What is it about?
We ran NASA FUN3D on PAW3’s IFCPT S-duct in three configurations (baseline, AIP rake, flow-control vanes), first on the workshop-supplied grids and then with output-based off-body mesh adaptation aimed at total pressure at the AIP. Without the rake in the geometry, recovery is overpredicted relative to the test. With the rake included, adapted FUN3D matches the measured recovery to about 0.03% (0.9781 vs 0.9784), versus a 7.5% miss on the finest workshop grid for that configuration. Adaptation also tightens the estimate of the grid-converged recovery.
Featured Image
Why is it important?
It is a clean demonstration that adjoint/output-based adaptation, plus geometry that actually includes the probes, beats just throwing more uniform points at an S-duct. For inlet CFD, the mesh and the hardware you omit are as important as the turbulence model.
Perspectives
Mike Park’s adaptation machinery in FUN3D plus Pointwise-built starting grids. This is the paper I still point to when someone says the workshop grids should have been enough.
Zach Davis
Space Dynamics Laboratory
Read the Original
This page is a summary of: IFCPT S-Duct Grid-Adapted FUN3D Computations for the Third Propulsion Aerodynamics Workshop, July 2017, American Institute of Aeronautics and Astronautics (AIAA),
DOI: 10.2514/6.2017-4835.
You can read the full text:
Contributors
The following have contributed to this page
