Variability in the predicted maximum acceleration of non-structural components anchored to concrete buildings subjected to earthquake demands

Abstract

Non-structural components (NSC) are an important part of reinforced concrete (RC) buildings. Their damage constitutes a large portion of the losses encountered by these structures after moderate-severe earthquakes, as well as after more frequent ones. Most, if not all, of the current standards for the design of acceleration-sensitive NSC make use of floor acceleration response spectra, thereby neglecting the influence of the anchorage hysteresis behaviour on their response. Only a few studies have included such an effect on the numerical evaluation of the maximum acceleration reached by NSC anchored to RC buildings. These studies showed that, for single-anchor connections, if the shear hysteresis of the anchorage is included, the maximum acceleration of the NSC and the maximum force of the anchor itself can significantly exceed the design demands computed with design codes. Moreover, the latest study on this topic showed that a very large degree of variability is encountered in the results, depending on the earthquake record used, and of the type of anchor considered. This paper focuses on such a variability – threated as epistemic uncertainty - in the predicted numerical results, showing how, for achieving a given probability of exceedance threshold, the design values might be much larger than those anticipated by the codes or even than the average numerical results. It is shown that if the tolerance gap of the anchorage is reduced or filled in, such a variability/uncertainty decreases considerably.