Air Cleaner Shell Noise Reduction with Finite Element Shape Optimization

In this paper, finite element shape optimization is used to determine the optimum air cleaner shape and rib design for low shell noise. Shape variables are used to vary the height and location of rib elements, as well as vary the shape of the air cleaner surfaces. The optimization code evaluates each design variation and selects a search direction that will reduce surface velocity. Sound power radiation is calculated for each optimized design using an acoustic code.

Large reductions in shell noise were achieved by optimizing the shape of the air cleaner surface and rib design. Optimization of the rib pattern alone yielded a local optimization, as opposed to a global optimization that represented the best possible design.