Fully constraint sandwich panel with distributed load

Hello,
Thought someone more familiar with PrePoMAx could help me shed some light on what is going wrong. I’m merely getting the ‘job failed’ answer. I am trying to model a fully constraint sandwich panel. I have tied the skins to the core. I set all bc to 0 on all 4 sides: u=0, v=0 and w=0 and also all rotations θx =0, 𝜃𝑦 =0 and θ𝑧 =0. The pressure is distributed equally on one surface. I’m using solid elements and have tried a number of different mesh densities with no luck. I have a i9 processor and 64GB RAM; the model is quite small.
How to get this really simple model running? Scratching my head for hours now :s Many thanks!
Frank
PS, tried to upload my input file but the forum does not let me as a new user … :face_with_open_eyes_and_hand_over_mouth:

This is the model: Sandwich 1000x245 hc30 ts15.pmx - Google Drive

The setup looks good and it runs to completion with much coarser mesh for me. CalculiX is quite sensitive to large meshes (especially around 1 mln nodes and more). I would just use hexahedral elements for this simple shape, it will give you much smaller mesh.

Sandwich 1000x245 hc30 ts15 hexa.pmx (6.5 MB)

Thank you for looking at it! Afraid, even with your file and changes to hex elements, I get this error when running the file: Job failed - no results exist. I’m trying both, Prepomax v2.2.0 and v2.2.6 dev; same issue :frowning:

Strange, this one should work. What are your settings in the Calculix tab ?

Below settings in the screenshot. I have tried all different solvers, any number of processors etc to no avail. This is the last part of the analysis file with a number of warnings but understand that these should not cause failure of the job:
(…)
*WARNING in gentiedmpc:
DOF 3 of node 344 is not active;
no tied constraint is generated
*INFO in gentiedmpc:
failed nodes are stored in file
Analysis-1_WarnNodeMissTiedContact.nam
This file can be loaded into
an active cgx-session by typing
read Analysis-1_WarnNodeMissTiedContact.nam inp
Job failed - no results exist.
Process elapsed time: 2.488 s

Also tried to import the file in previous versions but sadly it does not seem downwards compatible.

You should set the number of processors to the number of CPUs your computer has. Try with a very simple model like just the core of this sandwich structure.

The warnings are normal when using tie constraints.

Thanks, I will try… just tried to set up the model from scratch in v1.4.0 and this ran for coarse meshes but failed for more refined ones… I’ll try to download a fresh copy and then run your file again with the same number of processors as the chip got. I will update here but many thanks already for your help and suggestions. That is much appreciated!

I’ve tried a new downloaded copy of the software and it made no difference. I have tried to set the number of processors to the max cores of the chip (24) and it made no difference. Tried to simply it even further as suggested to a single skin but the job keeps failing with no apparent reasons. Really not sure what else could be tried :s
I have tried to use the Calculix launcher to see more details why a job may fail but didn’t see any other information in there either (may have overlooked something though)…

You can try exporting the input file and running it with a standalone CalculiX solver: Using standalone CalculiX

Maybe try something really basic in PrePoMax like a cube and see how it works for different meshes.

¿Could you try installing Prepomax in a folder without privilages like C:\Users.?

Change installation folder

Yes, tried that already from running it from the desktop or from the download folders … So far, I only got v1.4.0 running :s

Really not sure what is going on… managed to run v2.2.0 on my laptop which has an i5-1035G4 and 8GB RAM with above input files without problems. The desktop with an i9-14900K and 64GB RAM keeps having the same error over and over again… Both systems run on Win11, laptop on home and desktop on pro, each latest version. Haven’t managed to run it in the Calculix standalone yet; need to get a bit more familiar with it first… Is there anywhere, where additional errors could be found or any other place to get an indication of what is going wrong? Are there any dump files etc?
PS: I can run files created in v1.4.0 in v2.2.0 without an issue, just the files created here do not run…

Just to confirm, last stable version to prepare and run simulations for myself is v1.4.1. I can run simulations prepared in these versions in the v2s but anything I set-up in any of the v2s keeps failing to run. I will stick with v1.4.1 for now; but thank you for looking into it and I will try again with a future update and let you know if that worked any better.

This is really a strange behavior. I do not think the PrePoMax version impacts the solution. However, the older PrePoMax version uses an older CalculiX solver. You can try changing the CalculiX solver between PrePoMax versions by simply replacing the entire CalculiX subfolder.