Standard Mesh Fails (Machine ran out of memory)

Hi everyone,

I am working on a project about the flow of a fan.
The link of the project is the following: SimScale

I have run various successful meshes and simulations in different geometries.
However for some reason, I can’t mesh the geometry of the simulation called “No MRF”.

This is the message that appears after each attempt.

It is strange because it is a geometry that I have meshed in the past and the process was not even expensive in core hours.

Can anyone help?


Hey Miltos!

This means that the computed mesh was too big for the automatically selected machine.

Please try with a lower mesh refinement and see if you can have a successful run.

1 Like

Thanks for your reply!

I tried it with a lower mesh refinement and it failed again.

The mesh that I am trying to generate “costs” less than 5 core hours.
The loading bar reaches 70% in a minute and it stays there until this error appears.
So I think that this message doesn’t make sense.

Thanks for the test and letting me know.

I will copy the project and try it to see if I can replicate the error.

Doesn’t seem like a small mesh.

I tried to run it and cut it after 3 hours.

Thanks a lot for your effort.

The mesh has been generated for this geometry. It costed 2.3 core hours.

Then I copied the mesh and tried to generate it for this geometry.

The only differences between them is that the second geometry has a surface behind the blades and the nozzle is gone.

So I assumed that the meshing cost would be similar.

In any case thanks again for your efforts!

Hi @miltos980,

Just to double check, what is the purpose of having a physical wall right behind the blades? This is going to fully block the flow.

On a side note, the memory issues are likely caused by small gaps between the additional wall and the rest of the structure.


1 Like

Hi RicardoParis,

Initially I used the MRF method to simulate the flow that is generated from the fan.

Then I wanted to see if different results come, when I have the following approach.

Thanks for your reply!