Академический Документы
Профессиональный Документы
Культура Документы
By
TRB 2003
Abstract
This paper investigates various aspects of the structural behavior of doweled joints, including load
transfer, in rigid airport pavement systems using nonlinear threedimensional finite element methods.
The finite element models include two concrete slab segments with dowels connecting them. The
concrete slab and supporting layers are simulated by continuum solid elements to enhance the
accuracy of the simulation. Solid elements can capture the severe deformation gradients in the
concrete slab in the vicinity of wheel loads, allow the modeling of non linear behavior in the
supporting layers and the modeling of frictional contact interfaces between the concrete slab and
supporting layers. These features have not been considered in classical approaches. The structural
behavior of the doweled joint is investigated for various design and loading conditions, including: (1)
tire pressure, (2) dowel spacing, (3) slab thickness, (4) dowel looseness and (5) multiple wheel loads.
The amount of load transfer can be obtained directly from the shear force in the Timoshenko beam
elements that simulate dowel. According to the finite element results, 15 to 30% of the applied wheel
load is transferred to the adjacent slab segment by the dowels. This number varies in accord with
design and loading conditions. In addition, 95% of the transferred shear force is carried only by the
nine or eleven dowels which are closest to the applied load.
Key words
3D finite element analysis, doweled joint, airport pavement, load transfer, dowel looseness
TRB 2003
1. INTRODUCTION
A rigid airport pavement system is composed of numerous discrete concrete slabs, longitudinal and
transverse joints, and dowels. Longitudinal joints are provided for construction convenience and
transverse joints are provided to control cracks caused by thermal deformation and drying shrinkage of
the concrete slab. Despite those benefits, the joint often reduces the load carrying capacity of the
concrete slab near the edge and results in pavement damage under repeated wheel loads [1, 2]. Field
experience has demonstrated that dowel load transfer systems are among the most effective means of
increasing the load carrying capacity of rigid pavements. A dowel connects concrete slabs and
transfers wheel load across the joint primarily through shear force. For rigid airport pavements, the
importance of doweled joints is much greater than for ordinary highway pavements because the
applied load level of airport pavements is much higher than that of ordinary ones and the
consequences of interslab faulting is much greater.
The doweled joint has been employed in rigid pavements since the early twentieth century. A
great deal of research has been devoted to assessing the amount of load transfer across a doweled joint.
An intact joint is known to transfer more wheel load to adjacent concrete slabs than a damaged joint.
Unfortunately, there is no way to directly measure the shear force in a dowel with available sensor
technology. Various indirect measures have been developed to estimate the load transfer over doweled
joint. Among them, the displacementbased load transfer efficiency (LTE) has been widely used. The
LTE is defined as the ratio of displacement of the unloaded slab to that of the loaded slab at a joint.
Although LTE can easily be measured in the field with a Falling Weight Deflectometer, it does not
correlate well with actual load transfer across a joint. Rather, it gives an implication of the magnitude
of damage at the joint due to the pumping and funnelling (dowel looseness).
3
TRB 2003
This paper reports on an analytical investigation of load transfer across doweled joints under
various loading and design conditions using 3D FE(Finite Element) models. The parameters
investigated include (1) load level, (2) dowel spacing, (3) concrete slab thickness, (4) multiple wheel
loads and (5) dowel looseness. For parametric analysis, FE models were constructed with two concrete
slab segments composed of solid elements with Timoshenko beam elements to simulate the dowels.
This approach enhances the accuracy of FE solution with solid elements simulating the
concrete slab and supporting layers. Solid element can capture severe deformation gradients in the
concrete slab under multiple wheel loads, which is impossible with classical approaches using
Kirchhoff plate elements [3]. Further, solid elements used in supporting layers count on heterogeneous
material properties of each layer. Hence, this approach can provide more accurate displacement field,
which affects on the stress response of the concrete slab, than the classical approaches with Winkler
foundation [3]. We modeled the frictional contact interface between the concrete slab and supporting
layers [3]. In addition, the FE mesh density can be easily varied with regard to the stress gradient to
improve mesh efficiency. Above all, this approach can directly evaluate the amount of load transfer
across doweled joint by computing the shear force in the beam elements. Therefore, one can observe
dowel shear force distributions for each case and determine the number of engaged dowels as well.
TRB 2003
joint and gradually propagates inside the concrete slab. As a result, the dowels are often free to deform
until they touch undamaged surrounding concrete.
In approach (2), the dowels are simulated by elastic spring elements directly connected to
plate elements over the joint. Therefore, the dowel cannot resist bending. The amount of transferred
shear force is determined by
V = K
where V is the dowel shear force, K is the spring constant and D is the relative displacement between
loaded and adjacent concrete slabs.
Approach (3) uses Timoshenko beam elements to simulate the dowels, but they are indirectly
connected to plate elements by elastic springs. Approach (2) and (3) can simulate dowel looseness
through the elastic deformation of spring elements. However, the behavior of the doweled joint is
dominated more by the artificial spring constant than it is by the mechanical properties of the dowel
and concrete slab. The contact force acting between the beam (dowel) and plate elements (slab) is
determined by the artificial elastic spring constant. Further, these approaches always require
calibration of the artificial spring constants with FWD test measurements. Often, calibrated spring
constants show a wide range of variation, from 21 to 10000kPa (from 3 to 1500ksi) [8]. Hence, the
simulation of dowel looseness is not well bounded by physical observation.
Approach (4) is suitable for simulation of both intact and damaged dowel joints because it
uses continuum solid elements for both the dowel and concrete slab. It simulates their interaction
through frictional contact. The detailed stress and strain distribution within a dowel and the interaction
between the dowel and the concrete slab can be observed from this approach. Further, pavement
damage can be simulated by using plastic constitutive models for the concrete in the vicinity of the
dowel or by specifying the funnel geometry at the outset. Despite these advantages, the problem size
becomes too large to be solved on todays computational platforms. The dimensions of the dowels are
much smaller than those of the concrete slab. Hence, a much finer mesh is necessary to simulate the
dowel and concrete casing, while a coarser mesh is adequate to model the farfield behavior of the
concrete slab and other parts of pavements. An adequately refined mesh leads to a huge problem size,
especially if the model is composed of multiple concrete slabs. The large problem size either prohibits
or limits our ability to perform parametric analysis.
The objective of this paper is to understand the behavior of doweljointed rigid airport
pavement systems with both intact and loosed joint with dowel load transfer. Timoshenko beam
elements [9, 10] were selected to simulate dowels. They were directly connected to continuum solid
elements, which simulate the concrete slabs, for intact joint. The same approach was used in 2D plain
strain analysis for rigid highway pavement in the MNROAD project [11]. The entire length of the
dowel is simulated by seven beam elements so that the rotation field can be adequately resolved.
5
TRB 2003
Therefore, the load transfer action does not include artificial springs. The dowel shear force is directly
transferred to concrete slab.
The gap contact algorithm was employed for loosed joint simulation. The gap contact allows a
physical gap between concrete slab (solid elements) and dowel (beam elements). In order to simplify
the simulation, the looseness is represented by the size of a gap and is only assumed to be present on
the adjacent concrete slab - not the loaded one. This approach requires two separate models for intact
and loosed joints but it demands much smaller problem size than approach (4) because a refined 3D
solid mesh is not required to simulate the dowels.
TRB 2003
construction. The contribution of outer dowels to the global pavement behavior was expected to be
negligible because they are far from the applied wheel load. The numerical verification of this
assumption will be discussed in the following section. From experimental results in the literature, one
can also find that most of the load transfer is achieved by a few dowels near the applied load [14, 15].
A linear elastic constitutive model was used for the concrete slab because the stress under the
wheel load was expected to be far less than the strength of concrete. A linear elastic constitutive model
was also used for the subgrade layer even though they are composed of granular material. Again, the
computed stress was very small and always in compression. The Mohr-Coulomb elastoplastic
constitutive model was used for the cementtreated base layer, and a 1365kPa (198psi) cohesion limit
was used to define the yield condition, which is approximately equivalent to a 1650kPa (240psi)
tensile strength [3]. In addition, a frictional contact interface exists between the concrete slab and the
cementtreated base layer in order to simulate uplift and sliding. It allows discontinuous deformation
through the depth and attenuates unrealistic tensile stress developed by layered elastic analysis [3].
The FE parametric analysis was performed with commercial FE software ABAQUS [16].
TRB 2003
Figure 2(b) shows the relative displacement between loaded and adjacent slab segments along
the joint. The largest positive relative displacement is observed at the origin (symmetry line), where
the wheel load is applied, and a small negative relative displacement is observed at the corner. The
relative displacement distribution is almost identical to the dowel shear force distribution along the
joint shown in Figure 2(a). In addition, the negative relative displacement at the corner corresponds
with the negative shear force at last dowel. Approximately 330cm (130inch) from center or near the
second outer dowel, the sign of the relative displacement changes from positive to negative. This
means less displacement is observed on the loaded slab than on the adjacent slab after this point,
because the loaded slab always has a larger curvature than the adjacent slab. Further, maximum
displacement always occurs beneath the wheel load as does the maximum relative displacement along
the joint. In addition, the negative dowel force vanishes if the last dowel is taken out.
Figure 2(c) shows the ratio of load transfer (the amount of transferred load by the dowels
divided by the amount of applied wheel load). This ratio increased with the increase of applied tire
pressure, while the size and location of pressure load remained identical for all three cases. That means
more wheel load can be transferred to the adjacent concrete slab if wheel load pressure increases. This
phenomenon is also evident in the dowel shear force distribution. The inner seven dowels of the
2970kPa case carried almost five times more shear force than those of the 740kPa case. However, the
total amount of applied load was only four times more. Only 3.5% of the shear was carried by the
outer sixteen dowels for the 2970kPa case, while 6.2% of the shear was carried by them for the
740kPa case. The high wheel load increases the load transfer ratio. Meanwhile, it also increases the
demand on a few inner dowels beneath the wheel load, which may cause more damage to the joints
and eventually lead to pavement failure.
TRB 2003
total amount of transferred load does not change much in terms of relative location between wheel
load and dowel, if the spacing remains the same.
Figure 3(b) shows that only five or six dowels were engaged for both 61cm spacing cases,
while nine were engaged for the 30cm spacing case. This result suggests that the size of the region
containing engaged dowels does not change with dowel spacing; only the distribution of shear forces
varies. This issue will be discussed further in Section 4.3. The last column shows the summation of the
contribution for internal dowels (until dowel No. 8). For the 30cm and 61cm EVEN spacing cases, the
summation value exceeds 100% because the negative shear forces (up to 0.5%) from the outer dowels
are not included.
The last column of the table in Figure 3(c) shows normalized maximum tensile bending stress
of loaded and adjacent concrete slabs. Those of the 30cm and 61cm ODD spacing cases were almost
identical to each other, but that of 61cm EVEN spacing case was quite different. The difference
suggests that the stress response of the concrete slab is more sensitive to the relative location between
applied wheel load and dowels than the dowel spacing. The wheel load was located between two
dowels in the 61cm EVEN case and created more deformation on the loaded concrete slab segment.
As a consequence, this extra deformation caused more bending stress in the loaded slab segment. In
contrast, less deformation was observed on the adjacent slab and correspondingly less stress was
generated in the adjacent slab. Friberg anticipated such behavior in his paper, and our numerical
results support his observation [14].
TRB 2003
30cm thickness case, the summation values exceed 100% because the negative shear force
contributions (up to 1.7% for the 69cm case) from the outer dowels are not included.
The last column of the table in Figure 4(c) shows the normalized maximum tensile bending
stresses of loaded and adjacent concrete slab segments with respect to those from the 43cm thickness
case. As a consequence of the stiffness increase for a thick concrete slab, a much reduced tensile
bending stress was observed for the 56 and 69cm cases. In fact, a thick concrete slab provides two
significant benefits: higher load transfer and lower maximum tensile stress of concrete slabs.
TRB 2003
cases always occur at the edge. That means one large deformation basin develops under the tritandem
wheel loads, as if the wheel loads were applied at the interior of the concrete slab. In fact, dowels
provided partial continuity to the discrete concrete slab segments by transferring shear forces. On the
contrary, two separate deformation basins were observed under the dual and tritandem wheel load
from the undoweled pavement analysis [17]. From Figure 5(c), the magnitude of displacement
increased relatively little from the dualtandem to the tritandem cases compared to others because the
internal wheel loads are, again, carried more by the supporting layers than by dowels. In addition,
uplift was observed on the opposite side of the adjacent slab, and the magnitude was proportional to
the amount of transferred load. This uplift would be restrained by gravity loading, if an additional slab
were to exist next to the adjacent slab and if they were connected with dowels.
Figures 5(e) and 5(f) show the maximum stress and displacement results from the FE models
with and without dowels along the joint, respectively. All values have been normalized by those from
the single wheel load case. Maximum stress and displacement data for the single wheel load case are
listed at Figure 5(g). From the observation of pavement models with and without dowels, one can see
that the dowel reduces the magnitude of maximum compressive and tensile bending stress by
approximately 13% and 16%, respectively, for every wheel load case. Further, maximum
displacements were reduced by 7% in doweljointed pavement results. From the wheel load
interaction analysis of the single slab segment model, a surface tensile bending stress zone existed
between two edge wheels and the two or four internal wheels for dual and tritandem gear cases [17].
Such a zone exists because two edge wheel loads dominate the behavior of the entire structural
system. On the contrary, the surface tensile bending stress zone vanishes in a doweljointed rigid
pavement because its dominance is reduced by the dowel load transfer system. Nevertheless, global
stress and displacement contours are quite similar for both models with and without dowels, as one
can see similarity in normalized maximum stress and displacement and their locations for the two
models.
TRB 2003
deformation on loaded concrete slab than the single wheel load case. Pumping and other joint damages
were not included in this analysis.
The variation of gap size makes significant changes on the behavior of load transfer. Wider
gap reduces the amount of transferred load over the joint. Figure 6(b) and 6(c) show the result of
dowel looseness study for single wheel load case. The amount of load transfer reduced from 22% to
0%, while the gap between concrete slab and dowel reduced from 0.00254mm to 0.254mm. The
number of engaged dowels are reduced from five(same as intact joint) to zero. In the meantime,
l=
Eh 3
2
12 1 v
0.25
maximum bending stress also varies. Due to the reduced load transfer, wider gap increases the
maximum bending stress in loaded concrete slab up to 16%, while it decreases the maximum bending
stress in adjacent slab. From the observation on joint, traditional LTE varied from 95%(intact joint) to
60%(0.28mm).
Figure 6(d) demonstrates dowel shear force through the joint for tri-tandem wheel load case.
For small gap between concrete and dowel(up to 0.25mm), the third dowel from the center, which is
located under the wheel load, makes the largest load transfer contribution. This phenomena change
after gap size 0.5mm because the maximum relative displacement between loaded and adjacent slabs
occurs on the center of joint. Figure 6(f) shows the maximum relative displacement, which makes the
largest load transfer contribution, occurs beneath wheel load for gap size 0.25mm case. On contrary, it
occurs at center line for gap size 0.76mm case. At the corner of concrete slab, it is found that the
displacement of loaded slab is always smaller than the adjacent slab. Once again, it is because the
curvature of loaded slab is greater than adjacent slab. Figure 6(e) shows the amount of transferred load
and normalized maximum tensile bending stress. Transferred load reduces with the increase of gap
size between slab and dowel. After the gap size of 0.76mm, the amount of transfer load is negligible.
From the observation of maximum tensile bending stress in Figure 6(e), one can infer the maximum
bending stress of concrete slab can be magnified up to 18% due to the damaged joint.
TRB 2003
dowel from the point of loading, to zero at a distance of 1.8l. No dowels beyond that point influence
the moment at the load point. This observation implies that most of load transfer should occur within
1.8l of the loading.
Figure 7(a) illustrates the schematic deformed shape of a concrete slab along the joint under
single edge wheel load. According to the previous FE results, the relative displacement between
loaded and adjacent slabs determines the magnitude of dowel shear force. Further, the magnitude
decreases as the distance from the wheel load increases. From numerical results in Section 4, the
extent of engaged dowels varied only with the slab thickness. Tire pressure and dowel spacing did not
change the extent of engaged dowels. Spacing did change the number of engaged dowels, but they still
stayed within the same distance, if the thickness of the concrete slab was constant. Seven, nine, eleven
and thirteen engaged dowels were identified from the thicknesses 30, 43, 56, and 69cm, respectively.
From Figure 7(b), one can find that the location of the last effective dowel demonstrated a good match
with Fribergs 1.8l distance observation. Here, the radius of relative stiffness was computed based on
the subgrade reaction modulus measured by plate loading test simulation with axisymmetric FE model
[3]. The last row of this table shows the amount of load transfer by the engaged dowels, and they were
very close to 100%. Hence, one can conclude that numerical results support Fribergs observation.
8. CONCLUSIONS
This paper has investigated the load transfer and structural behavior of doweled joints with respect to
the variation of load level, dowel spacing, slab thickness, dowel looseness and landing gear
configuration. Timoshenko beam elements were used to simulate dowels. The ratio of load transfer
was predicted from 18% to 30% with respect to above parameters. In contrast, the number of engaged
dowels depended only the slab thickness. For the 43cm slab thickness, nine engaged dowels achieved
almost 99% of the entire load transfer. This behavior was independent of the variation in load level
and dowel spacing.
From the multiple wheel load analysis, the load transfer ratio decreased with an increase in
applied wheel load. The two or four internal wheel loads (from dual and tritandem landing gear) are
applied away from the joint and, therefore, make a small contribution to load transfer. From the
comparison between models with and without dowels, the dowel load transfer action reduces
maximum tensile bending stress up to 20%. Further, the dominance of the edge wheels, identified
from single slab analysis, was attenuated by the dowel load transfer mechanism. Hence, dowels indeed
contribute to better durability of rigid airport pavement systems. Dowel looseness magnifies maximum
bending stress up to 18% for the worst case. Through the FE analysis, small looseness gap between
concrete slab and dowel makes a significant change in the behavior of concrete pavement.
13
TRB 2003
9. REFERENCES
1. Huang, Y. H., Pavement analysis and design, Prentice Hall, Eaglewood Cliffs, New Jersey, 1993,
pp. 186-187.
2. Tayabji, S. D. and B. E. Colley, Improved pavement joint, Transportation Research Record 930,
TRB, National Research Council, Washington, D. C., 1983, pp. 69-78.
3. Kim, J., Threedimensional finite element analysis of multi-layered system: Comprehensive
nonlinear analysis of rigid airport pavement systems, Ph.D. Thesis, Department of Civil
Engineering, University of Illinois at Urbana-Champaign, 2000.
4. TabatabaieRaissi, A. M., Structural analysis of concrete pavement joints, Ph.D. Thesis,
Department of Civil and Environmental Engineering, University of Illinois at UrbanaChampaign,
Illinois, 1978.
5. Huang, Y. H., A Computer package for structural analysis of concrete pavements, Proceedings,
Third International Conferenceon Concrete Pavement Design and Rehabilitation, Purdue
University, West Lafayette, Indiana, 1985, pp. 295-307.
6. Guo, H., J. A. Sherwood, and M. B. Snyder, Component dowelbar model for loadtransfer
systems in pcc pavements, Journal of Transportation Engineering, ASCE, Vol. 121(3), 1995, pp.
289-298.
7. Shoukry, S. N, 3D finite element modeling for pavement analysis and design, Proceedings. The
First National Symposium on 3D Finite Element Modeling for Pavement Analysis and Design,
Charleston, West Virginia, 1998, pp. 1-92.
8. Ioannides, A. M. and G. T. Korovesis, Analysis and design of doweled slabongrade pavement
systems, Journal of Transportation Engineering, ASCE, Vol. 118, 1992, pp. 745-768.
9. Bathe, K.J., Finite element procedures, 2nd ed., PrenticeHall, Inc., New Jersey, 1996, pp. 234251.
10. Cook, R. D., D. S. Malkus, and M. E. Plesha, Concepts and applications of finite element analysis,
3rd ed., John Wiley & Sons, New York, 1989, pp. 278-280.
11. Zhang, Z., H. K. Stolarski, and D. E. Newcomb, Development and simulation software for
modelling pavement response at Mn/ROAD, Minnesota Department of Transportation, Minnesota,
1994.
12. Hjelmstad, K. D., J. Kim, and Q. H. Zuo, Finite element procedures for threedimensional
pavement analysis, Proceedings, Aircraft/Pavement Technology, ASCE, Seattle, Washington,
1997b, pp. 125-137.
13. Hjelmstad, K. D., Q. H. Zuo, and J. Kim, Elastic pavement analysis using infinite elements,
Transportation Research Record 1568, TRB, National Research Council, Washington D.C.,
1997a, pp. 72-76.
14. Friberg, B. F., Design of Dowels in Transverse Joints of Concrete Pavements, Transactions,
ASCE, Vol. 105, 1940, pp. 1076-1095.
14
TRB 2003
15. Foxworthy, P. T., Concepts for the development of a nondestructive testing and evaluation system
for rigid airfield pavements, Ph.D. Thesis, Department of Civil Engineering, University of Illinois
at UrbanaChampaign, Illinois, 1985.
16. ABAQUS Theory Manual and Users Manual, Hibbitt, Karlsson & Sorensen, Inc., Pawtucket,
Rhode Island, 1994.
17. Kim, J., K. D. Hjelmstad, and Q. H. Zuo, Threedimensional finite element study of wheel load
interaction. Proceedings, Aircraft/Pavement Technology, ASCE, Seattle, Washington, 1997,
pp.138-150.
15
TABLE 2
Thickness
(cm)
Elastic Modulus
(MPa)
Poissons
Ratio
1. Concrete slab
43
27,600
0.15
2. Cement-Treated Base
20
13,800
0.20
3. Subbase
30
345
0.35
4. Soil subgrade
760
55
0.45
22,800
145
140
1,480
34.7
55.4
175 cm
345 cm
Dowels
Concrete Slab
Plane of
Symmetry
Supporting
Layers
35.5 @ 3 = 106.5
30.5 @ 9 = 274.5
Plane of Symmetry
Timoshenko
Beam elements
40.0
Shear Force (kN)
Wheel Load
0.7
0.0
---0.5
---4.0
381
Coordinate (cm)
(a) Shear Force Distribution of Dowel at Joint
Difference (mm)
.075
Tire
Pressure
0.0
---.025
Coordinate (cm)
381
Applied Transferred
Load
Load
Ratio (%)
740
56
11.7
21.0
1480
112
26.2
23.5
2970
224
56.6
25.3
FIGURE 2 Results of Single Wheel Load Case under Load Level Variation
24.5
30 cm Spacing
61 cm Spacing ODD
61 cm Spacing EVEN
381
Coordinate (cm)
(a) Shear Force Distribution of Dowel at Joint
Dowel Number
Location (cm)
30
61
91
122
30.9 20.4
8.9
3.3
1.2
0.5
53.5
20.1
% of Shear
---2.2
Wheel Load
30 cm
61---O cm
41.0
61---E cm
2.8
8.4
0.3
0.1
100.3
99.9
0.3
1.0
Total
---0.1
100.6
Normalized Stress
Applied
Load (kN)
Transferred
Load (kN)
Ratio
(%)
Loaded
Adjacent
30
112
26.2
23.5
1.00
1.00
61---O
112
21.9
19.4
1.01
1.00
61---E
112
21.7
19.6
1.09
0.93
20.0
Shear Force (kN)
Wheel Load
43 cm Slab Thickness
56 cm Slab Thickness
69 cm Slab Thickness
381
Coordinate (cm)
(a) Shear Force Distribution of Dowel at Joint
Dowel Number
Location (cm)
30
61
91
122
% of Dowel Shear
---2.2
30 cm Slab Thickness
Total
30 cm
38.2 21.6
6.8
1.6
0.3
0.1
0.1
0.1
99.4
43 cm
30.9 20.4
8.9
3.3
1.2
0.5
0.3
0.1
100.3
56 cm
28.6 19.3
9.1
4.0
1.8
1.0
0.6
0.2
100.6
69 cm
25.0 18.2
9.6
4.8
2.4
1.5
1.1
0.4
101.0
Thickness of
slab (cm)
Normalized Stress
Applied
Load (kN)
Transferred
Load (kN)
Ratio
(%)
Loaded
Adjacent
30
112
21.9
19.6
1.40
1.35
43
112
26.2
23.5
1.00
1.00
56
112
32.8
29.4
0.72
0.81
69
112
33.8
30.2
0.55
0.63
22.2
---4.4
Tandem
Wheel Load
Symmetry Line
Adjacent
Slab
Loaded
Slab
Joint Line
Single Wheel Load
0
381
Coordinate (cm)
Wheel
Load
Applied
Load
Trans.
Load
Ratio
(%)
Single
112
26
23.5
Tandem
224
49
21.8
Dual-Tan.
447
76
17.0
Tri-Tan.
671
91
13.5
Displacement (mm)
---2.54
Coordinate (cm)
381
Displ. (mm)
0.25
---2.54
---762
Single
Tandem
Dual-Tandem
Tri-Tandem
0
Coordinate (cm)
(d) Deformed Shape along Symmetry Line
FIGURE 5 Results of Multiple Wheel Load Cases (contd)
762
1.38
1.47
1.79
3.51
1.26
1.32
1.61
2.84
1.02
1.01
1.27
1.80
1.00
1.00
1.00
1.00
1.65
1.64
1.79
3.41
1.47
1.47
1.61
2.87
1.11
1.12
1.27
1.82
1.00
1.00
1.00
1.00
Tri-Tandem
Wheel
Dual-Tandem
Wheel
Tandem
Wheel
Single
Wheel
1.40
1.46
1.86
3.46
1.27
1.30
1.66
2.89
1.01
0.99
1.28
1.81
1.00
1.00
1.00
1.00
Loaded
Adjacent
[ c] max
2588
2268
1089
[ t] max
[ t] max
2151
.2052
1800
.2002
1096
.1986
.0716
.0665
.0650
w] max
1.00
1.00
1.00
1.00
[ t] max
[ t] max
[ w] max
Location of
Maximum
Bending Stress
[ c] max
Wheel
Loads
Adjacent Slab
Loaded Slab
Gap from
Dowel
Looseness
22.2
0.0025 mm Gap
0.025 mm Gap
0.051 mm Gap
0.127 mm Gap
0.254 mm Gap
Single
Wheel Load
Intact Joint
---4.4
0
381
Coordinate (cm)
(b) Shear Force Distribution of Dowel at Joint for Single Wheel Load
Gap btw. Slab and
Dowel (mm)
Normalized Stress
Loaded
Adjacent
0.0025
112
24.6
22.0
1.00
1.00
0.025
112
16.5
14.8
1.02
0.94
0.051
112
12.6
11.3
1.05
0.88
0.127
112
4.2
3.8
1.11
0.71
0.254
112
0.0
0.0
1.16
0.58
22.2
Shear Force (kN)
Tri---tandem
Wheel Load
Intact Joint
0.0025 mm Gap
0.025 mm Gap
0.25 mm Gap
0.76 mm Gap
1.02 mm Gap
---4.4
0
381
Coordinate (cm)
(d) Shear Force Distribution of Dowel at Joint for Tri--tandem Wheel Load
Normalized Stress
Loaded
Adjacent
0.0025
671
85.7
12.8
1.00
1.00
0.025
671
70.2
10.5
1.00
0.99
0.25
671
43.4
6.6
1.05
0.90
0.76
671
12.6
1.9
1.16
0.65
1.02
671
9.7
1.4
1.18
0.49
Adjacent Slab
Loaded Slab
0
Coordinate (cm)
Gap = 0.76 mm
Displacement (mm)
Displacement (mm)
---2.54
---0.76
Gap = 0.025 mm
---2.54
381
Coordinate (cm)
381
10
Wheel Load
1.8l
Loaded Slab
(M ) MAX
Adjacent Slab
(M ) MAX
+
30
43
56
69
1.8l (cm)
118
153
185
216
Location of Last
Effective Dowel (cm)
122
152
183
213
98.2
98.5
99.0
100.2
11