Validity of models for frictional heating in sliding isolators

This paper investigates the validity of theories for the prediction of the temperature increase at the sliding interfaces of sliding isolators. The prevailing theory considers the isolator to be indefinite in depth and the heat flux generated at the sliding interface(s) to be infinite in extent. In reality, the depth is finite and small, and bounded by an insulating boundary, whereas the heat flux is bound by the circular area of contact. New theories are presented in which the restrictions of the past theory are relaxed. The resulting solutions have been implemented in program OpenSees for single and for triple friction pendulum isolators but are very complex and time-consuming to be used in response history analysis with many isolators. Analysis of the response of sliding isolated structures with these models demonstrates that consideration of the finite depth of the isolators and of the limited extent of the heat flux does not have any important effects on the calculated isolator displacement demands in either short-duration or long-duration ground motions. It is shown that the finite depth has some effect on the peak temperature at the sliding interface and, more importantly, on the duration of the high temperature in long-duration ground motions. This is of some importance in the use of these theories for the development of testing specifications for isolators with unusually thin end plates.