See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/355969856

Modelling of the stresses and strains distribution in an RCC pavement using

the computer code 'Abaqus'

Article · January 2010

CITATIONS READS

0 39

4 authors, including:

Mustapha Zdiri Mongi Ben Ouezdou

Université de Tunis, Ecole Nationale Supérieure d’Ingénieurs de Tunis National Engineering School of Tunis, University of Tunis El Manar
24 PUBLICATIONS 105 CITATIONS 172 PUBLICATIONS 1,737 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Mongi Ben Ouezdou on 14 August 2023.

The user has requested enhancement of the downloaded file.

 Electronic Journal of Structural Engineering, 9 (2009)

Modelling of the stresses and strains distribution in an RCC pavement

using the computer code "Abaqus"

M. Zdiri*
Civil Engineering Laboratory, National Engineering School of Tunis, BP 37, Tunis-Belvédère 1002, Tunisia.
Laboratoire Génie Civil et Environnemental - Ecole des Mines de Douai, rue Charles Boursel -59508 France
N. Abriak
Laboratoire Génie Civil et Environnemental - Ecole des Mines de Douai, rue Charles Boursel -59508 France
J. Neji
Applied Mechanics and Systems Research Laboratory - Polytechnic School of Tunis, Tunisia.
M. Ben Ouezdou
Civil Engineering Laboratory, National Engineering School of Tunis, BP 37, Tunis-Belvédère 1002, Tunisia.

*
Email: [email protected]

ABSTRACT: The Roller Compacted Concrete (RCC) pavements are built employing the Portland cement.
The design of the rigid RCC pavements was developed gradually through methods of design. These methods
were established by various organisms for the determination of the necessary thicknesses of roadways. In this
study a numerical 3D modelling was used by introducing to the computer code "Abaqus" the behaviour law of
the RCC. Several other methods of 2D modelling were applied for the determination of the stress and the
strains in the RCC slab such as the Closed-form Formulas and the Portland Cement Association (PCA)
method. Finally the results of 3D modelling are compared with those obtained by the various other methods.
The comparison shows good correspondences although the 3D modelling gives results slightly lower than
those given by the 2D methods in stresses.

1 INTRODUCTION analysis, computer programs were developed by us-

ing the finite element method such as ILLI-SLAB
The determination of the pavement thicknesses was (Tabatabaie and Barenberg 1980), WESLIQID
based purely on the experiment until the 1920 years. (Chou 1981), J-SLAB (Tayabji and Colley 1986),
These experiments were developed also with the FEACONS-IV (Choubane and Tia 1995), IS-
time. The rigid RCC pavements, like those of con- LAB2000 (Khazanovich et al. 2000) and
ventional concretes, can be analysed by the plate's WESLAYER (Huang, 2004). The main advantage of
theory. The plate's theory supposes that the concrete these approaches is the evaluation of the critical load
slab is a medium thickness plate with plane sections transfer phenomena and the stress distributions in
before and after strains. If the wheel is placed close the rigid pavements like the RCC slabs. Moreover,
to the edge of the slab, then only the plate's theory Channakeshava et al. (1993), Masad et al. (1996)
can be used for the rigid pavements. The plate's the- and William and Shoukry (2001) employed the finite
ory or the layers theory can be used when the load is element methods in order to analyze the behaviour
applied to the slab centre. The existence of the joints of concrete pavements. Using Abaqus software,
in the rigid pavements makes the use of layers theory Zaghloul et al. (1994), Uddin et al. (1997) and
inapplicable (Jeuffroy and Sauterey, 1989). Darter et al. (2001) developed models for the study
The design methods were developed by various of rigid pavements and its foundation. Davids et al.
organisms for the determination of the necessary (2003) studied, also, the effect of dowel looseness on
thicknesses of pavements (Gauthier et al., 2004). the structural response of jointed concrete pave-
The analytical solutions, developed thereafter, vary ments. The study employed 3D finite element mod-
from the "Closed-form Formulas" to the complex els developed with EverFE.
derivations which are valid for the determination of In the present study, a numerical 3D modelling
the stress and the strains in the rigid pavements. But was used by introducing, to the computer code
with the development of the powerful finite element "Abaqus", the behaviour law of the RCC. Other
method, we notice a significant evolution in the methods of 2D modelling were applied for the de-
analysis of the rigid pavements. Various finite ele- termination of the stress and the strains in the RCC
ment models have been developed for analysing the slab such as the Closed-form Formulas and the Port-
behaviour of concrete pavement systems. For these land Cement Association (PCA) method. Finally, the
37
 Electronic Journal of Structural Engineering, 9 (2009)
2 cm Joint
Contact surface of S
results of 3D modelling were compared with those o 1
loading
obtained by the various other methods. 3

RCC Slab 1 RCC Slab 2

2 NUMERICAL MODELING 3D BY THE

4m
COMPUTER CODE "ABAQUS 6.7" Circulation direction

2.1 The 3D modelling principle 7m S 7m

In the present model, a solid foundation was used,

which is more realistic although this modelling re-
Figure 1. Model adopted for the RCC - plan view
quires much more memory for the resolution. The
solid foundation is more realistic than the liquid Tire
foundation, because the deflection in any nodal point



2
3
depends not only on the force in this node but also of o
RCC Slab1
the forces in all the other nodes. Moreover, all the
Gravel Foundation Layer
analytical solutions of the 2D models are based on
the proposal that the slab and the foundation are in Support Ground Layer
perfect contact (Coquand, 1989). With the advan-
tages of calculations of the numerical methods by
computers, the developed analyses are based on a
partial contact between the layers (Huang, 2004). 4m
This work includes, mainly, a 3D modelling by the Transverse section S-S
use of the computer code of the finite elements
"Abaqus 6.7" in order to understand, with more pre-
Tire loading
cision, the distribution and the evolution of the 2 Circulation direction
1 2 cm Joint
stresses and displacements in the entire RCC slab. o
RCC Slab1 (H,E,ν) RCC Slab2 (h,E,ν ) h
These results were shown especially for many load- Gravel Foundation Layer (hf, Ef,ν f) hf
ing position, in top and bottom fibres and in the in-
terface with the sub-base. In the numerical 3D mod- Support Ground Layer (hs, Es,νs)
hS
elling by finite elements using Abaqus 6.7, the
authors choose various options of manual and auto-
matic incrementing with an automatic tolerance of
convergence. The control parameters of the man- b) Elevation view
agement of the numerical analysis of the problem are
adjusted automatically with a low manual adjust-
Figure 2 (a & b). Model adopted for the RCC
ment. The convergence criteria are also adjusted dur-
ing the analysis to ensure a precise solution.
Basing on the literature, the Friction ratio
In this numerical approach, two RCC slab sepa-
Slab/Foundation was taken equal to 1.5 and the Fric-
rated by a joint, were modelled. Both rest on a gravel
tion ratio Foundation/ground equal to 1.1. The dif-
sub-base suitably compacted. The whole also rests
ference between the top and bottom RCC slab tem-
on a ground support. The geometries and the me-
peratures was assumed to be equal to 11°C (ARA,
chanical properties of material were introduced.
Inc., ERES Division, 2003).
These introduced parameters were: the elastic
modulus E, the Poisson's ratio ν and the admissible 2.2 Meshing and boundary conditions
stresses of tensile and compression. All dimensions
The boundary conditions were selected in order to
are finite in 3D; the diagrams of the model are pre-
approach to the real boundary conditions. The gravel
sented in Figures 1 and 2.
sub-base and ground support layers were fixed in the
The data of the adopted model are the following:
two directions 1 and 2. The two RCC slabs were
-Two RCC slabs of (4.00 m ×7.00 m), h= 20 cm,
fixed on one face in directions 1 and 2. The gravel
E = 31000 MPa, ν=0.22 sub-base and the two slabs were free in direction 3
- Gravel foundation layer of (4.00 m × 14.00 m), and they can move in this direction. Only, the
hf= 30 cm, Ef = 155 MPa, νf=0.35 ground support was fixed in the direction 3 (Figure 1
- Ground layer of (4.00 m × 14.00 m): hs= 130 cm, and 2). Meshes based on hexahedrons with eight
Es = 50 MPa, νs=0.45 nodes were chosen. The size of the meshes was also
optimized in order to obtain a compromise between
the time and the stability of calculation (Figure 4). It

38
 Electronic Journal of Structural Engineering, 9 (2009)

was found that a fine mesh (10 by 10 cm) is required

4m
in the RCC slab supporting the load. However a rela- RCC slab Circulation direction

tively coarse mesh (30 by 30 cm) was chosen in the Contact Surface
gravel sub-base, ground support and RCC slab not of loading
supporting the load. It did not significantly affect the
precision of the stress and displacement prediction. 7m
Moreover, all the preceding analytical solutions
of the 2D models are based on the proposal that the
slab and the foundation are in perfect contact. In the c) Load in the corner of the slab
present study, two assumptions can be considered in
this modelling. The first consists in supposing that
the interaction between the layers is without friction,
which means that the layers are stuck. The second
consists in considering a friction ratio between the
RCC slab and the gravel foundation. Another fric-
tion ratio was taken between the sub-base and the
support ground.

Slave d) 3D Model for interior loading case

Surface Figure 4 (a, b, c & d). Different positions of load applied to the
Master
3D Model
Surface

In their studies, ARA and ERES (ARA, Inc.,

ERES Division, 2003) indicated that load transfer ef-
Figure 3. Contact graph and surfaces interactions. ficiency (LTE) at the transverse joints has only a
limited effect on the distribution of the RCC bottom
2.3 Various cases of loadings surface stresses. For this reason, this phenomena was
not taken into accompt in the present modelling.
Three cases of loading of the RCC slab were studied; To assure adequate simulation of the tire–pavement
an interior loading (load in the centre), a load in the contact area, tire footprints were modelled using rec-
corner and a load at the edge of the slab; but the ap-
tangular shapes with equivalent contact areas as
plied load were a static loading. The applied load
was of 65000 N/tire which is equal to 740000 Pa. showed in Figure 5.
The three cases of loading are schematized in Figure
0.6 Lc=25cm

4: Area=0.5227 Lc2
0.6 Lc

0.3 Lc
Contact Surface
RCC slab of loading
Lc 0.871 Lc=35cm
4m

a) Real contact surface of tire b) Equivalent contact surface

Circulation direction Figure 5 (a & b). The adopted contact surface of tire.

7m
2.4 The RCC behaviour
To suitably simulate the behaviour of the RCC be-
fore and after cracking (rupture), the option "Con-
a) Interior loading (Load in the centre of the slab)
crete Smeared Cracking" was used by introducing
the necessary parameters to the definition of the be-
RCC slab
Circulation direction haviour law of the RCC (Figure 6)
4m

Contact Surface
of loading 3 MODELLING RESULTS

In order to exploit suitably and to develop results of

7m
the 3D modelling using the computer code by finite
elements "Abaqus 6.7", the distribution of the stress
b) Load in the edge of the slab and displacements in the loading position and in the

39
 Electronic Journal of Structural Engineering, 9 (2009)

Stress-strain curves of the RCC interfaces was studied. The distributions in the RCC
35
M1 M2 M3 M4
slab, the foundation and the ground support are
30
schematized on the graphs in forms of colour curves.
25
3.1 Stress distributions
Stress (MPa)

20

15
The results of this modelling and the distributions of
the stress and the strains in the RCC slab, the foun-
10
dation and the ground are presented in Tables 1 and
5 Figure 7
0
0 0,0004 0,0008 0,0012 0,0016 0,002 0,0024 0,0028 0,0032 0,0036 0,004 Table 1. The obtained results of Stress by the 3D modelling
strain (m/m) Stress Interior load Corner load Edge load
(MPa) Top Bot- Top Bot- Top Bot-
Figure 6. Representation of the relation stress-strain of the RCC σ11 1.720 1.51 2.310 1.790 2.570 2.210
σ22 0.735 0.263 0.781 0.387 0.761 0.443
σ33 1.690 1.490 1.280 1.280 0.537 0.420
σ12 0.266 0.266 0.518 0.518 0.657 0.657
a) Top Fibres σ13 0.221 0.221 1.690 1.550 0.772 0.772
σ23 0.264 0.264 1.94 1.590 0.328 0.328
as presented in Figure 7 and 8.
Tables 1 and Figure 7 give the results obtained by
the 3D modelling of the RCC pavement in the case
of only one applied load. The extreme values of the
stresses and the strains appear on the top and bottom
fibres of the RCC slab. The maximum values of the
stresses are in direction 1. They are recorded in the
loading case of "Load at the edge". Compressive
stresses of 2.57 MPa was obtained at top fibres of
the slab and, in the same time, tensile stresses of
b) Bottom Fibres 2.21MPa were obtained at bottom fibres. Shear
stresses are extreme also in the plan 1-3. The maxi-
mum value is equal to 1.69 MPa and it is in the load-
ing case of a "Load at the corner".
- It is in the case of the edge loading that the stresses
in direction 1 and displacements in direction 2, are
the highest compared to the other cases of loading
(from 1.2 to 1.5 times).

3.2 The displacements distributions

The distributions of the strains in the RCC slab, the
foundation and the ground are presented in Tables 4
and schematized by the graphs in Figure 8.
c)
c) Section
Section
Table 2. The obtained results of Displacements by 3D model
Corner load Interior load Edge load
(×10-5m)
Deflection

Top Bottom Top Bottom Top Bottom

Fibres Fibres Fi- Fibres Fi- Fibres
bres bres
U1 3.21 3.21 2.70 2.70 6.43 10.3

U2 143 143 40 40 126 126

U3 1.76 2.36 3.14 3.14 4.47 10.6

Figure 7(a, b & c). Example of Stress distribution for the case
of the interior loading.

40
 Electronic Journal of Structural Engineering, 9 (2009)

4 APPLICATION OF THE 2D MODELLING TO

a) Corner THE RCC PAVEMENT

In 2D modelling several formulas, used for the study

of the stresses and the strains in the rigid pavements,
were applied such as the Closed-form Formulas and
the PCA method.
4.1 Corner loading
For the loading in the corner, the Westergaard equa-
tion which proposes a successive approximation
method was used. It gives the stress σ and the de-
flection ∆ on the bottom surface of the slab in the
form 1 and 2 (Huang, 2004):

 0.6

b) Interior 3 P   a 2  
σ = 2 1− (1)
h   l  
 

P   a 2 
∆= 1.1 − 0.88  (2)
kl 2   l 
 

Where l is the radius of relative stiffness (m), a is

the contact radius (m), P is the concentrated loading
(N), h is the thickness of the RCC slab (m) and k is
the reaction modulus of the foundation (MPa/m).
l can be determined by the formula:
0.25
 Eh 3 
c) In edge
l= 2  (3)
 12( 1 − ν )k 
Where E: elastic modulus, ν: Poisson's ratio.
For the studied cases, P= 65000 N, h= 0.20 m and
three cases of foundation were treated: k=35, k=95
and k= 250 MPa/m. Table 3 summarizes the results
for these three treated cases.
The formulas 4 and 5, suggested by Ioannides et
al. in 1985, were also used. In these equations, the
finite element method was applied to evaluate the
solution of Westergaard (Huang, 2004). The results
are summarized in Table 3:
3P   c  
0.72

σ =    
1 − (4)
Figure 8(a, b & c). Displacements distribution for the three h 2   l  
cases of the loading
P   c 
In all loading cases, displacements in direction 2 ∆= 1.205 − 0.69  (5)
kl 2   l 
are higher compared to those in direction 1 and 3. In
this direction 2, deflections in the loading case at the
corner (1.43×10-3 m) are higher than those in the Where c can be taken: c=1.77×a
other loading cases (0.40×10-3 m in the interior and
1.26×10-3 m in the edge). Generally, in directions 2 4.2 Interior Loading
and 3, the displacements, in the case of Edge load-
For the interior loading case, the method suggested
ing, are higher than those in the other cases of load-
by Westergaard in 1926 was adopted to determine
ing.
the stress and the deflection in rigid slab under circu-
lar load of radius which is represented by the formu-

41
 Electronic Journal of Structural Engineering, 9 (2009)

las 6 and 7 (Huang, 2004). The results are summa- 4.3 Loading in the edges
rized in Table 4:
The equations of the stresses and the deflections
3( 1 + ν )P l
σ = ( Ln + 0.6159 ) (6) used in this case, due to the edges loadings, are pro-
2πh 2 b
posed by Westergaard in 1948 (Jeuffroy and Sau-
terey, 1989). These equations represent the solutions
P  1   a   a  
2
for the maximum stresses and deflections produced
∆= 1 +  Ln  − 0.673   (7)
by loads applied to the edges of a rigid slab. They
8 kl 2  2π   2l   l  
were corrected after that by Ioannides and al., to be
Table 3. Results of the stresses and deflections for the corner presented in the following forms 9 and 10:
loading case by two methods
EFoundation (MPa) 35 155 500 3( 1 + ν )P   Eh 3  4ν 1 − ν 1.18( 1 + 2ν )a 
σ=  Ln  + 1.84 − + +  (9)
CBR 6 32 90 π ( 3 + ν )h 2   100 ka
4
 3 2 l 
ERCC (MPa) 31000 31000 31000
k (MPa/m) 35 95 250
0.91 0.71 0.53 2 + 1.2νP  ( 0.76 + 0.4ν )a  (10)
l (m) ∆= 1 − 
σ Westergaard (MPa) 3.14 2.81 2.46 Eh 3 k  l 
σ Ioannides (MPa) 3.22 2.83 2.404
∆ Westergaard (mm) 2.216 1.595 0.625 The maximum tensile stress at bottom fibre of the
∆ Ioannides (mm) 2.281 1.627 0.718 slab edge, caused by the truck load, is calculated by
the model "ALIZE" (Jeuffroy and Sauterey, 1989)
Moreover the procedure of design "Design of air- through the following formula 11.
port concrete pavement" established by the Portland
Cement Association (PCA) was applied. It leads to 0.572 P l (11)
σ= ( 4 log   + 0.359 )
determine, according to the thickness of the RCC h2 b
pavement, the tensile stress in bottom fibre of the
pavement in the slab centre which is caused by the The results are summarized in Table 5:
load of a given axle (PCA, 1984), (Gauthier et al.,
2004). The calculation of the stress was achieved by Table 5. Results of the stresses and deflections for the edges
the following relation 8: loading case by two methods

EFoundation (MPa) 35 155 500

0.316 P l
σ= 2
( 4 log   + 1.069 ) (8) a (m) 0.175 0.175 0.175
h b ERCC (MPa) 31000 31000 31000
K MPa/m) 35 95 250
The results are summarized in the following Table l (m) 0.91 0.71 0.53
4: σ Westergaard Pa) 3.9 3.27 2.70
σ ALIZE (mm) 3.21 2.83 2,59
Table 4. Results of the stresses and deflections for the interior ∆ Westergaard (mm) 1.307 0.714 0.416
loading case by two methods
EFoundation (MPa) 35 155 500
Interpretation: These results show that the stresses
a (m) 0.175 0.175 0.175
b (m) 0.163 0.163 0.163
are maximum in the edge loading than in the corner
ERCC (MPa) 31000 31000 31000 and in the interior loading, whereas the deflections
K (MPa/m) 35 95 250 obtained in the edge loading are lower than those in
l (m) 0.91 0.71 0.53 the corner load cases.
σ Westergaard (MPa) 2.61 2.14 1.98
σ PCA (mm) 2.46 2.02 1.82
∆ Westergaard (mm) 0.434 0.23 0.181 5 APPLICATION OF THE DANTU METHOD
FOR THE RCC PAVEMENT
Interpretation: The deflection due to the load in the
corner is higher than the deflection in the slab centre; In this section, the method of Dantu (Coquand,
this is true only when there is no load transfer by the 1989) was applied for the study of the maximum
joints to the corner. In the case of load transfer by stresses in a slab of limited size and where load does
the joints to the corners the stresses will be lower not occupy necessarily the centre. This method is
whereas the deflections remain without variations. based on the optical study of the strains of a reduced
model of elastic material which is the cork. The
equations of the moments T as function of z are de-
fined by the following formulas 12, 13, 14 and 15:

42
 Electronic Journal of Structural Engineering, 9 (2009)

- Interior load case: but these results remain also different to those given
T = P( 1 + σ )
1
(12) by the Dantu method (Coquand, 1989). The 3D
3.60 + 12 × Z modelling gives lower stresses than those given by
the other methods. It gives also displacements higher
- Edge load case: than those given by the Westergaard method (Table
T = P( 1 + σ )
1
(13) 7).
1.70 + 4.1 × Z + 10.3 × Z ² - For the edge loading cases, the Westergaard, Ioan-
nides, and Dantu methods give very close results in
- Corner load case: stresses. The 3D modelling gives lower stresses to
1
(14)
those given by these methods. However it gives dis-
T = P( 1 + σ )
7 − 21 × Z + 115 × Z ² placements higher than those given by the Wester-
gaard method (Table 7).
Where:
E f (1 − ν ² ) Table 7. Results of the stresses and deflections for three loading
a
Z= 3 6 (15) cases by different methods
h (
E 1 −ν f ² ) Corner Interior Edge
Loading Loading Loading
Methods
The tresses were calculated by the equation 16 and σ ∆ σ ∆ σ ∆
the results are summarized in Table 6: (MPa) (mm) (MPa) (mm) (MPa) (mm)
6T Westergaard 2.81 1.595 2.14 0,23 3.27 0.714
σ= (16) Ioannides et
b × h2 al.
2.83 1.627 - - -
PCA - - 2.02 - - -
Table 6. Results of the stresses for three loading cases by ALIZE - - - - 2.83 -
Dantu method.
Dantu 1.204 - 1.78 - 3.261 -
EFoundation (MPa) 35 155 500
3D Modelling
a (m) 0.175 0.175 0.175 1.79 1.43 1.51 0.40 2.21 1.26
by "Abaqus"
K (MPa/m) 35 95 250
l (m) 0.91 0.71 0.53
Z 0.16 0.28 0.42
σ (interior) (MPa) 2.25 1.78 1.40 7 CONCLUSION
σ (edge) (MPa) 4.63 3.261 2.29
σ (corner) (MPa) 1.88 1.204 0.683 Through the present study the following conclusion
were drawn:
Interpretation: In the edges loading case, values of - The comparisons of the stresses and displacements
moments and stresses were found higher than those predicted by several models with those obtained by
in the other cases because fatigues are clearly higher Abaqus code 6.7 show good correspondences al-
at the edge than in the interior and corner. though the present 3D modelling gives results
slightly lower than those given by the other methods
in stresses.
6 COMPARISON AND INTERPRETATION OF - The numerical 3D modelling method using com-
THE RESULTS OF THE VARIOUS METHODS puter code "Abaqus 6.7" is concluded a reliable
method for the determination of the stress and dis-
The RCC pavement studied in this work are rigid placements in the RCC slab. The differences be-
types in which the RCC slab bends elastically under tween this method and the 2D modelling are due to
the loads. It transmits also the stresses and thus dis- the difference in the adopted assumptions and the
tributes them on a large surface of the foundation. friction ratios.
This principle is visualized by the graphic distribu- - The results of this study showed that the use of the
tions showing that the vertical stresses in direction 2 assumptions of the partial contact between the lay-
which arrive at the foundation are generally very ers: RCC slabs, Gravel foundation and the ground
low. support, is more realistic then the perfect contact.
- For the corner loading case, the methods of - Compared to the past researches, this study contain
Westergaard and Ioannides et al. give very similar several particularities such as the introduction of the
results in stresses. The 3D modelling gives results behaviour low of the RCC and the analysis of two
lower than those given by these methods in stresses RCC slabs subjected to traffic loading.
and with lower displacements (Table 7) also. How- - This work confirms that the use of Abaqus soft-
ever, the results of Dantu method remain different ware has a great potential as a powerful tool for a 3D
from those given by the others methods. modelling of the RCC pavement.
- In the interior loading case, the methods of Wester-
gaard and PCA give very close results in stresses,
43
 Electronic Journal of Structural Engineering, 9 (2009)

REFERENCES [11] Huang Y. H., (2004), Pavement Analysis and Design, 2nd
Edition, Kentucky University USA Prentice Hall, pp. 147-
[1] Abaqus guide Version 6.7, (2007), Computer Software for 166.
Interactive Finite Element Analysis by Hibbitt, Karlsson & [12] Ioannides A. M., Thompson M .R. et Barenberg E. J.,
Sorensen, Inc. Pawtucket, RI. (1985), Westergaard solutions reconsidered, Transportation
[2] ARA, Inc., ERES Division, (2003), Guide for Mechanistic- research record 1043 Transportation research board pp.13-
Empirical Design of new and rehabilitated pavement struc- 23.
tures [13] Jeuffroy G. and Sauterey R., (1989), Cours de routes,
[3] Canadian Cement Association, (1984), Thickness design for Chaussée en béton de ciment, Presse de l'Ecole Nationale
concrete highway and street pavements, Canadian Edi- des Ponts et chaussées Paris France, pp. 49-79.
tion/metric, Engineering Bulletin, EB209.03P, 48 pages. [14] Khazanovich L., Yu H.T., Rao S., Galasova K., Shats E.
[4] Channakeshava C., Barzegar F. and Voyiadjis G., (1993), and Jones R., (2000), ISLAB2000, Finite Element Analysis
Nonlinear FE Analysis of Plain Concrete Pavement with Program for Rigid and Composite Pavements, User’s
Doweled Joints, Journal of Transportation Engineering, Guide, ERES Consultants, Champaign, Illinois.
Vol. 119, N.5, pp.763-781. [15] Masad E., Taha R. and Muhunthan B., (1996), Finite-
[5] Choubane B. and Tia M., (1995), Analysis and Verification Element Analysis of Temperature Effects on Plain-Jointed
of Thermal-Gradient Effects on Concrete Pavement, Journal Concrete Pavements, Journal of Transportation Engineer-
of Transportation Engineering, Vol.121, No. 1, pp. 75-81. ing, Vol. 122, No. 5, pp. 388-398.
[6] Chou Y.T., (1981), Structural Analysis Computer Programs [16] Tabatabaie A. M. and Barenberg E.J., (1980), Structural
for Rigid Multicomponent Pavement Structures with Dis- Analysis of Concrete Pavement Systems, Journal of Trans-
continuities: WESLIQUID and WESLAYER. Technical portation Engineering, ASCE, National Research Council,
Report GL-81-6. Vicksburg, MS: U.S. Army Engineer Wa- Washington, D.C. Vol. 106, No. 5, pp. 493-506
terways Experiment Station. [17] Tayabji S.D. and Colley B.E., (1986), Analysis of Jointed
[7] Coquand R., (1989), Routes, Chaussées rigides et chaussées Concrete Pavements, Fedral Highway Administration,
souples, 8eme Edition, Eyrolles, Paris France. McClean VA, Report No. FHWA/RD-86/041, February
[8] Darter M., Khazanovich L., Snyder M., Rao S. and Hallin 1986.
J., (2001), Development and Calibration of a Mechanistic [18] Uddin W., Noppakunwijai P. and Chung T., (1997), Per-
Design Procedure for Jointed Plain Concrete Pavements. formance Evaluation of Jointed Concrete Pavement Using
Proceedings, 7th International Conference on Concrete Three-Dimensional Finite-Element Dynamic Analysis,
Pavements, Orlando, FL. Transportation Research Board, 76th Annual Meeting,
[9] Davids W.G., Wang Z. M., Turkiyyah G., Mahoney J. and Washington, D.C., Paper No. 1414.
Bush D., (2003), 3D Finite, Element Analysis of Jointed [19] William G.W. and Shoukry S. N., (2001), 3D Finite Ele-
Plain Concrete Pavement with EverFE2.2, Transportation ment Analysis of Temperature-Induced Stresses in Dowel
Research Record, No 1853, National Research Council, Jointed Concrete Pavements. International Journal of Ge-
Washington D.C, pp. 92-99 omechanics, Vol 1, No 3, 2001, pp. 291-308.
[10] Gauthier P. and Marchand J., (2002), Conception et réalisa- [20] Zaghloul S. M., White T. D., Vincent P. D. and Drian C.,
tion de revêtements en béton compacté au rouleau au Qué- (1994), Dynamic Analysis of FWD and Pavement Response
bec, Association Béton, Québec (ABQ) Québec, 2004 pp. Using a Three-Dimensional Dynamic Finite Element Pro-
63-87. gram, Nondestructive Testing of Pavements and Back cal-
culation of Modulus (2nd Volume), ASTM STP 1198,
American Society for Testing and Materials, Philadelphia.

44
View publication stats