I’ve experimented with explicit dynamic, simple model were lead ball hits steel plate. In results i got about 300 time steps. When time to time I switch steps PPM crash. And when try to start animation error massage appears and after crash may happen, especially it happens with “Section view” turn on, or deformation Automatic type. .
I can confirm that, it happens to me with all results regardless of the model size (just get the error message without crash)
Perhaps section view should be disabled when running animations. Instead, hiding selected elements (like in Abaqus - often better than section view anyway) could be added in the future.
In previous PPM 2.0.0 animation worked with “section view” turned on, in animation “scale factor”. Now in 2.2.0 it doesn’t.
I have a prove, I recorded video:
it’s sad regress for me
I can also confirm that it worked until v2.1.1
Hi, I’ve just downloaded the new PrePoMax 2.2 and your model. For me the animation works perfectly in regular and in section mode, but there is a trick. For section mode it only worked when I utilized the GPU render, for non-section view it worked in both ways (rendered by CPU and GPU). Hope this helps somewhat.
Thanks for trick!
I use AMD Radeon HD 7400 Series and looks like GPU render no very fit for me (i don’t how turn it on).
And in this file [sharik5.pmx] (sharik5.pmx - Google Drive) just turning on “Section view” causes PPM crash.
May it be because I use still use Win10 1607, and now something lacking in OS. But other function seems works proper and PPM 2.2.0, I guess, architecturally not very far from 2.1.0
And it would be better if everything works from box without dancing with tambourine.
A bug in PrePoMax prevents the section view from working normally in some cases. If you get a window with the error message, the problem is with PrePoMax. If the program crashes, the problem is usually the graphical library.
I will try to fix this problem, but using a large number of animation frames will always result in crashing the PrePoMax due to the problems in the graphical library to which I have no access.
1 Like
I was able to fix the PrePoMax bug, but this still does not help with your model.
After some testing, I think the problem is the excessive deformation of your model with elements turned inside out, for which the VTK cannot compute the section cut and throws an error.
If I use small deformation scaling like Automatic 0.25, it works, but for larger scaling factors, it crashes.
I am using the file:
2 Likes
Then here the aforementioned hiding of selected elements could help. It was recently implemented in FreeCAD FEM.
Another thing is that explicit dynamics procedure in CalculiX is really problematic and output frequency usually needs to be reduced significantly (in most cases hundreds of frames are not needed).
True 
In this case, the problem with the PrePoMax visualization is that while PrePoMax only displays the outer surface of the mesh, the outer surface is pushed inside the material while the elements get inverted, and a new outer surface should be computed. However, I think such behavior is non-physical and will not cause problems in general.
I wonder why this analysis hasn’t failed due to excessive elements distortion / negative jacobians (inverted elements) then. I guess CalculiX doesn’t check that during the simulation which is a major difference from Abaqus and another thing to be very careful about when using the explicit dynamics procedure in CalculiX.
1 Like
just for curiosity I create similar scheme but made of tetrahedral finite elements. And “Section view” normally works (except animation of course). Looks like crash cause Box elements in mesh
Yes, the unphysical distortion (inversion) of tetrahedral elements. So remeshing (also with hexahedrons) or changing the analysis conditions helps. The original analysis should be stopped by CalculiX in the first place.
1 Like