Repository navigation
Swept wings fall 8-11% below the vortex-lattice Oswald efficiency, and a 25% swept wing does not converge #416
Description
Agreeing with the Python VSM v2.3.0 (#415) leaves swept wings well below the vortex-lattice reference in Gaunaa, Li & Pirrung 2026 (zenodo 10.5281/zenodo.18222326), and the 25 % swept wing does not converge.
Oswald efficiency error against the paper's vortex-lattice results, from its appendix wings. Inviscid, 120 sections, 10° twist about the local tangent, chord elliptic in projected span, on #415's head 242880a:
| wing | #415 | the paper's lifting-line model |
|---|---|---|
| 25 % dihedral | +2.45 % | −0.03 % |
| 25 % anhedral | −2.34 % | −0.47 % |
| 15 % sweep | −8.22 % | +5.61 % |
| 10 % sweep + 25 % anhedral | −8.02 % | −0.47 % |
| 25 % sweep | −11.08 %, solver status FAILURE | +1.06 % |
Measured by the task on #318 and reported on #400, which plots them. A kite is swept and curved, so this is the case that matters.
Not known yet:
- The appendix does not fully state the geometry, and the chord direction on the swept wings is an assumption.
- Whether Python gives the same numbers on these wings has not been run. It matches on the nine reference cases of The default solve now gives the same results as the Python VSM v2.3.0 #415, none of them these wings.
What I would do: settle the swept geometry first, then run these wings through Python to see whether the gap is shared. If it is, closing it means leaving Python's rule for the paper's, and that choice is the owner's, because #410 asks for the two codes to agree. The failure at 25 % sweep is separate and worth fixing either way.
Activity