DEVELOPMENTS IN COMPUTATIONAL DYNAMICS TO ENABLE LARGE-SCALE REAL-TIME HYBRID SIMULATION FOR ADVANCING PERFORMANCE-BASED EARTHQUAKE ENGINEERING

Abstract eng:
Hybrid simulation is a method where a structural system is discretized into two domains, namely, an analytical substructure and an experimental substructure. The two domains are linked through their common degrees of freedom of the system, where the former substructure is modeled numerically (typically using the finite element method) and the later substructure modeled physically by constructing a test specimen in the laboratory. The necessity to physically model a part of the structural system stems from the lack of a computational dynamic model that can accurately model that portion of the system. By including both substructures the complete structural system is accounted for in the simulation. Hybrid simulation is performed in real-time (and referred to herein as real-time hybrid simulation (RTHS)) to account for load-rate dependency of the system and thereby obtain accurate and reliable results under dynamic loading. RTHS is a new technique that had gained considerable interest by researchers for experimentally evaluating the performance of structural systems under dynamic loading, particularly that involving earthquake loading conditions. The advancement of RTHS will enable the experimental evaluation of new forms of structural systems to be assessed, including innovative structural designs that increase the resiliency of modern society. To achieve accurate results in a real-time hybrid simulation the integration scheme used to integrate the equations of motion must be robust, stable, and of higher order since the time step is limited by the speed of the servo-hydraulic control system used in the simulations (which typically limit the time step to be no smaller than 1/1024 sec.). This lecture will present recent developments in explicit integration algorithms that enable RTHS of nonlinear complex structural systems subject to earthquake loading to be achieved. The development and implementation of explicit unconditionally stable algorithms possessing controlled numerical damping are discussed. The interaction of the algorithms and effects of experimental error and noise in restoring forces, nonlinear states in the substructures, and servo-hydraulic actuator delay compensation is presented and means to minimize these effects to arrive at an accurate simulation are discussed. Examples of RTHS of steel and reinforced concrete structures subject to strong ground motions from earthquakes are provided to illustrate the success of the algorithms in advancing RTHS.

• Export as: BibTeX | MARC | MARCXML | DC | EndNote | NLM | RefWorks
• Add to your basket