Investigation of Element Size Effect on the Nonlinear Behavior of Flanged Shear Walls

Completion and development of reliable analytical models using finite element method could
help investigation and prediction of the actual structure response results. Analysis of each
model in finite element software needs meshing, in which computer results are dependent
specifically to geometry and dimensions of the elements, called "mesh dependence". Finding
a strategy for independency of the results to meshing is tangible. For the mentioned purpose
and also to investigate the role of finite elements meshing in flanged shear walls, finite
element software was used. Different meshings of shear walls (tested by Vecchio and
Palermo) were analyzed and studied. The results of analyses with different meshs showed
different ultimate strengthes and lateral displacementes. Different shapes of mesh create
various results, which are indicated in the finite element model. By increasing of the size of
mesh, the final force was increased and the lateral displacement was decreased, which
presents a rigid model. On the other hand, by decreasing of the dimension of mesh, a stiff
model was seen. So, it is a need to create well proses to analyze and evaluate the flanged
section of shear walls with finite element model. Getting suitable accuracy of finite element
model, the mentioned concrete shear wall (vecchio and Palermo) was modeled by ANSYS.
3D SOLID65 element was employed for modeling of shear wall structures. This element is
capable of cracking in tension and crushing in compression. In concrete applications, for
instance, the solid capability of the element may be used to model the concrete while the rebar
capability is available for modeling of reinforcement behavior. After calibration, optimum
forms and dimensions are recommended. As an illustration, an idea was presented, by which
flanged shear wall could be analyzed more carefully in ultimate strength and ductility. This
analysis showed that the results of squared mesh are closer to the fact. For example, this type
of meshing 6% error in ultimate strength and ductility in accordance to lab Specimen,
presented the closer responses. Furthermore, investigation on the optimum size of the mesh
showed that if the mesh has the same size of the thickness of the connecting element (Shear
Wall Web), the results will have very high accuracy. For instance, squared meshes with same
dimension of meshes equal to web thickness, rather than those with half dimension of that led
to 1% of lateral resistance, which is closer to experimental test. It is possible that web
thickness is 150 mm, thereby, we can use mesh sizes of 150mm, 75mm or 50mm. However,
in order to obtain ultimate loads accurately, the mesh size of 150mm seems reasonable.
Square meshes have four degrees of freedom. If the size of square is chosen to be the same as
the web thickness, nodal forces induced in the web would be proportionate. For thischallenge, a flanged section reinforced concrete shear wall tested was selected to confirm the
web thickness square theory. This shear wall was modeled by finite element program. The
results of analysis showed accuracy in the investigated theory. In this study, the web thikness
square theory has indicated 8% error in ultimate strength.