Discussion Forum

Hello,

I am a PhD student in UCL, London and am working with SST model, trying to do a CFD study on a 2D rotating disc. So far, I am having trouble getting a converged solution for low viscosities (eg. 0.001 Pa.s, 0.000894 Pa.s). I have tried using k-e solution as an initial guess, but it still won't converge.

Once I went beyond the 'extremely fine' standard physics induced mesh and made my mesh much finer, I get the 'Very ill-conditioned preconditioner.
The relative residual is more than 1000 times larger than the relative tolerance.
Returned solution has not converge'.

I was wondering if you had any tips for solving SST model, and in particular with regards to the problem I have here. For the k-e model, I performed a continuation auxiliary sweep starting from high viscosity to get the final viscosity solution of 0.000894 Pa.s, and then using the last solution to compute SST for 0.000894 Pa.s. My boundary conditions seem to be well defined, and these are the following settings I am using:

Step 1: (k-e): using an Auxiliary Sweep of
visc_factor so that we have the final solution at the desired
viscosity value (0.000894 Pa.s) for the k-epsilon Interface

Step 2 (wall distance initialisation) : Variables not solved for settings, I set it to use the final solution from
the Auxiliary Sweep in Step 1.

Step 3: Both the variables solved and not solved for to "Solution", "Step 2, Stationary".

I have attached the file.

--
MSH