Road Deterioration, Works

Effects and Calibration

Presenter: William D. Paterson, World Bank

HDM-4 Tutorial
9 August 2001, Seattle, WA
Road & Pavement Classification
Road Pavement Pavement Surface
Category Class Type Type
Paved Bituminous AMGB AM
AMSB
STGB ST
STAP
Concrete JPCP
JRCP
CRCP
Block

Unpaved Gravel

Earth

HDM-4 4
Modeling Framework
 For each pavement Type there is a generic model which
describes how the pavement deteriorates

 To take account of the different behaviour of a particular

pavement Type constructed with different materials, the
coefficients of the generic model depend on the different
combinations of the materials

 After maintenance treatments the generic pavement type

can change
Bituminous Pavements
Deterioration Modelling
Pavement Classification System
(Bituminous Pavements)

Surface Type Base Type Pavement Type

Asphalt Mix Granular Base AMGB
Asphalt Base AMAB
Stabilised Base AMSB
Asphalt Pavement AMAP
Surface Treatment Granular Base STGB
Asphalt Base STAB
Stabilised Base STSB
Asphalt Pavement STAP
Pavement Classification System
Surface Type Surface Material
 Asphalt concrete
AM - Asphalt Mix  Hot rolled concrete
 Polymer modified asphalt
 Rubberised asphalt concrete
 Soft Bitumen mix (cold mix)
 Porous asphalt
 Stone mastic asphalt

ST -  Cape seal
 Double bituminous surface dressing
Surface Treatment  Single bituminous surface dressing
 Slurry seal
 Penetration macadam
 1998 (STAP) 25 mm Surface
New Surfacing Dressing

1994 (AMAP) 50mm AC Overlay 50 mm AC Overlay

100 mm
50 mm AC
Previous Surfacing Surfacing
1988 (AMGB) 50mm AC Surfacing

200mm Granular Roadbase

150mm Granular Sub-base

Subgrade (CBR 8%)

Figure 1.1Case Study PROJ2: Details of existing pavements and pavement type resets
Pavement Strength
 The strength of bituminous pavements is characterised
by the adjusted structural number - SNP
 The SNP applies a weighting factor, which reduces with
increasing depth, to the sub-base and sub-grade
contributions so that the pavement strength for deep
pavements is not over-predicted.

 SNPS = SNBASUS + SNSUBAS + SNSUBG

Seasonal and Drainage Effects
 The average annual SNP used in the models is derived
from SNPd and SNPw, and the lengths of the dry and
wet seasons
 SNP = fs*SNPd
 Drainage effect on pavement strength is modelled
through the changes in the drainage factor DF [1
excellent - 5 very poor]
 f = f (MMP, DF, ACRA, APOT)
Construction Quality
 Poor construction quality results in greater variability
in material properties and performance
 Relative compaction of the base, sub-base and selected
subgrade layers - COMP
 Construction defects indicator for bituminous
surfacings - CDS (based on binder content 0.5 brittle,
1.0 normal, 1.5 soft)
 Construction defects indicator for the base - CDB
based on gradation of material, aggregate shape, etc. (0
no defects, 1.5 several defects)
 Area of Rut
Cracking depth

t1 t1
Time Time

Water Lower strength Faster

ingress deformation

Uneven
Surface Uneven
surface
Further
Spalling cracking
ROUGHNESS Shear

Potholes
Patches
Patches
Cracking
 Two types of cracking:
 Structural cracking - modelled as ‘All’ and ‘Wide’
cracking
 Transverse thermal cracking
 (Reflection cracking to be included)
 All Cracking Initiation
 ICA=Kcia{CDS2*a0exp[a1SNP+a2(YE4/SN2)]
+ CRT}


ICA time to cracking initiation, in years

CDS construction defects indicator for
bituminous surfacings

SNP structural number of pavement

YE4 annual number of ESALs, in millions/lane
 Kcia calibration factor for cracking initiation

CRT cracking retardation time due to
maintenance
 Crack Initiation
Crack Initiation

100
90
80
Percent Area of Cracking

70
60
50
40
30
20
10
0
0 5 10 15 20
Years

Kci = 1.00 Kci = 1.80 Kci = 0.55

 Crack Progression
Crack Progression

100
90
Percent Area of Cracking

80
70
60
50
40
30
20
10
0
0 5 10 15 20
Years

Kcp = 1.0 Kcp = 2.0 Kcp = 0.4

 Rut Depth
 Four components of rutting:


Initial densification

Structural deformation

Plastic deformation

Wear from studded tyres

 RDM=min[(RDO+RDST+RDPD+RDW), 100]
 Structural Deformation

Structural Deterioration - With Cracking

Rut Depth

Densification - No Cracking

Initial Densification (First Year)

Age
 Plastic Deformation
 RDPD = Krpd CDS3 a0 YE4 Sha1 HSa2


RDPD incremental increase in plastic
deformation in analysis year, in mm

CDS construction defects indicator for bituminous
surfacings

Sh speed of heavy vehicles, in km/h

HS total thickness of bituminous surfacing, mm
 Krpd calibration factor
 Roughness
 RI = Kgp [RIs + RIc + RIr + RIt ] + RIe

 RIe = Kgm m RIa


RI total incremental change in
roughness during analysis
year, in m/km IRI

m environmental coefficient
 Kgp calibration factor
 Concrete Pavements
Deterioration Modelling
 Surface types upon which the concrete RD models are based

Surface type Description

JP Jointed Plain concrete pavement - without load transfer dowels

JP Jointed Plain concrete pavement - with load transfer dowels

JR Jointed Reinforced concrete pavement

CR Continuously Reinforced concrete pavement

 Distress modes modelled in HDM-4

No. Distress mode Units of measurement Pavement surface type

1 Cracking Percent of slabs cracked JP

Number per mile JR

2 Faulting inches JP and JR

3 Spalling Percent of spalled joints JP and JR

4 Failures Number per mile CR

5 Serviceability loss Dimensionless JR and CR

6 Roughness Inches per mile (or m/km) JP, JR and CR

Structural Characteristics
 The principal data for predicting the deterioration of
concrete pavements:
 Properties of materials
 Percentage of reinforcement steel
 Drainage conditions
 Load transfer efficiency (across joints, and between
slabs and shoulder)
 Widened outside lanes
Cracking - JP Pavements
 Transverse cracking (% of slabs cracked) is modelled
as a function of cumulative fatigue damage in the slabs
and:
 Cumulative ESALs
 Temperature gradient
 Material properties
 Slab thickness
 Joint spacing
Faulting
 The average transverse joint faulting is predicted as a
function of:
 Cumulative ESALs
 Slab thickness
 Joint spacing and opening
 Properties of material
 Load transfer efficiency
 Climate/environment (FI, PRECIP, DAYS90)
 Base type
 Widened outside lanes
Roughness
 For JP concrete pavements, roughness is calculated as
a function of faulting, spalling and cracking
 For JR and CR concrete pavements, roughness is
calculated as a function of PSR
Calibration of Pavement
Relationships
 Level 1 – Desk-based study & minor field
work
 Level 2 – Based on field survey data of
sample of pavement sections
 Level 3 – Based on long-term pavement
performance data to reconfigure predictive
relationships
Pavement - Sensitivity Ranking
Deterioration Factor Typical Impact Net Impact Sensit-
Values Elastic’y % ivity
Age-environment 0.2-5.0 0.20 10 High
Cracking Initiation 0.5-2.0 0.25 6 High
Cracking Progression 0.5-2.0 0.22 6 High
Rut depth progression 0.5-2.0 0.10 3 Low
Roughness progression 0.09 0.8-1.2 1 Low
Potholing progression 0.3-3.0 0.03 2 Low
Raveling Initiation 0.2-3.0 0.01 1 Low
Pavement Models - Level 1
Calibration

Estimate unmeasured parameters from
table, e.g. environmental ‘m’

Estimate calibration factors through
secondary information (tables, etc.)

Verify that average predicted condition is
similar to current condition
Pavement Models - Level 1
Verification
 Simulation of Past

take sample of roads with historical data
(traffic, design, etc.)

predict deterioration from construction to
current age (using HDM or s/sheet)

compare results
 Average predicted condition should be
similar to current condition
16
Pavement Models - Level 2
Calibration
 Age-environment:
 sample5 sections x 2-4 climatic zones
 compute m from IRI, NE, SNC, AGE
 Crack Initiation:
 15
sections per surface-climate group
(prefer “low” distress)
 Kci = mean Obs’d / mean Predicted
 Crack progression: similarly
Level 2 Calibration of Environment coefficient ‘m’
for Roughness Progression Prediction

Sect- Surf- NE Est Init

ion Climate ace Age-R IRI (ESA m) SNC IRI Est. m Kge
1 Arid-temp AM 12 2.32 3.00 4.45 1.8 0.0139
2 Arid-temp AM 17 2.16 1.70 5.02 1.8 0.0089
3 Arid-temp AM 8 1.96 1.20 4.87 1.8 0.0075
4 Arid-temp AM 10 2.24 3.00 4.90 1.8 0.0159
Arid-temp AM 0.0116 0.503
5 Subh-temp AM 15 2.88 4.30 5.12 1.8 0.0266
6 Subh-temp AM 21 4.10 6.50 4.99 1.8 0.0337
7 Subh-temp AM 9 3.12 9.40 5.56 2.0 0.0386
8 Subh-temp AM 11 3.76 7.20 4.65 2.2 0.0361
Subh-temp AM 0.0337 1.467

m = {ln [RIt ] - ln [RI0 + 263 NE (1+SNC)^-5]}/ AGER

Adaptation of IQL-3 to IQL-2 input
Raveling to "Enduit" Score

120

100
Area of Raveling (%)

80 E-0
E-1
E-2

60 E-3
E-4
E-5
40 Means

20

0
0 1 2 3 4 5 6
"Enduit" (E) score
Pavement Models - Level 2
Calibration (cont)
 Rut depth progression (low)

20 x 200m sections (30 for thin & thick AP),
50% with mean RD> 6mm

use prediction equation & calculate geometric
means of Obs & Pred
 Ravelling initiation (low)
 Potholing Initiation (low)
Pavement Models - Level 2
Calibration (cont)
 General Roughness Progression (low)

4-yr series of reliable roughness data, 20+
sections (>10 / pavement type)

determine incremental data

calculate Predicted using algorithm

compare Obs & Pred, check residuals
Pavement Models - Level 3
 Controlled Long-term Studies

collects detailed data over time on traffic,
roughness, deflections, condition, rut depths

sections must be continually monitored

long-term (5 yr) commitment to quality data
collection
 Advanced statistical analysis & modelling

17
Works Effects
Road Works Classification
Works Works Class Works Type Works
Category Activities
Maintenance/ Routine Routine Pavement
Preservation Maintenance (Reactive)
Drainage
Routine Miscellaneous
(Cyclic)
Periodic Preventive Treatment
Maintenance
Resurfacing
Rehabilitation
Reconstruction
Special Emergency
Winter
Development Improvement Widening
Realignment
Off-carriageway
Construction Upgrading
New section

HDM-4 4
Works Standards
S ectio n A

M ain ten an ce M ain ten an ce Im p ro vem en t E tc.

S tan d ard M 1 S tan d ard M 2 S tan d ard I1

[O peration 1]+ [In tervention criteria ]+ [E ffects]+ [C ost]= "W orks item 1 "

[O peration 2]+ [In tervention criteria ]+ [E ffects]+ [C ost]= "W orks item 2 "
 Defining Road Works
 Intervention Criteria
 scheduled at a fixed time or intervals, multiple points
in time
 responsive, condition, strength, speeds, flows, accidents
 Limits: time, roughness, traffic
 Design
 pavement structure, road geometry, road type and class
 Effects
 pavement strength, condition, history, road use
 Duration and Costs

HDM-4 5
Works Effects and Costs
Note:
Road Cay = Variable at the beginning of year y
variable Cby = Variable at the end of year y

Responsive Cb1
Cbo = Ca1
intervention
Cao

Cb4

Cb2 = Ca3

Ca2

1 2 3 4 5 Years
Scheduled
intervention

Percent
total
costs

100

1 2 3 4 5 Years
Bituminous Pavements
Works Effects
 Pavement type resets after maintenance works

Works activity Existing pavement type

AMGB AMSB AMAB AMAP STGB STSB STAB STAP

Routine works AMGB AMSB AMAB AMAP STGB STSB STAB STAP

Preventive AMGB AMSB AMAB AMAP STGB STSB STAB STAP

treatment

Reseal STAP STAP / STAP STAP STGB STSB STAB STAP

1
STSB

Overlay AMAP AMAP / AMAP AMAP AMGB AMSB AMAB AMAP

1
AMSB

Inlay AMGB AMSB AMAB AMAP STGB STSB STAB STAP

Mill & replace to **AP **AP **AP **AP N/A **SB **AB **AP
intermediate
surface layer

Mill & replace to **GB **SB **AB **AP **GB **SB **AB **AP
base
Source: NDLI (1995
Drainage Works
 Drainage maintenance is modelled through its effect on
pavement strength, SNP
 f = (MMP, DF, ACRA, APOT)
 After drainage works DF is reset based on DFbw and
DMCF (drainage maintenance cost factor)
 DMCF is defined as the ratio of the annual cost of
drainage works performed to the annual cost required
to maintain the drainage system in excellent condition
Resealing
 Resealing without shape correction can repair surface
distress but cause little change to roughness or
structural strength
 Resealing with shape correction can achieve some
reduction in roughness through the filling of
depressions and repair of damaged areas
 Not performed if AGE2 < user-specified minimum
 Preparatory works applied and quantities calculated
separately, roughness reset to RIap
Overlay
 Adds structural strength, SNP
 Not performed if AGE3 is less than the user-specified
minimum, or if AGE2 < 4 or if AGE1 < 2
 Preparatory works (patching and edge repair) are
applied
 Resets surface distresses to zero, rutting to
0.15*RDMbw (by default)
 Resets roughness as a function of roughness before
works and the overlay thickness
 Resets pavement age, previous cracking, TD, SFC, etc.
Widening
 This includes lane addition and partial widening, and it
is assumed that these operations will not alter the road
alignment
 Specified by the new road type (speed-flow), road class,
increase in width or number of lanes, pavement type
for the entire section, pavement details of the widened
part of the carriageway, etc.
 Additional works include re-surfacing the existing
carriageway, patching, crack sealing
 The modelling parameters are the weighted average of
the original pavement and the pavement widening
Upgrading
 Involves pavement upgrading and geometric
improvements
 Changes the existing surface class to another surface
class of a higher performance grade
 Specify new length, width, road geometry, pavement
details, road type, road use, RD factors, etc.
 Downgrading??
Concrete Pavements
Works Effects
 Maintenance works for concrete pavements

Works Works type Works activities Pavement surface type

class
JP JR CR
Routine Routine Vegetation control, line marking, drain   
maintenance cleaning, etc.

Load transfer dowels retrofit 

Preventive Tied concrete shoulders retrofit  

treatment
Longitudinal edge drains retrofit  

Joint sealing  

Slab replacement 

Periodic Full depth repair  

Restoration
Partial depth repair 

Diamond grinding  

Bonded concrete overlay   

Rehabilitation
Unbonded concrete overlay   

Reconstruction Pavement reconstruction   

 Maintenance works applicable to CR concrete carriageway

Works type Works activity / operation ID code Ranking Unit cost

2
Reconstruction Pavement reconstruction REC 1 per m
2
Unbonded concrete overlay UOL 2 per m
Rehabilitation 2
Bonded concrete overlay BOL 3 per m
2
Restoration Full depth repair FDR 4 per m
 0 .4 0

0 .3 5

0 .3 0
A v e ra g e J o in t F a u ltin g (in )

0 .2 5

0 .2 0

0 .1 5

0 .1 0
O rig in a l
D o w e ls P la c e m e n t
0 .0 5

0 .0 0
0 .0 0 5 .0 0 1 0 .0 0 1 5 .0 0 2 0 .0 0 2 5 .0 0 3 0 .0 0 3 5 .0 0 4 0 .0 0 4 5 .0 0 5 0 .0 0

C u m u la tiv e E S A L (m illio n s p e r la n e )

F ig u r e 4 .1 R e tr o fittin g lo a d tr a n s fe r d o w e ls e ffe c t o n tr a n s v e r s e jo in t fa u ltin g

in J P c o n c r e te p a v e m e n ts w ith o u t d o w e ls
 100%

90%

80%
T ran sverse crack in g of slab s [% ]

70%

60%

50%

40%

30%

20% O rigin al
4 0% S lab s R ep lacem en t
10%

0%
0 .0 0 5 .0 0 1 0 .0 0 1 5 .0 0 2 0 .0 0 2 5 .0 0 3 0 .0 0 3 5 .0 0 4 0 .0 0 4 5 .0 0 5 0 .0 0
C u m u lative E S A L [m illio n s p e r la n e ]

F ig u re 5 .1 S la b re p la c e m e n t e ffe c t o n tra n s v e rs e c ra c k in g o f s la b s in J P
c o n c re te p a v em e n ts
 3 0 .0 0 J R C P - c r a c k in g d e t e r io r a t io n : f u ll d e p th r e p a ir

2 5 .0 0
T ra n s v e rs e C ra c k in g [ Nº / m i ]

2 0 .0 0

1 5 .0 0

1 0 .0 0

5 .0 0
O rig in a l
5 0 % F u ll D e p th R e p a ir
0 .0 0
0 .0 0 1 0 .0 0 2 0 .0 0 3 0 .0 0 4 0 .0 0 5 0 .0 0 6 0 .0 0
A g e [ y e a rs ]

F ig u r e 5 .4 F u ll d e p th r e p a ir e ffe c t o n d e te r io r a te d t r a n s v e r s e c r a c k in g in J R
c o n c r e t e p a v e m e n ts
Can We Believe HDM-4 Output?

 Yes, when calibrated

 HDM has proved suitable in a range of
different countries and conditions
 As with any model, need to carefully
scrutinise output against judgement
 If predictions are unexpected, review (a)
data, (b) calibration, (c) your judgement!
23