| description abstract | Roughness caused by deposition, erosion, and additive manufacturing can significantly affect gas turbine efficiency. Previous research has often examined film cooling performance under limited roughness configurations, resulting in inconclusive findings. In this study, film cooling performances under various upstream roughness conditions were investigated to simulate the roughness-affected film cooling performance of the suction side and the endwall. Three roughness heights (k/D = 0.1, 0.2, and 0.4) and shapes (ks/k = 0.17, 0.67, and 1.95) were selected to cover a wide range of surface roughness characteristics. Three blowing ratios were examined (M = 0.5, 1.0, and 1.5). The weakly roughened surfaces (ks /k = 0.17) showed improved cooling effectiveness as k/D increased. Meanwhile, the moderately and severely roughened surfaces (ks /k = 0.67 and 1.95) showed a decrease in cooling effectiveness with increasing k/D at M = 0.5 and 1.0 but an increase at M = 1.5. Cases with shallower and higher roughness elements at M = 1.5 outperformed the smooth plate. Subsequently, a similarity hypothesis for film cooling effectiveness was proposed. At all blowing ratios, the scaled cooling effectiveness profiles converged around the smooth plate results for ks /D < 0.391, encompassing common turbine roughness scales, including irregularly roughened surfaces. Deviations emerged at ks /D = 0.782, and they were correlated with the deteriorated regions observed at various blowing ratios. Ensemble-averaged scaled cooling effectiveness exponentially grew with increasing roughness scale for all blowing ratios, and an empirical expression based on the smooth plate result and roughness scale was proposed (R2 > 0.97). Finally, the experimental results and fitting correlations of cooling effectiveness were compared. The results demonstrated that the proposed similarity hypothesis potentially facilitated the fast prediction of the roughness-affected film cooling performance. | |
