Seismic rocking simulation of unreinforced masonry parapets and façades using the discrete element method

Abstract

When subjected to high-intensity out-of-plane seismic loading that results in the threshold value of static stability being exceeded, unreinforced masonry parapets and façades form mechanisms that exhibit rigid body rocking motion before collapse. The discrete element method was used to perform rocking simulations which were compared to experimental observations and analytical models to obtain relationships that enable the stiffness proportional Rayleigh damping parameter (β) to be estimated from the coefficient of restitution. The seismic performance of parapets and façades was compared using incremental dynamic analysis, and similar performance was observed between the compared models, confirming the validity of the developed expressions to obtain β. The gap size generated after failure between the façade and the return wall was parametrized and the outcome revealed the hazardous situation that arises when the connectivity between the uncollapsed façade and the adjacent return walls is damaged during an earthquake. Pulse-like accelerations, generally associated with near-fault earthquakes, caused collapse of parapets and façades at substantially lower accelerations when compared to earthquakes with oscillations that were evenly distributed in time. It was concluded that when assessing the seismic vulnerability of a rocking façade located near a fault, a capacity equal to that of a free-standing block should be used, validating the assumption made in some seismic assessment guidelines of not considering the effect of return walls when assessing the façade capacity. For façades located far from a fault a range of earthquakes that represent the site conditions needs to be included in any dynamic capacity assessment.

A multi‐level approach to predict the seismic response of rigid rocking structures using artificial neural networks

“…As stated, η is an independent input variable with values 0.72 ≤ η ≤ 0.99 based on experimental results. 57,58 Finally, recorded earthquake ground motions from the PEER database 59 are employed for the nonlinear dynamic analyses. Each record consists of three components.…”

“…Importantly, this work investigates the contribution of the coefficient of restitution to rocking response. As stated, η is an independent input variable with values 0.72 ≤ η ≤ 0.99 based on experimental results 57,58 . Finally, recorded earthquake ground motions from the PEER database 59 are employed for the nonlinear dynamic analyses.…”

Chattering: an overlooked peculiarity of rocking motion

“…Thus, for values of η within the range of practical interest in rocking dynamics (i.e. η > 0.75 [24]), a cr can be considered independent of ω.…”

Experimental and analytical studies of infill wall with sliding joints considering a door opening

“…In recent years, semi‐rigid steel frames have been a favorable option used for seismic applications due to their exceptional deformation capability and remarkable energy dissipation ability, while such frames have relatively low lateral stiffness and strength 1–3 . The intervention of partitioned infill walls, however, would disturb the expected seismic response of the frames because of the premature failure of the susceptible infill wall 4–7 .…”

“…In recent years, semi-rigid steel frames have been a favorable option used for seismic applications due to their exceptional deformation capability and remarkable energy dissipation ability, while such frames have relatively low lateral stiffness and strength. [1][2][3] The intervention of partitioned infill walls, however, would disturb the expected seismic response of the frames because of the premature failure of the susceptible infill wall. [4][5][6][7] Moreover, severe deformation of the inevitable door openings on the infill wall might aggravate such undesirable effects, given the absence of standardized door opening numbers and locations.…”