[CMIUCBS]

Hi everyone.

I'm modeling a boiling steady state flow in a 12 m horizontal circular tube with ID=50 mm and OD=70 mm based on a papers geometry with the mass flow of 0.3 kg/s and under a Uniform heat flux and since the water enters the tube as a saturated liquid I'm using the non-eq RPI boiling model for my simulation to consider vapor convection and I have tried both K omega sst and K epsilon realizable as my turbulence model.

So far the results of the simulation for both turbulence models are near the results of the paper but the problem is my residuals are way to big to consider the simulation converged. (I have monitored few variables and they are stable)

here is a picture of my residuals for K omega sst and you can see that the residuals are not converging and fluctuating.

same thing happened when I used K-e realizable model; continuity and e were fluctuating around 1e-3 and the rest were around 1e-4.
so my question is that how can I obtain full convergence?
Thank You in advance and sorry for my bad grammar.

[h_ankur]

have you solved the issue?

[CMIUCBS]

Hi.
In many complex cases like this one, the decrease in residuals is something u never gonna see.

At the end, I changed the coupled scheme to simple, used k-e model instead of k-w, adjust the URFs based on existing papers and monitored the convergence of the simulation by tracking four variables: Volume Fraction, Pressure Drop, Mass Flow Rate, and Average Wall Temperature. I considered the simulation complete once these variables reached a steady state. The residuals, were around 1e-6, except for continuity which was around 1e-3.


Hope that helps.

[h_ankur]

Hello,
I am trying the same geometry with transient case but I am getting reverse flow at the exit of the tube and then my solution is getting diverged. Rest all the boundary coditions are same.
Can you help me in this issue? It would be of great help.

[CMIUCBS]

I think the divergence is related to the URFs. Reduce the URFs first and then increase the transient iterations at least for the first few time steps. Or even better, run a steady case and use the solution as the initial condition.
As for the reverse flow, it will disappear after a 1-2 k iteration but since you are doing a transient simulation I suggest you add to the length of your tube. If your tube is about 5 m, add another 0.5 m and then get the results from your original 5 m. The influence of the reverse flow on your original geometry will diminish.
+ Set the reverse flow T as the saturation temperature.