Professional Documents
Culture Documents
Abstract— Flexible pavement structures are very complex For the transformed section shown in Fig. 1 the equivalent
systems usually consist of multi-layers with each layer having thickness “he” can be calculated as follows:
different properties (elastic modulus and Poisson’s ratio). In
order to simplify these complex systems for stress and strain h13 E1 he3 E 2
; or
T
calculations, Odemark has developed a method to transform
these multi-layer systems into an equivalent one-layer system 1 v12 1 v 22
with equivalent thicknesses but one elastic modulus. This method
has been used in many advanced research studies and design
(2)
methods including the newly developed Mechanistic-Empirical
Pavement Design Guide (MEPDG). This paper investigates the E1 1 u 22
accuracy of the Odemark method and presents a methodology to
he h1 3
E 2 1 u12
ES
increase its accuracy. A two-layer system with different modular
ratios and thicknesses was extensively analyzed. The results
showed that a correction factor must be used with Odemark’s
method in order to produce highly accurate stress and strain
results. This correction factor is not constant and depends not
For the case of a two-layer system with equal Poisson’s
ratio, the equivalent thickness can be calculated using the
following formula:
only on the modular ratio and the thickness of the layer but also E1
on the depth of interest. he h1 3 (3)
E2
Keywords; Odemark; stress; stiffness; elastic modulus, MEPDG Comparing stresses and strain calculated using the
Odemark’s method with those from the elastic theory led to the
I. INTRODUCTION conclusion that they are relatively different. In order to achieve
a better agreement between Odemark’s method and the elastic
Odemark has developed an approximate method to theory, a correction factor “f” was applied to the above
A
calculate stresses and strains in multiplayer pavement systems equation as follows.
by transforming this structure into an equivalent one-layer
system with equivalent thicknesses but one elastic modulus. E1
This concept is known as the method of equivalent thickness he f h1 3 (4)
(MET) or Odemark’s method. MET assumes that the stresses E2
and strains below a layer depend only on the stiffness of that Researchers reported that the value of the correction factor
layer. If the thickness, modulus and Poisson’s ratio of a layer is “f” depends on the layer thicknesses, modular ratios, and the
IJ
changed, but the stiffness remains unchanged, the stresses and number of layers in the pavement structure. Furthermore, they
strains below the layer should also remain (relatively) mentioned that using a value of 0.8 to 0.9 for “f” leads to a
unchanged. According to Odemark, the stiffness of a layer is reasonably good agreement between the two methods [2].
proportional to the following term [1]:
h3 E
1 v2
(1) h1 E1 1 he E2 2
where:
h = thickness of the layer
E = elastic modulus E2 2 E2 2
v = Poisson’s ratio
Figure 1. Odemark’s Transformation of a Layered System
T
using the falling weight deflectometer (FWD) [2]. Senseney
and Mooney also used this method in the characterization of a
two-layer soil system using lightweight deflectometer [3]. The
newly developed mechanistic empirical pavement design guide
(MEPDG) implemented this method to transform a multi-layer
pavement system into an equivalent one layer system. This
of interest define the effective length of the stress pulse. In MEPDG, for any pavement layer, the effective length of
Because the slope of the stress distribution shown in Fig. 2 the stress pulse is computed at the effective depth (Zeff). The
is a function of the stiffness of the layer and since there is no effective depth is the transformed depth at which the loading
present relationship exists to relate them together, a multi-layer frequency is needed. The effective depth for the transformed
pavement system is transformed into an equivalent one layer section (as shown in Fig. 4) is calculated with the help of
system in order to estimate the effective length. The Equation. 8 [6]:
transformed section using MET is shown in Fig. 3. The
transformed section has the modulus of the subgrade and has n 1
E E
an equivalent thickness of he. In MEPDG, for simplicity, the Z eff hi 3 i hn 3 n (8)
stress distribution for a typical subgrade soil is assumed to be at i 1 E SG E SG
45 degree as shown in Fig. 4. Using this stress distribution, the
effective length can be calculated at any depth within the
transformed pavement system.
T
for each thickness were analyzed. A Poisson’s ratio of 0.35 was ratio. The results showed good agreement only for the vertical
assumed in all computations. Fig. 5 shows the applied load and stresses calculated at the interface between the two layers when
the properties of the two layer system used in the analysis. using f of 0.8 to 0.9. However, at any depth other than the
interface between the two layers the results showed a
significant difference between the two solutions. This means
9000 lb that the correction factor f is also dependent on the depth.
Z
120 psi
2a = 9.772 in
ES h
In order to verify that, the “Solver” function in Excel
spreadsheet was used for each modular ratio, to calculate the f
factor at each depth such that:
vzi – vzti = 0 (9)
E1, where:
vz = vertical stress calculated from a two-layer
E2, system at depth zi using KENPAVE.
vzti = vertical stress calculated from Odemark
Figure 5. Two-Layer Pavement System used for the Analysis transformed depth zti (One layer system) using
Boussinesq or KENPAVE.
A
First a linear elastic solution was performed on the two-
layer structure using the KENPAVE software to calculate the The results showed that f depends not only on the modular
vertical and radial stresses at different depths measured from ratio and the thickness of the upper layer in the two-layer
the surface of the upper layer under the centerline of the load. pavement system but also on the depth of interest.
Then Odemark transformation concept was used to convert the
two-layer problem into one layer with equivalent thicknesses Figs. 7 through 9 depict the relationship between the
and one modulus. A comparison between stresses calculated correction factor “f” and depth at different modular ratios for
the investigated two layer system with h1 = 6, 10, and 15 inches
IJ
1.00
0.90
0.40
0.30
0.20
0.10
0.00
0 2 4 6 8 10 12
Depth, Z (in)
Figure 7. Relationships between the Correction f and Depth (Z) for the system with (h1 = 6 in.)
T
1.00
0.90
0.70
0.60
0.50
0.40
ES Interface between layers 1 & 2
E1/E2 = 16.66
E1/E2 = 33.33
E1/E2 = 50.00
E1/E2 = 66.66
0.30
0.20
0.10
0.00
0 5 10 15 20 25
Depth, Z (in)
Figure 8. Relationships between the Correction f and Depth (Z) for the system with (h1 = 10 in.)
A
1.00
0.90
y = -5E-05x3 + 0.0024x2 - 0.0146x + 0.7033
0.80 R² = 0.995
Correction Factor (f)
Layers 1&2
y = 0.0001x3 + 0.0003x2 - 0.0074x + 0.4241
0.60 R² = 0.9975 y = 0.0001x2 - 0.0096x + 0.9712
R² = 1
E1/E2 = 3.33
Figure 9. Relationships between the Correction f and Depth (Z) for the system with (h1 = 15 in)
1.00
0.90 thickness of the upper layer did not eliminate the effect of layer
0.80 thickness. However, there seems to be a general relationship
0.70 that relates the f value for each layer, in a two-layer system, to
0.60 the modular ration and the depth.
0.50 Figures 7 thru 9 show also that for the cases with modular
0.40
ratios higher than 3.33 the value of f asymptotes to 0.85, 0.8,
0.30
and 0.79 for the pavements systems with h1 = 6, 10, and 15 in
0.20
y = 0.0012x3 + 0.0225x2 - 0.1252x + 0.6244 respectively. For the 3.33 modular ratio, the f at the interface is
0.10 R² = 0.9992
0.89, 0.87, and 0.85 for the systems with h1 = 6, 10, and 15 in
0.00
0 1 2 3 4 5 6 respectively. It can be concluded from these results that, in a
Depth, Z (in) two-layer system, for different modular ratios, f in the range of
0.8 to 0.9 yields vertical stresses that are relatively close to the
Figure 10. Relationship between Depth and Correction Factor for Layer-1
ones from theory of elasticity at the interface between the two
(h1=6 in., E1/E2=50) layers.
Fig. 13 presents an example of the relationship between the
vertical stresses calculated at different radial distances, for
1.00
different depth values, for the two-layer system with (h1 = 10
Corretion factor for layer-2
0.90
in) and a modular ratio of 16.67. Theses vertical stresses were
T
0.80
0.70 calculated for a two-layer problem using KENPAVE. This
0.60 two-layer system was then transformed using Odemark’s
0.50 method with the correction factor f as a function of depth and
0.40 the vertical stresses at the transformed depths (Zt) were
0.30 calculated. This is shown on Fig. 14. Comparing the values of
0.20
y = 0.0068x2 - 0.1653x + 1.5915
the vertical stresses from both methods resulted in agreement
as shown in Fig. 15.
0.10
0.00
5 6 7 8
Depth, Z (in.)
9
ES
R² = 0.9997
10 11
120
110
100
90
Leff @ Z = 1 in
11.772 in
Z = 1 in
Z = 3 in
Figure 11. Relationship between Depth and Correction Factor for Layer-2 80 15.772 in
Z = 5 in
Leff @ Z = 3 in Z = 7 in
70 Z = 9 in
19.772 in
60 Leff @ Z = 5 in
centerline of the load for the transformed system and the two-
Leff @ Z = 9 in
20
0
-24 -20 -16 -12 -8 -4 0 4 8 12 16 20 24
Radial Distance, in
150
Equivalent Transformed Section
125 Figure 13. Relationship between Vertical Stresses and Radial Distances at
Different Depths (E1/E2 = 16.67), Two-Layer Solution
Calculated Stress, psi
120
100
110 Zt = 1.24 in
IJ
100 Zt = 3.47 in
75 Zt = 6.15 in
90 Leff @ Zt = 3.47 in
16.712 in
Zt = 10.11 in
50 80
Vertical Stress, psi
Zt = 17.27 in
70 Leff @ Zt = 6.15 in
22.072 in
25 60
50 Leff @ Zt = 10.11 in
Line of Equality 29.992 in
0 40
0
Figure 12. Comparison of the Two-Layer and the Equivalent One-Layer -24 -20 -16 -12 -8 -4 0 4 8 12 16 20 24
Pavements Computed Stresses after Appling the Developed Correction Radial Distance, in
REFERENCES
140 [1] Ullidtz, P., (1987), Pavement Analysis, Development in Civil
Engineering, Vol.19, Amsterdam, the Netherlands.
Vertical Stress, psi (Odemark One- Layer
120
[2] Subagio, B., Cahyanto H., Rachman, A., and Mardiyah, S., “Multi-Layer
Pavement Structural Analysis Using Method of Equivalent Thickness,
100
Case Study: Jakarta-Cikampek Toll Road,” Journal of the Eastern Asia
Society for Transportation Studies, Vol. 6, pp. 55 - 65, 2005.
80
[3] Senseney, C., and Mooney, M., “Characterization of a Two-Layer Soil
Solution)
T
Correction Factor (f as a Function of Depth) for the system with h1 =10 in. and Predict Alligator Fatigue Cracking Distress based on AC Dynamic
E1/E2 = 16.67 Modulus,” In Transportation Research Record, Journal of the
Transportation Research Board, No. 2095, Transportation Research
board of the National Academies, Washington, DC, 2009, pp. 115-124.
V. CONCLUSIONS [8] Sotil, A., “Use of the Dynamic Modulus E* Test as Permanent
Based on the analyses conducted in this research, the Deformation Performance Criteria for Asphalt Pavement Systems”,
following conclusions were highlighted: Ph.D. Dissertation. Department of Civil and Environmental Engineering,
Arizona State University, Tempe, AZ, December 2005.
ES
In order to get a good agreement between the stresses
and strains calculated using Odemark’s concept and
those from theory of elasticity; a correction factor f has
to be introduced. This correction factor was found to
be a function of the layer thickness, depth and modular
ratio.
The study showed a good agreement between the
vertical stresses at the interface between the two layers,
in a two-layer system, calculated using the theory of
elasticity and Odemark’s concept when using a
correction factor (f) in the range of of 0.8 to 0.9 which
A
agrees with the other literature studies.
However, at any other depth within each layer, this
correction factor is not a constant value. It was found
that this correction factor varies with the change in the
depth of interest. The study showed that, the points (Z
values) in the first layer follow a 3rd degree polynomial
relationship with the correction factor (f) for each
IJ