A VERSATILE HIGH PERFORMANCE TESTING FACILITY TOWARDS REAL-TIME DYNAMIC HYBRID TESTING
A. Reinhorn,M. Bruneau,A. Whittaker,M. Constantinou,S. Thevanayagam,S. Chu,M. Pitman,K. Winter
Abstract:The University at Buffalo (UB) node of the George E. Brown Jr. Network for Earthquake Engineering Simulation (NEES), funded by the National Science Foundation (NSF), consists of two moveable six-degree-of-freedom shake tables, capable of accommodating specimens up to 40 m (120 feet) long and weighing 100 tons (220,000 lbs), high-performance, high-capacity dynamic actuators networked to model-based data fusion processing and control systems, and a multimillion dollar laboratory expansion, to include an expanded strong floor, reactions walls, and a trench for the moveable shake tables. This equipment is fully integrated in order to achieve a Real-Time Dynamic Hybrid Testing (RTDHT) system. INTRODUCTION RESEARCH VISION Testing of very large full-scale specimens and models is currently possible in a few earthquake engineering laboratories worldwide. However, full-scale laboratory seismic testing of entire civil engineering structures (e.g., cable-stayed bridges, multistory office buildings, industrial facilities, and pipeline distribution systems) is not likely in the near future due to the prohibitive costs that would be associated with such testing. Not only would the materials, labor, and time associated with full-scale testing exceed available research resources, but the testing of extremely large size specimens and entire structures might even be counter-productive, making it difficult to study localized or specific problems within the 1 Professor, 2 Postdoctoral Associate, 3 Lab Manager, 4 Project Manager, Dept. of Civil, Structural and Environmental Engineering, University at Buffalo (SUNY), Buffalo, NY 14260. Ph: (716) 645-2114 x2400 SIMULATED STRUCTURE FULL OR NEAR FULL SCALE TESTED SUBSTRUCTURE SHAKING TABLES (100 ton) REACTION WALL INTERFACE FORCES ACTIVE FEEDBACK FROM SIMULATED STRUCTURE APPLIED BY ACTUATORS AGAINST REACTION WALL Fig.1. Real-Time Hybrid Seismic Testing System (Substructure Dynamic Testing) complex system. Powerful and damaging earthquakes frequently provide full-scale testing of real structures, in uncontrolled experiments of sorts, but adequately instrumenting such structures to generate the data necessary for research is also prohibitive, particularly given the unknown and often long time intervals between large earthquakes at any given site. It is believed that the best approach to experimentally generate the data needed for the development of reliable and accurate models of behavior is to compliment testing of large-scale models with innovative testing methods, that make it possible to conduct complementary tests simultaneously, and seek to supplement such experiments with real-time interactive computational analyses for better understanding of whole systems. The intent is to dynamically test large structures, or substructures using shaking tables, while simultaneously applying actively controlled dynamic forces at the boundary of the specimens -forces that simulate in real time the behavior and interactions of the rest of the structure (Figure 1). Conceptually, this allows a researcher to focus on specific problems in the most realistic conditions using emerging computational power in tandem with control systems. Such procedures and set-ups significantly extend the testing capabilities by integrating large-size physical components into virtual complete systems of unlimited size and configuration. Experimental capabilities must be sufficient to work at the scale necessary to ensure that credible results can be generated, but need not be oversized. Furthermore, to ensure that new problems and needs can be addressed, and to prevent premature obsolescence of an experimental facility, a high degree of experimental flexibility is required. Finally, to tackle increasingly complex problems and accelerate the time from research to implementation, the experimental infrastructure must be highly integrated with computational and model-based simulations (or even complementary experimental work), not only locally, but also at remote sites to allow broad inclusive cross-disciplinary and multi-disciplinary teams of researchers to collectively contribute toward the same goal. The motivation underlying the development of the above Real-Time Dynamic Hybrid Testing (RTDHT) was not to build an experimental facility that would have the largest shake-tables or other associated equipment, but rather to build the most versatile large-scale earthquake engineering facility. This versatility is achieved by combining state-of-the-art experimental equipment, on-line experimental control methods, and the expertise of earthquake engineering researchers at University at Buffalo (UB). This will result in the development and implementation of: (i) new experimental techniques and approaches in earthquake engineering; (ii) new earthquake-resistant design concepts and systems; (iii) analytical and computational methods supported by experimental data, and (iv) network-based collaborative research activities and sharing of data. NEXT-GENERATION EXPERIMENTAL CAPABILITIES AND INSTRUMENTATION Key components of the effort to implement the aforementioned capabilities include: 1. Expansion of the Structural Engineering and Earthquake Simulation Laboratory (SEESL) at UB in order to accommodate the new NEES equipment (at a cost of $9,000,000, funded by the State of New York), 2. Installation of dual shake tables, dynamic actuators, and a high-capacity hydraulic system, 3. Development of high-performance structural control systems, 4. Development of networked tele-experimentation capabilities using modular and expandable teleobservation and teleoperation equipment. Fig. 2: Schematic of New Shake Table in Trench These items combine to make possible a RTDHT facility that is modular and highly flexible, and that can test multiple configurations of full-scale components simultaneously with integrated real-time numerical simulations to investigate the seismic behavior of large structural systems in ways not currently possible. The expansion of SEESL, housed in Ketter Hall on the Amherst Campus of UB, will include a greatly enlarged strong floor area, large reaction wall, and a trench for the moveable shake tables. A tele-participation room, equipped with high-resolution digital video and Internet2 connections is also planned as part of the laboratory expansion. This will permit broader sharing of experimental information in a setting proper to real-time observation and interaction. As part of the equipment upgrade for SEESL, two new 6-DOF shake tables will be purchased. Both will possess the following characteristics: maximum horizontal table accelerations of 1.15g, velocities of 1.25 m/sec, and strokes of ±0.15 m (i.e. standard limits used in experimentation, and usually above the real life probable maxima, except for stroke in some instances). In addition, each table has a 50 metric tons maximum capacity, and a 20 metric tons nominal capacity at which the maximum dynamic performance can be achieved. The tables will be moveable, and located in a common trench (Figure 2). As a result, the two tables can be located directly next to each other, or anywhere up to a maximum distance of 33m (100 ft) from each other (center-to-center), accommodating test specimens up to 40m (120 ft) in length. The re-positioning of the tables within the trench can be accomplished in less than 3 days. The tables can be operated fully in-phase, or in any other way to provide correlated or fully independent multiple support excitations. Additional dynamic actuators can also be added within the trench to provide additional points of static or dynamic excitations either vertically or horizontally, depending on the needs of specific experiments. The new tables will be capable of operating at up to 100 Hz, making it possible to investigate unresolved issues relating to the seismic performance of large, stiff, non-structural types of equipment. The experience of UB researchers indicates that such performance is needed in order to generate the knowledge required for seismic qualification of equipment and systems according to emerging standard protocols. Associated controllers, an additional hydraulic power supply and distribution system, and accessories necessary to operate this equipment, are also included. Another key element of the SEESL upgrade is the acquisition of three high-performance dynamic actuators (having ±1000 kN capacity, ±500mm stroke, 1.0 m/s max. velocity, and 800gpm servovalves), and two static ±2000 kN ±500mm stroke actuators. A flexible controller system is included with software to conduct in addition to quasi-static step-by-step testing also pseudo-dynamic testing. Source code is included such that more complex structural configurations, higher levels of sub-structuring, and other more advanced formulations of pseudo-dynamic testing can be developed and implemented. A digital control system also provide the fully flexible platform needed to develop new approaches in structural testing using real-time control, such as the RTDHT system or the effective force control technique (EFCT). High-performance hydraulic power supply and distribution system necessary to operate this equipment are also included. Local Network for Testing and Analysis: The networked tele-experimentation system will use modular and expandable teleobservation and teleoperation equipment. As illustrated by Figure 3, the UB node of the NEES network is structured around an integrated system for data collection from experiments, and computational model adaptation that serves as the center and gateway for tele-observation, -processing, and -control, and interfaces with the advanced experimental equipment described in the previous appendix. The system includes a basic numerical computational center (knowledge accumulator) with parallel processing capabilities for real-time data processing and control. It receives data from an experiment through advanced sensors, namely digital vid