Bug using STL geometry?

Hi all,

I’m trying to simulate a 3D bed, similar to the tutorial example, however I’ve been finding difficulties.
In this case, I modified the geometry, cell size and bed particles, but following the very same steps that those in the tutorial. The simulation totally diverges and in the first output I see this weird cells outside of the domain.

At first, I though it would be just a visualization problem, but I found other strange outputs as well which I believe are the reason that makes the simulation diverge.

Any idea?

Looks like bad intersections with the stl file @jeff.dietiker

I’ve been doing some tests, also following DEM Particles with unrealistic high velocity - #8 by gbotha which suggested to reduce stl_small_angle. So far what I found to solve it is to increase the resolution of the primitive cylinder to 200 (example was 30).

the STL intersection likes nice triangles (good aspect ratios). It has a particularly hard time with long, skinny triangles.

Does the pre-processor report any ignored stl facets?

Not really, the FACETS_IGNORED.stl file is empty and the otput shows Number of ignored facets = 0

I attached the files of the simualtions in case it helps.geometry_0003.stl (32.4 KB) geometry_0003.stl (32.4 KB) MRI-3D_cylinder.mfx (16.8 KB)

so any clue on how to create cylinders? because aspect ratio in cylinders will be always huge, specially in internal flow

thanks!

The issue here is your STL cylinder is closed with top and bottom caps and you apply a wall BC on the entire STL file. You need to either only select the side wall (not the caps) when you define the wall BC or extend the cylinder so it sticks out of the top and bottom planes. I suggest extending the cylinder’s height, by setting the height to 0.35 instead f 0.30 m. This will give you a clean intersection and no conflict with the plane inlet and outlet BCs you have already defined.

Thanks for the point Jeff, that is usually something I do (following tutorial advice) and I already noticed the trouble of the top and bottom caps. I forgot to adjust it in this case (so many tests done…sorry).

However, I’m going forward and I’m trying to simulate a cylindrical bed with an internal. I tried to make the geometry using the mfix tools and creating a stl using other software and I found weird outputs in any case.

This is an example of two cylinders treated separately and created using mfix tools, but making and union of them also produces rare results.

geometry_0003.stl (32.4 KB) geometry_0004.stl (19.5 KB) MRI-3D_cylinder_2.mfx (18.3 KB)