TY - GEN

T1 - From Measurement To Simulation A Review Of Aerodynamic Investigations Of Real Rough Surfaces By DNS

AU - Kurth, Sebastian

AU - Cengiz, Kenan

AU - Wein, Lars Carsten

AU - Seume, Jörg Reinhart

PY - 2022

Y1 - 2022

N2 - Improving the understanding of the effect of roughness on wallbounded flows is crucial for further improvements ofturbomachinery and wind turbines. To this end, direct numerical simulation (DNS) with the immersed boundary method (IBM) is a powerful tool, if the surface topography of the real rough surface is known. Throughout the processes from geometric measurements to simulation, many inherent uncertainties can induce misleading aerodynamic results. This study reviews the three main steps for investigating real rough surfaces: measurement, data processing, and simulation. Taking a used blade surface from an aero engine as a sample both, an optical and a tactile surface measurement technique are compared as to their weaknesses and advantages. It is shown that the measured roughness depends on the resolution of the measurement. A concept, used for extracting the surface roughness by filtering out the superimposed blade curvature after the measurement, is introduced. Mapping the processed rough surface onto the simulation mesh lowpass filters the roughness. Three DNS IBM with the same roughness but different mesh resolutions are used to determine this sensitivity of the simulation results to the resolution of the mesh. For the studied surface, even a coarse mesh grants good results

AB - Improving the understanding of the effect of roughness on wallbounded flows is crucial for further improvements ofturbomachinery and wind turbines. To this end, direct numerical simulation (DNS) with the immersed boundary method (IBM) is a powerful tool, if the surface topography of the real rough surface is known. Throughout the processes from geometric measurements to simulation, many inherent uncertainties can induce misleading aerodynamic results. This study reviews the three main steps for investigating real rough surfaces: measurement, data processing, and simulation. Taking a used blade surface from an aero engine as a sample both, an optical and a tactile surface measurement technique are compared as to their weaknesses and advantages. It is shown that the measured roughness depends on the resolution of the measurement. A concept, used for extracting the surface roughness by filtering out the superimposed blade curvature after the measurement, is introduced. Mapping the processed rough surface onto the simulation mesh lowpass filters the roughness. Three DNS IBM with the same roughness but different mesh resolutions are used to determine this sensitivity of the simulation results to the resolution of the mesh. For the studied surface, even a coarse mesh grants good results

U2 - 10.33737/gpps21-tc-213

DO - 10.33737/gpps21-tc-213

M3 - Conference contribution

BT - Proceedings of Global Power and Propulsion Society (GPPS), Xi'an, China

ER -