Seismic Collapse Potential of Reinforced Masonry Buildings

Professor Benson Shing

Buildings designed according to current codes in the US are expected to have a low probability of collapse in an extreme seismic event. In specific, ASCE 7 targets a collapse probability of not greater than 10% in a 2,500-year event. To develop effective design specifications to achieve this goal, reliable analytical tools are essential for assessing the collapse potential of a building design. Simulation or prediction of collapse is especially challenging for shear wall structures. Depending on the reinforcing details, the aspect ratios of wall components, and the interaction of various structural elements in the system, the behavior of a reinforced masonry or concrete wall structure can vary from very brittle to ductile with vastly different failure mechanisms. In an on-going research, SE graduate student, Andreas Koutras, and Prof. P. Benson Shing have developed refined 3-D finite element models to capture the inelastic seismic response of reinforced masonry buildings through collapse in detail. The models account for geometric as well as material nonlinearities, including the cracking and crushing of masonry, the possible buckling and fracture of reinforcing bars, the bond slip and the dowel action of reinforcing bars, as well as the possible inelastic action of horizontal diaphragms and their connections with walls. This entails the development and implementation of new material models in LS-DYNA. In a parallel effort, graduate student, Jianyu Cheng, is developing simplified models that are computationally more efficient for the assessment of collapse potential of reinforced masonry buildings using Incremental Dynamic Analyses. The simplified models are calibrated with results of detailed finite element analyses. With funding from the NSF NHERI program, several single-story reinforced masonry wall systems will be tested to collapse on the outdoor shaking table at the Englekirk Structural Engineering Research Center to verify the computational models.