Modeling of Joints and Effect on Crash Simulation

Abstract This paper reports the findings of a study that was conducted to evaluate the effect of joints and their proper modeling technique in the crash analysis. Joints such as universal joints, lower control arm connections, etc. are generally not characterized in details in the current CAE models used for crash simulation. In the past, assumption was that crash is a short duration event and therefore accurate modeling of the joint kinematics is relatively less important for crash simulation. However, review of many recent crash tests have shown that some of the critical joints in the vehicle may undergo severe joint motion during crash affecting the crash modes and the responses in a significant way. Especially, in some offset crash tests, tire was found to rotate inward and caused significant toe-board intrusion. Improper modeling of the mechanical joints sometimes can lead to inaccurate prediction and jeopardize the design verification process. To study the effect of joints on crash responses, detailed models for the major joint types in the vehicle were developed. Four types of joints — spherical, revolute, cylindrical and universal joints, have been identified and investigated since they are used extensively in the design of vehicle’s chassis. In particular, attention was focused on accurately simulating the kinematics of the joints under various loading and impact conditions. Component modeling of the four joints was investigated first to ensure the proper motion of the different joints individually. A number of analyses were conducted with the models for different impact conditions. The joint models, thus verified, were incorporated into sub-system models to verify the responses of multiple components and joints. Finally full vehicle simulations, with and without these joints, were performed and compared with physical test to prove out the joint modeling methodologies. Both car and truck chassis were included in the full system study to examine the robustness of the models. The results showed that the inclusion of the joints in the CAE models can affect the simulated responses significantly.