Defect identification in electro-mechanical assembly using finite element method and modal assurance criterion

Abstract

Globally automotive industry is going through a critical transition from gasoline vehicle to electric and hybrid electric vehicle. These changes have challenged industry with completely different vehicle driveline architecture. These systems have more electro-mechanical components which need to be critically designed for the vibration, thermal, structural and shock related requirements. Design for vibration endurance has been a major requirement and robust matured design of these systems may take several design iterations. Finite element models give us freedom to simulate and analyze design performance before actual prototype building. Simulation results can be predicted more accurately if model fidelity is strong enough. To have strong model fidelity, modal test data correlation is an important step towards robust design. In this paper, study on Test-FE model correlation has been performed for an electro-mechanical assembly using Modal Assurance Criterion. During the entire correlation study, it was assumed that the DUT was undamaged and entire focus of the work was towards improving FE model fidelity. But obtained results pointed the inconsistency between the FE model and test results at a specific region which later has been captured as a defect in the DUT. The results of this paper can help in identifying anomalies in actual product samples and in taking decision on testing other samples.