Local Modal Parameter Identification Based On The Least Complex Squares Exponent Method And Applications

In the automotive and other fields of engineering, the modal analysis is the main method to identify the dynamic mechanical structure inherent properties, and as an important measure of modal analysis, the experimental modal analysis is widely used in the practical engineering application. However, in the process of test, it often appears that the added mass of test equipment such as sensors and cables lead to the situation that the measured FRFs is inconsistent. To solve this problem, a new local solve method for modal parameters estimation based on the least square complex exponent method(LSCE) is presented in this paper. Firstly, taking rectangular thin plate as the research object, this paper designs two experiments, applies the method in identifying the FRFs respectively which is consistent and inconsistent caused by added mass of sensors and gives comparison analysis of the results. It shows that the method is not only applicable to the FRFs which is consistent, but also can accurately identify modal model when the FRFs is inconsistent. Moreover, taking automobile brake disc as an example, this study applies this method to a multiple root modal of complex structures and discusses its applicability. Finally, it realizes the algorithm successively by MATLAB and the use of parameterized control makes the algorithm more practical.In the case of the rectangular thin plate structure with no multiple roots, this paper firstly uses the global solve and local solve based on the LSCE to identify its modal model respectively. Contrast with the test results by global solve, the natural frequency acquired by local solve has an error under 0.9% with standard tests and the modal shapes are consistent with the theoretical ones. The conclusion can be drawn that the local solve method based on LSCE can accurately identify the modal model for the inconsistent FRFs. Then focusing on the automobile brake disc structure with multiple roots, this study discusses the method’s correctness in multiple roots structure. The experimental results show that this method cannot identify the multiple roots of structures. But compared with the global solve, the natural frequency identified in every peak of frequency by the local solve has a minimal error that satisfied the engineering applications such as natural frequency on-line detection system. Lastly, using the MATLAB software to program this algorithm, and compared with the existing commercial software recognition results, it verifies the correctness of the independent program and lays the foundations that the method can be applied to the natural frequency on-line detection system and improves its accuracy, stability and scope of application.The test results and practical applications that are given in this paper all show that the local solve method based on the LSCE can quickly and effectively eliminate the impact of inconsistent FRFs caused by added mass. Moreover, the natural frequency identification results can satisfy the needs of the natural frequency on-line detection system and the results of the algorithm programmed by MATLAB are accurate, reliable that can be applied to engineering practice.