Optimization Strategies Applied to Exhaust System Design

ArvinMeritor is using optimisation methods to shorten development times for automotive muffler systems.

One such method that has been used is optimisation via genetic algorithms. An analysis is presented of the performance tuning of the genetic algorithm including the importance of elitism, mutation rate and new blood. The genetic algorithm is shown to be well suited to application in exhaust systems optimisation, being able to consistently return performance improvements over conventionally designed mufflers.

More traditional optimisation strategies are also presented with an application to manifold design using the RSM ‘hill climbing’ model [1]. This analysis uses sound quality metrics to search for a manifold primary pipe arrangement which would yield the highest possible score against a ‘sporty sound’ benchmark [2].