Uncertainty in the numerical modelling of masonry triplet tests under dynamic loading

Abstract

Masonry structures have been broadly used worldwide for centuries. These structures are likely to be subjected to seismic movements, malicious or accidental blast and/or impact loading. Expanding the current body of knowledge of how masonry structures perform under such conditions and developing reliable and robust modelling techniques is essential to improve both the efficiency and safety of the design and retrofitting of such structures. A great deal of research is currently ongoing to understand the behaviour of masonry under shock and impact loading, and this is proving to be a challenging endeavour. Masonry construction on the whole is surrounded by a high degree of variability ranging from the heterogeneity of the materials used, the degree of workmanship during construction and the uncertainty regarding the physical and mechanical properties of the brick-to-mortar interface. Masonry is known to experience a dynamic enhancement of its strength properties when subjected to impact loading and dynamic increase factors (DIFs) have been used to adjust static masonry properties accordingly when subjected to this type of loading. These DIFs are derived from sparse experimental tests and their use can be severely limited to the conditions of the tests performed, and the results obtained can carry a high degree of uncertainty. This paper considers the uncertainty present at the brick-to-mortar interface, by using Monte Carlo simulations, when subjected to dynamic loading in a standard triplet test using LS-DYNA. The results of the modelling have been compared, contrasted and discussed for use in a larger research project on the robust characterisation of masonry structures when subjected to blast and seismic loading.