A 57 (2002) 159–176.
7. Moatimid, G. M., Non-linear instability of two dielectric viscoelastic fluids, Can. J. Phys. 82
(2004) 1109–1133.
8. Liu, Z. and Liu, Z., Linear analysis of three-dimensional instability of non-Newtonian liquid
jets, J. Fluid Mech. 559 (2006) 451–459.
9. Liu, Z. and Liu, Z., Instability of viscoelastic liquid jet with axisymmetric and asymmetric
disturbances, Int. J. Multiph. Flow 34 (2008) 42–60.
10. Joseph, D.D. and Liao, T., Potential flows of viscous and viscoelastic fluids, J. Fluid Mech. 256
(1994) 1–23.
11. Funada, T., Joseph, D. D. and Yamashita, S., Stability of liquid jet into incompressible gases and
liquids, Int. J. Multiph. Flow 30 (2004) 1177–1208.
12. Joseph, D. D., Beavers, G. S. and Funada, T., Rayleigh-Taylor instability of viscoelastic drops at
high weber numbers, J. Fluid Mech. 453(2002) 109–132.
230 Mathematical Modeling for Intelligent Systems

13. Funada, T. and Joseph, D. D., Viscoelastic potential flow analysis of capillary instability, J. Non-
Newtonian Fluid Mech. 111(2003) 87–105.
14. Awasthi, M. K., Asthana, R. and Agrawal, G. S., Viscoelastic potential flow analysis of the sta-
bility of a cylindrical jet, Sci. Iran. 21 (2014) 578–586.
15. Lamb, S. H., Hydrodynamics, NewYork: Cambridge University Press (1994).
16. Funada, T., Saitoh, M., Wang, J. and Joseph, D. D., Stability of a liquid jet into incompressible
gases and liquids: Part 2. Effects of the irrotational viscous pressure, Int. J. Multiiph. Flow, 31
(2005) 1134–1154.
17. Wang, J., Joseph D. D. and Funada, T., Viscous contributions to the pressure for potential flow
analysis of capillary instability of two viscous fluids, Phys. Fluids, 17 (2005) 052105.
18. Wang, J., Joseph, D. D. and Funada, T., Pressure corrections for potential flow analysis of capil-
lary instability of viscous fluids, J. Fluid Mech. 522 (2005) 383–394.
19. Joseph D. D. and Wang, J., The dissipation approximation and viscous potential flow, J. Fluid.
Mech. 505 (2004) 365–377.
15
Evaporative Kelvin–Helmholtz Instability
of a Porous Swirling Annular Layer

Shivam Agarwal and Mukesh Kumar Awasthi

Babasaheb Bhimrao Ambedkar University

CONTENTS
15.1 Introduction ........................................................................................................................ 231
15.2 Problem Statement ............................................................................................................. 232
15.3 Boundary and Interfacial Conditions ............................................................................. 233
15.4 Dispersion Relationship .................................................................................................... 235
15.5 Nondimensional Form ...................................................................................................... 237
15.6 Results and Numerical Discussions................................................................................ 239
15.7 Conclusions......................................................................................................................... 243
Acknowledgment ........................................................................................................................ 243
References..................................................................................................................................... 243

15.1 Introduction
The K-H instability appears when there is a velocity shear at the interface of two fluids.
The K-H instability is not only restricted to a water surface but is evident through other
natural phenomena such as the ocean, Saturn’s band, and Jupiter red spot.
Hsieh [1] first studied the K-H instability between two fluids including heat transport at
the interface and found that when the liquid is colder than the vapor, heat transport has a
destabilizing nature. Nayak and Chakraborty [2] extended the analysis of the K-H instabil-
ity in the cylindrical geometry between the vapor–liquid phases admitting heat transport
at the interface. They compared results for the plane and cylindrical geometries, which
revealed that plane configuration is more stable. Lee [3] analyzed the nonlinear stability of
the streaming liquid-vapor interface in the presence of heat transport.
The viscous potential flow (VPF) theory is a theory which deals the fluid motion as irro-
tational but viscosity is nonzero. Funada and Joseph [4] examined the instability of viscous
fluids in a horizontal rectangular channel and they considered the normal irrotational
viscous stresses into the model. It was observed that the VPF analysis agrees with the
experimental data. Awasthi and Agrawal [5] studied K-H instability in cylindrical geom-
etry. This analysis was also based on the VPF theory. Funada et al. [6] applied the stability
of a viscous liquid jet surrounded by a viscous fluid is investigated using the VPF theory.
The effect of heat transport in the VPF analysis of K-H instability at the cylindrical inter-
face was included by Asthana et al. [7]. They showed that transport of heat destabilizes
the interface but the presence of viscosity delays the instability. Awasthi et al. [8] analyzed
DOI: 10.1201/9781003291916-15 231
232 Mathematical Modeling for Intelligent Systems

the nonlinear theory of the same problem discussed by Asthana et al. [9] and found that
nonlinear effects reduce the range of stability. The effect of the porous medium on the
K-H instability was examined by Awasthi et al. [10,11]. It was established that perturbation
travels faster in a viscous medium than in a porous medium.
The swirling of cylinders also affects the instability criterion in the cylindrical configu-
ration. Fu et al. [12, 13] studied the swirling effects on interfacial instability admitting heat
transport. They considered nonviscous fluids for their study and found that swirling has
stabilizing nature. Awasthi et al. [14–16] considered the swirling effect on the VPF analysis
of stability problems including heat transport.
The K-H instability in an annular porous zone bounded by two nonflexible cylinders
is investigated in this work. The medium porosity and permeability are assumed to be
constants. The inner cylinder is stationary, whereas the outer cylinder swirls at a constant
angular velocity. The interface is responsible for the transmission of heat along with mass
from one phase to the other. The fluids are viscous, and their dynamic viscosities vary.
The mathematical equations are solved using the VPF theory, and the stability is tested
using the normal mode approach. The marginal stability curves are drawn as well as the
stability/instability criteria are explained.

15.2 Problem Statement
The porous annular region is considered with constant porosity ε 1 and permeability ki .
This region is enclosed by two cylinders of radii ai and bo ( ai < bo ) . The inside cylindrical
boundary is immovable while the outside cylinder is swirling with velocity rΩ, where Ω is
angular velocity. The fluid with density ρ ( i) , viscosity µ ( i) , and its vapor with density ρ ( o) ,
viscosity µ ( o) lie in this porous annular region and is separated by an interface r = R. The
surface tension and temperature at the interface are denoted by σ and T ( o ), respectively.
The fluid in the region ai < r < R is flowing with velocity Vi and the temperature r = ai is
represented by T ( a ). The velocity of the fluid in the region R < r < bo is Vo and the tempera-
ture at r = bo is represented by T (b ).
We use the Darcy–Brinkman model to incorporate the porous medium effects into the
analysis. The continuity equation and the Navier–Stokes equation for the considered prob-
lem can be expressed as

∇⋅q = 0 (15.1)


ρ  ∂q 1   µ  µ 
+ ( q ⋅ ∇ ) q  = −∇p − q + ∇ 2 q (15.2)
ε 1  ∂t ε 1  k i ε 1


where µ represents the fluid viscosity, p denotes pressure, and q represents the fluid
velocity.
We are interested in studying the temporal stability of the considered flow and there-
fore, a small perturbation is applied to the basic state of flow and interface takes form as
r − R = ξ (θ , z, t), where ξ (θ , z, t) represents the free-surface deflection.

E ( r , θ , z, t ) = ξ (θ , z,t) + R − r = 0 (15.3)
Evaporative Kelvin-Helmholtz Instability 233

The normal unit vector at the interface can be obtained as

 ˆ 1 ∂ξ ˆ ∂ξ ˆ 
er − eθ − ez
∇E  r ∂θ ∂ z 
n̂ = = (15.4)
∇E  1 ∂ξ
2
∂ξ 
2
 1 + 2   +   
 r  ∂θ   ∂ z  
 

In purpose of answering the mathematical problems, the potential flow theory is applied,

q j = ∇φ ( j ) (15.5)

∇ 2φ ( j ) = 0, where ( j = i, o) (15.6)

∂2 1 ∂2 1 ∂ ∂2
∇2 = + + +
∂r 2 r 2 ∂θ 2 r ∂r ∂z 2

The potential functions can also be represented as in their most basic state.

φo( j ) = Vj z ( j = i, o) (15.7)

15.3 Boundary and Interfacial Conditions

Fixed cylindrical surface r = ai and r = bo, normal velocities vanish and, therefore,

 ∂φ ( i )
 ∂r = 0 at r = ai

 (15.8)
 ∂φ ( o )
 ∂r = 0 at r = bo

The interfacial circumstances will be defined at r = R. The mass transport equation in

mathematical form as

 ∂E   ∂E 
ρ (i)  + ∇φ ( i ) ⋅ ∇E  = ρ ( o)  + ∇φ ( o ) ⋅ ∇E  (15.9)
 ∂t   ∂t 

 ∂E  1
 ∂t + ∇φ ⋅ ∇E  = Lρ ( i) S (ξ )
(i)
(15.10)


Here S (ξ ) represents the heat flux and L stands for the released latent heat.
The heat flux in the interior phase is achieved as

−K i (T ( a ) − T ( i ) ) R(log( a ) − log(R))
i (15.11)
234 Mathematical Modeling for Intelligent Systems

The heat flux in the outer phase expressed as

−K o (T ( o ) − T (b ) ) R(log(R) − log(bo )) (15.12)

Hence, the total heat flux is

1  K o (T ( o ) − T (b ) ) K i (T ( a ) − T ( i ) ) 
S(ξ ) = − (15.13)
( R + ξ )  log(bo ) − log(R + ξ ) log(R + ξ ) − log(ai ) 


Expanding the total heat flux about ξ = 0,

ξ2 ξ3
S(ξ ) = S(0) + ξ S′(0) + S′′(0) + S′′′(0) + ........ (15.14)
2 2

Considering the only linear terms, we have

S(ξ ) = S(0) + ξ S′(0) (15.15)

From the symmetric condition S(0) = 0 and, therefore,

K o (T ( o ) − T (b ) ) K i (T ( a ) − T ( i ) )
= =P (15.16)
R ( log(bo ) − log(R)) R ( log(R) − log( ai ))

equation (15.10) can be expressed in its linear form as

 ∂φ ( i ) ∂ξ ∂ξ ∂φo( i ) 
ρ (i)  − − ⋅ = βξ (15.17)
 ∂r ∂t ∂z ∂ z 

P  log ( bo ai ) 
here β =  
LR  log ( R ai ) log ( bo R ) 
The interfacial normal stress is balanced by the interfacial tension. This equation can be
expressed as

 ∂2 φ (i) 
(p( o) − p( i) ) −
1
ki
( )
∇φ ( o ) µ ( o) − ∇φ ( i ) µ ( i) − 2  µ ( o)

∂2 φ (o)
∂r 2
− µ (i)
∂r 2 
− ρ ( o)Ωo2 Rξ

 ∂2 ξ 2 1 ∂2 ξ 2 ξ 
= −σ  + 2 + (15.18)
 ∂z 2
R ∂θ 2 R 2 

As the flow is irrotational, the pressure can be computed by Bernaoulli’s equation and,
therefore, above equation takes form as

 ( o)  1 ∂φ ( o ) ∂φo( o ) 1 ∂φ ( o )  1 ( o) ( o )  1 ∂φ ( i ) ∂φo( i ) ∂φ ( i ) 1 
ρ 
 ε 1 ∂z ∂z
+ 
ε 1 ∂t  ki
+ (
µ φ − µ ( i)φ ( i ) − ρ ( i)  )
 ε 1 ∂z ∂ z
+
∂t ε 1 


2  ( o) ∂2 φ ( o) ∂2 φ (i)   ∂2 ξ 1 ∂2 ξ ξ  
+  µ − µ (i) 2 
+ σ  2 + 2 2 + 2  = ρ ( o)Ωo2 Rξ  (15.19)
ε1  ∂r 2
∂r   ∂z R ∂θ R  
Evaporative Kelvin-Helmholtz Instability 235

15.4 Dispersion Relationship
We are seeking the solution of the above equations in the context of the normal mode
technique. Using the normal modes technique, the surface elevation ξ (θ , z, t) may be
written as

ξ = B exp ( imθ − iω t ) exp(ikz) + c.c. (15.20)

here B represent perturbation’s amplitude, k is the wavenumber, m is a positive number, ω

is the growth rate, and c.c. represents a complex conjugate.

1 β 
φ (i) =  ( i)
+ ikVi − iω  BH ( i) ( kr ) exp ( imθ − iω t ) exp ( ikz ) (15.21)
kρ 

1 β 
φ (o) =  ( o)
+ ikVo − iω  BH ( o) ( kr ) exp ( imθ − iω t ) exp ( ikz ) (15.22)
kρ 

here

I m ( kr )K m′ ( kai ) − I m′ ( kai )K m ( kr )
H ( i) ( kr ) = ,
I m′ ( kR)K m′ ( kai ) − I m′ ( kai )K m′ ( kR)

I m ( kr )K′m ( kbo ) − I m′ ( kbo )K m ( kr )

H ( o) ( kr ) =
I m′ ( kR)K m′ ( kbo ) − I m′ ( kbo )K m′ ( kR)

P  log ( bo ai ) 
β=  
LR  log ( R ai ) log ( bo R ) 

Using the expression of equations (15.20)–(15.22) in equation (15.19) we have

D(ω , k ) = C0ω 2 + (C1 + iD1 )ω 2 + (C2 + iD2 ) = 0 (15.23)

1 (i) (i)
where C0 =
ε1
[ ρ H (kR) − ρ(o) H (o) (kR)

2k ( o)
C1 =  ρ Vo H ( o) ( kR) − ρ ( i)Vi H ( i) ( kR) 
ε1  

β 2
D1 =  H ( i) ( kR) − H ( o) ( kR)  +  µ ( i) k 2 G( i) ( kR) − µ ( o)k 2G( o) ( kR) 
ε1   ε1  

1 (i) (i)
+  µ H ( kR) − µ ( o)H ( o) ( kR) 
ki  
236 Mathematical Modeling for Intelligent Systems

k 2 (i) 2 (i) β  µ (i) µ(o) 

C2 =  ρ Vi H ( kR) − ρ ( o)Vo2 H ( o) ( kR)  −  i H ( i) ( kR) − ( o) H ( o) ( kR) 
ε1   ki ρ

( )
ρ 

+
(
σ k 1 − m2 − k 2 R 2 ) − 2k β  µ( ) G( ) (kR) − µ( ) G( ) (kR) − ρ( )Ω Rk
2 i
i
o
o o 2
 (i) (o)  o
R 2
ε1 ρ ρ 

 µ (i) (i) µ(o)  k

D2 = −2 k 3  VG
i ( kR) − VoG( o) ( kR)  −  µ ( i)Vi H ( i) ( kR) − µ ( o)Vo H ( o) ( kR) 
 ε1 ε1  ki

V V 
− β k  i H ( i) ( kR) − o H ( o) ( kR) 
ε
 1 ε 1 

 m2  1
Here G( i) ( kR ) =  1 + 2 2  H ( i) ( kR ) −
 k R  kR

 m2  1
G( o) ( kR ) =  1 + 2 2  H ( o) ( kR ) −
 k R  kR

So, from equation (23)

D(ω , k ) = ω 2 + ( P1 + iQ1 )ω + ( P2 + iQ2 ) = 0 (15.24)

C1 D C D
where P1 = , Q1 = 1 , P2 = 2 , Q2 = 2
Co Co Co Co
Put ω = ω r + iω i in equation (24)

(ω r2 − ω i2 ) + ( p1ω r − q1ω i ) + p2 = 0 (15.25)

ω r ( 2ω i + q1 ) + ( p1ω i + q2 ) = 0 (15.26)

Eliminating ω r between equations (15.25) and (15.26), we get

B0ω i4 + B1ω i3 + B2ω i2 + B3ω i1 + B4ω i0 = 0 (15.27)

5 2 1 2
where B0 = 1, B1 = 2 q1 , B2 = −p2 + q1 + p1
4 4

1 3 1 2 1 1 1
B3 = − p2 q1 + q1 + p1 q1 , B4 = − q22 + p1 q1 q2 − p2 q12
4 4 4 4 4

For marginal stability curve ω i = 0 and, therefore,

q22 − p1 q1 q2 + p2 q12 = 0 (15.28)

Evaporative Kelvin-Helmholtz Instability 237

Then, we get

(
 k 6 ( i) ( o)2 ( i)
( ) ) + 4 βεk {H ( ) ( kR ) H ( ) ( kR )
4

( )
2 2
 4 ε 3 ρ µ H ( kR ) G ( kR ) − ρ ( o) µ ( i) H ( o) ( kR ) G( i) ( kR )
2
(o) i o
3
 1 1

(
× ρ ( i) µ ( o)G( o) ( kR ) − ρ ( o) µ ( i)G( i) ( kR ) + 4 )} k4
ε 1 ki
2 {
ρ ( i) µ ( o) H ( i) ( kR ) H ( o) ( kR ) G( o) ( kR ) − ρ ( o) µ (i)
2 2

× H (1) ( kR ) H ( o) ( kR ) G( i) ( kR ) + } β 2 k 2 (i) (i)

ε 13 ( 2
(
2
) (
ρ H ( kR ) H ( o) ( kR ) − ρ ( o) H ( i) ( kR ) H ( o) ( kR ) ) )
+
k2
ε 1 ki
2 ( 2
(
2
)
ρ ( i) µ ( o) H ( i) ( kR ) H ( o) ( kR ) − µ ( i) ρ ( o) H ( o) ( kR ) H ( i) ( kR )
2
( ))
2
+2
β k 2 (i) (o) (i)
ε 12 ki
{ρ µ H ( kR )

) }U
 β  µ (i) µ(o) 
( ) (
2 2
× H ( o) ( kR ) − ρ ( o) µ ( i) H ( o) ( kR ) H ( i) ( kR ) 2
rel =   ( i) H ( i) ( kR ) − ( o) H ( o) ( kR )
 i  ρ
k ρ 

+2
k 2 β  µ (i) (i)
G ( kR ) −
µ(o) (o)
G ( kR )
 σk 1− m − k R
−
2 2 2 
+ ρ ( o) Ωo2 Rk 
( )

ε1  ρ ( i) ( o)  2
ρ  R 

 k 2 (i) (i)
× 2
 1ε
(
µ G ( kR ) − µ( o)G( o) ( kR ) +
β
ε 1
)
H ( i) ( kR ) − H ( o) ( kR ) ( )
2
 µ (i) (i) µ(o) (o) 
+ H ( kR ) − H ( kR )  (15.29)
 ki ki  

Here U rel = V2 − V1.

15.5 Nondimensional Form
If we assume characteristic length H = bo − ai and characteristic velocity V, other param-
eters in the dimensionless form are

βH ˆ Vi ˆ Vo ωH ρ (i) µ (i)
ς = kH, βˆ = ( o ) , Vi = V , Vo = V , ω̂ = V , ρ = ( o ) , µ = ( o ) ,
Vρ ρ µ

µ σ ˆ = R, H2 ρ ( o) HV VH
ν= , σ̂ = ( o) , ℜ o Pˆi = , Re = ( o)
, Ro = ,
ρ ρ HV 2
H ki µ Ωo

Here Re denotes the Reynolds number, Ro denotes Rossby number, and ν represents the
kinematic viscosity.
Hence, the nondimensional form of equation (15.23) is

D(ωˆ , kˆ ) = Cˆ 0ωˆ 2 + (Cˆ 1 + iD

ˆ 1 )ω̂ 2 + (Cˆ 2 + iD
ˆ 2) = 0 (15.30)
238 Mathematical Modeling for Intelligent Systems

where

Cˆ 0 =
ρ  (i) ˆ 1 ˆ ς )
H (ℜ o ς ) − H ( i) (ℜ
 o 
ε1 ρ 
2ς  ˆ ( o) ˆ ˆ ς )
Cˆ 1 = Vo H (ℜ oς ) − ρVˆi H ( i) (ℜ 
ε1 
o

ˆ
ˆ ς )  + 2ς µˆ G( i) (ℜ̂ ς ) − 1 G( o) (ℜ
ˆ ς )  + Pi µ  H ( i) (ℜ ˆ ς )
2
ˆ 1 = 1  βˆ H ( i) (ℜ
D ˆ ς ) − βˆ H ( o) (ℜ ˆ ς ) − 1 H ( o) (ℜ
ε1 
o o
 ε 1 Re  o
µ̂
o  Re  o
µ
o 
  

Ĉ2 =
ς 2  ˆ 2 (i) ˆ ˆˆ
ˆ ς )  − β Pi ν H ( i) (ℜ
ρVi H (ℜ oς ) − Vˆ o2 H ( o) (ℜ ˆ ς ) − H ( o ) (ℜ
ˆ ς ) +
ˆ 2ς 2
ˆ 1 − m2 − ℜ
σς o ( )
ε1   Re  
o o o
2
ℜ̂ o

2ς 2 βˆ  ( i) ˆ
2
− ˆ ς ) −  1  ℜ
ν G (ℜ oς ) − G( o) (ℜ ˆ ς
Re ε 1  o
  Ro  o

ˆ
ˆ 2 = − 2ς  µ̂ Vˆ i G( i) (ℜ
ˆ ς ) − 1 Vˆ G( o) (ℜ̂ ς )  − ς Pi  µVˆ H (i) (ℜ
3
D ˆ ς ) − Vˆ H ( o) (ℜ
ˆ ς )
Re  ε 1  Re  i 
o o o o o o
ε1 
ˆ ˆ
ˆ  Vi H (i) (ℜ
− βς 
 ε 1
o
ε1
ˆ ς 
ˆ ς ) − Vo H ( o) ℜ
o 

( )
Nondimensional form of equation (15.29) is

( ( )) H (ℜˆ ς ) − µρ (G (ℜˆ ς )) H (ℜˆ ς ) + 4 εβςRe H (ℜˆ ς ) H (ℜˆ ς )

 ς 6 ρ  (o) ˆ 2 2 2 ˆ 4
(i) (i) (o) (i) (o)
 4 ε 3 Re2  G ℜ oς o o o 3 o o
 1  1

( ( )) + 4 ες ReP ( ρ H (ℜˆ ς ) H (ℜˆ ς )G (ℜˆ ς ) − µ H (ℜˆ ς ) H (ℜˆ ς )

ˆi
( )
4
ˆ ς − µG( i) ℜ
× ρG( o) ℜ ˆ ς (i) (o) (o) 2 (i) (o)
o o 2 2 o o o o o
1

( )) βˆ 2ς 2
( ) ( ( )) ( ( )) ( )
ˆ ς  + ς P̂i ρ H ( i) ℜ
( ( )
2 2
ˆ ς  (i) ˆ (o) ˆ
2
ˆ ς
2
ˆ ς
×G( i) ℜ o +  ρ H ℜ oς H ℜ oς − H (i) ℜ o H (o) ℜ o  ε Re2 o
ε 13 1

( ( )) − µ ( H (ℜˆ ς )) H (ℜ̂ ς ) + 2 ςε βReP  ρH (ℜˆ ς )( H (ℜˆ ς )) − µ ( H (ℜˆ ς ))

2 2 2 ˆˆ 2 2
ˆ ς
× H (o) ℜ 2 (i) (o) i (i) (o) (i)
o o o 3 o o o
1

( )) ( ( ) ( )) + 2 ες Reβ (νG (ℜˆ ς ) − G (ℜˆ ς ))

 βˆ Pˆ 2 ˆ
× H ( o) ℜ̂ oς Uˆ rel
2 ˆ ς − H (o) ℜ
=  i ν H (i) ℜ ˆ ς (i) (o)
  Re
o o
1
o o

(1 − m ) ˆ   ς µ  G( i ) ℜ
( ) ( ) ( ( ) ( ))
2
ˆ
σς ˆ2 + 1  ℜ
2
ˆ ς − 1 G( o) ℜ̂ ς  + β̂ G( i) ℜ
ˆ ς − G( o) ℜ
ˆ ς
− 2
− ς 2ℜ o   oς   2  o o  ε o o
ℜ̂ 2o Ro  ε 1 Re  µ 1

2
P̂ µ 
Re 
o ( )
µ
ˆ ς 
ˆ ς − 1 G( o) ℜ
+ i  G( i ) ℜ o  

( ) (15.31)
Evaporative Kelvin-Helmholtz Instability 239

( )
2
Here Uˆ rel
2
= Vˆ2 − Vˆ1
In the case of viscous medium (ε 1 = 1, P̂i → 0), the relation of Awasthi et al. [15] can be
recovered from equation (15.31). If swirling is absent, we can get the relation of Awasthi [7].

15.6 Results and Numerical Discussions

The parametric values for the numerical calculation are taken as

µ ( i) = 0.00001 poise , ρ ( i) = 0.001 gm / cm3 , σ = 72.3 dyne / cm,

µ ( o) = 0.01 poise , ρ ( o) = 1.0 gm / cm3 , ε 1 = 0.3, ki = 0.0003 cm2

The effect of Reynolds number, viscosity ratio, Rossby number, heat transfer coefficient,
porosity, etc. The arrival of instability is depicted in the following graphs.
The porous medium effect on the instability of the interface is illustrated in Figure 15.1.
The marginal stability curves for the relative velocity of the interface are plotted in this fig-
ure. In this figure, we plotted asymmetric disturbances, i.e., m = 1. The lower portion of the
curve shows a stable area, while the above portion represents an unstable area. It is noted
from the figure that the stable area is more when the medium is porous and, therefore, one
can say that the perturbation expands more rapidly in the pure flow than the porous flow.
In the case of a porous medium, the flow area is less in comparison with a viscous medium
and, therefore, perturbations do not grow as quickly as they grow in the viscous medium.
As a result, the interface in a porous medium is more stable than in a viscous medium.
Now we are looking to study the swirling effect on the instability of the vapor–water
interface. Figure 15.2 shows the response of the Rossby number (Ro) on the instability of
the interface through marginal curves of relative velocity. It is detected from this figure
that as Ro increases, stability range decreases. The Rossby number Ro is defined as the
ratio of inertia force to the Coriolis force, i.e., Ro is inversely proportional to the swirling.

FIGURE 15.1
Marginal stability curves for viscous and porous medium (β̂ = 1, m = 1).
240 Mathematical Modeling for Intelligent Systems

FIGURE 15.2
Effect of Rossby number when (β̂ = 1, m = 1, Re = 100, ε 1 = 0.3, Pˆi = 1/ 0.0003).

FIGURE 15.3
Effect of Reynolds number (β̂ = 1, m = 1, Ro = 0.1, ε 1 = 0.3, Pˆi = 1/ 0.0003).

Hence, the swirling works across the development of perturbation and stabilize the inter-
face. As the Rossby number Ro increases, the swirl decreases and hence we concluded that
the swirl induces stability.
For various values of the Reynolds number, we plotted the growth of perturbations in
Figure 15.3. The stability range reduces as the Reynolds number grows, indicating that
the Reynolds number causes instability at the system’s interface. The Reynolds number is
defined as Re = ρ ( o) HV µ ( o) . If the Reynolds number increases, the viscosity of the outer
fluid ( µ ( o) ) decreases, and consequently the resistance of the fluid flow at the interface
decrease. Hence, the flow becomes unstable. If the density of the outer fluid ( ρ ( o) ) increases,
the Reynolds number also increases showing that ρ ( o) has destabilizing nature.
The variation of marginal stability curves of relative velocity for heat transport constant
()
β̂ is shown in Figure 15.4. It should be noted from the figure that as β̂ increases, the range
of instability increases, which is in the conclusion that the heat transfer at the interface
Evaporative Kelvin-Helmholtz Instability 241

FIGURE 15.4
Effect of heat transport (Re = 100, m = 1, Ro = 0.1, ε 1 = 0.3, Pˆi = 1/ 0.0003).

acts to encourage the perturbations. As the heat transport through the interface increases,
more heat is conveyed to a point in a trough and less heat is passed aside from that point.
Therefore, the amount of evaporation increases at the trough and it passes to the other side
through the interface. This added evaporation will grow the amplitude of the perturba-
tions and interface move toward instability. A similar result was achieved by Fu et al. [12]
and Awasthi et al. [9]. One can say here that the heat transport effect does not change even
in the presence of the porous medium, swirling, and/or viscosity of the fluids.
Figure 15.5 compares the marginal stability curves of relative velocity for distinct values
of porosity ε 1 . Since porosity is defined as the ratio of the volume of voids over the total
volume and therefore, more porosity implies more void space. If a perturbation gets more
void space, it rotates and, therefore, travels slower than the less void space. Hence, if the
porosity of the medium increases, the void space increases, and perturbations grow slow.
A similar observation has been obtained from Figure 15.5. We recognize that more void
volume will form a stable arrangement.

FIGURE 15.5
Effect of medium porosity (Re = 100, m = 1, Ro = 0.1, Pˆi = 1/ 0.0003).
242 Mathematical Modeling for Intelligent Systems

In Figure 15.6, we have drawn marginal stability curves for relative velocity for par-
ticular values P̂i . The range of stability enlarges on increasing the value of P̂i , and this is
an indication that P̂i induces stability into the system. Since Pˆi = H 2 ki , i.e., P̂i is inversely
proportional to the permeability (ki ). So, we can say that permeability has a destabilizing
character. Hence, if the material has a greater ability for fluid flow, the interface may get
easily unstable.
The response of the viscosity ratio of two fluids µ = µ ( i) µ ( o) has been examined in
Figure 15.7. We take a vapor–water system for this study and, therefore, µ directly depend
on vapor viscosity, while it varies inversely with water viscosity. It is detected from the
figure that as the viscosity ratio grows, the range of instability also grows. On increasing
µ, µ ( i ) increases while keeping µ ( o ) constant. The increase in vapor viscosity impedes the
flow and, therefore, perturbations travel faster. The vapor viscosity has a destabilizing
character. If outer fluid viscosity increases and vapor viscosity is fixed, the viscosity ratio
decreases, hence outer fluid viscosity induces stability.

FIGURE 15.6
Effect of medium permeability (Re = 100, m = 1, Ro = 0.1, ε 1 = 0.3).

FIGURE 15.7
Effect of viscosity ratio (Re = 100, m = 1, Ro = 0.1, ε 1 = 0.3).
Evaporative Kelvin-Helmholtz Instability 243

15.7 Conclusions
The K-H instability at the cylindrical interface of two viscous fluids is examined. We
achieve a second-order polynomial in the growth rate parameter and the imaginary part
of the growth rate parameter is illustrated to investigate the effect of physical parameters
like porous medium, heat/mass transfer, Reynolds number, and Rossby number. Our
main result indicates that the vapor–water interface is more stable in a porous medium
than the  viscous medium. We found that heat/mass transfer grows the perturbation
growth at the interface. The outer fluid viscosity has stabilizing nature, while inner fluid
viscosity destabilizes the system. The medium porosity shows a stabilizing nature while
medium permeability has destabilizing nature. The outer fluid density also has destabiliz-
ing nature.

Acknowledgment
The author S. Agarwal is thankful to CSIR, New Delhi, India (File No. 09/961(0013)/2019-
EMR-I) for their financial assistantship to carry out this work.

References
1. Hsieh, D. Y., Interfacial stability with mass and heat transfer, Phys. Fluids, 21 (1978) 745–748.
2. Nayak, A. R. and Chakraborty, B. B., Kelvin–Helmholtz stability with mass and heat transfer,
Phys. Fluids, 27 (1984) 1937–1941.
3. Lee, D. S., Nonlinear Kelvin–Helmholtz instability of cylindrical flow with mass and heat
transfer, Phys. Scr., 76 (2007) 97–103.
4. Funada, T. and Joseph, D. D., Viscous potential flow analysis of Kelvin–Helmholtz instability
in a channel, J. Fluid Mech., 445 (2001) 263–283.
5. Awasthi, M. K. and Agrawal, G. S., Viscous potential flow analysis of Kelvin–Helmholtz insta-
bility of cylindrical interface, Int. J. Appl. Math. Comput., 3 (2011) 130–141.
6. Funada, T., Joseph, D. D. and Yamashita, S., Stability of liquid jet into incompressible gases and
liquids, Int. J. Multi. Flow, 30 (2004) 1177–1208.
7. Awasthi, M. K., Study on Kelvin–Helmholtz instability of cylindrical flow with mass and heat
transfer through porous medium, J. Porous Media, 17 (2014) 457–467.
8. Awasthi, M. K. and Asthana, R., Nonlinear Study of Kelvin-Helmholtz instability of cylindri-
cal flow with mass and heat transfer, Int. Commun. Heat Mass Transfer, 71 (2016) 216–224.
9. Asthana, R., Awasthi, M. K. and Agrawal, G. S., Viscous potential flow analysis of Kelvin–
Helmholtz instability of cylindrical flow with heat and mass transfer, Heat Transfer Asian Res.,
43 (2014) 489–503.
10. Awasthi, M. K. and Asthana, R., Viscous potential flow analysis of capillary instability with
and mass transfer through porous media, Int. Commun. Heat Mass Trans., 40 (2013) 7–11.
11. Awasthi, M. K., Effect of viscous pressure on Kelvin−Helmholtz instability through porous
media, J. Porous Media, 19 (2016) 205–218.
12. Fu, Q. F., Jia, B. Q., Yang, L. J., Stability of a confined swirling annular liquid layer with heat and
mass transfer, Int. J. Heat Mass Trans., 104 (2017) 644–649.
244 Mathematical Modeling for Intelligent Systems

13. Fu, Q. F., Deng, X. D., Jia, B. Q. and Yang, L. J., Temporal instability of a confined liquid film
with heat and mass transfer, AIAA J., 56, (2018) 2615–2622.
14. Awasthi, M. K., Rayleigh–Taylor Instability of swirling annular layer with mass transfer,
J. Fluids Eng., 141 (2019) 071202.
15. Awasthi, M. K., Sarychev, V. D., Nevskii, S. A., Kuznetsov, M. A., Solodsky, S. A., Chinakhov,
D. A., Krampit, M. A., Kelvin-Helmholtz instability of swirling annular layer with heat and
mass transfer, J. Adv. Res. Dyn. Control Syst., 11 (2019) 86–96.
16. Awasthi, M. K. and Devi, M., Temporal instability of swirling annular layer with mass transfer
through porous media, Spec. Top. Rev. Porous Media Int J., 11 (2020) 61–70.
Index

advanced locking systems 62 flow sensors 9

agility 80 fuzzy DEMATEL method 85
application software 3, 26
asymptotic behaviours 205 Galerkin Scheme 171
authentication and access Gaussian–Legendre quadrature 176
control 22 generalized eigenvalue condition 163
automated lightings 62 generalized heptagonal fuzzy number 89
availability 22 glucose monitoring. 65
azimuthal velocity 204 Green Chemistry 21

Bassi Rebay 175 Hall effect sensors 9

Bernaoulli’s equation 234 hand hygiene monitoring. 65
Blockchain technology 69 HANITM 134
Brain–Computer Interfaces 76 harsh environment sensors 10
health organizations 83
cloud service provider layer 19 heart-rate monitoring 65
cold booting 42 heat flux 234
complete memory dump 47 heat-transfer coefficient 194
composite velocity 195 heptagonal fuzzy number 88
COMSOL 203 hibernation Files 45
confidentiality 22 hidden layer 12
configuration 51 highly reliable networks 8
conformable time-fractional 134 hybrid cloud(s) 80, 82
convective instability 165
cost and scale effectiveness 20 illumination 119
cost-savings 80 industrial automation 6
critical assessments 164 influencing factors 102
critical Rayleigh–Darcy number 166 infrastructure as a service 80
cryptocurrency 67 infrastructure scalability 20
cylindrical configuration 232 input layer 12
integrity 22
Darcy–Brinkman model 232 interfacial normal stress 234
data dumper 44 interpersonal influencing factors 106
day night sensors 62 intrapersonal factors 106
dedicated hardware 42 inverse conformable laplace 139
DEMATEL method 86
deploy globally 80 Jeffrey fluid-soaked 163
DG discretization 175 Jeffery parameter 166
diffusion impacts 178
digitization 9 kernel drivers 45
diverse estimates 165 kernel memory dump 47
drift current 123 K-H instability 231

elasticity 20, 80 latent heat 233

error estimation 140 Legendre polynomial 174
linguistic variables 86
firmware 3 location independence 20
Fisher–KPP 172 low power sensors 10
245
246 Index

magnetic sensors 9 Processor Dumper Utility 44

magnet RAM capture 51 pseudo-spectral scheme 172
MCC privacy needs 23 public cloud 79
mediator device 3
methodology for DumpIT 52 quasi-linearization technique 189
methodology for FTK Imager 51
miniaturization 9 radiological departments 78
mixed DG construction 173 Rayleigh–Darcy number 165
mobile network layer 19 Reynolds number 237
mobile user layer 19 Rossby number 237
monitoring the activities 62
multiclouds 80 secondary variables 202
multisensor modules 10 semiconductor 122
skin-friction 193
Newton’s routine 163 skin friction coefficient 189
NIDTM 134 small memory dump 48
normal modes technique 235 software as a service 80
software crash dumps 45
operating system 3, 26 software debuggers 45
operating system injection 45 software emulators 43
optimal homotopy analysis 150 special hardware bus 42
orthogonal properties 163 speed tracking machines 63
output layer 12 stretching plate 198
symmetric component 220
Péclet number 164
photoelectric sensors 9 tangential velocity profile 205
plastic sheet extrusion 185 Telemedicine System 83
platform as a service 80 temperaments 101
platform reliability 26 time-fractional 150
PMDump 44 toll gates 63
Poisson’s Equation 121 trapezoidal fuzzy number 87
porosity 232 triangular fuzzy number 87
Prandtl number 188
presence sensors 9 viscoelastic cylindrical jet 221
privacy and data protection 26 viscoelastic fluid 215
private cloud 80 viscoelastic pressure 220