Damage detection based on stiffness identification in time domain for poly‐linear hysteresis using sparse modeling

Abstract

The physical parameter identification of the dynamic mechanical model, such as stiffness identification, provides some valuable information for detecting post‐earthquake damage to building structures. During an earthquake, tracking changes in stiffness on the skeleton curve can be effective. It would also be helpful as an approximate evaluation if the method could be applied to structures with arbitrary poly‐linear hysteresis characteristics. However, previous research has not identified any poly‐linear hysteresis characteristics in a unified formulation. This paper proposes a new stiffness identification method in the time domain for building structures with any poly‐linear hysteresis characteristics using sparse modeling. The validity of the proposed method was investigated through numerical simulations and a full‐scale shaking table test.