MODELING THREE-DIMENSIONAL TIMBER LIGHT-FRAME BUILDINGS

A nonlinear finite-element model (LightFrame3D) is presented in this paper to study the perfor- mance of 3D timber light-frame buildings under static loading conditions. The uniqueness of the model is the implementation of a mechanics-based representation of the load-deformation characteristics of individual panel- to-frame nail connections in the diaphragm system. This approach requires as input basic material properties and static load-deformation characteristics of the connections. The model can analyze a light-frame building with varied material and structural components and combined loading conditions. Either load control or dis- placement control can be used as input history. The model was verified and tested by theoretical and experimental means and good agreements were achieved. The model can provide information on the hysteresis behavior of structures under cyclic loading and the torsional response of eccentric 3D buildings. Low-rise residential houses and small industrial and com- mercial buildings in North America are conventionally light- frame structures using wood-based materials. Typically, these are composed of 2D systems (i.e., shear walls, floors, ceilings, and roofs) and are highly indeterminate. These wood structural systems are generally believed to perform well under seismic loading when carefully constructed, which could be attributed to the high strength-to-weight ratio of timber as a building material, the redundancy of the whole system, and the ductility of connections. The structural integrity of wood frame buildings under the action of natural hazards is not necessarily guaranteed, as was shown in past earthquakes and hurricanes, especially in mul- tiple-story buildings with asymmetrical geometry. For many years, a large amount of experimental and analytical work has been done to understand the structural behavior of wood- based, light-frame systems. The work, to a great extent, has been limited to the study of 2D structural components, such as shear walls, roof and floor diaphragms, and metal connect- ors and fasteners, under static monotonic or cyclic loading. Experiments on full-scale light-frame houses have seldom been done due to the high costs and test demands. The knowl- edge obtained thus far about the structural behavior of com- plete wood buildings is mainly derived from construction prac- tice and a few experimental studies on major structural components. Analytical methods were also developed by a few research- ers to predict the structural performance of an entire building. Chehab (1982) developed a linear seismic analysis of a typical wood frame house. Even though the element mesh and input properties were coarse, the results captured the effects ob- served in earthquake-damaged houses, including torsional ef- fects resulting from a nonsymmetrical arrangement of the shear walls. Gupta and Kuo (1987) developed a simple linear elastic building model containing seven ''superelements'' and nine global degrees of freedom to analyze the building tested by Tuomi and McCutcheon (1974). The superelement was based on the shear wall model from their previous work