The Earthquake Engineering Online Archive

Continuous hybrid simulation with geographically distributed substructures

Mosqueda, Gilberto; Stojadinovic, Bozidar; Mahin, Stephen A.

UCB/EERC-2005/02, Earthquake Engineering Research Center, University of California, Berkeley, 2005-11, 168 p (400 E177 2005-02)

The hybrid simulation test method is a versatile technique for evaluating the seismic performance of structures by seamlessly integrating physical and numerical simulations of substructures into a single model. This method has advanced considerably since its inception 30 years ago, moving from ramp-hold specimen loading to more realistic continuous and real-time loading histories. Further advancements are necessary to fully utilize the potential of the method. Specifically, larger and more complex structures can be tested by integrating advanced analytical modeling tools and physical models using a geographically distributed hybrid simulation testing method. A control system is presented that supports the implementation of computationally demanding hybrid simulation algorithms including: continuous algorithms, real-time algorithms, and hybrid testing with geographically distributed substructures. The controller is based on an event-driven scheme, as opposed to a real-time clock-based scheme, that supports the implementation of continuous algorithms on distributed models where network communication, numerical integration and other tasks may have random completion times. The advantage of an event-driven approach is that logic can be included to minimize, if not eliminate, the adverse effects of random completion times on the stability and accuracy of the test. This procedure is demonstrated by computing the earthquake response of a two-story shear building model with two remote physical substructures connected using the Internet. A rigorous investigation is carried out to evaluate the validity and reliability of the results from the hybrid simulations with geographically distributed substructures. The test results are compared with the results from a conventional local testing configuration and a pure numerical simulation. Additionally, a method is developed to estimate the reliability of the simulation results based on energy errors that accumulate in the experimental substructures. The proposed error indicators can be used to monitor experimental errors during a hybrid simulation and provide a measure of confidence that assesses the quality of the results. (EERC-2005-02)

Available online: http://nisee.berkeley.edu/documents/EERC/EERC-05-02.pdf (3 MB)