Improved Large-Eddy Simulation Validation Methodology: Application to Supersonic Inlet/Isolator Flow

A process was demonstrated that enables more meaningful comparisons of experimental and large-eddy simulation results of the same flowfield. This approach was motivated by recent studies, which show the importance of modeling the measurement device or process when making comparisons of this nature. In the current case, a large-eddy simulation of a Mach 5 inlet/isolator flow is the subject of the validation process, a flow that was previously validated using particle image velocimetry measurements. In the present study, the particle image velocimetry measurement process is emulated by adding particles with inertia to the large-eddy simulation and using synthetic particle image velocimetry to mimic the process by which velocity data are extracted. Analysis of the particle response reveals modifications to the underlying flow consistent with the original particle image velocimetry measurements, including a drastic decrease in rms velocities in the vicinity of shockwaves and weaker velocity gradients throughout the flow. The synthetic particle image velocimetry is found to capture the sampled particle velocities quite well but does impart a significant bias error to the regions near the walls, also consistent with the experimental data. Furthermore, this process suggests improvements that could be made to any new experiment to achieve higher accuracy and thus improve the validation process.