Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.
NITF help
Posted Nov 7, 2023, 12:56 p.m. EST Fluid & Heat, Studies & Solvers, Physics Interfaces Version 6.1 3 Replies
Please login with a confirmed email address before reporting spam
Hey All,
I have been recently trying to solve a heat transfer and fluid flow model with a time dependent study, but I keep getting NLfail. It failed 8 times until the time step was 2.5225e-5 seconds. Which is so small that my study has only reached 0.02 seconds after 26 hours (And I'd like the study to reach a few hours at least).
I have read on other forum posts that the NLfail happens when the mesh is not refine enough. But I cannot find Comsol documentaion to back that up. Is this true? Or are there others things that can contribute to NLfail too?
I have already trie refining my mesh and followed all the steps mentioned in this article: https://www.comsol.com/support/knowledgebase/1127
Is there anything else that people have found works? I would post my model but the model is a project for work so I cannot.
Thanks, Odette
Please login with a confirmed email address before reporting spam
There are a lot of possible causes of this. If you've already tried the steps in the article you cited, then I'm guessing you have some kind of error in your setup that you are simply overlooking. Have you been able to get a simpler model that is otherwise similar to work? If you can simplify your model to the point it is non-proprietary (e.g., replace your complicated geometry with just a cube and/or cylinder, for instance, and maybe change your time scale to 0 to 1 second, modify and/or rename all your materials, etc.), and then if it still exhibits the same issues as before, you could post that model here and just might receive some useful advice.
-------------------Scientific Applications & Research Associates (SARA) Inc.
www.comsol.com/partners-consultants/certified-consultants/sara
Please login with a confirmed email address before reporting spam
Hi Robert,
Thank you so much for your reply, apologies for not getting back until now.
I have created a simplified version of the model which I can now post. The NL fails have (for the most part) stopped but I have 2 different curious issue.
The solid domains on my part seem to get crazy hot (4830 degC after 20 seconds), even though they are the container of a pipe with constant flow. If the water is entering the pipe at 20 degrees celcius, shouldn't this cool down the solid part?
The fluid appears to be travelling in the opposite direction that I want it to (i.e. out the inlet and in the outlet). At first it was not doing this, but I cannot figure out which condition is affecting it.
I'm a comsol newbie so I am sure these issues are on me and not anything actually wrong. If anyone could please look over my model (especially the BCs), I would be very grateful :)
Thanks so much, Odette
Attachments:
Please login with a confirmed email address before reporting spam
Hi again,
I think the issues have to do with my fluid flow boundary conditions. The velocity of the fluid is very low, which explains why the temperature is reaching high values.
I have changed the boundary conditions so that there is a pressure condition (90 psi) on the inlet and a velocity condition (3.157 m/s or 40 L/min) on the outlet. But now I am back to the convergence issues, with NL fails and T fails. Not sure why because the problem seems rather simple (fluid through a pipe with a heat source).
I attached the model with said changes. Please advise, I have researched the issue extensively but no luck.
Thank you, Odette
Attachments: