How reliable is the method of comparative sources?

Radiation and scattering from vibrating bodies can be expressed by a source system that is placed inside the radiator and whose source strengths are determined appropriately. The method of comparative sources (MCS) provides an estimate of these source strengths from the near-field data. The MCS is reviewed by applying it to cylinders with elliptical cross section. This geometry is suitable to check the method for two reasons. First, it is general enough to demonstrate methodical difficulties which arise from its spatially varying field impedances as a result of different surface curvatures. On the other hand, it is not too complicated to allow analytical treatment by Mathieu functions. Therefore, it is possible to generate exact solutions of the sound field from different, more and less efficiently radiating velocity distributions. These solutions are compared to results obtained by the MCS, in which various source systems and different weighting functions are used. This way, it is demonstrated that the relative error in the surface velocity approximation generally is not proportional to the error in the radiated sound power. Moreover, it is shown that certain weighting functions may be advantageous and that it is more difficult to predict inefficient than efficient radiators by this method. a)Now at Daimler-Benz Group, ATG Research and Technology.