External noise of motorcycles is composed of engine noise, tire noise, muffler noise, etc. The biggest noise source among them is muffler noise, and improvement of muffler noise reduction technology is urgently needed. At present, muffler noise is measured using a microphone when driving an actual vehicle on a test course or chassis dynamometer. Therefore, the design of the muffler requires a lot of man-hours and cost. In addition, the effect of the reduction measures on each noise source cannot be clearly understood, because the noise from the whole vehicle is measured in this test using an actual vehicle. In order to efficiently evaluate the effects of noise reduction measures, it is effective to use simulation technology. However, at present, it is limited to determine the optimal design for each part of a vehicle. From this background, it is necessary to improve exhaust noise prediction technology using numerical analysis. However, exhaust noise prediction using 3D simulation is not realistic because of expensive software and high computational load, although it is useful for the detailed investigation of noise source from small areas. On the other hand, 1D simulation software on the market is inexpensive, but it is also not useful for muffler manufacturing companies, because models that deal with complex phenomena such as internal combustion engines are charged and calculations cannot be made easily by complex model structure. Therefore, in this study, sound pressure level from a single-cylinder small motorcycle before and after the muffler was predicted from exhaust pressure calculated by 1D simulation of engine exhaust flow using OpenWAM, an open source gas dynamics code for internal combustion engines. In addition, the validity of these calculations was investigated by comparing the results of these calculations with the test results of an actual vehicle measured using the accelerated noise test method in accordance with the UN Regulation No. 41. These results showed that the 1D simulation using OpenWAM could reproduce the test results of a real vehicle and it was confirmed that this method could predict exhaust noise easily at low cost.

Mr. Akira Takemoto, Osaka Sangyo University, JAPAN Dr. Eng. Hiroyuki Houzu, National Traffic Safety and Environment Laboratory, JAPAN Prof. Dr. Daisuke Kawano, Osaka Sangyo University, JAPAN