No abstract

In this study, as a state of the art testing method, real-time hybrid simulation (RTHS) is implemented and verified with a shake table for education and research. As an application example, the dynamic behavior of a tuned liquid damper (TLD)-structure system is investigated. RTHS is a practical and economical experimental technique which complements the strengths of computer simulation with physical testing. It separates the test structure into two substructures where part of the structure for which a reliable analytical model is not available is tested physically (experimental substructure) and coupled together with the analytical model of the remaining structure (analytical substructure). The implementation of RTHS involves challenges in accurate control of the experimental substructure as well as the synchronization of the signals. The details of the hardware and the software developed and the steps taken to improve the controller are discussed in this paper so that the implementation of RTHS is properly introduced. The accuracy has been verified using tracking indicators as well as using the response obtained from a spring-mass oscillator and TLD system. The shake table used in this study is available in over 100 universities around the world. In this paper, the implementation of RTHS is provided with sufficient details to enable easy introduction of this testing method wherever a similar shake table is available. This additional functionality will not only provide a new research tool, but it will also facilitate classroom demonstrations to improve how students understand new concepts in structural dynamics and earthquake engineering.

Tuned liquid dampers (TLDs) are dynamic vibration absorbers used in suppressing structural vibration under wind and seismic loads. They are easy to design and implement with low cost and low maintenance. However, due to their highly nonlinear behavior, it is difficult to establish representative models for TLDs that are accurate for a wide range of operations. In this paper, a new numerical model (finite volume method/finite element method (FVM/FEM method)) is introduced by simultaneously using finite volume and finite element approaches to represent fluid and solid domains, respectively. In order to assess the accuracy of the FVM/FEM results a state of the art experimental technique, namely real-time hybrid simulation (RTHS), is used. During the RTHS the response from the TLD is obtained experimentally while the structure is modeled in a computer, thus capturing the TLD–structure interaction in real-time. By keeping the structure as the analytical model, RTHS offers a unique flexibility in that a wide range of influential parameters are investigated without modifications to the experimental setup. This is not possible in traditional shake-table dynamic tests where a physical model of the structure needs to be built and tested together with the TLD. As a result, the verification of the numerical models for TLD–structure interaction available in the literature only consider a smaller, restricted dataset. In this study three numerical models from the literature are selected and together with the FVM/FEM developed here, the accuracies of these four models are assessed in comparison with RTHS results that consider a wide range of influential parameters. Results show that the proposed FVM/FEM model can accurately predict TLD behavior in both sinusoidal and ground motion forces and Yu’s model is the most accurate among the investigated simplified models.

No abstract