[taillanm]

Hello!
I am modeling a case of a sphere-cylinder body (like a bullet) (2D axissymmetric). My ultimate goal is to make studies about transition.

My case: Mach of 1.5, coupled flow, free stream inlet, pressure outlet. I have no problems about the physics (as I refined manually along the shock with volumetric controls and ran the case, no problems... very time consuming to make manual mesh controls for the shock)
I followed step by step the methodology to adapt (refine) my mesh ("automatically") near the shockwave in front of my body, from this tutorial: https://steve.cd-adapco.com/articles...pturing-shocks

But I didnt succeed..

My mesh settings: polygonal + prism layers (20, at 1e-6 first cell, 3mm thick) I am working with absolute values for mesh sizes. Default: Target surface size: 0.01m Minimum surface size: 1e-5m Surface growth rate: 1.2 I have one volumetric control to refine around my body (5e-3m).

My field functions:
0-CellWidth : pow(${Volume}, 1/3)
0-DenGrad : mag(grad(${Density}))*CellWidth
0-Refine : (${WallDistance} < 0.04) ? 0 : (${Dengrad} > 0.05) ? ${CellWidth}/2 : 0)


The condition of wall distance is there because density gradients are also present near the boundary layer and I don't need refinement there..
I checked while the simulation is running, the cells supposed to be divided by two in length are highlighted, but when I remesh (after extracting my table), the refinement doesn't appear..
Am I wrong somewhere.. ?

[taillanm]

the field function for the cell width is not valid, it should ressemble

some constant * pow(Volume/Position[1], (1/2))

but the problem still persists