In car body development the computational evalution of mid-frequency interior acoustics is a well established tool. With a complex FE model it is possible to detect and to remedy the acoustic hot spots, and to dimension the panel sheets, beads, and damper pads. The validation with measured transfer functions shows clearly the adequateness of this approach. Due to the high numerical effort for this acoustic computation, optimization was seldom effordable. The recent development of numerical solvers like AMLS (Automated Multi-Level Substructuring) has changed this situation. First trials with classical gradient based optimization strategies were deflating due to the high non-linearity of the response surface. In alternative, using a self-adaptive evolutionary algorithm showed to be much more promising. In this contribution, the results from a multi-criteria evolutionary optimization of standard car body acoustics are presented. The acoustic level induced by engine excitation was reduced remarkably while the mass of the body was also minimized. The final obtained Pareto front illustrates the relation between the obtainable level reduction and the required additional masses.     «

In car body development the computational evalution of mid-frequency interior acoustics is a well established tool. With a complex FE model it is possible to detect and to remedy the acoustic hot spots, and to dimension the panel sheets, beads, and damper pads. The validation with measured transfer functions shows clearly the adequateness of this approach. Due to the high numerical effort for this acoustic computation, optimization was seldom effordable. The recent development of numerical solv...     »