A research effort sponsored by Yamaha Motors, LTD. was undertaken to investigate exhaust system and exhaust silencing concepts for two-stroke engines. A complete one-dimensional gas dynamic computer program was developed for power-tuned exhaust systems. This program provides great insight into the actual nonlinear wave behavior of both conventional and innovative power-tuned exhaust systems. In addition, the simulated outlet flow rate has been coupled with the spherical wave equation in order to predict sound levels at any distance from the engine.An overall engine model has been developed for simulation of the internal gas dynamic behavior throughout the engine. This model includes the combustion and induction aspects of the engine cycle as well as a simplified exhaust system model. Engine power and torque can be predicted. The main use of the engine model is in comparative studies of engine parameter variations.Finally, both analytical and experimental programs are described for simulation and testing of candidate mufflers and muffling techniques. Since it is desirable to avoid using absorptive material in the exhaust systems of oil burning engines, reactive mufflers were studied primarily. The simplest analytical approach consisted of using the flow output of the complete exhaust system model as an input to linearized acoustic filter models, neglecting all “back” effects of the filter. Using the simple source equation, sound levels from the acoustic filters can be predicted. The accuracy of this procedure is not good but is quite useful for comparative studies.A second approach includes the “back” effect of the muffler on the exhaust system. This is accomplished by coupling the gas dynamic equations to the muffler equations and solving simulatneously. Although accuracy is improved, the relatively expensive gas dynamic program must be run for each muffler configuration tested.One-dimensional nonlinear reactive mufflers were incorporated directly into the gas dynamic program. The results showed that acoustic circuit analysis was often optimistic in noise attenuation prediction.