[cgswjs]

Hi everyone,
I’m new to star-ccm+ and I really need someone to help me at this point. After watching tutorials on YouTube and reading materials from google, I figured out most of my project and I got a good residual plot until something happened at interaction 250. Here’s what I’ve done: I repaired surface and there’s no piercing. I set up auto mesh and surface control all different regions. I did mesh diagnostic and removed bad cells. Cell quality remediation was turned on as well. My project is a double piped heat exchanger with hot water in the inner tube and cold air in the annulus. I set velocity inlet for annulus air and mass flow inlet for tube water. Both outlets were set to pressure outlet. I got a nice temperature and velocity contour before residual went sky high. After iteration 250, the residual went up to 1e4 and I stopped the algorithm. Can someone help me identify the problem here? I’ve tried lower the URF but it didn’t seem to work. I really don’t know what else to do. Btw, advancing prism layer mesh was used for both fluid regions and I assumed the boundary layer thickness was 1mm.

[Nadaaaa22]

I don’t have many knowledge in the software, for the mesh you can use mesh sensitivity to assure that your mesh is ok. For the divergent it may due the back flow. If you have a fluid domain you may try to make it larger I don’t know if this would be helpful.

[cgswjs]

Originally Posted by Nadaaaa22

I don’t have many knowledge in the software, for the mesh you can use mesh sensitivity to assure that your mesh is ok. For the divergent it may due the back flow. If you have a fluid domain you may try to make it larger I don’t know if this would be helpful.

Yes I think it caused by reversed flow since at interation 251 it appeared a lot reversed flow in air. How can I fix this reversed flow issue without changing the geometry? What do you mean by make it larger? Since I need to verify this CFD model with an experiment set up, I can't just change the model.

[Nadaaaa22]

How did you add the fluid domain of air?.

[cgswjs]

Originally Posted by Nadaaaa22

How did you add the fluid domain of air?.

Create a fluid continua and in region property tab assign it to the region

[steve lee]

Diverge problem is quite complecated.. It is very hard to give you any advice about this kind of thread. Sometime mesh problem is right, but sometimes are not.

Check what factor diverged in your residual first

[RKE]

Good Morning!

The description you provide suggests to me that your problem may be due to numerical instability occurring at some point where the velocity field changes rapidly (say at a turn, flowing past a baffle, etc.). That has happened to me when modeling agitator blades as the pass in front of a baffle in a stirred tank.

I have been able to correct that by specifying relaxation factor ramps (i. e., progressively changing relaxation factors) for the various solvers.

Maybe some variation of these adjusments could be of help to you.

[cgswjs]

Originally Posted by RKE

Good Morning!

The description you provide suggests to me that your problem may be due to numerical instability occurring at some point where the velocity field changes rapidly (say at a turn, flowing past a baffle, etc.). That has happened to me when modeling agitator blades as the pass in front of a baffle in a stirred tank.

I have been able to correct that by specifying relaxation factor ramps (i. e., progressively changing relaxation factors) for the various solvers.

Maybe some variation of these adjusments could be of help to you.

I'm using stead solver right now and yes I have 6 turbulent generators in my annulus region so I guess it could be the problem. Do you suggest me to use implicit solver?

[RKE]

Yes, I suggest using under-relaxation in the solvers. The system provides some initial ramp settings to try first.