This study involves the investigation of the sensitivity of the torsional stiffness of a bolt-together heavy truck chassis to variations in tight/loose conditions for various combinations of individual bolts or groups of bolts. The behavior of the chassis, which features several hundred threaded fasteners as the primary means of connecting the cross members to the longitudinal beams, is studied using the Hierarchical Evolutionary Engineering Design System (HEEDS) in connection with the finite element program (ABAQUS) for performing a multiple-run analysis. A novel analysis approach involving a systematic segregation of the bolts is introduced.
First, the bolts that are used for connecting the components of the chassis are divided into various sets. Each set represents a number of individual bolts in each joint between the longitudinal beams and the various cross members. In order to identify the relative significance of each joint location to the chassis torsional stiffness, the entire set of bolts in each joint is turned “on” or “off” to simulate whether the joint is tight or all loose, respectively. The effect of various permutations of tight and loose joints is investigated. Second, individual bolts are turned “on” or “off” within the identified critical joint(s) in order to simulate whether the individual bolts are tight or loose, respectively. In order to verify the numerical trends, an experiment focusing upon a simplified chassis is conducted.
Both numerical and experimental studies are carried out. For the chassis design considered in this study, those bolts oriented parallel (as opposed to perpendicular) to the chassis length most influence the torsional stiffness. Experimental results indicate that the amount of system contact gaps (a.k.a. “play”) tends to be more sensitive to those bolts oriented parallel to the chassis length than those bolts oriented perpendicular to the chassis length.