Seismic Performance and Generalized Damage Risk of Modular Cross Laminated Timber Houses Across Ten United States Cities

ABSTRACT Housing shortages in the United States and the pursuit of sustainable and resilient communities have positioned cross‐laminated timber (CLT) housing as a promising alternative due to the ability to design for modularity, ease of construction, and low embodied carbon performance. However, the seismic performance of CLT houses had not been thoroughly evaluated. This study assessed the seismic risk and performance of five modular CLT house designs for a range of seismic hazards. The study investigated three research questions: (i) the conditional probability of collapse under the maximum considered earthquake (MCE) intensity, (ii) the variability of collapse fragility curves across the U.S., and (iii) the unconditional probabilities of experiencing generic non‐structural damage and collapse within 50 years for ten representative sites across the U.S. The houses were designed using platform‐constructed CLT shear walls as the lateral‐force‐resisting system, which is a code‐compliant construction method. The houses were symmetric rectilinear buildings with similar shear wall distribution in the two orthogonal directions. The analyses were conducted using a performance‐based earthquake engineering framework. Ground motion datasets were established for three seismic regions: (i) Western U.S. without pulses, (ii) Western U.S. with pulses, and (iii) Central and Eastern U.S. without pulses. The numerical models were developed in OpenSeesPy with idealized nonlinear spring models representing the wall behavior. Multiple stripe analysis was employed to calculate the collapse fragility curves and inter‐story drift distributions. Generic damage fragility curves and the seismic hazard curves for 10 sites across the United States were used to estimate the 50‐year non‐structural damage and collapse probabilities. Results indicate that ground motion characteristics across the U.S. do not significantly impact the collapse fragility curves. Additionally, the houses meet the code‐level targets for both conditional and unconditional probability of collapse at MCE intensity. Intentional over‐design had no substantial impact on reducing generic non‐structural damage probability, so it should be avoided. Instead, the shape of the site‐specific hazard curves had the largest effect on non‐structural damage probability. In summary, the consistent behavior, low collapse risk, and acceptable non‐structural damage probability of modular CLT houses make them a reliable, resilient, and high‐performing seismic housing alternative for any region of the U.S.