The application of genetic algorithms to optimising the design of an engine block for low noise

A procedure to optimise finite element models of engine structures for low noise using genetic algorithms was investigated. Experiments were performed on a simple engine block model with 1800 degrees of freedom to study the effects of changing the control parameters. The procedure was then applied to the optimisation of a concept level model with 13500 degrees of freedom. Optimisation of this model yielded a reduction the predicted noise of 1.7 dB(A) for no mass increase and 2.3 dB(A) for an increase in mass of 5 per cent.