Numerical Simulation Of Sound Wave Propagation Problem With Complex Boundary And Moving Boundary

At present,computational acoustics is widely used in various fields,and acoustic simulation has been very developed in acoustics,especially acoustic simulation of static vocal tract(boundary).However,in the study of speech production,the vocal tract of human is more complex,especially in the continuous pronunciation process,often accompanied by real-time changes in the shape of vocal tract,so the complex and moving boundary problems which this process follow have not yet a good solution.Therefore,it is very important to explore the sound wave propagation in the process of human speech production in real and accurate way.Numerical methods of acoustic simulation have mesh-based methods and meshfree methods,mesh-based methods mainly contain boundary element method,finite element method and finite-difference time-domain method.These traditional mesh-based methods in acoustics has been a great success.In particular,finite-difference timedomain method is widely used in acoustics because of its high efficiency and accuracy.However,this method is often accompanied by mesh reconstruction and mesh distortion when dealing with complex and moving boundary problems.At the same time it is easy to cause a larger computational error,and the computational process is quite time consuming.In order to solve this problem,this paper combines finite-difference time-domain and immersed boundary method to model and then apply it to the acoustic analysis of the vocal tract.Immersed boundary method is an important method in the field of computational fluid dynamics.It is mainly used to deal with the motion of fluids under complex structure and various moving boundary problems.In the acoustic model which is consist of finite-difference time-domain and immersed boundary method,the immersed boundary method is to express the interaction between the vocal tract boundary and acoustic waves by adding force in the acoustic governing equations and then calculate them on the simple Cartesian meshes,parameters on meshes(such as pressure and velocity,etc.)are alternately calculated by the finite-difference time-domain method.In order to validate the hybrid numerical method in solving the problems of complex and moving boundary,this paper has carried on the acoustic simulation experiments of Mandarin vowels,and has validated the numerical model And the validity of the model is applied to the numerical simulation of the dynamic change of the vocal tract,and the model can be used to solve the acoustic problem with complex and moving boundary.