Learning Forum

[Vasanth]

Transient simulation convergence

I am trying to simulate flow over bluff body with k-epsilon turbulence model. It's a transient simulation and the solution is converging at every time step. As you can see, residuals are not changing at all with number of iterations. How can I know if the solution is converged? I searched for similar query in forum but the answers are not very satisfactory or explanatory. It will be helpful if anyone can shed some light on the problem of convergence.

Thanks.

[Essence]

Hello,

What time step size (seconds) did you use? Try to lower it down.

[Vasanth]

I was using dt=0.025. Simulations are converging even with this time step at each time. I was just wondering how to know solution reached final value? Should I take average of the flow properties and see if they are changing or is there any other way?

[Essence]

Hello,

I can see that according to the convergence criteria you have set the solution is converging. But please do not judge the convergence only by residuals, as residuals provide information about conevrgence partly. For proper information, check the net mass and heat (if applicable) flow rates in the Fluxes. Of course, usage of monitors/report plots is very essential in detetming the convergence.