Probabilistic Analysis for Resolving Fatigue Failures of the Connecting Rod Oil Hole
This paper describes an application of ANSYS Probabilistic Design System (PDS) as a superior method to efficiently identify the relative influences of random input variables on fatigue stress and to optimize the chosen random input variable to achieve the desired reliability. The axial oil-hole of a connecting rod in a reciprocating air conditioning compressor acts as a stress raiser and can lead to early fatigue failures. A parametric 3-D finite element model of the connecting rod with a hollow wrist pin was developed to study the fatigue stress. Contact elements were placed between the rod and the pin. The service stresses were simulated by pushing the pin against the connecting rod. A high predicted fatigue stress at an observed failure origin verified the model. A probabilistic analysis was carried out with three independent random input variables: the wrist pin bearing ID, the wrist pin OD and the hollow wrist pin bore ID. A macro file was created by a computer program to relocate selected nodes, by which the dimensions defined by the random input variables were automatically adjusted during each analysis loop. The maximum fatigue stress at the failure location was the random output parameter. This random output parameter was explicitly related to the random input variables using the Response Surface Method. The sensitivity analysis revealed that the wrist pin hollow ID had the most effect on the fatigue stress. Additional PDS runs were performed to optimize the pin ID and tolerance based on the estimated probabilities of failure using the Monte Carlo simulated cumulative distributions of the maximum fatigue stress. The analysis showed that the fatigue failures could be eliminated by reducing the wrist pin hollow ID, which was also the most cost effective fix.