A Statistical Approach Towards Connecting Rod Optimization

This paper discusses a structured and scientific use of statistical method, Design of Experiments (DoE), in conjunction with classical optimization techniques for the optimization of connecting rod weight. Using the response surface methodology, a transfer function was obtained to find the parametric relationship between the contributing factors and the response (weight). This transfer function was the objective function for minimization optimization problem. A multi-variable constrained optimization code was written in Matlab to find the minimum weight subject to the constraints of minimum fatigue factors of safety at critical locations and under buckling load. Sequential quadratic programming (SQP) algorithm was used for this constrained optimization. It was concluded that considerable weight saving (10.2 %) can be obtained using this approach and this method can be an efficient and cost effective way of optimization of connecting rod.