[skiingengineer]

Hey guys, question for you. I have a cross sectional assembly of a cartridge heater inside two halves of a simulated nuclear fuel rod that I am modelling and I'm having issues with it. I'm trying to get the embedded mesh to apply through the entire assembly. BUT no matter what I do I can only get the two outermost regions to use the embedded mesh. Here is a the geometry in question with one of the meshes I've been working on:



Here are a few regions hidden:


You can see the inner regions are all pure poly mesh with no thin meshing (despite trying to force it in all regions by specifying custom thin mesher specs in all regions)

A few more details about the model. This is an all solid energy simulation with all the assembly parts having different thermal contact resistances between their interfaces. To keep the mesh conformal I have just one mesh continua for all 10 regions.

If anyone has any recommendations I'd be grateful. Also if you need more details please let me know.

[me3840]

What version are you using? The embedded thin mesher is old region-based terminology for the parts-based thin mesher. Newer versions of the thin mesher are much better.

The aspect ratio of the parts you have there isn't very large in the middle, the thin mesher may have trouble generating cells there. Have you looked into directed meshing this geometry?

[skiingengineer]

Originally Posted by me3840

What version are you using? The embedded thin mesher is old region-based terminology for the parts-based thin mesher. Newer versions of the thin mesher are much better.

The aspect ratio of the parts you have there isn't very large in the middle, the thin mesher may have trouble generating cells there. Have you looked into directed meshing this geometry?

I'm actually using 9.06 which is pretty old, I'm trying to finish off a few simulations I have left for my Master's so I didn't want to go changing versions if I didn't have to. (Unless I'm shooting myself in the foot with it being overly outdated)

That was what I assumed was the problem because it really isn't "thin". I wasn't sure where to go next but I'll definitely try directed meshing (haven't used it before). The main thing I'm shooting for is an evenly distributed mesh (in terms of the mirrored pairs of regions i.e. the top and bottom outer shells having the same number of cells or very close to it) and to have a conformal mesh so all the heat transfer across the interfaces are correct. So far those two things have been mutually exclusive for me.

I'll work through some directed mesh examples and redo this simulation and report back. I have more complicated versions of this model with helical heating elements but I want to get this one working first.

Thank you for the response I really appreciate your time.

[skiingengineer]

One issue I've been running into now is that most of the regions in these simulations have significant surface repair/modification and because of that STAR isn't allowing me to use a directed mesh ... is there a way around that? (besides having all un-modified CAD files)

I had a lot of issues importing the CAD for these sims and getting lots of pierced faces. So I oversized some of the parts in the CAD and subtracted them from each other then imprinted and that was a quick fix for the pierced faces.

Thanks again for the help/advice...

[me3840]

No, there isn't a way around that. But this geometry is very simple, it should be pretty easy to draw in CAD.

[skiingengineer]

Originally Posted by me3840

No, there isn't a way around that. But this geometry is very simple, it should be pretty easy to draw in CAD.

Okay, that's what I figured. Oh well ... Thank you!

I actually have all the CAD done already in SolidWorks, the issue I was having was when I imported it into STAR as a parasolid, I have issues with tons of pierced faces between most of the parts hence all the surface repair.

I'll double check my imprinting settings and try again...

[me3840]

I'm not sure about your version, but at some point 3D-CAD had an imprint operation in it, which will maintain the CAD but still do the imprint. That may help.

[skiingengineer]

Originally Posted by me3840

I'm not sure about your version, but at some point 3D-CAD had an imprint operation in it, which will maintain the CAD but still do the imprint. That may help.

THAT is the missing link I think... I was just reading in a tutorial that they explicitly tell you to use 3D-CAD imprint for directed mesh and not surface repair imprint (What I did)