and practically feasible APJ design that mitigates stress and strain
concentrations under operating thermal and traffic loads. The
premise of the paper is that improved APJ performance can be
obtained by minimizing stress and strain demands within the APJ,
a task achieved through parametric simulations. The parametric
studies described herein involve four sets of alternative geometric
parameters as shown in Fig. 2: interface between pavement and
APJ Set IF, gap plate thickness Set GT, gap plate edge detail
Set GE, and gap plate width Set GL. Based on the information
obtained from the parametric studies, an improved APJ design is
proposed and its superiority over the traditional design is demon-
strated.
Finite-Element Model Development, Calibration,
and Validation
Two-dimensional, plane strain finite-element models developed
by Park et al. 2010 are used in the investigation of APJ re-
sponse. The typical APJ design 500100 mm of widthdepth
with 20010 mm gap plate suggested by the Bridge Joint As-
sociation 2003 is used as a prototype. Eight-node isoparametric
elements are used to mesh the concrete deck, pavements, main
APJ region, and gap plate as shown in Fig. 3. The developed
models are run on ABAQUS V6.7.Since the mesh contains corners, the stress level at those loca-
tions will rise as the element size is decreased. To ensure that
meaningful results are achieved, the mesh is graded such that the
element size in critical zones, where cracks are expected to ini-
tiate, is 1 mm and quantities of interest within each element are
averaged over the element domain. The chosen 1-mm element
size is deemed appropriate for representing the processes that lead
to direct or fatigue crack initiation. This size is similar to that
used in fracture studies conducted by Kim et al. 2009, albeit
they used a particle element model rather than a finite element
model as used herein. As noted in Park et al. 2010, the objective
of this effort is not to model crack initiation and propagation in
the APJ material itself, which involve complex phenomena asso-
ciated with time and temperature dependencies as well as healing
of asphalt. Rather, the intent is to evaluate stress and strain mea-
sures that are deemed indicative of the propensity for direct or
fatigue cracking in the main APJ body.
To present the results and facilitate comparisons between vari-
ous configurations in the future, the results of the simulations are
plotted along the path AFDBCE-ECBDFA shown in Fig. 4.
Points D and E along with corresponding points E and D
represent the tips of the debonded interface. Park et al. 2010
showed that the most critical points are located along this bound-
ary. Due to symmetry, only one-half of the model is employed in
thermal movement analysis, whereas, in case of moving traffic
load, a full model is used because the loading condition is not
symmetric. The main quantities of interest are maximum principal
stresses p and strains p and are computed 0.5 mm away
from the boundary, which coincides with the center of the small-
est elements at the critical points.
While APJ material is comprised of aggregate and pure as-
phalt, the model employed herein assumes that the material is
homogenous. As a viscous, temperature-dependent material, as-
phalt exhibits a stiffer and stronger response at high strain rate or
cold temperatures compared to response under slow strain rates or
high temperatures. Therefore, the model for the APJ material used
in this work is a time- and temperature-dependent model, which
consists of two parallel networks: an elastic-plastic network and
an elastic-viscous network. The behavior of the elastic-plastic net- 桥梁沥青塞接头英文文献和中文翻译(2):http://www.751com.cn/fanyi/lunwen_16694.html