Study On Flow Regeneration Noise Of Internal Combustion Engine Muffler

Flow noise regeneration is a major factor effect on muffler attenuation performance of internal combustion engine, the quality of attenuation performance has direct relationship with noise size on the work of whole vehicle, and it relates to environment pollution and energy consumption. When flow velocity is high enough, the attenuation value of muffler even can become negative value, noise will increase, and muffler can turn into noise amplifier. With the high speed development of internal combustion engine, flow regeneration noise may be most important. So researching on flow noise regeneration of muffler is performed, occurring mechanism of the flow noise will be discovered, and quantity description and prediction of the flow regenerated noise will be realized. These have an important realistic meaning for optimizing internal flow field of muffler and controlling the flow noise and improving attenuation performance.Based on two-microphone transfer function method and muffler transmission loss with no flow, a method of measuring flow noise regeneration of muffler is suggested, and a special test equipment is built, and the procedure of measuring incident sound power within the tube is written. Correctness and anti-interference capabilities of the procedure are certificated by theory examples, estimation of the method are analyzed. Results indicate that the factors including the flow velocity and circumstance temperature and deviation of microphone spacing and the complex reflection ratio have less effect on the procedure. Comparison of the finite element method (FEM) and the boundary element method (BEM) is performed to calculate muffler transmission loss with no flow. It shows that results of FEM and BEM are the same, and they can agree with experiment results well.Using orthogonal experimental structure parameters of muffler work piece are determined. Results of orthogonal test chart are analyzed adopting extreme difference method. Relationship models on total sound power of flow regeneration noise and structure parameter and flow velocity are established utilizing dimensional analysis. Accuracy of the model estimation is researched, and confidence intervals are given, all the established models are highly significant under 0.01 levels. Regularity of the basic element of muffler effect on flow noise regeneration is discussed using the models. Results show that flow regenerated noise on the perforated tube muffler element changes strongly with the flow velocity at the range of studied velocity. When flow velocity reaches about 40 m/s, it becomes the strongest, and it becomes weak with the increasing of flow velocity, but flow regeneration noise on muffler elements of simple expansion chamber and inserted tube and perforated plate changes calmly.The spectrum characteristics of flow noise regeneration for the muffler element are discussed using numerical method of Computational Fluid Dynamics (CFD) and acoustic mode, the characteristic parameter of flow field distribution is provided, and relation model of flow regenerated noise model with it is established. Flow regeneration flow of muffler elements is analyzed using the spectrum of static pressure pulsation in the flow field, sound field does not need directly be solved, and foundation is provided for spectrum characteristic of sequence wide frequency band of flow regenerated noise. Static pressure pulsation in the internal flow structure is motivation source, when the position of main sound pressure of acoustic mode is the same as the position of flow regenerated noise, sound mode of the system of the muffler element is provoked, and flow regenerated noise is resonated, so flow regeneration noise becomes stronger. When structure parameters are constant, velocity core distribution characteristic and the side wall impact flow of inner chamber has independent of them, only numerical values are different, and there is a critical velocity on the side wall impact flow. The characteristic parameter of flow field distribution is found, this testifies the ability transforming from flow mechanical energy to sound energy, it has a concordance trend with the change of the regeneration noise. The relation model on flow regeneration noise and the characteristic parameter and inlet velocity is obtained using experimental results on the different velocity, the relationship between the total sound power and the characteristic parameter is cubic polynomial, and it is quartic power with inlet flow velocity.The total sound power level of flow noise regeneration from basic muffler element is predicted according to the total sound power model and BP neural network model and the relation model on flow regeneration noise and the characteristic parameter, and verification result is performed. It is shown that the three methods are all feasible and all precision is higher. The prediction method of the neural network is better convenience, only structure parameters and flow velocity are input in the trained forecast model, predicted result of the power level on flow regeneration noise is obtained easily, a new idea is provided for predicting flow regeneration noise of muffler. With the comparison of the former two prediction methods, the relation model perdiction on flow regeneration noise and the characteristic parameter is based on inner CFD of muffler, it does not need too many experiment data, so its scope of application may be more broadcast.