作者: Donald P. Rizzetta , Miguel R. Visbal
DOI: 10.2514/1.25916
关键词: Optics 、 Domain decomposition methods 、 Finite difference method 、 Turbulence 、 Mathematics 、 Mechanics 、 Reynolds number 、 Compact finite difference 、 Compressible flow 、 Boundary layer thickness 、 Direct numerical simulation
摘要: Direct numerical simulation was used to describe the subsonic flow past an array of distributed cylindrical roughness elements mounted on a flat plate. Solutions were obtained for element heights corresponding roughness-based Reynolds number (Re k ) both 202 and 334. The method sixth-order-accurate centered compact finite difference scheme represent spatial derivatives, which in conjunction with tenth-order low-pass Pade-type nondispersive filter operator maintain stability. An implicit approximately factored time-marching algorithm employed, Newton-like subiterations applied achieve second-order temporal accuracy. Calculations carried out massively parallel computing platform, using domain decomposition distribute subzones individual processors. A high-order overset grid approach preserved accuracy mesh system elements. Features flowfields are described, results computations compared experimentally measured velocity components time-mean flowfield, available only Re = 202. Flow about is characterized by two weak corotating horseshoe vortices. For 334, unstable shear layer emanating from top generated nonlinear unsteady disturbances sufficient amplitude produce explosive bypass transition downstream array. case displayed exponential growth turbulence energy streamwise direction, may eventually result transition.