Set this between the blade tip and the duct/housing inner wall. Everything within this radius of the axis rotates (hub + blades); the duct outside it stays stationary. The blue cylinder in the viewer shows the zone live.

⚠ Expensive — for an overnight run.
What it does. Normally the first cell sits far from the wall and the near-wall flow is estimated by a wall function — a flat-plate formula. This option instead puts the first cell inside the boundary layer so the solver computes the wall shear. On curved, thin or separating surfaces the wall-function estimate typically over-predicts skin friction, so drag comes out high and efficiency low. Turn this on when you need trustworthy drag, forces or efficiency rather than a quick comparison.
What changes. The first-cell height is computed from your current speeds to hit the target y⁺ (rotating parts use the helical resultant √(V∞² + (ωR)²), not V∞ alone). The layer count is derived to span the boundary layer, so it replaces the slider above — expect roughly 10–15 layers near y⁺ 5 and 20–25 near y⁺ 1, with a substantially larger mesh and hours, not minutes.
Before you rely on it. Set V∞ and RPM before meshing — changing them afterwards invalidates the sizing. After meshing check the layer coverage in Mesh details (low coverage means the layers failed to insert), and after solving check the reported y⁺ actually lands near your target.