A Simple Strategy for Dynamic Substructuring and its Application to Soil-Foundation-Structure Interaction
MetadataShow full item record
Dynamic substructuring (DS), also known as hybrid simulation, is an experimental method used to study the dynamic behavior of complex engineering systems. Models are created that consists of two parts actively interacting during the experiment, (i) a physical subsystem - an experimental component representing a portion of a system and (ii) a virtual subsystem - a computer model the remainder of the system. Since only the key components need to be physically constructed, this form of simulation results in reduced costs and more effective use of the laboratory equipment. Furthermore, using this approach, the properties of the virtual subsystem can be easily varied, providing for a means of readily performed parametric studies in the laboratory. During DS experiments, the interface conditions between the two subsystems are imposed using actuators and the response of the physical subsystem is measured using sensors and fed back to the computer model. Accurate control design is required to achieve a stable feedback system and ensure that the effect of the virtual subsystem is represented accurately during the experiment. DS has been extensively studied in recent years in the earthquake engineering field. The approaches conventionally used involve development of a tracking controller and delay compensators to account for actuator dynamics. Due to the complicated control design required in these approaches, DS has been limited to relatively modest configurations. In the current study, DS is approached using a strategy conceptually different from the conventional methods, aimed at challenging configurations such as the Soil-Foundation- Structure-Interaction (SFSI). The new strategy is characterized by simplicity of control design decoupled from the physical subsystem, and results in robustly stable and accurate testing. Experiments were performed using a specially designed versatile testbed to demonstrate the effectiveness of this strategy. Moreover, preliminary, proof of concept, experiments of the SFSI DS were performed, showing the feasibility of such experiments.