turbo_moc.meshing.wall_spacing module

First-cell height for a target y+ from the Phase-1 MoC expansion.

A rough, pre-CFD estimate: local turbulent flat-plate skin friction at the blade chord, with the flow state taken from the nozzle’s own isentropic solution (wall_final: rho, T, M per station). Properties come from jaxprop at (rho, T), so a two-phase (wet) expansion uses the mixture viscosity and the equilibrium speed of sound. Accurate to roughly a factor of two – the usual precision of a y+ estimate; check the actual y+ in the CFD.

Re_c = rho U c / mu, U = M a C_f = 0.026 Re_c^(-1/7) (turbulent flat plate, local; White) tau_w = C_f rho U^2 / 2, u_tau = sqrt(tau_w / rho) y = y+ nu / u_tau, nu = mu / rho

The station with the highest wall shear (smallest y) governs, so the target is met everywhere along the expansion.

Fluids without a CoolProp viscosity model (e.g. Novec649, MDM) fall back to the Chung et al. (1988) dilute-gas correlation from the critical constants (within ~1% of CoolProp for non-polar vapours, ~10% for polar ones such as R245fa; dipole and dense-gas corrections are omitted). In a two-phase state that is the vapour-phase value at the local temperature.