Experimental performance of floor mounted nonstructural components under seismic loading

Research attention to nonstructural damage has been awakened in recent years, as performance based earthquake engineering has advanced and loss estimations have shown that more than fifty percent of the damage following an earthquake is due to nonstructural components. Nonstructural components response importance was first considered over life safety concerns. The possibility of nonstructural components, including but not limited to floor mounted equipment, to shift or topple over during a seismic event has led the addition of specific code provisions for the anchorage of nonstructural components. Available methodologies for seismic design of nonstructural components anchorage are based on very simplistic equations. Using constant values for generalized sub-groups of components, together with a basic dynamic approach of spectral acceleration distribution throughout the structure height. Current design equations and variables are considered to lead to very conservative results, due to the simplicity and over-strength incorporated in anchorage design. Previous testing of nonstructural component anchorage has been performed without considering the contribution of the structural system of the component itself. In order to deepen the understanding of the seismic loading of nonstructural components, shake table tests were conducted with floor mounted components anchored to concrete and steel, using instrumented anchors. The shake table was subjected to input motions from several ground motions based on real recordings and broadband spectrum matching. The ground motion includes different excitation directions allowing the comparison of 1 vs 2 and 3 components of motion. This paper focuses on the component responses and anchor loads, comparing results to predictions using design equations and first principles and presents implications for design practice of floor mounted nonstructural components.