I have model that has a stagnation region downstream of a heat source. If left in 'automatic', CFX takes the highest temperature and applies it to the entire volume. I've noticed that the stagnation region will keep the solver from converging in a short amount of time due to high energy residuals.

One way I've stopped this is to setup an initial pick for the temperature that is equal to the inlet temperature, but this may not be the final temperature (although convergence is quick) and I'm back to getting 500+ iterations before convergence.

What kind of things can I setup to help CFX power through the stagnation region and converge my heat equations?

Hi,

What is the 'automatic' you discuss? Which option is that?

Glenn Horrocks

Hi Wooster,

It not so difficult. To speed up the convergence of the energy equation you can use a larger time step for the energy equation. A better option could be to switch of the fluids and turbulence equation for say 100 iterations with a large time step. Be careful. If in your case energy and flow is strongly coupled (free convection, compressible flow, etc) you must end with a calculation where all equations are solved simultaneously.

Regs, Gert-Jan

www.bunova.nl

You can switch of equation in the expert parameter section with "solve fluids = f" and "solve turbulence =f".

Regs, Gert-Jan

Glen,

When you initialize a model there is an option to set initial temperature to "automatic" or "automatic w/ value". From what I've seen, "automatic" picks the highest temperature in the model as the initial guess to the problem.

-M

Thanks for the help again Gert,

I'll try your first suggestion increasing the timestep. I had toyed with the option for a while but I'm not 100% sure how those timesteps help/harm the problem. I know the big picture idea from the manual but I'd like to know how I can "eyeball" an appropriate time step based on geometry and fluid properties if that is possible.

-M