Dear Aravinth Kamaraj,

I don't know whether disabling energy is possible in compressible flow simulations or not....I hope you are using a density-based solver. In that case, I think one needs to set density to Ideal Gas and get viscosity computed using Sutherland's law. One also needs to tweak the operating pressure for compressible flow simulations....I think there is some connection between the operating pressure and supersonic/initial gauge pressure options for compressible simulations.

You also need to have prior knowledge regarding couplings existing in the problem formulation. For example, 300K air flowing over a cylinder at 325K is one-way coupled since flow would only transport temperature as a passive (and diffusive) scalar. On the other hand, if very high density particles are being injected in air, the coupling between air flow and particle distribution is not one-way. Then, the equations for DPM, turbulence and flow should (probably) not be solved in segregated (separated) form.

Having said that, you can initially turn off DPM, k-ω, species transport etc. and see whether flow alone is converging or not. If not, then there are issues at the solution-setup level itself. Try reducing all solution methods to first order accuracy (initially) and see if the divergence problem is solved or not.

I realize that this may not be of much help....but a general suggestion is that you need to zero down to either the region of the domain/mesh or the portion of the numerical model that is causing solver divergence....and that can only be achieved by simplifying the simulation setup as much as possible.

Best regards.

Shaswat Saincher