New approach for identifying boundary characteristics using transmissibility

A novel approach is proposed for identifying boundary properties as a response model using transmissibility. This approach differs from those proposed in previous studies dealing with frequency response functions (FRFs) for joint identification. Transmissibility includes only response data, unlike FRFs that include force measurements. The boundary properties can be estimated by comparing the characteristics of the components under the free condition and connected to boundary conditions. When analyzing the components assembled compactly in the system for setting the shaker or measuring the impact force exerted on the component correctly, the proposed method could reduce the errors caused by an incorrectly measured force. The derived equation is verified using a discrete multiple degrees of freedom system with single boundary and multiple boundary conditions and by application to a beam, which is the simplest continuous structural form to validate the feasibility of the theory. The transmissibility defined by the apparent mass matrix is used for verifying the derived equation for identifying the boundary properties in the discrete system. However, when applying the equation to practical cases, as is the purpose of this research, the transmissibility matrix should be defined using only the response data. For this purpose, the accelerance matrix is modified slightly to the response matrix using the input as a unit force. This transmissibility matrix composed of response data is used for validating the equation in a continuous system. Furthermore, the effects of measurement noise are also investigated to assess the robustness of the method for application under practical conditions. Consequently, the proposed method could show reliable results by properly extracting the boundary properties in both cases. In many practical cases, this research is expected to contribute toward identifying the boundary properties in a complex system more conveniently compared to the method using FRFs.