Problems with meshing - pipe system results in a cube

Hi all,

To model a pipe system, I made a simple setup in Inventor, exported as STP file and uploaded it to SimScale. However, I get problems during meshing. Below, you see a section cut of the model in Inventor. When I try to mesh it, it either takes very long and tells me that the mesh is too complex or it results in a cube. The model does not seem complex to me.
Would the physical size of the model be a problem? And what would be a solution in that case? I expect that water flows through these pipes will be different when I downscale the model.


Hi @tdijkstra!

I will have a look at it later on and let you know what the issue is. Trying to mesh it myself as well.

Best,

Jousef

1 Like

Hi again @tdijkstra!

Just to let you know, the model you posted here is obviously not “watertight” that the Hex-dominant mesher needs to work properly. The one you have is open and should be closed so that no water could “flow out or into your model” (maybe that makes it a bit easier to grasp what watertight means here). The other model you uploaded called “Pipes (closed)” has two layers of surfaces which only should be only layer just to remain a consistent and clean modeling technique which also makes sure you do not run into trouble when meshing your geometry. Remember: A good mesh is more than half of the battle won :wink:

Simply fix your geometries and keep us up-to-date about the progress. Happy SimScaling!

Jousef

Hi Jousefm,

Thank you for your reply, although I had some other work to do in the mean time. Initially, I made a shell in Inventor, resulting in the double surfaces. The current version (Singular face) is a solid in Inventor. After uploading, I was able to set all the parameters. I will run the simulation later today.

1 Like

Sounds good @tdijkstra! :slight_smile:

Let me know how things go and do not hesitate to ask if there are any questions.

Best,

Jousef

That works so far, thanks! I just ran a first simulation, see below. However, in reality, the water flow is not filling the entire pipe system. Instead, a constant level is maintained, as shown in the photograph below. I found a similar case in another SS-project. How can I simulate this effect of a certain water level, instead of an enirely filled pipe? I think I should make use of air as a second material, but how would I set it up?

Hi @tdijkstra!

You defined the whole domain to be water which can also be observed in your screenshot. You would have to define a small domain/box that acts as a container for your fluid - define associated phases to it (Phase1 & Phase 0) and assign the box to water (whatever phase water has in this case). The link you posted demonstrates this quite good and I would use this as a template for your project. In your case you would have to create a box that is slightly smaller than your height of the container (in your post-processing picture the right side I assume).

Cheers,

Jousef

I tried it. Now, I don’t get an error anymore, but the project just keeps running. After it took almost an entire working day to reach 10% of a simulation, I cancelled it. What am I doing wrong?

Too%20long

Hey @tdijkstra!

Having a look at it later on and will give you an update as soon as I know more.

Best,

Jousef

Hi @jousefm,

Have you had time to take a look at it yet? I compared my project to the waterfall. I changed a few things - possibly I did not setup the mesh correctly in the simulation settings. A new simulation is running now. I’ll keep you updated!

TD

Hi @tdijkstra,

my run stopped at 21% due to violation of the CFL condition. Have you tried running it already? If not then I can run it with more cores on my account to save you some core hours.

Best,

Jousef

Hi Jousef,

Hm, did that happen during the last hour - so after my changes? It is currently running, but after 50 minutes only at 2%. Do you mind running it for me? Then I’ll cancel mine.

TD

Sure @tdijkstra!

Just post the most recent link here and then I will copy and run it for you.

Best,

Jousef

Thank you, Jousef!
Does this work?

TD

1 Like

Works @tdijkstra!

Which simulation is the most recent one you adapted? I assume the last one?

Cheers,

Jousef

Indeed, @jousefm! It’s the one I renamed ‘Multiphase (lastest)’ and the simulation is called ‘Run 3 (lastest)’.

1 Like