This research project considers buildings and structures which are damaged or fail during a strong earthquake. The behavior of damaged or failed structures often becomes highly nonlinear and dependent upon the history of loading. A common technique that is used to describe the dynamic behavior of such structures, including the nonlinear and history-dependent characteristics, is to use a mathematical model with parameters that change as the load is applied. The parameters are estimated using some system identification technique in conjunction with the recorded response of the structure. Many identification techniques have been developed but a comparison and evaluation of the various techniques is lacking. The purpose of this project is to compare and verify the various identification techniques. As a result, one or more of the techniques will be used to assess the damage in a structure. In particular, an attempt will be made to relate the changes in the identified parameters to the damage as it accumulates in a structure. The goal is to develop a mathematical model that will assist in making decisions regarding the safety, or reliability, of a structure and the need for repair or demolition. The design of a structure for earthquake resistance includes consideration of the interaction between gravity loads and lateral displacements. This interaction, known as the P-Delta effect, refers to a tendency for unidirectional yielding of the structure, which can lead to dynamic instability and collapse of the structure. A methodology to consider P-Delta effects in the design of seismic resistant structures is lacking. The objective of this research is to develop such a methodology to consider the P-Delta effect and dynamic instability. The approach is based on the definition of the inelastic spectrum. The premise is that P-Delta effects can be included through spectrum modification, in a way that is similar to the methods used to incorporate the effects of ductility and damping.
