Generalized estimation methods of turbulent fluctuation of high-speed flow with single-pixel resolution particle image velocimetry

The methods that measure the turbulence statistics distribution of high-speed flow with high spatial resolution using particle-image-velocimetry images are proposed, and their performances are verified. Two methods are proposed, and the problems caused by blurring the particle shape due to the high-speed movements are resolved. While the conventional method approximates the unique particle shapes with a circular distribution for image pairs, the first proposed method approximates the different particle shapes with an ellipse for first and second images of pairs and reduces the effects of the blur in the flow direction of the particles. Meanwhile, the second proposed method treats the general particle shapes such as the blur of temporally changing laser intensity and adopts the deconvolution analysis using a Fourier transform. Synthetic particle images were created and a supersonic jet test was performed, and the proposed methods were evaluated to be superior to the previous method for the estimation of turbulent fluctuation using those data.