CFD
Not air. Not Fluent. Not TESPy. Not the 20k foil spin. Not a green on 2. Not a green on 6.
Nozzle Euler Giles
| Item | Status |
|---|---|
| C | 201 vs 208 |
| α | 70.7 vs 75 |
| h0 | 1158 held |
| ρ | on 22.19 |
| Residual | short of −8 |
Architecture held — no new vane family.
Rotor rotating-frame Euler 80k
| Item | Status |
|---|---|
| α | 72.5 |
| dump P | 14.14 vs 13.89 |
| C_out | 210 vs 120 |
| Δh0 | 38 vs 61 |
| rms | −1.40 stalled |
| Metal | same 13 / 0.40 / Ø48 |
Mixing-plane LAST GOOD (cold start, outer 94)
| Item | Status |
|---|---|
| Outer_Iter | 94, 14:24 CEST, SIGTERM freeze |
| nozzle TE C | 181.5 vs 208 |
| α | 69.4 vs 75 |
| rotor exit C | 266.4 vs 120 |
| P | 14.02 vs 13.89 |
| Δh0 | +56.5 vs Euler 40.4 kJ/kg |
| rms[Rho] | −2.17 / −2.03 vs −8 |
Supersedes the first-200 card (OUTER 200, 20:46 CEST: nozzle C 200.7 / α 70.6, rotor C 279.7 / P 14.06, Δh0 +51.5, rms −3.33 / −2.90) — history, not the current freeze.
Continue after that flipped Work to −7 kJ/kg (rotor adding h0) — killed. The −7 flipped field is not the current result and is not published as current.
Viscous SST
| Item | Status |
|---|---|
| SST | NaN at iter 0 |
| ks | 5 μm unsigned |
Solver vs design (item 6 — card, not a green)
| Question | Card |
|---|---|
| Duty (CFD export) | Euler Δh0 40.4 kJ/kg (aero/MEANLINE.md 40.42), dump P 13.89, α2 75°, C2 208. 75.9 is dead (old η_ts×101 totals); 61 was static h2−h3 mis-labeled Δh0. |
| Freeze vs duty | Frozen outer 94: Δh0 +56.5 vs Euler 40.4, not 61. Extra is 2D exit swirl: the midspan mesh dumps at r ≈ 10 mm with C_out 266; Japikse exducer is axial (Cθ3 = 0). Swirl cap |C3| ≤ 120 gives Δh0 ≤ 49.5; the rest is the dirty residual (rms ~−2). Solver/mesh miss, not a recut. |
| 19 vs 13 pitch | CENTRIPETAL pitch scaling is on both rows. SU2 recomputes n_blades 19.869 (nozzle metal 19) / 14.3165 (rotor metal 13) — the 2D wrap (atan pitch), not a missing scale factor, not a blade-count recut. AXIAL kind gave n_blades ~159 — do not go back. |
| Viscous SST first invalid | nozzle/su2_viscous_r1.log (same r2): NaN before Inner_Iter 0, no rms row — wall-coefficient (y+) did not converge, and ks = 5 μm is unsigned; SU2 8.5 forbids WALL_ROUGHNESS with STANDARD_WALL_FUNCTION. Solver card, not a TE recut. η_tt 0.802 stays unlabeled. |
Keep stage/last_good/. Do not iterate it — the blind continue flipped Work to −7 (killed history, not current). No new SST launch. 61 stays history: the old mis-labeled static drop, not the freeze comparator.
Metal locked
| Item | Status |
|---|---|
| stator | 19 / TE 0.35 frozen |
| rotor | 13 / finish TE 0.40 / tip Ø48 |
No recut. SU2 is not run from this site. No VTK or restart files are hosted.
360° gallery — fine 80 (unsigned, chip 6 OPEN)
| Item | Status |
|---|---|
| Mesh | NPOIN 61352 / NELEM 264860 — gallery_fine.su2, do not overwrite coarse gallery.su2 |
| Run | ITER 80 fresh start; rms[Rho] −3.84 (need −8); rms[RhoE] +2.22; no NaN; finished 03:35 CEST |
| mass conservation | FAIL — Σ −2.03 kg/s; |Σ ṁ|/|inlet| 10.31 |
| ṁ split (unsigned) | 0.016–1.98× vs uniform across the 19 throats — do not export ṁ |
| 4×Ø5 dump | 5.6× inlet (1.105 vs 0.197) — this iterate’s Ø5 is the dump; the holes stay OPEN, not a 25 MPa plug. Hardware OPEN (20k cold). 25 MPa plugs later |
| Pt / α vs 75° | Pt is the 22.19 MPa Riemann outlet, not a gallery loss. α not extracted — T0 ~641 K vs 923 K BC, not a kinematics source |
| Chip 6 | OPEN — do not continue OUTER on this 80-iter field; not last-good |
Coarse NPOIN 16150 + continue 400 (rms[Rho] −3.32→−3.82, energy still +) — killed. Hardwar stopped that field. Not last-good, not to iterate, not a green. Same voice as the mixing-plane −7 continue.
Iter-40 first pass stays history: split 0.19–1.82×, ΣØ5 0.85 vs inlet 0.19 — that was the open-hole dump. No recut 13 / 0.40 / Ø48 / 19 / TE 0.35 / 4×Ø5.