Civil Engineering Reference
In-Depth Information
between the stiffener and the girder top flange, loosening of the bolts con-
necting the cross bracing to the stiffener, and supplementing a stiffener plate
opposite to the original stiffener side. The study has shown that the connec-
tion plate addition and loosening of bolts alternatives were effective in
reducing induced strains and stresses in the web gap region. An inverse rela-
tionship between web gap height and induced strains and stresses with the
shortest web gap height resulting in the highest strains due to increased
bending by diaphragm forces in the web was also shown. The authors
concluded that expressions developed to relate vertical stresses and relative
out-of-plane displacements combined with measurements of out-of-plane
displacements by transducers can be utilized for the prediction of induced
stresses at other critical web gap regions of the bridge and at critical locations
in the web gaps of similar bridges. The authors used the solid modeling
option available in ANSYS [6.10] for node and element generation due
to the complexity in geometry and details of the structural members of
the bridge. The ANSYS shell element (SHELL63) was adopted for model-
ing. SHELL63 has four nodes, each with six degrees of freedom, and is capa-
ble of modeling bending and membrane behavior. The bridge investigated
consisted of four spans with the web gap located near the central pier. Traffic
loads acting on the two outer spans, that is, remote from the web gap, had
less critical effects on the differential deflection between the exterior and
adjacent girders and hence on the out-of-plane distortion of the web gap
near the middle pier. Therefore, only the middle two spans were included
in the coarse finite element model. The weld connecting the girder flanges
to the web was modeled with shell elements that have variable thickness. To
connect the nodes corresponding to the plate elements of the bridge deck
and girders, rigid link elements were used. This was defined in ANSYS using
constraint equations with the nodes along the flange and the deck labeled as
master and slave nodes. The ends of the two spans representing the cut sec-
tions near piers opposite to the central pier were modeled by imposing fixed
boundary conditions at these locations. The support provided by the central
pier was modeled as a roller support that restrained the displacement in the
direction perpendicular to the plane of the deck. The web gap under inves-
tigation was located near the central pier. The size of the elements in the web
gap region of the coarse model did not coincide with the spacing of the strain
gages that was utilized in the field test and would not allow direct compar-
ison between the analytic and field test strain results. Hence, a submodel near
the vicinity of the web gap region was built to ensure better accuracy in cap-
turing the local distortion behavior. The submodel of the web gap region
Search WWH ::
Custom Search