Grant: Research Tools and Instruments - Category 1 (<$150,000)

GENERAL: Commercial ground vehicles have to pass durability tests before entering the market. Such tests are costly and time consuming. Accelerated durability testing consists of tests designed to quantify the life characteristics of a vehicle component under normal use conditions by testing the unit at higher stress levels to accelerate the occurrence of failures. However, it should not incur unrealistically high load that might alter failure mechanism. Accelerated tests are conducted on a Multi-Axis Simulation Table (MAST). Development of such tests involves: (1) generating accelerated loading profiles, and (2) implementing such loading accurately. RESEARCH PROGRAM: We have developed accelerated loading profiles successfully. When implementing them, we found that the output loading from MAST deviates from the desired one, which can result in changes in failure mechanism of test units. We identified the causes to be the unknown dynamics of test units and the poor robustness of the controller used by MAST. We plan to develop robust controllers, i.e., in spite of the unknown test unit dynamics, the actual loading is close to the desired one. However, most theories for robust control design are model-based, i.e., the model of the system must be available. Due to the complexities of MAST and test units, it is not feasible to model such a system. We propose to develop data-driven approaches. Our research on nonlinear dynamics via the time series analysis laid the foundation, but will be extended. A Single-Axis Simulation Table (SAST) is crucial for our fundamental research to produce the data and to test our controllers. SAST has been designed and funding is requested here for purchasing the equipment. SIGNIFICANCE: The research can significantly improve the accelerated durability testing, which empowers manufacturers to bring their vehicles to the market quickly and economically. Our research on developing data-driven approaches to nonlinear dynamics is especially effective to complex systems. With the advancement of sensor technologies, development of such approaches becomes imperative.