Realistic shear capacity assessment of infill frames: Comparison of two numerical procedures

Abstract

Masonry infilled reinforced concrete frame structures (MIRCFS) feature in many seismically prone areas of the world regardless of socio-economic status. In several regions of the world rapid economic growth prompts increased flow of population to urban areas and concrete frame buildings represent a very short term cost effective means of providing large quantity of high density affordable housing with comparably modest land use. The seismic vulnerability of MIRCFS has been noted in numerous seismic events since the 1967 Adapazari earthquake. Prior to the implementation of the second generation of seismic codes of practice most MIRCFS were simply designed to resist only gravity loads. It is this considerably large proportion of building stock, together with a relatively large number of buildings with inadequate reinforcement detailing, which currently represent a high risk in regions of medium and severe seismic action. By using two different numerical modelling techniques, and comparing results with experimental evidence, this paper sets out to identify the range of the various parameters that influence masonry infill–concrete frame interaction. The aim is to provide a more realistic model for the lateral load redistribution between infill and frame in the post cracking regime, and to quantify relative values of stiffness, capacity and ductility which lead to brittle shear failure of the system. In the following sections, after a brief review of previous relevant work carried out in the two numerical environments chosen, the methodology used to model shear behaviour in both the commercial ALGOR v19.3 analysis package and the open source DRAIN3DX program is first set forth. Numerical analyses are then carried out and results shown and validated against experimental results for a single bay, double bay and double storey frame configurations. Where shear failure is determined, a strategy for shear force redistribution is proposed and implemented to show the ultimate behaviour of the MIRCF. Results are compared and discussed with experimental evidence. In terms of peak capacity results are within 7% of experimental values, and post peak behaviour is matched accordingly. Similar progressions of shear failure were demonstrated in both modelling approaches. Shear failure of the concrete members correspond well to experimental evidence both in terms of location and magnitude of applied load. Failures verify observations recounted in reconnaissance reports though the frame geometry has been confined to two dimensions. Three-dimensional modelling would have the advantage of incorporating omni-directional loading however this would increase the run time and make the models too heavy for commercial use. Two dimensions allows for quick identification of meaningful results whilst still retaining accuracy as can be seen from the correlation between physical and analytical modelling.