Various Errors in Substructure Testing 2001-01-1404
A frequency domain substructure method is employed for deriving the motion transmissibility equation on a coupled structure. Various errors in its experimental implementation are investigated by numerical calculations in order to identify the most important source of errors so that future effort can be more focused. The interfaces with and without the tangential and rotational DOF's are taken into account when the DOF deficiency error is studied. By decomposing the DOF deficiency error into different components, it is shown that all six interfacial DOF's must be taken into account for general structures. The bias, noise, and underflow errors are investigated and their relative importance is compared with the DOF-deficiency error. It reveals that these three errors are moderate in nature and can be handled with traditional methods while there is no simple way to compensate for the DOF deficiency error. However since there are no mature methods to find rotational FRF from tests, the DOF deficiency error presents a real hurdle to experimentally use the substructure methods. Subsequent efforts should be targeted on how to extract the full set of interfacial frequency response functions that include both translational and rotational DOF's.