Numerical and experimental investigations of the seismic response of light-frame wood structures
Christovasilis, Ioannis Periklis
MetadataShow full item record
In support of the performance-based seismic design procedures for light-frame wood structures, developed within the NSF-funded NEESWood Project, a dual study with experimental and analytical components was conducted. In the context of the experimental investigation, a full-scale, two-story, light-frame wood townhouse building was tested on the twin relocatable tri-axial shake tables operating in unison, at the University at Buffalo UB-NEES site. The test structure was designed according to modern US engineered seismic design requirements (ICBO 1988) and constructed according to applicable practices in the 80's in California. Four different test phases were conducted, associated with additional components in the building configuration, as the test structure initially featured only the structural shear walls considered in the design, and progressively interior (gypsum wallboard) and exterior (stucco) wall finishes were installed. The main objectives were to benchmark the dynamic characteristics and the seismic performance of a low-rise townhouse building with realistic dimensions under various base input intensities, representative of both ordinary and near-field ground motions in southern California, and to investigate the effect of non-structural components on the seismic response of the test structure. The test structure performed well under both DE and MCE levels of shaking, satisfying the collapse prevention objective, inherent in code-compliant seismic design. Moreover, the test results demonstrated the beneficial effect of wall-finishes on improving the seismic response of the structure, increasing the stiffness and the strength of the individual shear walls. The analytical task focused on the development, implementation and validation of a novel numerical framework, suitable for nonlinear inelastic, static and dynamic two-dimensional (2D) analysis of light-frame wood structures. The 2D building model is based on a sub-structuring approach that considers each floor diaphragm as rigid body with three kinematic, and potentially dynamic, degrees-of-freedom (DOF). A sub-structure model is developed for each individual single-story wall assembly that interacts with the adjacent diaphragms and generates the resisting quasi-static internal forces. The 2D shear wall model takes explicit consideration of all sheathing-to-framing connections and offers the option to simulate: (i) deformations in the framing members, (ii) contact/separation phenomena between framing members and diaphragms, and (iii) any anchoring equipment (i.e. anchor bolts, holdown devices), typically installed in light-frame shear walls to develop a vertical load path that resists overturning moments. Corotational descriptions are used to solve for displacement fields that satisfy the equilibrium equations in the deformed configuration, accounting for geometric nonlinearity (large rotations – small deformations) and P-Δ effects. These attributes result in a nonlinear element capable of capturing the lateral response of shear walls up to their complete failure and, thus, the side-sway collapse of the structure. To validate the proposed numerical framework, a number of simulation examples are presented, based on existing experimental results from pseudo-static tests of single- and two-story full-scale shear wall specimens, as well as shake-table tests of a single-story full-scale structure. These examples demonstrate the capability of the model to accurately simulate load paths in the structure and successfully predict variations in strength, stiffness and energy dissipation properties of the lateral-load-resisting system. The complete set of numerical analyses presented in this study illustrates the versatility of the proposed sub-structure model to provide reliable response predictions for structural systems incorporating different geometric configurations, anchorage conditions and gravity loading.