A New Approach for Prediction of Crankshaft Stiffness and Stress Concentration Factors

This paper introduces a new approach based on a statistical investigation and finite element analysis (FEA) methodology to predict the crankshaft torsional stiffness and stress concentration factors (SCF) due to torsion and bending which can be used as inputs for simplified crankshaft multibody models and durability calculations. In this way the reduction of the development time and effort of passenger car crankshafts in the pre-layout phase is intended with a least possible accuracy sacrifice.

With the designated methodology a better approximation to reality is reached for crank torsional stiffness and SCF due to torsion and bending compared with the empirical approaches, since the prediction does not depend on the component tests with limited numbers of specimen, as in empirical equations, but on various FE calculations. The predicted values deviate mostly in ±11% range from FEA values of real cranks for torsion stiffness and SCF due to torsion, and maximum ±20% for SCF due to bending. Moreover the quality values of the multiple regression models are in 99% range, which leads to the conclusion that the models are stable and the regression fitting is very good. Therefore the equations generated with this approach can be further used for the prediction of crank torsional stiffness and SCF due to torsion and bending. An additional aim of the study is to comprehend and reveal the share of each influencing parameter on the investigated results. As a result of a Pareto analysis it is found that the crank pin outer diameter retains the highest effect on the torsional stiffness, the fillet radius on the SCF due to torsion and web length on the SCF due to bending. These outcomes are valid for the ranges that are defined for this study which should cover the passenger car inline crankshafts.