Dynamic Seismic Response of Nonlinear Displacement Dependent Devices versus Testing Required by Codes: Experimental Case Studies

Passive energy dissipation systems are one of the most resilient solutions to mitigate the seismic risk of structures. In case of strong motions, they can confine the eventual damages into easily replaceable anti-seismic devices. The performance characteristics of nonlinear displacement dependent devices (NLD) shall be defined by the force-displacement cyclic behavior, as well as the expected number of cycles related to both the duration of the earthquake and to the fundamental frequency of the structural systems. The aims of this paper are the comparison between the dynamic results of two different experimental campaigns performed on NLDs included in dissipative bracing systems and the assessment of the reliability of quasi-static testing procedures proposed by current seismic codes for type tests and factory production control tests. The number of cycles under the design earthquake of hysteretic dampers were experimentally evaluated through shaking table testing. Two experimental case studies of a two-story steel frame and of a three-story post-tensioned timber frame both with bracing systems including flexural steel dampers, hysteretic dampers (HDs), and U-shaped flexural plates (UFPs) respectively, were analyzed. Controlled-displacement tests of NLDs were performed considering quasi-static loading procedures specified by codes. Shaking table tests were carried out considering almost the same seismic sequence composed by a set of seven natural earthquakes at increasing peak ground acceleration (PGA) levels. More than one hundred inelastic cycles were experimentally recorded from dynamic tests before the failure of devices in both cases. In line with American standards testing requirements, the number of cycles at the design PGA level, estimated from shaking table tests and from non-linear dynamic analyses, shows a decreasing trend with the increase of ductility demand.