Need help with A CPU water pump simulation (Basic)

Project Support
1

Hi there, we are currently having problems defining the inner flow region between two pipes of which one is inlet and one is outlet. This is a basic bottom half of a water cooled CPU Unit. Any suggestions to fix this? Using a CHT transfer case for this setup. I will probably need some more guidance along the way, I’m quite new to all of this.

Thank you .

Capture23
Capture231061×731 85.8 KB

2

Hi @abersiks!

Please always share the project link with us if you need any support. In the meantime I am tagging the @cfd_squad to jump in once you posted your project here.

Best,

Jousef

3

Thank you very much. Let me try give the link now.
I’ve just sent the URL I am currently on, would that work?

Thanks again

abersiks

4

Is it normal for a meshing process to be longer than 15 minutes?

5

Hi @abersiks!

I saw that you have a fineness of 0 for your mesh? That seems a bit odd, maybe you try something like 2-3 or keep it as 5 and see how long it will take you for the coarse one.

Best,

Jousef

6

Thank you, I’ll try that next. I assumed the more coarse the less time to evaluate?

7

I changed it to 5 now, I think I forgot to toggle physics based meshing, let me do it again

8

Definitely but maybe 2-3 should be coarse enough, if you have any struggled with Core Hours, feel free to tag the PowerUsers and they can help you out with meshing :slight_smile:

Keep us posted!

Jousef

9

Thank you very much! I’m not too sure how to tag PowerUsers.
Is there a physics component that hasn’t been prepared in the simulation that you can see? It says I shouldn’t mesh until that is sorted out.

Thanks for all the help! :smiley:

10

Could you post the message here? Just give it a spin and you can see if you’re satisfied with the mesh afterwards (assuming the workbench lets you mesh the geometry). PowerUser are tagged by writing @power_users. :wink:

Cheers,

Jousef

11

I tried changing the mesh again, I realized I used the mesh without physics again. So I have to retry that. The process seems to still be taking very long, and I still have the suggestion for fixing the simulation setup before the meshing. However, everything seems to be well defined.

If I want to address the power_users do I tag them here like this? @power_users. They should be able to see this then. :sweat_smile:

Thank you for all your help Jousef

12

It just seems that I cannot toggle the physics mesh, when all the parameters should be correct. It also seems to be stuck at 50% consistently when I mesh it without that option. @power_users

13

Any progress here @abersiks? Otherwise I would give it a spin later on.

Best,

Jousef

14

@jousefm Unfortunately not. The mesh did finally form last night taking a long time and using too many resources. Now to run a simulation is impossible because of the run time required. So I was advised to simplify the whole geometry of which I have done.

If you could maybe take a shot at it, that would be great! Thank you!

All the best
abersiks

15

Hey @abersiks!

One thing I just noticed is that you have threads on your parts which most likely causes the geometry to have a very refined mesh in that area, could you maybe remove the threads (anyway they do not contribute that much to the analysis) and give the mesh a try? I believe that it would mesh faster then (that’s my hope) :slight_smile:

All the best!

Jousef

16

Thank you. You were probably right about just needing to discard the threads. We just made a new much more simplistic model in the end. I was wondering is there a way to export temperatures and or any data from simscale? I did see you did mention something on a thread about paraview. Is there not a more direct way?

Thanks for all the help :smiley:
Regards

abersiks