Turbulence phenomenon can be said to be the most common and most complex phenomenon in nature. Because he is very much in space and time, the current computer capabilities and numerical methods are directly viable NS equations (ie DNS) through value to the NS equation (ie DNS) to obtain accurate values. Solution is impossible, currently studying this phenomenon is mainly based on Renault Average Equation (RANS) and a large vortex simulation (LES) method, the scale can be simulated depends on the grid scale used, and the smaller scale is required to help various The SGS (Sub Grid Scale, the Joge Scale) model is approximated. It was originally used for turbulence research. Although the famous name, I didn't actually use it, :-) Later, due to the development of the computer, other finite differences, finite volume, and finite element methods were used to study, especially The development of parallel technology makes some simple turbulent phenomena such as the direct numerical simulation of the same-sex turbulence, although these simplified turbulent phenomena is far from true flow, but it has enormited awareness. effect. When I started to get into the turbulence, I feel that I am mysterious, I don't want to go, it is messy, unlike the laminar flow. And those turbulence mode is so complicated, huh, huh. However, it is still very interesting to gradually go. Think about it, the simplest cylinder is around, when the arrow speed is relatively small, there is a symmetrical vortex, now gradually increase the speed, the card eddy session; then these falling vortices are irregular, Finally, turn into turbulence. It is enough to study for a while on this seemingly simple issue. The above is a bit of personal, and the supplementation is correct.
For turbulent processing, in the engineering calculation of CFD, it is mainly used to use the Renault time average equation, in which a non-stable control equation is averaged. In this process, unknown amounts such as time mean including pulsation product are introduced, resulting in an unacceptable of the equation. Therefore, all turbulence model problems are for the purpose of closing the equations, during the time flow, the primary processing has no change, and the second item will appear additional items containing the pulsation value. For this additional process, BoussinesQ proposed Boussi to assume that there is a simplification of the problem, which is a commonly used turbulent viscous coefficient method. In the next time, the wave of Sydney can assume, thereby appearing zero square, one The turbulence model, such as equation, second equation, etc., although these turbulence model equations are unconnected in physical sense, they can better handle problems in engineering applications in engineering applications. personal opinion for reference only
