Fin Resolution Sensitivity Study#

Pre-T2.6 gate check: UMR fin resolution at candidate LBM grid spacings.

Physical geometry#

  • Vessel diameter (domain): 9.4 mm

  • Body radius: 0.87 mm, length: 4.1 mm

  • Fin outer radius: 1.42 mm

  • Fin length: 2.03 mm, width (circ. arc): 0.55 mm, thickness (radial overlap): 0.15 mm

  • Helix pitch: 8.0 mm (MIME-ANO-002: assumed)

Key geometry insight#

The 6 fins (2 sets x 3) each span 2.03mm axially in a 4.1mm body. Fin spacing = 0.68mm < fin length = 2.03mm, so fins overlap axially. The correct distinguishability metric is angular bands per z-cross-section (not z-bands).

The thinnest free-standing fin dimension in the code is the circumferential arc length = fin_width = 0.55mm at the body surface. The fin_thickness parameter (0.15mm) controls a radial overlap zone with the body, not a free-standing thin feature.

Resolution comparison#

Resolution

dx (mm)

Circ. arc (lu)

Radial (lu)

Angular bands

Fin nodes

Solid %

64x64x82

0.1469

1.4

2.5

3

428

1.233%

96x96x124

0.0979

2.0

5.5

3

1395

1.195%

128x128x164

0.0734

2.6

6.0

3

3469

1.227%

192x192x246

0.0490

4.1

10.0

3

11265

1.236%

Minimum viable resolution: 128^3 (dx = 0.0734 mm, circ. arc = 2.6 lu, radial = 6.0 lu)

Parameter interpretation note#

The code uses fin_width (0.55mm) as circumferential arc and fin_thickness (0.15mm) as a radial overlap with the body. The paper’s “fin thickness” likely refers to the blade thickness (circumferential), which would make the fin thinner than the code implements. If the interpretation is revised so that the circumferential extent = 0.15mm instead of 0.55mm, the resolution requirements increase significantly (0.15mm / dx at 128^3 = 2.0 lu — marginal). This should be resolved before T2.6.