2008 年 56 巻 655 号 p. 363-369
The present paper reports the result of a numerical study on efficient solution-adaptive grid refinement algorithms for simulating unsteady flows with discontinuities. We implement a flow feature detection algorithm originally developed for quadrilateral grids by Aftosmis and Kroll (AIAA Paper 91-0124, 1991) to achieve efficient adaptation for triangular grids for simulating strongly unsteady flows with moving discontinuities such as shock waves and contact discontinuities. In their algorithm, the detection of strong shock waves is isolated from that of other smooth flow features, and the threshold for grid refinement is automatically tuned based on the statistics of grids in the smooth flow feature detection step. We illustrate the performance of the algorithm with numerical results for two benchmark problems (shock tube and double Mach reflection problems) of two-dimensional unsteady flows.