Improving Loss Estimation for Woodframe Buildings. Volume 2: Appendices

This report documents Tasks 4.1 and 4.5 of the CUREE-Caltech Woodframe Project. It presents a theoretical and empirical methodology for creating probabilistic relationships between seismic shaking severity and physical damage and loss for buildings in general, and for woodframe buildings in particular. The methodology, called assembly-based vulnerability (ABV), is illustrated for 19 specific woodframe buildings of varying ages, sizes, configuration, quality of construction, and retrofit and redesign conditions. The study employs variations on four basic floorplans, called index buildings. These include a small house and a large house, a townhouse and an apartment building. The resulting seismic vulnerability functions give the probability distribution of repair cost as a function of instrumental ground-motion severity. These vulnerability functions are useful by themselves, and are also transformed to seismic fragility functions compatible with the HAZUS software. The methods and data employed here use well-accepted structural engineering techniques, laboratory test data and computer programs produced by Element 1 of the CUREE-Caltech Woodframe Project, other recently published research, and standard construction cost-estimating methods. While based on such well established principles, this report represents a substantially new contribution to the field of earthquake loss estimation. Its methodology is notable in that it calculates detailed structural response using nonlinear time-history structural analysis as opposed to the simplifying assumptions required by nonlinear pushover methods. It models physical damage at the level of individual building assemblies such as individual windows, segments of wall, etc., for which detailed laboratory testing is available, as opposed to two or three broad component categories that cannot be directly tested. And it explicitly models uncertainty in ground motion, structural response, component damageability, and contractor costs. Consequently, a very detailed, verifiable, probabilistic picture of physical performance and repair cost is produced, capable of informing a variety of decisions regarding seismic retrofit, code development, code enforcement, performance-based design for above-code applications, and insurance practices