Elmod PDF

HIGHWAYS DEPARTMENT
GUIDANCE NOTES
ON
BACKCALCULATION OF LAYER MODULI
AND
ESTIMATION OF RESIDUAL LIFE
USING
FALLING WEIGHT DEFLECTOMETER
TEST DATA
Research & Development Division RD/GN/027A

June 2009
TABLE OF CONTENTS
1 INTRODUCTION 6
2 BASIC THEORY 8
2.1 PAVEMENT STRUCTURE 8
2.1.1 TOP LAYER 10
2.1.2 SUB-BASE 10
2.1.3 SUBGRADE 10
2.2 BACKCALCULATION OF PAVEMENT LAYER MODULI 11
2.2.1 BOUSSINESQ'S EQUATIONS 12
2.2.2 ODEMARK'S METHOD OF EQUIVALENT THICKNESS 13
2.2.3 SURFACE MODULUS 14
2.2.4 SUBGRADE NON-LINEARITY 15
2.2.5 ITERATION 15
2.2.6 WESTERGAARD'S EQUATIONS FOR RIGID PAVEMENT 16
2.2.7 LOAD TRANSFER ACROSS JOINT FOR RIGID PAVEMENT 17
2.3 ESTIMATION OF RESIDUAL LIFE 18
2.3.1 SEASON AND TRAFFIC 18
2.3.2 TEMPERATURE EFFECTS 18
2.3.3 MATERIALS 18
2.3.4 LOAD 18
2.4.5 ESTIMATION 19
2.4 SECTIONING 19
2.4.1 SECTIONING OF PAVEMENT 19
2.4.2 CUMULATIVE SUMS OF DEFLECTION 19
2.4.3 CHARACTERISTIC DEFLECTION 20
3 PREPARING FOR THE ANALYSIS 21
3.1 ELMOD INSTALLATION 21
3.2 PARAMETER AND TEMPERATURE FILES 21
3.2.1 SEASONS 22
3.2.2 MATERIALS 22
3.2.3 LOADS 30
3.3 FWD RAW DEFLECTION DATA FILES 31
3.4 PAVEMENT MODEL AND LAYER THICKNESS 31
3.5 ANNUAL AVERAGE TRAFFIC LOADING 32
4 ANALYSIS STEPS 33
4.1 OPEN FWD RAW DATA FILE 33
4.2 INPUT PAVEMENT STRUCTURE 34
RD/GN/027A Page 2 of 73
4.3 BACKCALCULATE EFFECTIVE MODULI 35
4.4 ESTIMATE RESIDUAL LIFE 37
4.5 CHECK RESULTS 37
4.6 REVIEW DEFLECTION DATA 38
4.7 REVIEW CUMULATIVE SUM OF DEFLECTIONS 39
4.8 REVIEW EFFECTIVE MODULI 40
4.9 REVIEW RESIDUAL LIFE 42
4.10 REVIEW AND ARCHIVE RESULTS 42
5 ANALYSING A FLEXIBLE PAVEMENT STRUCTURE 43
5.1 OPEN RAW DATA FILE 43
5.3 BACKCALCULATE EFFECTIVE LAYER MODULI 44
5.6 R
EVIEW DEFLECTION DATA 49
5.7 R
EVIEW CUMULATIVE DEFLECTIONS 50
5.8 EDIT PAVEMENT STRUCTURE 51
5.9 RE-BACKCALCULATE EFFECTIVE LAYER MODULI 53
5.10 REVIEW EFFECTIVE LAYER MODULI 54
6 ANALYSING A RIGID PAVEMENT STRUCTURE 63
6.1 OPEN RAW DATA FILE 63
6.3 BACKCALCULATE EFFECTIVE LAYER MODULI 64
6.6 R
EVIEW CUMULATIVE DEFLECTIONS 68
6.7 REVIEW EFFECTIVE LAYER MODULI, MODULI OF SUBGRADE REACTIONS
AND LOAD TRANSFER 69
LIST OF FIGURES
FIGURE 2.1 – TYPICAL FLEXIBLE PAVEMENT MODEL

FIGURE 2.2 – TYPICAL RIGID PAVEMENT MODEL
FIGURE 3.1 – SEASONAL CHARACTERISTICS

FIGURE 3.2 – ASPHALT FATIGUE LINES
FIGURE 3.3 – VARIATION OF ASPHALT MODULI WITH TEMPERATURE
FIGURE 3.4 – ASPHALT NEW MIX PROPERTIES
FIGURE 3.5 – ASPHALT OLD MIX PROPERTIES
FIGURE 3.6 – CONCRETE PROPERTIES
FIGURE 3.7 – SUBGRADE FATIGUE LINE
FIGURE 3.8 – SUBGRADE PROPERTIES
FIGURE 3.9 – STANDARD AXLE LOAD CONFIGURATION
FIGURE 4.1 – OPEN SINGLE FILE

FIGURE 4.2 – ELMOD STRUCTURE WINDOW
FIGURE 4.3 – ESTIMATE MODULI WINDOW
FIGURE 4.4 – FIX LAYER MODULI
FIGURE 5.1 – INPUT STRUCTURAL DATA

FIGURE 5.2 – BACKCALCULATE EFFECTIVE LAYER MODULI
FIGURE 5.3 – SELECT SECTION AND DROP FOR ANALYSIS
FIGURE 5.4 – INPUT MATERIALS AND LOADING
FIGURE 5.5 – DEFLECTION ALONG TEST RUN AFTER INITIAL ANALYSIS
FIGURE 5.6 – LAYER MODULI AFTER INITIAL ANALYSIS
FIGURE 5.7 – RESIDUAL LIFE AFTER INITIAL ANALYSIS
FIGURE 5.8 – DEFLECTION ALONG TEST RUN WITH SUSPICIOUS DATA REMOVED
FIGURE 5.9 – CENTRE DEFLECTION AND CUMULATIVE SUM
FIGURE 5.10 – INPUT SECTION 1 STRUCTURAL DATA
FIGURE 5.11 – INPUT SECTION 2 STRUCTURAL DATA
FIGURE 5.12 – BACKCALCULATE EFFECTIVE LAYER MODULI
FIGURE 5.13 – SECTION 1 LAYER MODULI
FIGURE 5.14 – SECTION 2 LAYER MODULI
FIGURE 5.15 – SECTION 1 RESIDUAL LIFE
FIGURE 5.16 – SECTION 2 RESIDUAL LIFE
FIGURE 5.17 – OPEN ‘DMS’ FILE IN EXCEL
FIGURE 5.18 – SELECT THE DELIMIT SIGN
FIGURE 5.19 – CHECK TAB AND COMMA BOXES
FIGURE 6.1 – INPUT SECTION STRUCTURAL DATA FOR RIGID PAVEMENT

FIGURE 6.2 – BACKCALCULATE EFFECTIVE LAYER MODULI FOR RIGID PAVEMENT
FIGURE 6.3 – INPUT MATERIALS AND LOADING (RIGID)
FIGURE 6.4 – DEFLECTION ALONG TEST RUN ON RIGID PAVEMENT
FIGURE 6.5 – GEOPHONE ARRANGEMENT FOR TESTING JOINTS AT RIGID PAVEMENT
FIGURE 6.6 – CENTRE DEFLECTION AND CUMULATIVE SUM ON RIGID PAVEMENT
FIGURE 6.7 – LAYER MODULI FOR RIGID PAVEMENT
FIGURE 6.8 – EFFECTIVE CONCRETE MODULUS AND MODULUS OF SUBGRADE
REACTION (RIGID)
FIGURE 6.9 – DIFFERENTIAL DEFLECTION AND LOAD TRANSFER RATIO
FIGURE 6.10 – RESIDUAL LIFE (RIGID)
1 INTRODUCTION
These Guidance Notes (GN) describe the standard procedures for estimating pavement layer effective
modulus and residual life using data acquired with the Dynatest 8002 Falling Weight Deflectometer
(FWD) and the accompanying Evaluation of Layer Moduli and Overlay Design (ELMOD) software.
This GN is written in close relation with the ELMOD software. Hence, these GN are not intended for
other evaluation software. By following these standard procedures, the data is analysed in a consistent
manner and can be used to compare the relative conditions of pavement sections in the road network.
It is expected that as more experience are gained in FWD testing and analysis, the Benkelman Beam
(BB) test will be phased out. This GN is to be read in conjunction with the accompanying
‘RD/GN/026 - Guidance Notes on Falling Weight Deflectometer Field Survey’ that stipulates the
FWD deflection data collection procedures.
Highways Department (HyD) has been using the BB test to evaluate the structural condition of
pavements. However, the BB test has the following limitations:
(a) It cannot be used for rigid pavements.
(b) The loading time for the BB test is longer than normal traffic loading times and the deflection
may be different under actual loading conditions.
(c) It cannot measure the shape of the deflection basin.
(d) It cannot analyse individual pavement layers.
(e) The BB test results are analysed according to TRRL Laboratory Report No. 833, which is
based on empirical results in the UK and may not be appropriate for local conditions.
(f) The BB test is time consuming, requiring road closure and traffic diversion.
The FWD test is considered more suitable to evaluate the structural condition of the
pavement for the following reasons:
(a) It can be used to test both flexible and rigid pavements.
(b) The loading time is comparable to that of normal traffic load, and the resulting stress, strain
and deflection correspond well with those under normal traffic.
(c) It can measure the shape of the deflection basin.
(d) It can analyse the condition of individual pavement layers.
(e) The FWD test and analysis parameters can be tailored for local materials.
(f) The FWD testing times are relatively short, road closures are usually not necessary except for
high speed roads.
In view of the limitations of the BB test, the more advanced Falling Weight Deflectometer (FWD) test
was introduced under the Road Testing Programmes consultancy in 1991 to test around 600 km of
roads in the territory. In this study, both flexible and rigid pavements were tested by FWD and
structural analysis of the data were carried out. Coring and laboratory tests were carried out to
ascertain the pavement thicknesses and material properties. Fatigue characteristics of flexible material
were obtained from indirect 4-point bending tests for two mixes available at the time of the study.
Data from these material properties are included in these GN for the establishment of the parameter
file of ELMOD software.
In 1994, a Highway Maintenance Management System Pilot Scheme consultancy Study was
commissioned to develop a pilot pavement management system for the Shatin District. A DYNATEST
FWD, together with the ELMOD analysis program, was acquired by HyD in association with project.
The FWD has been used experimentally to test the pavement. By the end of 1998, R&D Division
conducted a study with a view to promulgate the use of the FWD as a standard pavement structural
evaluation test and to phase out the BB test.
The FWD test to be conducted at the project level. The results should be considered at the project
level. The results should be considered in conjunction with other inspection and test data in order to
determine the most appropriate treatment method. Selection of the treatment method requires
engineering judgement. The overlay thickness automatically calculated by ELMOD4 should not be
directly used for rehabilitation design purposes. Overlays, if deemed to be most appropriate, should be
designed in accordance with the HyD Pavement Design Manual when it is published.
The procedures are illustrated with the use of ELMOD Version 4.4 (ELMOD4) and may need to be
updated when there is a major change in ELMOD. Periodical updates are required when the properties
of local pavement materials are better known and more analysis results are obtained to expand our
database.
Section 2 of these notes covers briefly the basic theory used by ELMOD in FWD data analysis.
Section 3 explains the preparation work required before carrying out the analysis. Section 4 describes
the detailed analysis steps in using ELMOD4. Section 5 illustrates the analysis steps for flexible
pavements with an example. Section 6 illustrates the analysis steps for rigid pavements with an
example. Appendix A is a list of references for further information.
2 BASIC THEORY
2.1 PAVEMENT STRUCTURE
The general principle in pavement modelling is that layers of similar properties should be combined
into one layer. The thickness of individual layers should be determined in accordance with Section 4.2.
The pavement model to be used for analysis should be determined as outlined below.
A flexible pavement is normally modelled as a 3-layer structure with all asphalt materials combined
into one top layer, the sub-base as the second layer, and the subgrade as the third layer. A rigid
pavement is normally modelled as a 2-layer structure with the concrete slab as the top layer and the
sub-base combined with the subgrade to form the second layer.
Sometimes the above standard models may not give reasonable results, e.g., the sub-base modulus is
lower than the subgrade modulus. The reason may be due to a non-linear subgrade modulus, normally
increasing with depth. The subgrade modulus may also be stress dependent, usually the modulus
increases as the deviator stress on the subgrade decreases. Hence the subgrade modulus measured
under the load centre is smaller than that measured at some distance away from the load. A third
reason may be the presence of an effective rigid layer at a certain depth in the subgrade. To achieve a
more reasonable backcalculation result, the subgrade may be further divided into 2 layers, with a layer
just beneath formation level of thickness 500 – 1000 mm on top of a semi-infinite bottom layer.
Overseas experiences indicate that the critical conditions in the flexible pavement structure are the
horizontal tensile strains at the bottom of the top layer and the vertical strains at the top of the
subgrade. A typical flexible pavement model is shown in Figure 2.1. The critical conditions in a rigid
pavement structure are more difficult to determine because of the presence of discontinuities at slab
joints. A typical rigid pavement model is shown in Figure 2.2. The modelling rules for the different
pavement layers are described in further details in Section 4.2.
WHEEL LOAD
ASPHALT LAYER
CRITICAL STRAIN POSITION
SUB-BASE
CRITICAL STRAIN POSITION

SUBGRADE
Figure 2.1 – Typical Flexible Pavement Model
WHEEL LOAD
CONCRETE SLAB
SUBGRADE
Figure 2.2 – Typical Rigid Pavement Model
2.1.1 TOP LAYER
For flexible pavement, all asphalt layers are combined into one layer. Note that non-structural
surfacing layers such as friction courses should be excluded from the total thickness.
For rigid pavement, the concrete slab is modelled as the top layer. For a composite pavement
where a thin asphalt friction course covers a concrete slab, the asphalt layer should not be
considered as a structural layer and only the thickness of the concrete slab should be used.
The top layer is designated as Layer 1, with thickness H1 and modulus E1.
2.1.2 SUB-BASE
For flexible pavement, the granular sub-base is modelled as a separate layer. This is
designated as Layer 2, with thickness H2 and modulus E2.
For rigid pavement with a granular sub-base, the difference in magnitude between the
concrete slab modulus and the underlying layers is sufficiently large such that the difference
in modulus between the granular sub-base and subgrade becomes insignificant. The sub-base
should be combined with the semi-infinite subgrade into one layer and designated as Layer 2
with modulus E2.
If a lean concrete sub-base is used, it is assumed that the lean concrete has cracked into pieces
of size 300mm or above. This size is equal to or larger than the size of the FWD loading plate
and the deflections in the neighbourhood of the loading plate cannot be correctly evaluated
with elastic theory. Therefore the sub-base should also be combined with the semi-infinite
subgrade into one layer and designated as Layer 2 with modulus E2.
2.1.3 SUBGRADE
The subgrade is normally modelled as one semi-infinite layer. In the 2-layer model, the
sub-base is combined with the subgrade and designated as Layer 2, with modulus E2. In the
3-layer model, the subgrade is designated as Layer 3, with modulus E3.
In a 4-layer model, the additional subgrade layer of thickness 500mm to 1000mm just below
formation level is designated as Layer 3, with thickness H3 and modulus E3 and the
underlying subgrade is designated as Layer 4, with modulus E4. However, ELMOD requires
that a ratio E2/E3 be fixed for a 4-layer model. For granular materials, ELMOD assumes the
ratio E2/E3 = 0.2 * H2 0.45 by default.
The use of a 5-layer model is normally not recommended. At some locations, a rigid layer, e.g.
rock head, may exist at a known depth below the pavement structure. In such cases, the depth
to the rigid layer should be entered into ELMOD.
2.2 BACKCALCULATION OF PAVEMENT LAYER MODULI

There are various software programs available for estimating the stiffness moduli of the pavement
layers. Different programs may apply different principles. ELMOD uses an approximate method
based on Boussinesq’s equations and Odemark’s method of equivalent thickness to estimate the layer
moduli. This Section outlines the basic theories used.
2.2.1 BOUSSINESQ’S EQUATIONS
The ‘Radius of curvature’ method adopted by ELMOD makes use of the Boussinesq’s
equations. Boussinesq developed a set of equations to calculate the stress, strain and
displacement conditions in a homogeneous, isotropic, linear elastic semi-infinite space under
a point load[9]. The stress, strain and displacement conditions under a uniform load can be
found by integration. At the depth ‘z’ below the centreline of a uniform circular load ‘σ0’ with
radius ‘a’, the stress, strain and displacement are given by the following:
σz = σ0 x {1– 1/[1+(a/z)2]3/2}
σr = σt = σ0 x {(1+2μ)/2 – (1+μ) / [1+(a/z)2]1/2 + (1/2) / [1+(a/z)2]3/2}
εz = (1+μ) x σ0/E x {(z/a) / [1+(z/a)2]3/2 – (1–2μ) {(z/a) / [1+(z/a)2]1/2 –1}}
dz = (1+μ) x σ0 x a/E x {1 / [1+ (z/a)2]1/2 + (1–2μ) x {[1+(z/a)2]1/2 – z/a}}
R = E x a/[(1–μ2) x σ0] / {1+[1+ 3/2/(1-μ)] x (z/a)2} x [1+(z/a)2] 5/2
εr = z / 2 / R
where σz = vertical stress;

σr = radial stress;
σt = tangential stress;
εz = vertical strain;
εr = horizontal strain;
dz = vertical displacement;
R = radius of curvature;
Ε = modulus;
and μ = Poisson ratio.
Note that the horizontal strain at the bottom of a bituminous layer, often the critical strain in
the pavement structure, can be found by first calculating the radius of curvature of the plane at
the bottom of Layer 1.
2.2.2 ODEMARK’S METHOD OF EQUIVALENT THICKNESS
Boussinesq’s equations are only applicable to a homogeneous layer. In practice, most
pavement structures are not homogeneous but are layered systems. Odemark developed an
approximate method to transform a system consisting of layers with different moduli into an
equivalent system where the thicknesses of the layers are altered but all layers have the same
modulus. This is known as the Method of Equivalent Thickness. The transformation assumes
that the stiffness of the layer remains the same, i.e.
I x E / (1 – μ2) remains constant

where I = moment of inertia;
E = layer modulus;
and μ = Poisson ratio.
Since I is a function of the cube of the layer thickness, the equivalent thickness transformation
for a layer with thickness = h1, modulus = E1, and Poisson ratio μ1 into a layer with equivalent
thickness = he, modulus E2, and Poisson ration μ2 may be expressed as follows:
h13 x E1 / (1 – μ12) = he3 x E2 / (1 – μ22); or

he = h1 x [E1 / E2 x (1 – μ22) / (1 – μ12)] 1/3.
Since this is an approximate method, an adjustment factor ‘f’ is applied to the right hand side
of the above equation to obtain a better agreement with elastic theory. The value of ‘f’
depends on the layer thicknesses, modular ratios, Poisson ratios and the number of layers in
the pavement structure. Furthermore, the Poisson ratio for all pavement materials can be
assumed to be the same, usually equal to 0.35. The equivalent thickness equation can
therefore be expressed as:
he = f x h1 x [E1 / E2] 1/3.
To analyse a multi-layer pavement structure with known layer moduli, the layers can be
successively transformed into an equivalent system with a homogeneous layer modulus equal
to the modulus of the semi-infinite subgrade layer by applying Odemark’s method.
Boussinesq’s equations can then be applied to calculate the stress, strain and displacement
conditions within the equivalent layered system. In analysing FWD data, the process is
reversed by using the surface displacements measured at varying distances under a plate load
to ‘backcalculate’ the moduli of individual pavement layers.
The backcalculated modulus is often called the effective modulus because the value
represents the effect of the layer within the whole pavement structure. This may be different
from the modulus obtained if the layer is evaluated in isolation, such as in the case of testing a
cored sample in the laboratory.
2.2.3 SURFACE MODULUS

The surface modulus is the ‘weighted mean modulus’ of the semi-infinite space calculated
from the surface deflection using Boussinesq’s equations. The surface modulus at a distance
‘r’ roughly reflects the surface modulus at the same equivalent depth z = r. If the subgrade is a
linear elastic semi-infinite space, the surface modulus should be the same at varying distances.
If a stiff layer is present, the surface modulus at some distance should become very large.
ELMOD begins the process by estimating the subgrade modulus using the outer deflections
since these are almost entirely controlled by the subgrade. The change in moduli with varying
distances from the load centre is used to check whether a stiff layer is present at some depth.
This can be checked by calculating the ‘surface modulus’ as follows:
E0(0) = 2 x (1 - μ2) x σ0 x a / d0(0) and
E0(r) = (1 - μ2) x σ0 x a2 / [ r x d0(r) ] (valid for r > 2a)
where E0(r) = surface modulus at distance r;

μ = Poisson ratio of the subgrade (normally = 0.35);
σ0 = uniform stress on the plate;
a = radius of the loading plate;
r = distance from the centre of load;
and d0(r) = surface deflection at distance r.
Note that the equation for E0(r) is only valid for r > 2a.
2.2.4 SUBGRADE NON-LINEARITY
If a stiff layer is not detected, ELMOD calculates the subgrade non-linearity coefficients ‘C’
and ‘n’ using the following equation:
E0 = C x (σ1 / σ) n
where E0 = surface modulus;
σ1 = major principal stress;
σ = reference stress, normally 160 MPa;
C = constant;
and n = negative constant.
Normally ‘C’ decreases almost linearly with the increase in moisture content. ‘n’ may be
taken as a measure of the non-linearity. If n is zero, the subgrade is linear elastic and as n
decreases, the non-linearity becomes more and more pronounced.
2.2.5 ITERATION
With the ‘Radius of Curvature’ method, ELMOD uses the centre deflection and the curvature
of the deflection basin under the loading plate to determine the moduli of the top layer, and
the intermediate layers if they are present. The subgrade modulus under the load centre is
adjusted according to the estimated stress level, and the outer deflections are checked. If
adjustments are necessary, the layer moduli are then recalculated.
The ‘Deflection Basin Fit’ method goes one step further by closely matching the calculated
deflection profile and the measured deflection profile. The percentage difference between the
calculated value and the measured value can be specified as the convergence criteria in the
iteration.
2.2.6 WESTERGAARD’S EQUATIONS FOR RIGID PAVEMENT
For rigid pavement, a special problem occurs due to the discontinuities at the concrete slab
joints. Elastic theory may be used for the slab centre but is not applicable close to a slab joint
or corner. Westergaard developed an approximate method based on the assumption that the
subgrade does not transfer shear stress[10]. This implies that the reaction of the subgrade on the
slab is equal to the deflection multiplied by a constant ‘k’ called the modulus of subgrade
reaction. ‘k’ can be calculated as follows:
k = 0.54 x Es / he
where Es = modulus of subgrade;
and he = equivalent thickness of the concrete slab with respect to the subgrade.
Based on the above, Westergaard developed equations that were later modified by Ioannides,
Thompson and Barenberg[11] to calculate the stresses and deflections for a concrete slab as
follows:
σI = {3P x (1 + μ) / 2h2} x { [ln (2L / B) + 0.5 – τ] / π + 1 / 32 x (B / L) 2}

dI = [P / (8kL2) ] x {1 + (1 / 2π) x [ln (a / 2L) + τ – 1.25] x (a / L) 2}
σE = {3P x (1 + μ) / [(3 + μ) x πh2] } x {ln (E x h3 / 100ka4) + 1.84 – 4μ / 3
+ (1 – μ) / 2 + 1.18 x (1 + 2μ) x (a / L)}
dE = [P x (2 + 1.2μ) 0.5] / { (Ekh3) 0.5 } x [1 – (0.76 + 0.4μ] x (a / L) }
σC = (3P / h2) x [1 – (C / L) 0.72]
dC = P/kL2 x (1.205 – 0.69C / L)
where σI = maximum bending stress when load is in interior position;

dI = deflection at centre of slab;
σE = maximum bending stress when load is at edge position;
dE = deflection when load is at edge position;
σC = bending stress when load is at corner position;
dC = deflection at corner position;
P = uniformly distributed single wheel load;
h = concrete slab thickness;
E = concrete modulus;
μ = Poisson ratio of concrete;
k = modulus of subgrade reaction;
a = radius of loaded area;
C = side length of a square loaded area;
B = (1.6a2 + h2) 0.5 – 0.675h for a < 1.724h;
B=a for a > 1.724h;
L = {Eh3 / [12k x (1 – μ2)] 0.25;
and τ = Euler’s constant (= 0.5772).
In the above equations, ‘interior position’ means the load is at a considerable distance from
the edges and ‘edge position’ means the load is at the edge but at considerable distance from
any corner. While the equations for ‘interior position’ are reasonably accurate when the
smallest slab dimension is greater than 3L, Westergaard’s equations are most useful for edge
and corner loadings.
2.2.7 LOAD TRANSFER ACROSS JOINT FOR RIGID PAVEMENT

ELMOD uses the Odemark-Boussinesq method to calculate the layer moduli at the centre of
the slab. The layer moduli are then used with Westergaard’s equations to calculate the load
transfer across joints. The load transfer efficiency factor ‘j” is defined as follows:
j = 1 – (dj – dj’) / (de – de’)
where dj and dj’ are deflections of adjacent slabs at any point across the joint;
and de and de’ are deflections at the same point if the joint transferred no load.
ELMOD can plot on screen the graphs of moduli of each layer, the non-linear properties of
the subgrade material ‘C’ and ‘n’, the modulus of concrete slab ‘E1’ and modulus of subgrade
reaction ‘k’, the differential deflection ‘Dd’, and the percentage load transfer efficiency ‘LT%’
across a joint. A small ‘Dd’ and a high ‘LT%’ indicates better joint integrity.
2.3 ESTIMATION OF RESIDUAL LIFE
2.3.1 SEASON AND TRAFFIC
Design traffic load in standard axles are required for the estimation of residual life. They can
be calculated in accordance with RD/GN/017 or the HyD Pavement Design Manual when it is
published. For ELMOD input, Section 3.5 shall be referred.
2.3.2 TEMPERATURE EFFECTS

Seasonal variation in traffic load affects the estimated traffic load used for estimation of
residual life. The moduli determined by ELMOD for a given test point naturally correspond
to the climatic conditions that happen to prevail during the testing. These conditions are not
likely to be representative of design conditions and the moduli must therefore be adjusted.
Asphalt material is sensitive to temperature changes. Pavement temperature at the time of

collecting the deflection data varies with time and season. Hence, a reference temperature
shall be adopted for comparsion and design purpose. The backcalculated asphalt moduli has
to be corrected to this reference pavement temperature before estimating the pavement
residual life. Besides, unbound material (granular sub-base and subgrade) is also sensitive to
seasonal and annual temperature variations.
2.3.3 MATERIALS
The general equation relating traffic loading to pavement deterioration, called the Transfer
Function, is as follows:
Permissible stress/strain = a x N –b x (E / Eref) c

where a, b and c = constants;
N = allowable stress/strain repetitions to fatigue/rutting failure;
E = modulus of the material; and
Eref = a reference modulus that converts the transfer function to
proper units.
Details of the values of constants a, b and c for different material are available in
Section 3.2.2.
2.3.4 LOAD
The 80kN standard axle load is adopted. All design traffic loads are expressed in
terms of equivalent standard axle load. For ELMOD input, Section 3.2.3 shall be
referred.
2.3.5 ESTIMATION
Backcalculated effective layer moduli are corrected for seasonal and temperature effect.
Critical strains / stresses are determined based on the design load and the assumed theory
adopted by each individual program. These critical strains / stresses at locations described in
Section 2.1 are compared with the failure characteristics of each material and a set of
allowable strain / stress repetitions of each material is obtained. Based on this data set, critical
load repetition values for layers are obtained. With the past and estimated future traffic load, a
set of estimated residual life of each individual pavement layer can be obtained for review.
Based on this data set, different maintenance options can be considered to restore the
pavement sections to desirable condition.
2.4 SECTIONING
2.4.1 SECTIONING OF PAVEMENT
The deflection data obtained from the FWD test should be reviewed to see if there are large
fluctuations or inconsistencies. The cumulative sums of the centre deflection can be used to
assist in deciding whether the length of the test run should be divided into different sections
for design purposes. In order to help observe if there are significant changes in trend along the
test run, method of cumulative sum can be adopted.
2.4.2 CUMULATIVE SUMS OF DEFLECTION

The cumulative sum A.d. at the ith station is defined as follows:
A.d. = Σδi − i μ
where Σδi = sum of deflections from the 1st station to the ith station inclusively;
i = number of stations from δ1 to δi inclusively;
and μ = mean deflection of the test run
From the cumulative sum, check if there are significant changes in trend (from upward to
downward or vice versa). A homogeneous section is one in which the cumulative sums
continue in the same upward or downward trend. A significant change in the upward or
downward trend indicates a change in section. Divide the test run into homogeneous sections
such that the trend is reasonably consistent within each section. This step is illustrated in
Section 5.7. If a former analysis has been carried out before the sectioning, the homogeneous
sections shall be re-backcalculated and analysed again. These steps are illustrated in Sections
5.8 and 5.9.
2.4.3 CHARACTERISTIC DEFLECTION
For design purposes, a set of characteristic values representative of the section has to be
selected. This is done statistically by choosing the set of values at the test station where the
centre deflection is at the 85 percentile of all the centre deflections in the section.
The effective moduli and residual life calculated for this test station can be used in the
subsequent rehabilitation design.
3 PREPARING FOR THE ANALYSIS
To get started in analysing the FWD data, the following are required:
a. A PC installed with the ELMOD program;

b. The customised HK042009.WPR parameter file and the OLD200.TEM and NEW200.TEM
temperature correction files for the old and new mix bituminous materials respectively;
c. The FWD raw data files acquired with the Dynatest equipment;
d. The pavement model and thickness of individual pavement layers of the pavement structure
being evaluated; and
e. The estimated average annual traffic loading for the road section being analysed.
3.1 E
LMOD INSTALLATION
The current version of ELMOD is Version 4.5.35. Copies of ELMOD licences will be provided to
relevant offices to perform the analysis. Relevant offices are required to provide a PC with minimum
32MB RAM and running Windows 98/2000/Me/NT/XP (English or Chinese) to install and run the
program[4].
3.2 PARAMETER AND TEMPERATURE FILES

A variety of tests have been conducted on samples of two different bituminous mixes in 1991 Road
Testing Programmes consultancy. Based on test results and the recommendations in the consultancy
reports, the HK042009.WPR parameter file and the OLD200.TEM and NEW200.TEM temperature
correction files are compiled and are available in R&D Division.
ELMOD uses 3 sets of parameters: seasons, materials and loads to estimate the residual life of the
pavement structure. Use ‘Parameter’ from the pull down menu to open the ‘HK042009.WPR’
parameter file. The ‘HK’ in the filename implies this parameter file is tailored for use under Hong
Kong conditions and the ‘042009’ in the filename implies the parameters are prepared in April 2009.
This file needs to be updated as new materials are introduced or knowledge on the properties of
existing materials and environmental effects is improved. The specific settings in HK042009.WPR are
described below.
3.2.1 SEASONS
ELMOD allows the user to set up seasonal variations in traffic loads. Currently, there is no
local data available for this variation and hence a single season of 52 weeks, with
characteristic week set at mid-year, is adopted. This setting needs to be reviewed in the future.
The ‘number of seasons’ is set to 1, the ‘No. of weeks in season’ is 52, the ‘Characteristic
week in season’ is 26, the ‘Date’ is entered in ELMOD as Jun 26, and the ‘% loads in season’
is 100.
Figure 3.1 – Seasonal Characteristics
Seasonal variations of unbound layers are also not set in Section 3.2.2.
3.2.2 MATERIALS
ELMOD uses the deterioration properties of pavement materials to estimate the damage under
traffic load. Default settings of ELMOD4 are changed to suit local conditions. Six types of
commonly used pavement materials are set in the parameter file HK042009.WPR:
Asphalt New Mix, Asphalt Old Mix, Concrete, Granular Sub-base, Lean Concrete Sub-base
and Subgrade.
They shall be amended when updated information on material is obtained. Besides, new
material properties can be added if new material or asphalt mix is adopted.
Asphalt New Mix refers to mixes designed to the new bituminous materials specifications
after the November 1988 Road Note 2 and Asphalt Old Mix refers to mixes designed to the
old specifications before this. The Road Testing Programme Consultants determined their
fatigue characteristics by carrying out laboratory tests on in-situ asphalt pavement materials.
The fatigue lines for the asphalt mixes are shown in Figure 3.2 below, where Flex New =
Asphalt New Mix and Flex Old = Asphalt Old Mix:
Figure 3.2 – Asphalt Fatigue Lines
The variation of the Asphalt New Mix modulus with temperature is given by the data in the
NEW200.TEM file and the variation of the Asphalt Old Mix modulus with temperature is
given by the data in the OLD200.TEM file. The temperature files are for an average layer
thickness of 200 mm. For average layer thickness exceeding 200mm, the variation of the
asphalt modulus with temperature does not deviate much as revealed from temperature files
for average layer thickness of 50, 100 and 400 mm. Hence, they are not included.
The variation of asphalt moduli with temperature is shown in Figure 3.3 below, where
Flex_new = Asphalt New Mix and Flex_old = Asphalt Old Mix:
Figure 3.3 – Variation of Asphalt Moduli with Temperature
The failure criteria for Asphalt New Mix is as follows:
Permissible strain = 175.44 μstrain * N –0.223 * (E / Eref) C

where N = number of load repetition;
E = modulus of Asphalt New Mix material;
Eref = 2100 MPa at 33.8o C;
and C = 0.
Figure 3.4 – Asphalt New Mix Properties
The failure criteria for Asphalt Old Mix is as follows:
Permissible strain = 272.85 μstrain * N –0.241 * (E / Eref) C

E = modulus of the Asphalt Old Mix material;
Eref = 4000 MPa at 33.8o C;
and C = 0.
Figure 3.5 – Asphalt Old Mix Properties
The Concrete properties are based on those adopted by the Highway Maintenance
Management Pilot Scheme Consultants. The failure criteria is as follows:
Permissible stress = 195 MPa * N –0.333 * (E / Eref) C

E = modulus of Concrete material;
Eref = 35000 MPa at 34o C;
and C = 1.
Figure 3.6 – Concrete Properties
The Subgrade properties are based on the Shell Pavement Design Manual[8]. Both the Road
Testing Program Consultants and the Highway Maintenance Management Pilot Scheme
Consultants adopted these properties. The fatigue line for the Subgrade is shown in Figure 3.7
below:
Figure 3.7 – Subgrade Fatigue Line
The failure criteria is as follows:
Permissible strain = 885 μstrain * N –0.25 * (E / Eref) C

E = modulus of the Subgrade material;
Eref = 160 MPa at 24o C;
and C = 0.
Figure 3.8 – Subgrade Properties
The Granular Sub-base properties and the Lean Concrete Sub-base properties are not well
known and are not entered. Since overseas experiences indicate that the conditions in the
sub-base layer are not critical, the residual life of the sub-base layer is not analysed.
3.2.3 LOADS
The 80kN standard axle load is adopted since all traffic loads are expressed in terms of
equivalent standard axle load. The load configuration parameters are as follows:
Name = Dual
Total load on gear = 40000 N
Dynamic / static load ratio = 1.0
Tyre pressure = 0.577 MPa
Wheel type = Dual
Wheel distance = 330 mm
% of loads = 100%
% total = 100%
Figure 3.9 – Standard Axle Load Configuration
3.3 FWD RAW DEFLECTION DATA FILES
The FWD raw data files are to be acquired according to RD/GN/026. The raw deflection data file is to
be transferred from the field laptop computer to a floppy disk, and in turn copied to the PC with
ELMOD installed.
3.4 PAVEMENT MODEL AND LAYER THICKNESS

The material and thickness of the individual pavement layer has to be known, normally by referring to
construction and maintenance records. If in-situ data is desirable, the thickness of the bituminous or
concrete layers can be measured by taking cored samples at selected locations. The thickness of the
granular sub-base layer can be assessed by conducting a dynamic cone penetrometer (DCP) test
through a cored hole. For more detailed information, trial pits may be excavated to facilitate visual
inspection of the layer thicknesses and pavement materials used. Alternatively, the use of ground
penetrating radar for non-destructive detection of layer thicknesses may be investigated in future.
Where coring and DCP tests are performed, the cored samples may be sent for laboratory tests to
determine their stiffness. The DCP test blows per mm penetration may also be related to the California
Bearing Ratio (CBR) to give an indication of the strength of the layers tested. The Consultants of the
Highway Maintenance Management Pilot Scheme[2] have adopted the equation developed by
Harrison[4] :
Log CBR = 2.55 – 1.14 Log (mm/Blow).
A similar equation was developed by Kleyn and Van Heerden and adopted by the TRRL:
Log CBR = 2.78 – 1.59 Log (mm/Blow).
The first equation is recommended by the consultant for local conditions.
3.5 ANNUAL AVERAGE TRAFFIC LOADING
The annual average traffic loading is the design traffic load (Cn) in standard axles calculated in
accordance with RD/GN/017 or the HyD Pavement Design Manual when it is published, divided by
the design life (n) in years of the road. In order to estimate the average traffic load in future, the
AADTd is to be taken as the annual average daily traffic from the Annual Traffic Census for the year
the FWD test is carried out. If the figure is not yet available, the figure from the previous year may be
used.
Quite often traffic data is not readily available for small roads. In such cases, the following AADTd
values may be assumed [1]:
Road Type Land Use AADTd

Minor Road - 100 – 1,800
Local Distributor Industrial 1,500
Local Distributor Residential / Mixed 5,600
District Distributor All 18,000
4 ANALYSIS STEPS
The FWD data analysis steps using ELMOD and the underlying theories are summarised in the
Sections below. The detailed operations for analysing a flexible pavement structure are illustrated with
an example in Section 5 and the detailed operations for analysing a rigid pavement structure are
illustrated with an example in Section 6.
4.1 OPEN FWD RAW DATA FILE

Copy the raw FWD test deflection data file from the field portable computer onto a floppy disk and
transfer the file to the analysis computer with ELMOD properly installed. Currently, data are acquired
with the Field Program Version 25 and the raw data files have a ‘.F25’ extension. Older data files
collected before the upgrade of the field program have a ‘.FWD’ extension. Either file can be opened
by ELMOD as a single file or a group of files can be processed by batch. Refer to the ELMOD
On-line Help File[4] for batch processing procedures.
FIGURE 4.1 – OPEN SINGLE FILE
4.2 INPUT PAVEMENT STRUCTURE
With the pavement model determined according to Section 2.2, the thickness of each layer can be
assessed and entered into the ELMOD. If a rigid pavement with tests across slab joints is being
analysed, check the ‘Use PCC Joint ID Numbers’ box such that load transfer across joints is analysed.
Refer to RD/GN/026, Appendix 1 for the testing method and test station naming convention.
Figure 4.2 – ELMOD Structure Window
ELMOD automatically assigns one section for the whole test run with the default values for ‘Section’
being ‘1’, ‘from’ being the chainage of the first test station, and ‘to’ being the chainage of the last test
station. The number of sections can be added or deleted with the buttons at the lower part of the
window.
ELMOD requires the layer thicknesses be entered and allows the moduli of individual layers to be
fixed. For Layer 1, the modulus can be manually entered together with a temperature so that the
program can adjust the stiffness of the layer at test temperature to a reference temperature.
For a 4-layer model, ELMOD requires a E2/E3 ratio to be fixed. This can be entered manually or a
default value = 0.2 x H20.45 can be entered by clicking the blank button to the right of the box.
If a rigid layer is known to exist at a certain depth, this can be entered in the ‘Max depth to rigid layer’
box. If a concrete pavement is tested at the joints, check the ‘Use PCC Joint ID Number’ box.
4.3 BACKCALCULATE EFFECTIVE MODULI
More detailed references to the theories adopted by ELMOD are given by Ullidtz[7]. The ELMOD
Help File[3] provides some additional but incomplete information on the internal operation of ELMOD.
Some of the information described below is obtained through numerous trials with the ELMOD
program. Sometimes the meaning of the program steps are not clearly documented and therefore the
internal operation is not known. Such information may have to be updated when more details are
available.
ELMOD provides two approaches of backcalculating the effective layer moduli, one being ‘Radius of
Curvature’ and the other being ‘Deflection Basin Fit’. Both approaches are based on the
Odemark-Boussinesq approximate method, the main difference being that the ‘Deflection Basin Fit’
approach carries out additional iterations until the calculated deflections matches the measured
deflections to within specified tolerance. The ELMOD Help File indicates that the radius of curvature
method is faster, often gives more believable moduli values, and comes closer to measured stresses
and strains in-situ. It is therefore recommended that analysis should be carried out using the ‘Radius
of Curvature’ method under ‘Backcalculation Mode’.
In backcalculating the effective moduli, ELMOD has a number of options to control the program. The
drops to be analysed can be selected together or individually be checking the boxes against each drop
number under ‘Select Drops’. Drop 1 is a seating drop and should not be analysed. If 4 drops are
recorded in accordance with the standard procedures stipulated in RD/GN/026, then backcalculation
for drops 2, 3 and 4 should be carried out. However, normally only Drop 4 needs to be further
analysed. If the test run is divided into more than one section, then backcalculation should be carried
out for all sections by checking the boxes against each section under ‘Select Sections’.
Figure 4.3 – Estimate Moduli Window
In the check measured deflection box, normally select the ‘None’ radio button. If ‘All’ is selected,
ELMOD will plot out the deflection against geophone distance for each test drop, together with a row
of geophone buttons. By examining the plot, the deflection data collected by each geophone can be
dropped from the analysis by clicking on the corresponding geophone button. If ‘Strange’ is selected,
the program will highlight test drops in which the calculated deflections are not uniformly decreasing
with increased geophone distance from load centre.
It is recommended that the options in the ‘Program Control Options’ box should be left blank. The
ELMOD Help File indicates that the radius of curvature method is faster, often gives more believable
moduli values, and comes closer to measured stresses and strains in-situ. It is therefore recommended
that analysis should be carried out using the ‘Radius of Curvature’ method under ‘Backcalculation
Mode’.
4.4 ESTIMATE RESIDUAL LIFE
ELMOD performs an overlay design by estimating the residual life of the pavement. It automatically
calculates the required overlay thickness for locations where the estimated residual lives fall below
the design life.
ELMOD requires the material used for each pavement layer to be entered. The overlay material used
should be ‘Asphalt New Mix’. ELMOD adjusts the backcalculated moduli for each season according
to the corresponding environmental conditions specified in the selected parameter file
‘HK042009.WPR’. The critical stresses and strains caused by the design loads on the pavement
structure are determined. ELMOD then makes use of the critical strains in each layer and the fatigue
curves defined in the selected parameter file for each material to calculate the allowable number of
standard axle loads. The number of loads is divided by the annual traffic to obtain the remaining life
in years. ELMOD then calculates the required overlay thickness at locations where the residual life is
smaller than the design life of the pavement. It should be noted that ELMOD does not take into
account past traffic loads when calculating the residual life since the pavement structural condition
over a road section, i.e. its effective layer moduli as determined by FWD test, has already included the
influence of past traffic.
Overlay is often not appropriate in the Hong Kong environment, and overlay for concrete pavement is
not recommended. Users should refer to the HyD Pavement Design Manual when it is published and
determine the required maintenance treatment independent of this overlay design from ELMOD.
4.5 CHECK RESULTS

After the initial run of the program, use the ‘Plot’ menu to plot the graphs of deflection, moduli, and
structural life. Analyse the results to check if there are anomalies. Delete suspicious data and re-run
the program to get new results.
The deflection data should be reviewed to see if there are large fluctuations or inconsistencies. The
cumulative sums of the centre deflection can be used to assist in deciding whether the length of the
test run should be divided into different sections for design purposes. The effective moduli can be
recalculated where necessary, and then checked to see if they have reasonable values. The residual life
needs to be re-estimated if the moduli are recalculated.
The more detailed steps after the initial run are described in the following sections.
4.6 REVIEW DEFLECTION DATA
According to RD/GN/026, 4 drops should be made at each test station. Drop 1 is used as a seating
drop and the results are not used for analysis. ELMOD has plot functions that can plot out individual
deflection patterns for each drop along the test run, but not more than 1 drop at the same time. The
deflection pattern should be plotted for each of the test drops 2, 3 and 4. Observe if there are
significant differences in deflection pattern between drops. Note down the chainage of any test station
where the deflection varies significantly between drops. The tests data at these stations may have to be
discarded or the test repeated to get more reliable results.
If the deflection patterns between drops are similar, it is recommended that the drop 4 data be used for
analysis since the contact between the loading plate and the pavement surface is closest for the last
drop.
For the drop selected for analysis, check if there are any test stations at which high fluctuation in
deflection magnitudes occur. Abnormally high deflections at a test station may be due to a local weak
spot and abnormally low deflections may be due to local hard ground or buried utilities. It is common
for FWD test data to exhibit variations from station to station along the length of the test run.
Experience and judgement is required to determine which points are suspicious. Record the chainage
at which the abnormal deflections occur. These points should be removed before further analysis is
carried out. Further tests and investigations on site may be required to determine the cause of the
abnormality.
To delete the suspicious data, exit the ELMOD application and open the data file with the Windows
Notepad or a text editor. Find the relevant data lines using the recorded chainage, delete those lines
and save the amended data file.
Exit Notepad or the editor and re-open the amended data file in ELMOD. Plot the deflection pattern
again to check that the suspicious data has been removed.
4.7 REVIEW CUMULATIVE SUM OF DEFLECTIONS
Sometimes the pavement conditions may vary along the test run, e.g., the pavement structure may
change. For rehabilitation design purposes, it may be desirable to divide the test run into different
sections. In order to help observe if there are significant changes in trend along the test run, ELMOD
uses the method of cumulative sum described in Section 2.1.2.
From the cumulative sum, check if there are significant changes in trend (from upward to downward
or vice versa). A homogeneous section is one in which the cumulative sums continue in the same
upward or downward trend. A significant change in the upward or downward trend indicates a change
in section. Divide the test run into homogeneous sections such that the trend is reasonably consistent
within each section. This step is illustrated in Section 5.7. The homogeneous sections shall be
re-backcalculated and analysed again. These steps are illustrated in Sections 5.8 and 5.9.
For design purposes, a set of characteristic values representative of the section has to be selected. This
is done statistically by choosing the set of values at the test station where the centre deflection is at the
85 percentile of all the centre deflections in the section. While ELMOD does not provide such a
function, this can be relatively easily done using a spreadsheet to import an ELMOD created file with
the same filename as the raw data file but with a ‘.DMS’ extension. The worksheet columns show the
number of sensors, the section number, the drop number, the test station number, the chainage, the
measured deflection for each sensor, the calculated deflection for each sensor, and the root mean
square error if the curve fitting option is used.
When the data is imported into the spreadsheet, simply use the PERCENTILE function to calculate
the 85 percentile value of the deflection measured by sensor 1 (the centre deflection). Choose the test
station with a centre deflection having the same value or the next highest value as the 85 percentile.
This step is illustrated in Section 5.12.
The effective moduli and residual life calculated for this test station can be used in the subsequent
rehabilitation design. Follow the procedures stipulated in RD/GN/017, or the Pavement Design
Manual when it is published, for the design. Do not simply rely on the overlay thickness automatically
calculated by ELMOD.
4.8 REVIEW EFFECTIVE MODULI
The backcalculated results may contain anomalies, e.g., sometimes E2 may be smaller than E3 or the
modulus of a particular layer may have an unreasonably large value. Apart from using the steps
described in Section 5.5 and 5.6 to remove suspicious test data and divide the test run into sections,
the interpretation of the backcalculated results still requires experience and expertise from the user. It
is expected that a database of effective modulus values will be built up in the Highway Maintenance
Management System as more and more tests and analyses are carried out in future.
For the time being, the statistical results from the 1991 Road Testing Program shown below may be
used as a reference in assessing whether the backcalculated effective moduli of different pavement
layers fall within a reasonable range. If the test results fall within the top or bottom 10% of the typical
ranges indicated, then the FWD data and backcalculated moduli may have to be examined more
carefully. For a detailed assessment, it is necessary to correlate the FWD results with other test data
such as in-situ tests or laboratory tests on cored samples from site. In case the backcalculated modulus
of a particular layer is deemed not reasonable, a modulus range for that layer should be adopted based
on other test results or experience. Click on ‘Options’ in the top menu and select the ‘Fix Layer
Moduli’. A window will pop up for the ‘E-min’ and ‘E-max’.
Figure 4.4 – Fix Layer Moduli
The E-min and E-max values can be selected with reference to the following table. Save the range
values and run the backcalculation step again.
Concrete Modulus (MPa) % Concrete Modulus (MPa) Cumulative %

< 15,000 13.0 >= 0 100.0
15,000 – 20,000 26.8 >= 15,000 87.0
20,000 – 25,000 14.6 >= 20,000 60.2
25,000 – 30,000 16.3 >= 25,000 45.6
30,000 – 35,000 13.8 >= 30,000 29.3
35,000 – 40,000 13.8 >= 35,000 15.5
> 40,000 1.7 > 40,000 1.7
Asphalt Modulus (MPa) % Asphalt Modulus (MPa) Cumulative %
< 1,000 14.4 >= 0 100.0
1,000 – 1,500 19.2 >= 1,000 85.6
1,500 – 2,000 20.6 >= 1,500 66.4
2,000 – 2,500 13.4 >= 2,000 45.8
2,500 – 3,000 11.1 >= 2,500 32.4
3,000 – 4,000 10.2 >= 3,000 21.3
> 4,000 11.1 > 4,000 11.1
Sub-base Modulus (MPa) % Sub-base Modulus (MPa) Cumulative %

< 100 15.5 >= 0 100.0
100 – 200 39.0 >= 100 84.5
200 – 500 21.8 >= 200 45.5
500 – 1,000 6.0 >= 500 23.7
1,000 – 5,000 11.4 >= 1,000 17.7
> 5,000 6.3 > 5,000 6.3
Subgrade Modulus (MPa) % Subgrade Modulus (MPa) Cumulative %

< 50 11.6 >= 0 100.0
50 – 100 41.5 >= 50 88.4
100 – 150 24.5 >= 100 46.9
150 – 200 14.1 >= 150 22.4
200 – 300 6.9 >= 200 8.3
> 300 1.4 > 300 1.4
In general, a sub-base modulus of less than 500 MPa represents granular material and a modulus
above 1,000 MPa represents lean concrete. For most tests, the ratios of granular sub-base modulus to
subgrade modulus range between 1 to 4, and generally can be considered to be around 2. The values
can be updated as more testing and analysis is carried out.
For the load transfer across joints in rigid pavements, the Road Testing Programme consultants
recommended that the joint condition and load transfer efficiency are classified as follows:
Joint Condition Load Transfer Efficiency (%)

Good 70 – 100
Fair 50 – 70
Poor Below 50
The non-linear coefficients ‘C’ and ‘n’ of the subgrade can be displayed. The non-linearity properties
of the pavement structure can be visualised by plotting the ‘surface modulus’ against equivalent depth.
4.9 REVIEW RESIDUAL LIFE
The residual life estimated by ELMOD may be used as one of the performance indices to assess the
relative maintenance priority of different sections of roads. ELMOD saves the calculated remaining
life for each layer at each test station in a file with the same filename as the raw data file but with a
‘.LNN’ extension where NN is a serial number from 2 to 99. The value of ‘NN’ is incremented by one
for each successive design run.
To better appreciate the conditions of the pavement layers, print out the residual life analysis results
using the print functions in ELMOD. Click on ‘Output’ at the top menu to enter the output window.
Click on ‘Print’ and then select ‘Overlay design results’ from the pull-down menu to print the residual
life results.
Alternatively, use a spreadsheet software such as Excel to import the ‘.LNN’ file. The worksheet
shows the chainage of each test station, the remaining lives of the top layer and the subgrade layers,
and their minimum is used as the residual life of the pavement structure. This analysis result provides
an indication on the layer to which maintenance works may be required.
4.10 REVIEW AND ARCHIVE RESULTS

In case there are dubious results in the above analysis steps, make the necessary adjustments to the
data and re-analyse the amended data. The maximum number of analysis runs that can be performed
on a data set is 98 before the old ‘LNN’ files have to be deleted. The results of each run are saved in
files that can be imported into a spreadsheet for further analysis independent of ELMOD. The final
analysis results should be archived both electronically in the native digital format and in printed
hardcopy together with a map showing the alignment of the test run. The data may then be entered
into the Highway Maintenance Management System when it is implemented to build up a geographic
database for future reference.
5 ANALYSING A FLEXIBLE PAVEMENT STRUCTURE
This Section illustrates the use of ELMOD to analyse a flexible pavement structure.
5.1 OPEN RAW DATA FILE

Invoke ELMOD. Click on ‘File’ in the top menu to select ‘Single File’ from the pull down menu.
Double click on a sample data file called “FLEXIBLE.FWD”.

Click on ‘Structure’ in the top menu to enter the pavement structure. A ‘Structural data’ window will
pop up. In ‘Section’, default values of ‘1’, from ‘0’ to ‘0.5’ are entered.
A 3-layer model is adopted for this example. The bituminous materials are combined into a single top
layer with thickness H1 = 160 mm. The granular sub-base has a thickness H2 = 250 mm. The subgrade
is not subdivided into two layers and its thickness is considered to be semi-infinite, there is no need to
input its thickness. The other fields are left blank.
Figure 5.1 – Input Structural Data
5.3 BACKCALCULATE EFFECTIVE LAYER MODULI
Click on ‘Moduli’ in the top menu to backcalculate the layer moduli. In the ‘Select Drops’ box, check
the ‘2’, ‘3’, and ‘4’ boxes so that deflection data collected during Drops 2, 3, and 4 are used for
analysis. In the ‘Select Sections’ box, check the ‘1’ box. In the ‘Check measured deflection’ box,
select ‘None’. Leave the ‘Program Control Options’ box blank. In the ‘Backcalculation Mode’ box,
select ‘Radius of curvature’. Click on the ‘Calculate’ button to perform the analysis.
Figure 5.2 – Backcalculate Effective Layer Moduli
When the backcalculation is completed, a window will pop up asking ‘Do you want to save your
results for viewing and printing?’ Click on the ‘Yes’ button. If a previous backcalculation has already
been carried out, a second window will pop up asking ‘Overwrite existing file? Select no to append.’
Click on the ‘Yes’ button to overwrite the existing file. A window with the message ‘Data has been
saved’ will pop up. Click on the ‘OK’ button to leave the window.
Click on ‘Design’ in the top menu to estimate the residual life. A ‘Select parameter file’ window will
pop up. Select the parameter file ‘HK042009.WPR’ and click the ‘Open’ button. A ‘Select Section’
window will pop up. Select Section ‘1’, and then select Drop ‘4’ to be analysed. Click on the ‘OK’
button.
FIGURE 5.3 – SELECT SECTION AND DROP FOR ANALYSIS
A ‘Materials and Loading’ window will pop up. Select the overlay material by highlighting ‘Asphalt
New Mix’ on the right hand box, then click on ‘Overlay’ in the left hand box. Since the pavement was
built using the old mix design, select ‘Asphalt Old Mix’ on the right hand box and click on Layer 1.
Select ‘Granular Sub-base’ on the right hand box and click on Layer 2. Select ‘Subgrade’ on the right
hand box and click on Layer 3. In the ‘Number of loads per year’ box, enter the predicted average
annual equivalent standard axle loads, 1285000 in this example. Click on the ‘OK’ button.
Figure 5.4 – Input Materials and Loading
A window with the message ‘Overlay design complete’ will pop up after ELMOD has completed
estimating the residual life and designing the overlay requirement. Click on the ‘OK’ button.
5.5 CHECK RESULTS
Click on the ‘Plot’ menu and a graphics window will be displayed. In the lower half of the window,
check the ‘Deflections (microns)’ radio button. Check all the sensor distance check boxes below. In
the ‘Drop No.’ box to the right, click the ‘4’ radio button. Click on the ‘Graph’ button near the centre
to show the deflection pattern along the length of the test run. The solid colour lines in the graph show
the measured deflections and the broken lines show the backcalculated deflections. The first 2 sets of
data on the right hand column shows the measured figures on top and the backcalculated figures
underneath in terms of number of test stations, the mean deflections and the standard deviations for
the deflections measured at each geophone. It is observed that the deflections of all geophones were
extremely low at chainage 0.123. The next 2 sets of data on the right hand column shows the
deflections at chainage 0.123.
To check the values at any point, simply click at the point in the graph and the chainage of the nearest
test station will be displayed at the bottom, together with the data on the right side of the window. To
see the mean values of a range of data, click and drag along the x-axis over the range of interest. To
clear the graphic, click on the ‘OK’ button to the right of the graph. To clear the data display, click on
the ‘CLS’ button at the bottom right of the window.
Figure 5.5 – Deflection along Test Run after Initial Analysis
In view of the abnormal deflection at chainage 0.123, the deflections should be checked against those
recorded by Drop 2 and Drop 3. Click on the radio button next to ‘2’ under the ‘Drop No.’ box on the
right. Then click on the ‘Graph’ button to see the deflection pattern of Drop 2. The process can be
repeated for checking against Drop 3. In this example, the abnormally low deflections are present in
all drops.
Click on the ‘Moduli (MPa)’ button and the graph of the 3 backcalculated layer moduli is shown. It
can be observed that the moduli of the sub-base and subgrade at chainage 0.123 were unreasonably
high.
Figure 5.6 – Layer Moduli after Initial Analysis
Click on the ‘Structural life (yrs)’ radio button and the estimated residual life graph is shown. It can be
observed that the estimated residual life varies greatly along the test run. While the mean life is 159
years, the maximum can be up to 10000 years and the minimum of 1.51 years occurs at chainage
0.326.
Figure 5.7 – Residual Life after Initial Analysis
In view of the very low deflection at chainage 0.123 and the corresponding unreasonably high moduli
of the sub-base and subgrade at that location, it is desirable to remove the dubious data at chainage
0.123 and perform the analysis again.
It should be noted that while the dubious data affects the mean value of backcalculated results, the
results at individual test stations are not affected.
5.6 REVIEW DEFLECTION DATA
The deflection data at chainage 0.123 should be removed as described in Section 3.6. Plot the
deflection again to check that the dubious data has been removed.
To close the graphic window, click on the ‘Quit’ button to the right of the graph.
Figure 5.8 – Deflection along Test Run with Suspicious Data Removed
5.7 REVIEW CUMULATIVE DEFLECTIONS
Click on ‘Plot’ in the top menu and uncheck all sensors except the 0mm sensor. In the ‘Drop No.’ box,
click on the radio button next to ‘4’ and then click on the ‘Graph’ button to plot the centre deflection.
Click on the ‘A.d.’ button to the right of the graphic to plot the deflection cumulative sum.
Figure 5.9 – Centre Deflection and Cumulative Sum
In the example, it appears that there is a significant change in trend at chainage 0.3. Therefore the test
run is divided into 2 sections, Section 1 from chainage 0 to 0.3 and Section 2 from chainage 0.3 to 0.5.
A change in section is also necessary when there is a change in construction details. Please note that
whether the test run should be divided into different sections and how many sections should be used is
largely a matter of engineering judgement.
Click on the ‘Quit’ button to leave the graphic window.
5.8 EDIT PAVEMENT STRUCTURE
pop up. In the ‘Section’ box, enter ‘1’, from ‘0’ to ‘0.3’.
A 3-layer model is adopted for this example. The bituminous materials are combined into a single top
layer with thickness H1 = 160 mm. The granular sub-base has a thickness H2 = 250 mm. The subgrade
is not subdivided into two layers and its thickness is considered to be semi-infinite, there is no need to
input its thickness. The other fields are left blank.
Figure 5.10 – Input Section 1 Structural Data
To input the structure for another section, click on the ‘Add’ button near the bottom. Go back to the
‘Section’ box, enter ‘2’, from ‘0.3’ to ‘0.5’. There is no change in layer thicknesses throughout the test
run and therefore the same data is entered. Click on the ‘Previous’ or ‘Next’ buttons to move from one
section to another. Click on the ‘Add’ or ‘Delete’ buttons to add more sections or delete existing as
necessary. Click on the ‘OK’ button’ to exit the window.
Figure 5.11 – Input Section 2 Structural Data
5.9 RE-BACKCALCULATE EFFECTIVE LAYER MODULI
Click on ‘Moduli’ in the top menu to backcalculate the layer moduli. In the ‘Select Drops’ box, check
the ‘2’, ‘3’, and ‘4’ boxes so that only the deflection data collected during Drops 2, 3, and 4 are used
for analysis. In the ‘Select Sections’ box, check both the ‘1’ and ‘2’ boxes so that both sections are
analysed. In the ‘Check measured deflection’ box, select ‘None’. Leave the ‘Program control options’
box blank. In the ‘Backcalculation Mode’ box, select ‘Radius of curvature’. Click on the ‘Calculate’
button to perform the analysis.
Figure 5.12 – Backcalculate Effective Layer Moduli
Click on the ‘Yes’ button to overwrite the existing file. Click on the ‘No’ button to append the new
results to the existing one. In this example, click ‘Yes’. Finally, a window with the message ‘Data has
been saved’ will pop up. Click on the ‘OK’ button to leave the window.
5.10 REVIEW EFFECTIVE LAYER MODULI
Click on ‘Plot’ in the top menu and under the ‘Section’ box at the lower middle of the screen, click on
the ‘1’ radio button. In the ‘Drop No.’ box at the lower right, click on the ‘4’ radio button. Click on the
‘Moduli (MPa)’ radio button to display the Section 1 layer moduli backcalculated from Drop 4 data.
See if the backcalculated modulus for the same layer vary significantly along the test run. Observe the
mean value and distribution of the layer moduli.
It can be seen that seemingly unreasonable results occur at chainage 0, where the backcalculated
modulus of the sub-base exceeds that of the asphalt layer.
Figure 5.13 – Section 1 Layer Moduli
Under the ‘Section’ box, click on the ‘2’ radio button to see the graph for Section 2. Again it can be
observed that the backcalculated moduli of the sub-base appear unreasonable at a couple of locations,
where it is either lower than the subgrade or higher than the asphalt layer.
Figure 5.14 – Section 2 Layer Moduli
The anomalies can sometimes be removed by modeling an additional layer of subgrade of 500 mm to
1000 mm just below formation level. At other times the anomalies remain even after much
remodeling. It may be useful to compare the dubious data against other test data, such as laboratory
test results on cored samples and DCP tests. Again engineering judgement is required to determine
whether additional tests are required or if the anomalies are acceptable.
In this example, the mean moduli values appear reasonable and the standard deviations are small. The
anomalies are accepted for further analysis.
‘Structural life (yrs)’ radio button to display the residual life values for Section 1 estimated from Drop
4 data. Observe that while the mean residual life is 142 years, the lowest residual life of 6.21 years
occurs at chainage 0.2.
Figure 5.15 – Section 1 Residual Life
Under the ‘Section’ box, click on the ‘2’ radio button to see the residual life of Section 2. Observe that
while the mean residual life is 52.1 years, the lowest residual life of 1.51 occurs at chainage 0.326.
Figure 5.16 – Section 2 Residual Life

From this example, it is observed that Section 1 has a higher mean residual life than Section 2. The
lower residual life of Section 2 corresponds to the larger deflections from chainage 0.326 to 0.376 and
0.426 to 0.45.
For design purposes, the condition at the test station that records the 85 percentile centre deflection
should be chosen to be representative of the section. In this case, refer to Section 3.7 to import the file
‘FLEXIBLE.DMS’ into a spreadsheet. Using MS Excel as an example, the steps are shown below:
Figure 5.17 – Open ‘.DMS’ File in Excel
Figure 5.18 – Select the Delimit Sign
Figure 5.19 – Check Tab and Comma Boxes
Click the ‘Finish’ button to complete the import process.
A portion of the spreadsheet, up to column F containing the measured deflection at geophone number
1 (centre deflection) is shown below. The test station at the chainage 0.3 is repeated in the original file
because it occurs in the boundary of both sections. The duplicate entry should be taken out when
calculating the 85 percentile value for the entire test run. Note that Rows 25 to 30 are manually
entered into the spreadsheet to calculate the 85 percentile centre deflection using the built-in
PERCENTILE function.
A B C D E F
1 Number of sensors Section # Drop # Point Chainage #1m
2 7 1 4 1 0 436
3 7 1 4 2 0.025 457
4 7 1 4 3 0.05 421
5 7 1 4 4 0.075 422
6 7 1 4 5 0.1 447
7 7 1 4 6 0.131 532
8 7 1 4 7 0.15 534
9 7 1 4 8 0.175 615
10 7 1 4 9 0.2 725
11 7 1 4 10 0.225 268
12 7 1 4 11 0.251 468
13 7 1 4 12 0.275 436
14 7 1 4 13 0.3 433
15 7 2 4 13 0.3 433
16 7 2 4 14 0.326 757
17 7 2 4 15 0.35 928
18 7 2 4 16 0.376 816
19 7 2 4 17 0.405 217
20 7 2 4 18 0.426 627
21 7 2 4 19 0.45 538
22 7 2 4 20 0.476 317
23 7 2 4 21 0.5 365
24
25 85 percentile centre deflection for section 1 = 550.2
26 85 percentile centre deflection for section 2 = 804.2
27 85 percentile centre deflection for entire test run = 725.0
28
29 Note: the 85 percentile for Section 1 can be calculated using the built-in
30 function = PERCENTILE(F2:F14,0.85).
In Section 1, the 85 percentile centre deflection value is 550µm and the test station with the nearest
centre deflection is at chainage 0.15. In order to check the residual life, refer to Section 3.9 and use
similar steps as shown above to import the file ‘FLEXIBLE.L2’ into a spreadsheet. Note that the
estimated residual life at this test station is 30 years and therefore no maintenance treatment is
required. However, since the estimated residual life at chainage 0.2 is only 6.2 years, some local
strengthening may be desirable.
Chainage Layer Remaining Layer Remaining Minimum Overlay

Life Life (years) Life Thickness
(years) (years) (mm)
0 1 30 3 30 30 0
0.025 1 30 3 30 30 0
0.05 1 30 3 30 30 0
0.075 1 30 3 30 30 0
0.1 1 30 3 30 30 0
0.131 1 30 3 30 30 0
0.15 1 30 3 30 30 0
0.175 1 30 3 30 30 0
0.2 1 6.2 3 21.4 6.2 35.1
0.225 1 30 3 30 30 0
0.251 1 30 3 30 30 0
0.275 1 30 3 30 30 0
0.3 1 30 3 30 30 0
In Section 2, the 85 percentile centre deflection value is 804µm and the test station with the nearest
centre deflection is at chainage 0.376. In order to determine the maintenance options, follow the steps
described in Section 3.9 to import the file ‘FLEXIBLE.L3’ into a spreadsheet. Note that while the
estimated residual life at chainage 0.376 is 6.4 years, failure is more likely to occur at chainages 0.326
and 0.35 that have residual lives of 1.5 and 1.6 years respectively. The residual life of the subgrade at
chainage 0.35 is only 6.9 years. Therefore, local full-depth reconstruction may have to be carried out
around chainages 0.326 to 0.376.
Chainage Layer Remaining Layer Remaining Minimum Overlay

Life Life (years) Life Thickness
(years) (years) (mm)
0.3 1 30 3 30 30 0
0.326 1 1.5 3 30 1.5 78.9
0.35 1 1.6 3 6.9 1.6 80.7
0.376 1 6.4 3 9.7 6.4 32.2
0.405 1 30 3 30 30 0
0.426 1 23.1 3 30 23.1 0
0.45 1 25.6 3 30 25.6 0
0.476 1 30 3 30 30 0
0.5 1 30 3 30 30 0
It should be noted that if the test run is not divided into 2 sections, then the 85 percentile centre
deflection value is 725µm and the test station with the nearest centre deflection is at chainage 0.2.
Since the estimated residual life at this test station is 6.2 years, in-lay strengthening may be
considered. However, local reconstruction may still be required at chainages 0.326 and 0.35.
Note also that in deciding on the maintenance treatment, structural pavement evaluation should be
considered together with other factors such as visual inspection results, surface roughness, traffic
condition, and engineering judgement.
6 ANALYSING A RIGID PAVEMENT STRUCTURE
This Section illustrates the use of ELMOD to analyse a rigid pavement structure.
6.1 OPEN RAW DATA FILE

Invoke ELMOD. Click on ‘File’ in the top menu to select ‘Single File’ from the pull down menu.
Select the raw data file to be analysed and click the ‘Open’ button. For illustration, this time a sample
data file called “ RIGID.F25“ is used. It should be noted that at the time of testing, the field program
was set to record only the last 2 drops. Hence only test data for 2 drops can be processed and both
Drop 1 and Drop 2 are used. It is expected that newly acquired FWD data will record all 4 drops in
accordance with the practice stipulated in RD/GN/026.

pop up. In the ‘Section’ box, enter 1, from 1.1 to 21.22. A 2-layer model is adopted for this example.
The concrete slab forms a single top layer with thickness H1 = 225 mm. The granular sub-base and the
subgrade is not subdivided into two layers and its thickness is considered to be semi-infinite, there is
no need to input its thickness. Note that since the test is carried out on rigid pavement and the load
transfer across joints is tested, check the box next to ‘Use PCC Joint ID Numbers’ at the bottom of the
window. The other fields are left blank.
Figure 6.1 – Input Section Structural Data for Rigid Pavement
6.3 BACKCALCULATE EFFECTIVE LAYER MODULI
Click on ‘Moduli’ in the top menu to backcalculate the layer moduli. The data file does not record the
test results of the first two drops, therefore the test results of Drop 3 and Drop 4 are recorded as Drop
1 and Drop 2 in the data file. In the ‘Select Drops’ box, check both the ‘1’ and ‘2’ boxes so that both
sets of deflection data are used for analysis. In the ‘Select Sections’ box, check the ‘1’ box. In the
‘Check measured deflection’ box, select ‘None’. Leave the ‘Program control options’ box blank. In
the ‘Backcalculation Mode’ box, select ‘Radius of curvature’. Click on the ‘Calculate’ button to
perform the analysis.
Figure 6.2 – Backcalculate Effective Layer Moduli for Rigid Pavement
Click on the ‘Yes’ button to overwrite the existing file. Click on the ‘No’ button to append the new
results to the existing one, useful for appending backcalculation results for different sections onto the
same file. Finally, a window with the message ‘Data has been saved’ will pop up. Click on the ‘OK’
button to leave the window.
Click on ‘Design’ in the top menu to estimate the residual life. A ‘Select parameter file’ window will
pop up. Select the parameter file ‘HK042009.WPR’ and click the ‘Open’ button. A ‘Select Section’
window will pop up. Select Section ‘1’ and Drop ‘2’ to be analysed. Click on the ‘OK’ button.
A ‘Materials and Loading’ window will pop up. Select the overlay material by highlighting ‘Asphalt
New Mix’ on the right hand box, then click on ‘Overlay’ in the left hand box. Select ‘Concrete’ on the
right hand box and click on Layer 1. Select ‘Subgrade’ on the right hand box and click on Layer 2. In
the ‘Number of loads per year’ box, enter the predicted average annual equivalent standard axle loads,
1200000 in this example. In the ‘Treatment before Overlay’ box, select ‘None’. Select ‘Concrete’ in
the top right-hand material box, then click in the empty box below ‘Slab at joint is’.
In the ‘Treatment before overlay’ box, click ‘None’ since asphalt overlay on concrete pavement is not
recommended. Click on the ‘OK’ button to start the design calculation.
Figure 6.3 – Input Materials and Loading (Rigid)
A window with the message ‘Overlay design complete’ will pop up after ELMOD has completed
estimating the residual life and designing the overlay requirement. Click on the ‘OK’ button.
6.5 CHECK RESULTS
Click on the ‘Plot’ menu and a graphics window will be displayed. In the lower half of the window,
check the ‘Deflections (microns)’ radio button. Check all the sensor distance check boxes below. In
the ‘Drop No.’ box to the right, check the ‘2’ radio button. Click on the ‘Graph’ button near the centre
to show the deflection pattern along the length of the test run.
Figure 6.4 – Deflection along Test Run on Rigid Pavement
Note that the deflection pattern shows a jagged pattern due to the tests being carried out alternately at
the joint position and then at the centre of the concrete slab. To check the data consistency with Drop
1, click on the radio button next to ‘1’ under the ‘Drop No.’ box on the right. Then click on the
‘Graph’ button to see the deflection pattern of Drop 1. The graph shows that the test patterns between
the 2 drops are highly consistent.
In this example, although the deflections of various geophones at chainage 5.1 appear to be
compressed when compared with others, their magnitudes do not differ significantly from others. The
data need not be deleted before reviewing other results. Graphs of moduli and residual life, which are
same as Figures 6.7 and 6.10 respectively, shall also be checked.
Note also that 9 geophones instead of 7 geophones are used. The geophones are spaced out such that
the geophones at distances 200mm (geophone no. 8) and 300mm (geophone no. 9) from the load are
placed behind the load while the other 7 geophones are placed in front of the load. This arrangement
allows test at transverse joints to be conveniently carried out. The geophone arrangement is as shown
below:
Figure 6.5 – Geophone Arrangement for Testing Joints at Rigid Pavement
To close the graphic window, click on the ‘Quit’ button to the right of the graph.
6.6 REVIEW CUMULATIVE DEFLECTIONS
Click on ‘Plot’ in the top menu and uncheck all sensors except the 0mm sensor. In the ‘Drop No.’ box,
click on the radio button next to ‘2’ and then click on the ‘Graph’ button to plot the centre deflection.
Click on the ‘A.d.’ button to the right of the graphic to plot the deflection cumulative sum.
Figure 6.6 – Centre Deflection and Cumulative Sum on Rigid Pavement
In this example, the cumulative deflection curve bottomed out at chainage 10.22. Again, whether to
divide up the entire test run into sections is a matter of engineering judgement. For illustration
purpose, this time the test run is to be analysed as a single section. Click on the ‘Quit’ button to leave
the graphic window.
6.7 REVIEW EFFECTIVE LAYER MODULI, MODULI OF SUBGRADE REACTIONS
AND LOAD TRANSFER
‘Moduli (MPa)’ radio button to display the Section 1 layer moduli backcalculated from Drop 2 data.
See if the backcalculated modulus for the same layer vary significantly along the test run. Observe the
mean value and distribution of the layer moduli.
Figure 6.7 – Layer Moduli for Rigid Pavement
Note that in this example, a small bump in layer moduli occurs at the chainage where the compressed
deflections are. However, the layer moduli for both the concrete slab and the combined sub-base and
subgrade layer are highly consistent. As seen on the right hand column, the standard deviations are
very small.
Click on the ‘E1 and k’ radio button to plot the concrete modulus and modulus of subgrade reaction.
Note that the ‘kc’ and ‘kj’ represents the moduli of subgrade reaction at centre and joint of the concrete
slab respectively, as if the subgrade does not transfer shear stress, like in a Winkler foundation.
Figure 6.8 – Effective Concrete Modulus and Modulus of Subgrade Reaction (Rigid)
Click on the ‘Dd and LT%’ radio button to plot the differential deflection across joints and the load
transfer ration. It can be seen that 21 tests across joints have been carried out and the mean differential
deflection is 9.05mm with a standard deviation of 1.91 mm. The load transfer ratio is high at 96.8%,
with a small standard deviation of 0.67 only. This indicates that the joints are in good condition.
Figure 6.9 – Differential Deflection and Load Transfer Ratio
Click on ‘Plot’ in the top menu and then click on the ‘Structural life (yrs)’ radio button to display the
estimated residual life values. It can be seen that the residual life of the concrete pavement as
calculated is consistently over 800 years. This indicates that the whole pavement structure is in good
condition and no reconstruction work is required.
Figure 6.10 – Residual Life (Rigid)

Similar to the example for flexible pavement, the results should be reviewed. In this case, the FWD
data appear fairly consistent and the results appear reasonable. Therefore, no re-analysis is required
and the results are archived electronically and on paper.
APPENDIX A – REFERENCES
[1] Acer Consultants (Far East) Limited, 1995. “Highway Maintenance Management Pilot
Scheme Volume A – Main Report”, Highways Department.
[2] Acer Consultants (Far East) Limited, 1995. “Highway Maintenance Management Pilot
Scheme Final Report Volume B4 – FWD Test Sections”, Highways Department.
[3] Dynatest International, 1997. “ELMOD Help File (electronic copy installed together with
ELMOD)”.
[4] Dynatest International, 1998. “ELMOD Quick Start Manual”.
[5] Harrison, 1989. “DCP Calibration (Indonesia)”. Australian Road Research, Vol. 19, No. 4.
[6] Stubstad, Baltzer, Lukanen, and Ertman-Larsen, 1994. “Prediction of AC Material

Temperatures for Routine Load/Deflection Measurements”. Proceedings of the 4th
International Conference on the Bearing Capacity of Roads and Airfields, Minneapolis, USA.
[7] Ullidtz, Per, 1987. “Pavement Analysis”, Elsevier.
[8] Shell International Petroleum Company Limited, London, 1978, “Shell Pavement Design
Manual – asphalt pavements and overlays for road traffic”.
[9] J. Boussinesq,1885, Applicaation des potentiels á l ’etude de l ’equilibre et du mouvement des

solides élastique.
[10] E. Winkler, 1867, Die Lehre von der Elasticität und Festigkeit, Prage Dominicus.
[11] A.M. Ioannides, M.R.. Thompson & E.J. Barenberg, 1985, Westergaard Solutions
Reconsidered, Transportation Research Record 1043.

Elmod PDF

Uploaded by

Copyright:

Available Formats

Elmod PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Elmod PDF

Uploaded by

Copyright:

Available Formats

HIGHWAYS DEPARTMENT

BACKCALCULATION OF LAYER MODULI

ESTIMATION OF RESIDUAL LIFE

FALLING WEIGHT DEFLECTOMETER

Research & Development Division RD/GN/027A

2.1 PAVEMENT STRUCTURE 8

2.1.1 TOP LAYER 10

2.2 BACKCALCULATION OF PAVEMENT LAYER MODULI 11

2.2.1 BOUSSINESQ'S EQUATIONS 12

2.2.2 ODEMARK'S METHOD OF EQUIVALENT THICKNESS 13

2.2.3 SURFACE MODULUS 14

2.2.4 SUBGRADE NON-LINEARITY 15

2.2.6 WESTERGAARD'S EQUATIONS FOR RIGID PAVEMENT 16

2.2.7 LOAD TRANSFER ACROSS JOINT FOR RIGID PAVEMENT 17

2.3 ESTIMATION OF RESIDUAL LIFE 18

2.3.1 SEASON AND TRAFFIC 18

2.3.2 TEMPERATURE EFFECTS 18

2.4.1 SECTIONING OF PAVEMENT 19

2.4.2 CUMULATIVE SUMS OF DEFLECTION 19

2.4.3 CHARACTERISTIC DEFLECTION 20

3 PREPARING FOR THE ANALYSIS 21

3.1 ELMOD INSTALLATION 21

3.2 PARAMETER AND TEMPERATURE FILES 21

3.3 FWD RAW DEFLECTION DATA FILES 31

3.4 PAVEMENT MODEL AND LAYER THICKNESS 31

3.5 ANNUAL AVERAGE TRAFFIC LOADING 32

4.1 OPEN FWD RAW DATA FILE 33

4.2 INPUT PAVEMENT STRUCTURE 34

4.3 BACKCALCULATE EFFECTIVE MODULI 35

4.4 ESTIMATE RESIDUAL LIFE 37

4.5 CHECK RESULTS 37

4.6 REVIEW DEFLECTION DATA 38

4.7 REVIEW CUMULATIVE SUM OF DEFLECTIONS 39

4.8 REVIEW EFFECTIVE MODULI 40

4.9 REVIEW RESIDUAL LIFE 42

4.10 REVIEW AND ARCHIVE RESULTS 42

5 ANALYSING A FLEXIBLE PAVEMENT STRUCTURE 43

5.1 OPEN RAW DATA FILE 43

5.2 INPUT PAVEMENT STRUCTURE 43

5.3 BACKCALCULATE EFFECTIVE LAYER MODULI 44

5.4 ESTIMATE RESIDUAL LIFE 45

5.5 CHECK RESULTS 46

5.8 EDIT PAVEMENT STRUCTURE 51

5.9 RE-BACKCALCULATE EFFECTIVE LAYER MODULI 53

5.10 REVIEW EFFECTIVE LAYER MODULI 54

5.11 REVIEW RESIDUAL LIFE 56

5.12 REVIEW AND ARCHIVE RESULTS 58

6 ANALYSING A RIGID PAVEMENT STRUCTURE 63

6.1 OPEN RAW DATA FILE 63

6.2 INPUT PAVEMENT STRUCTURE 63

6.3 BACKCALCULATE EFFECTIVE LAYER MODULI 64

6.4 ESTIMATE RESIDUAL LIFE 65

6.5 CHECK RESULTS 66

6.7 REVIEW EFFECTIVE LAYER MODULI, MODULI OF SUBGRADE REACTIONS

AND LOAD TRANSFER 69

6.8 REVIEW RESIDUAL LIFE 72

6.9 REVIEW AND ARCHIVE RESULTS 72

FIGURE 2.1 – TYPICAL FLEXIBLE PAVEMENT MODEL

FIGURE 3.1 – SEASONAL CHARACTERISTICS