Cyclic inelastic behavior of concrete filled sandwich panel walls subjected to in-plane flexure

PhD Student: Yasser Alzeni

Publication Year: 2015

Advisor: Michel Bruneau

Abstract: Research was conducted on concrete filled steel sandwich panel walls (CFSSP-Walls) in order to investigate the ductility and seismic performance of this structural system under in-plane flexure. The research focused onto walls which have an aspect ratio (height to cross section depth), h/W, over 2, such that the ductile behavior of the wall is induced by developing the plastic moment capacity of the cross section, M p . Accordingly, an experimental program was executed at the SEESL lab at University at Buffalo, followed by analytical investigation that involved plastic analysis, fiber analysis, and finite element analysis. The experimental and analytical results were used to develop seismic design recommendations for the CFSSP-Wall.

Previous studies on composite shear walls focused on walls developing the ductile behavior through shear yielding, with relatively little available work on walls developing flexural yielding, as reflected in existing design recommendations in AISC 341-10.

In the experimental program, 4 CFSSP-Walls specimens were tested under quasi-static cyclic loading (following the ATC-24 protocols) and had height to width ratio of 2.5 and 2.72 respectively. These walls were divided over two groups, namely: Group NB and B. Group NB consisted of walls with no boundary elements, where the specimens end flanges consisted of half round HSS in order to avoid stress concentration at the corner welds.

Group B specimens had full round HSS at their ends, serving as boundary elements. and again the two tested walls were different in their S/t ratio, which was taken equal to 25.6 in specimen CFSSP-B1 and 38.4 in specimen CFSSP-B2. Other differences from group B include the fact that specimen CFSSP-B2 tie bars were fillet welded to the web skin plates (instead of plug welded), and that fiber concrete was used in that specimen in an attempt to reduce the tension cracks in the concrete at ultimate behavior.

The four tested specimens were able to attain/exceed the expected plastic moment capacity and were able to sustain their load capacity up to a drift exceeding 3%, which emphasized the ductile behavior of this CFSSP-Walls.

The finite element method using ABAQUS 6.10EF2 explicit solver was then used to establish models simulating the behavior of the tested walls, and the calibrated models were further used to carry a limited parametric study investigating design parameters that were not covered in the experimental program. This included cases with S/t ratio of 50, D/t equal to 0.076 (E/Fy ), and finally the use of thicker concrete (taken equal to 0.35b instead of 0.2b, where b is the length of the steel web of the CFSSP-Wall.

Finally, the experimental and analytical results were used to develop design recommendations for CFSSP-B and CFSSP-NB walls, including expressions to limit the S/t and D/t ratios, and an expression to calculate the required diameter of the tie bars (based on the plastic deformation of the steel skin plate during local buckling). Simple plastic theory was used to derive expressions to calculate the plastic moment capacity of CFSSP-NB and CFSSP-B walls.