FE-analysis of bolted joints: Correlation between simulation and physical test data

Abstract

A correlation problem was found in a bolted joint when results obtained by physical test were compared to simulation results of a mudguard bracket mounted on a side skirt. The simulation predicts failure in another bolt than the bolt that fails in the physical tests. New rig tests have been performed together with new simulations to find the reason of why the correlation problem is present. The investigation implies that the existing method of evaluating bolted joints capture the overall behavior of the bolts in a good way. However, the new results indicate that the slip of the bolts affects the results to a big extent. The clamping forces together with the friction do play a big part of why the simulation results predicts another outcome of which bolt that fails, when compared to the physical test. The slip that occurs at some of the bolts during the testing causes wear of the clamped components and self-loosening of the bolts which causes great loss of clamping forces. The loss of clamping force is not accounted for in the existing simulation method to the same extent, high axial stresses and bending stresses are therefore obtained in reality which causes failure in the bolts. The slip that is present in the simulation for other of the bolts, do also affect the correlation problem in ways where high bending stresses are found in the simulations compared to the physical tests. The clamping force and the friction coefficient are the two main factors of when slip is initiated, which is difficult to capture in the simulations without measurements. When these two parameters are tuned according to measurements, improved correlation is obtained.