Development and Implementation of Effective Force Testing: A Method of Seismic Simulation

The objective of this research project is to develop and implement a test method, 'Effective Force Testing' (EFT), which enables real-time testing of large-scale structural systems subjected to earthquake loadings. This method has at least two desirable characteristics. First, it permits economical real-time dynamic testing of large-scale structural systems, which is important for structural details affected by strain rate effects. Second, it provides a means of testing the robustness and reliability of active, semi-active, and passive control systems on large-scale structures.

The EFT method provides capabilities which complement those of the other three primary types of earthquake simulation test methods: shake table, quasi-static, and pseudodynamic. Restrictions on the type of structures that can be tested are similar to those required by the pseudodynamic test method (i.e. lumped mass systems fixed to the laboratory floor). However, rather than using a displacement control algorithm, as is the case for the pseudo-dynamic test method, EFT uses a force control algorithm. The EFT method can be used to conduct real-time earthquake simulation studies because the effective forces to be applied to the limped mass structure are known a priori. They are the product of the respective story masses and the ground acceleration record. Consequently, the effective forces are independent of structural response.

A pilot study conducted on the EFT method on a linear elastic SDOF system was successful; but control problems associated with actuator/structure interaction were discovered. In this project, the investigation will be extended systems.

The project will be accomplished in two phases. The first phase consists of experimental and numerical simulation studies to investigate potential control problems expected with increased servovalve flow demands associated with taking a test structure into the nonlinear range of behavior. A source of control problems may be associated with servovalve/actuator nonlinearities not currently accounted for in the linear models used by the investigators. To detect these problems, tests will be conducted on a simple test structure designed to repeatably exhibit nonlinear behavior. The second phase of the research will consist of a proof-of-concept test on a replicate structure, which has been tested previously on an earthquake simulator to compare the results of the two earthquake simulation techniques.

An integrated interdisciplinary approach will be used throughout the project, which combines structural testing and analysis expertise from the Department of Civil Engineering faculty, and controls expertise from the Department of Aerospace and Mechanics faculty. Japanese collaborators on the project at University of Tokyo will focus on developing a real-time 'on-line' testing method, similar to pseudo-dynamic testing, including substructuring.

This project is supported under NSF 98-36, 'US-Japan Cooperative Research on Urban Earthquake Disaster Mitigation.'