EVALUATION OF GRID DEPENDENCY OF PYFR-BASED ILES ANALYSIS FOR AIRFOIL AT MARS ATMOSPHERIC CONDITION

Mars atmospheric aircraft often encounter low Reynolds number conditions ranging from 10SUP4/SUP to 10SUP5/SUP, resulting in complex flow characteristics over airfoils, such as the formation of a laminar separation bubble(LSB). Accurate prediction of the LSB is important as it directly impacts the aerodynamic performance of the aircraft. To accurately capture the LSB, a high-fidelity analysis is required. In this study, we conducted PyFR-based implicit large eddy simulation using GPU acceleration to identify the effect of various grid parameters on aerodynamic performance and flow structure. Grid tests were performed under incompressible conditions with a Reynolds number = 30,000 and angle of attack = 8°, a short LSB which showed a high sensitivity to grid resolution was observed. The grid parameter study of the near-wall grid resolution, spanwise grid resolution, spanwise domain size, and far-field boundary size were conducted. The convergence of the aerodynamic coefficients, the location and length of the LSB were compared. It was analyzed that the spanwise grid resolution had the smallest effect on the aerodynamic coefficient, and it was confirmed that a certain level or more of the near-wall grid resolution and spanwise domain size are required to properly predict the LSB.