A comparative study of hybrid simulation and dynamical substructuring system schemes for shake table tests

In shaking table experiments, the dynamics of the shake table is strongly affected by a specimen to be tested, particularly when the specimen is very heavy. This is also the case when substructuring experiments are implemented by the shaking tables. This study numerically examines two substructuring schemes applied in shaking table experiments. Hybrid simulation scheme, which is commonly used in substructuring experiments, directly uses the output of the numerical substructure as the input to the physical substructure. Dynamical substructuring system (DSS) scheme, which requires more complicated control design, determines the control input to minimize control error obtained by outputs of numerical and physical substructures. This study has found that, in shaking experiments, the stability of the hybrid simulation scheme is very sensitive to a pure time delay. It is also found that DSS scheme can provide much higher stability margin against the pure time delay than the hybrid simulation scheme. The effort to rather complicated controller design of DSS scheme will be rewarded at actual control practices.