抄録
A compression method using discrete wavelet transform (DWT) is developed for CFD result data. DWT, quantization, and entropy coding are used for the compression. The DWT divides three-dimensional flow field data into eight areas; one is the approximation part and the others are the detail part of its original data. This compression method is applied to large-scale flow field data, obtained by the Building-Cube Method (BCM). The BCM is a block-structured Cartesian mesh solver, and the computational domain is composed of assemblage of various size of building blocks. In the quantization process, quantization bit rate is variable according to the approximation/detail part set by the DWT and the block size set by the BCM. The computational result are successfully compressed, keeping the quality of the data.
