COLLAPSE FRAGILITY OF A 5-STOREY CLT STRUCTURE UNDER CHILEAN SUBDUCTION EARTHQUAKE RECORDS

Abstract

Mass timber structures have been used in North America, Europe, and Oceania, and it is currently evaluated in Latin America for mid-rise buildings in order to reduce the housing deficit and the contribution of the construction industry to greenhouse gas emissions. In highly seismic-prone areas, such as Chile, it is essential to provide resilient timber structures where structural and non-structural components adequately protect life safety and reduce earthquake-induced damage and repair costs. In this context, the Performance-Based Earthquake Engineering (PBEE) framework becomes a convenient alternative for the design of more efficient timber buildings and assessing the risk associated with seismic hazards, potential damage, and economical losses, supporting risk and sustainability-informed decision-making for designers. In this paper, the PBEE framework will be employed to assess the probability of collapse of a government-subsidized 5-story cross-laminated timber building designed under the Chilean seismic design code. However, as CLT structures are not considered in the current design code, the proposed value of the seismic response modification factor (i.e. R = 2.0) in the draft of the CLT national design standard is used. A probabilistic seismic hazard analysis (PSHA) was needed for hazard site characterization considering the subduction zone as the main source for the analysis. A database of ground motion records (SIBER RISK) was filtered and adjusted to the Conditional Mean Spectrum (CMS) for the selection of ground motion records compatible with the characteristic event resulting from the disaggregation. The modeling process of the building involved two variants: (i) a representative 2D planar elevation of the structure, and (ii) a full 3D representation of the building structure. As platform CLT structures exhibit nonlinearity in their wall-to-wall parallel connection and overturning restraint system (i.e. holdowns), the hysteretic response of those connections was modeled explicitly by calibrating the SAWS constitutive from test data through an error minimization procedure. The collapse probability analysis determined a probability of less than 0,1% at the spectral acceleration of the predominant period (at 1% in the 50-year probability level) for a building with the mentioned characteristics making it suitable for construction in Chile. Further research is needed to achieve loss estimation under the PBEE framework, such as quantifying damage fragility curves of representative engineering details for Chilean construction.