Proposed Real-Time Multi-Site Hybrid Testing System Applied to a Seismically Excited Building with Magneto-Rheological Fluid Damper

Abstract

The George E. Brown, Jr. Network for Earthquake Engineering Simulation (NEES), operational since October 2004, is intended to transform the nation's ability to carry out earthquake engineering research. NEES consists of fifteen shared-use equipment sites located throughout the U.S. that include a variety of state-of-the-art test equipment including large-scale hybrid test facilities, shaking tables, geotechnical centrifuges, a tsunami wave basin, as well as field equipment, and an information technology (IT) infrastructure linking these sites over high-speed internet. In 2004, the National Research Council (NRC) developed a report, at the request of the NSF, recognizing the potential of NEES to accelerate the pace of earthquake hazard mitigation and identifying research activities NEES should pursue (NRC, 2004). This report specifically identifies research activities involving tests at multiple equipment sites as a major focus of the geographically distributed NEES collaboratory. In addition to collaboration and data sharing between sites, NEES was charged to develop distributed physical and numerical simulation capabilities to conduct more complex experiments and simulation tests simultaneously at multiple equipment sites. This ability will accelerate the pace of earthquake hazard mitigation. Over the past 2 years of operation, the NEES community has focused on multi-site pseudo-dynamic testing including the Multi-site Online Simulation Test (MOST) and ensuing bench-scale mini-MOST demonstrations, and pre-NEESR zipper frames project (Leon, et al., 2004). Multi-site NEES activities are limited to pseudo-dynamic testing, that is, tests run at a much slower rate than real-time. The major challenge with distributed, or multi-site, testing is accommodating the large communication time delays present in sending data over large distances and over the internet. Local controllers and hybrid algorithms run at a rate of around 1000 Hz (e.g. the University of Colorado at Boulder NEES Fast Hybrid Test facility runs at a rate of 1024 Hz). 1000 Hz corresponds to a time step of 1 millisecond (msec). A larger time step, using the current methods and algorithms may result in less accurate simulation tests. it is generally accepted that communications over the internet would be limited to 75 Hz, over a 500 km distance. Assuming the test rate of 75 Hz is mainly a result of communications delays from both sending and receiving data, the communications time delay in one direction is on the order of 6 msec. Because of this limit in the rate of multi-site experiments, testing in real-time (which would allow for testing rate dependent physical components) can not be accomplished with the current methods and algorithms. NEES equipment sites such as the fast-hybrid testing facilities, shake tables, centrifuges, and tsunami wave basin were specifically built for fast and/or real-time testing. In order to demonstrate a potential method to achieve real-time multi-site testing, a simplified experiment is devised and implemented to simulate multi-site testing in real-time. This simplified experiment will seek to experimentally verify semiactive structural control of a seismically excited building. This paper proposes a system to conduct real-time multi-site experiments. A bench-scale experiment of a two-story building with semiactive control device located between the ground and first story is presented to illustrate this method, employing hybrid testing of a physical semiactive MR damper while simulating in real-time the seismic response of a two-story building model. This paper first proposes a system architecture that applies techniques from haptics to overcome simulation to experiment interface and delays anticipated in multi-site testing. Next, the test setup for experimental verification is described. Results from these tests are presented. Conclusions are drawn and future studies are identified.