[ranger85]

Hello all,

I have an interesting issue and I don´t know how to solve it.
For my pipe flow model with CHT I did a few first tests on the mesh where only the fluid domain is activated, it works out pretty good.
But when I run the entire model with the solid bodies around it, I do not reach my convergence criterion for the continuity equation.

I don´t understand what is wrong there, has anybody a clue about this?
Actually nothing changes in the fluid domain except temperature, but I am even not working with temperature-dependent properties.

Thanks for your help!

[LuckyTran]

The continuity model is a scaled residual, unlike the others. The continuity residual highly dependent on what happens during the first 5 iterations. I wouldn't use the continuity residual as a convergence check. It's not repeatable.

[ranger85]

Thanks for your quick reply!

But isn´t it curious, the mesh works out and continuity is only the problem if the solid bodies are activated, even if a solid body shouldn´t affect a flow field with temperature-independent properties right? Isn´t there any way to solve this issue?

[LuckyTran]

Let's say you setup the two cases identically with the same models, settings, and initial conditions. Now you start iterating. But since you have solid bodies now, your flow-field evolves differently during the first 5 iterations of your CHT case compared to your case without CHT. Let's say that eventually both solutions converge to the exact same fields. The difference in first 5 iterations alone is enough to affect the continuity residual because of the way it is calculated.

What I'm saying is, it is entirely within the realm of possibility that the continuity residual is acting up even when nothing is wrong with your case.

Also if you initialize your case with the final solution, the continuity residual will be stuck at 1 and never change (despite the case being perfectly converged).