Learning Forum

[RLester5]

The goal of this project is to simulate a gas entering a system and changing phase over time.

The issue I am running into is a floating point exception and after re-numbering the nodes I know for certain that this exception occurs at the inlet of the gas. 

The Geometry:

 

The gas will enter at the Gas inlet at the top of the picture. This gas will have already have traveled 13 feet along a similar tube so it should already be fully developed but will be moving very slowly because the only thing drawing this gas in is the temperature/pressure difference inside the condenser due to the liquid helium cooling the copper thus dropping the temperature inside the condenser thus "Cryo-trapping" the gas. 

Once the gas has made it into the condenser it should start to condense on the copper fins that protrude inside changing phase from a gas to liquid to solid. 

The goal is to watch this happen over time and monitor pressure and temperature changes with time as well as changing volumes of input gas. 

 

The Mesh:

Location of Floating point exception:

Mesh Details:

Setup Properties:

Initialization:

Calculation:

Patching:

Floating Point Exception Error:

 

I think I've looked into a lot of what could cause these errors such as shared topography, skewness, etc. But after continuous failures I am not sure how to fix this problem. I think the issue may no longer be with the mesh but in how I am defining the inlet itself.

Here is the setup:

I believe this is the correct way to set it up, however am unsure if I need to set a guess for an x-velocity and change to a velocity inlet.

 

Please let me know what you think, and if you need any more files please let me know!

 

Thank you for your time, and help.

 

Sincerely,

Ryan

[Rob]

Please plot the mesh image with walls on & interior off.   How do the fluids get out of the domain?  Given the limitations of VOF, why have you chosen that model?

[RLester5]

When you say walls on interior off, do you mean a wireframe view or do you just want to see the meshing at the inflation zones?

 

The idea is to put the gas in and freeze it onto those copper fins so the fluids would not leave the system until the cooling is turned off. 

And as to the limitations you mention I am not sure what those are exactly, I have picked that based on previous research done/videos watched online, but am more than open to suggestions on a better way to model it. 

[RLester5]

Warning: flonum_arg detected NaN. Trying to continue with it.

This is another error I am seeing. I believe it is associated with the floating point, but I noticed I didn't post it earlier.

[RLester5]

I don't believe I hit reply to you*

[Rob]

But if the materials are entering the domain where do they go? 

[RLester5]

@rwoolhou  the materials would enter, undergo this phase change for a period of time, and then I would increase the temperature to return everything to a gaseous phase and withdraw down the same pipe.

Timeline:

Travel down the pipe, into the condenser, stay here for ~8 hours undergoing phase changes, then the temp would increase again and gases be withdrawn

[Rob]

OK. You're going to need to be very careful with this one (it's not an easy application), and I'm not allowed to explain fully due to the rules we work under on here. 

Have a look at the phase change models in the documentation: also the limitations in VOF and Eulerian models. Think about where the mass/volume are going to go as you condense/evaporate the materials and how you'd account for that in a model. 

[RLester5]

Well that's unfortunate, I didn't know there were limitations for y'all. If there are any other forums you'd suggest that would be very helpful.

 

But okay, going off what you said:

In the documentation refers to the fluent manual I am guessing right? 

In terms of where the mass/volume are moving, I guess I am not thinking about this correctly, because I just want them to sit in the condenser, and from intuition I would guess condensation would start on the fins due to their large surface area. There may be some backflow up the pipe, but not much if any. 

There would be no evaporation. And the only way I know how to account for mass/volume changing would be in a multi-phase model.

 

If there are any examples you can have me look up that would be great to. I had watched one on youtube corresponding to oil and water mixing in a can to try and understand mixing. But I have yet to find anything like what I am doing. (which is frustrating to say the least) 

 

Thank you for your help so far, if there is anything else you can tell me I would be grateful!!

[Rob]

Yes, if there's phase change you will need to account for that. You may also need compressible fluids (gas) so the material has somewhere (space) to go! It may also make sense to model only one side of the system: read up in the Fluent manuals on wall thermal condition. 

Condensation isn't a simple model (as you're finding out). One option you have is to check with your supervisor as they may be able to contact us directly through support. That means we can do the check and so can answer questions fully. 

 

[RLester5]

@rwoolhou is there a way for me to privately contact you? I have spoken with my supervisor and our ANSYS contact and they mentioned a few solutions, that would enable you to continue to help me with this problem. Please let me know what the next steps should be.

 

[Rob]

You'll need your supervisor to log a support ticket: depending on region that'll be dealt with by the local team or partner.