Boundary element analysis of mufflers with an improved method for deriving the four-pole parameters

When a muffler is modelled by the boundary element method (BEM), the transmission loss (TL) can be evaluated by either the conventional four-pole method or the recently developed three-point method. The three-point method produces only the transmission loss (TL), and nothing else. On the other hand, the four-pole method has the advantage of retaining the transfer matrix of the muffler, which contains important parameters when the muffler is connected to another muffler or other components in the exhaust system. However, the major drawback of the conventional four-pole method is that it requires two separate boundary element runs due to the two different boundary conditions imposed on the outlet boundary. Therefore, it can take twice as long to get the TL when compared to the more efficient three-point method. In this paper, an improved method to derive the four-pole parameters for use in the BEM is introduced. Although two boundary element runs are still needed at each frequency, the improved method only solves the boundary element matrix once at each frequency. Therefore, it is as efficient as the three-point method. More importantly, the improved method also produces the four-pole parameters. The boundary element analysis is done by the direct mixed-body BEM. Numerical predictions are compared to experimental results for all test cases, including one with a mean flow effect. (C) 1998 Academic Press.