Sponsor: Internally Funded
PI: Trevor Hrynyk
Reinforced concrete (RC) flat plates are comprised of slabs supported by columns without the use of beams, column capitals, or drop panels. Today, their use is widespread due to their efficient load carrying capabilities and the architectural flexibility they provide. However, as a result of their minimized design, flat plate systems are prone to punching shear failure modes which occur as a result of high shear stress development in the regions of slabs immediately surrounding supporting columns.
One of two general finite element modeling approaches are typically employed to study the behavior of RC flat plates: i) models constructed explicitly with the use of three-dimensional solid elements, and ii) models constructed using some form of layered shell element which is usually developed on the basis of the plane sections assumption. Recent studies have shown that the use of three-dimensional solids can provide good estimates for the load-deformation response of slab-column subassemblies governed by shear critical failure modes; however, investigation of even a single slab-column connection with the explicit use of solids can be extremely costly, as fine meshes consisting of many degrees of freedom are usually required to obtain good results. Conversely, layered shell elements require significantly fewer degrees of freedom and, as such, are computationally more efficient. Unfortunately, typical shell element formulations either provide an extremely coarse representation of the out-of-plane (through-thickness) shear response or neglect out-of-plane shear effects entirely.
This research is focused on studying the viability and adequacy of using a combination of conventional solid and shear-capable layered shell elements for practical modeling of RC slab systems. The research consists of (1) incorporating a conventional solid element to an existing layered shell finite element program dedicated to the analysis of RC shells and plates, (2) developing an interface procedure between the solid and shell elements using multi-point constraints or, perhaps, a simple force balancing approach on the basis of the interface sectional response, and (3) validating the proposed finite element modeling procedure with experimental data available in the literature. The resulting procedure from this effort will provide a means of performing detailed analyses for full structural systems with limited computational effort and will serve as a practical tool for a broad range of design and assessment scenarios.