Local Structure Analysis In Turbulence And Vortex Detection

The statistical analysis of turbulent fluctuation signals can be divided into two categories, large amount sample analysis and local time-frequency analysis respectively. With the tool of wavelet analysis, the author did some time-frequency analyses on the basement of an experimental flow interacted between boundary-layer and wake. The conclusions are as follows, the vortexes formed by the wake's falling off mechanism are more than those by the boundary-layer in quantity, stronger in power and compacter in structure; the place of the wake's strongest shear produce well vortexes both in quantity and quality, but bust structures can be found in the boundary-layer occasionally; vortexes in the inverse energy cascade region are weak and incompact due to lack of vortex forming mechanism, and the linear scaling does not exist at small spatial scale because of strong fluctuation in the small scale. Meanwhile, the flow approach to locally isotropic and uniform turbulence if the flow condition is simple, and vortexes here are compacter than other place, vice versa.The wavelet theory is introduced in brief, then several applications in turbulence are referred including how strong intermittency makes the scaling exponent of velocity structure deviates from the K41 theory's linear scaling law when order-number is relatively large. The coherent structure is defined, we consider it's a large structure with a certain powerful fluctuation; then, a new method for coherent structure detection is put forward, i.e. the minimum of energy ratio criterion, which is more objective and meaningful than the previous methods.The last chapter introduces a new method of vortex detection, which is based on the streamline and streamline- cluster's topological characteristics, then several examples are used to verify the effectiveness of this method.