Made an Issue: Add Abaqus *EQUATION support · Issue #14 · Luecx/FEMaster · GitHub
This should be extremly easy, i can see if chatgpt can do that 
2 Likes
Thanks a lot .
I guess you already know or found out *EQUATION is what older softwares like Nastran called MPC (multi point constraint). It is useful in many instances for more general constrains and for special problems like periodic boundary conditions like in M Kraska example: CalculiX-Examples/RVE/Periodic/README.md at master · calculix/CalculiX-Examples · GitHub
here, the files to dowload to make it easier to reproduce:rve - Google Drive
1 Like
When can we add friction contact?
1 Like
latest release supported Abaqus/CalculiX keywords but solver running without any succeeded and reported as unknown.
If the solver supports the *STEP keyword now (it used to only support *LOADCASE) then maybe PrePoMax needs an update to account for that. Perhaps there are still some differences in the syntax though.
it seems the problem is related to FEMaster solver itself not working properly. A lot of Abaqus//CalculiX keywords commonly used already supported, even in material only partially read successfully for example Material, Density and Elastic is accepted but Expansion is not.
btw, i need to direct comparison with CalculiX but missing an element of linear hexahedral with incompatible mode (C3D8I) and hopefully it will be available, thank you @Luecx
Then it might be good to just report the issues directly in th FEMaster’s GitHub repo. The dev is active and should consider them.
I was getting this error using FeMaster v.2.8.0 and latest PPM:
Error: Analysis-1.inp:7873: Unknown keyword ‘*STEP’
It turns out FEMaster does understand *STEP. It just wasn’t reading my file as an Abaqus file. Look at this line in the solver window:
[INFO] Format : femaster
FEMaster can read two different input languages: its own native one, and the Abaqus/CalculiX one. By default it uses its own. In its own language *STEP doesn’t exist, so it complains. You have to tell it explicitly to read the file as Abaqus.
As a first (and failed) attempt was to go to Analysis > Edit and add --format abaqus after the file name in Executable arguments:
Analysis-1.inp --format abaqus
Run it again, and the *STEP error was gone, but a new one appear:
No variant of ‘BOUNDARY’ fits the upcoming data
If you open the .inp that PrePoMax writes and look for it, you find this:
*Boundary, op=New
** Name: Fixed-1
*Boundary
Internal_Selection-1_Fixed-1, 1, 6, 0
That first *Boundary, op=New is an empty line of instruction. PrePoMax writes it to clear any boundary conditions carried over from a previous step. CalculiX is fine with it being empty. FEMaster is not, it wants at least one line of data underneath. The same happens a few lines later with *Cload, op=New and *Dload, op=New.
As a workaround you could delete them by hand every time, but PrePoMax rewrites the file on every solve, so that gets old fast. Instead, make a small batch file that deletes them and then calls the solver for you.
*****************
Open Notepad, paste this, and save it as femaster_abq.bat in the same folder where FEMaster.exe is:
@echo off
setlocal
set “INP=%~1”
findstr /V /I /R /C:“^*[A-Za-z ]*, *op=New” “%INP%” > “%INP%.tmp”
move /Y “%INP%.tmp” “%INP%” >nul
“%~dp0FEMaster.exe” “%INP%” --format abaqus
Make sure Notepad saves it as .bat and not as femaster_abq.bat.txt. You don’t have to edit anything inside it. It finds FEMaster.exe on its own because it sits in the same folder.
Then, in Tools > Settings > Solver point in the FEMaster executable section to the .bat file, instead the .exe file.
Then it would work (at least for me)
Just one warning, the batch file edits the .inp before handing it over, so the file left in the Temp folder is not exactly what PrePoMax wrote. That doesn’t matter in normal use because PrePoMax regenerates it every time you solve, but keep it in mind if you are inspecting the input file.
1 Like
many thanks for hints about format definition is required and mandatory, running the bat files thrown some error message in plate analysis of unrestrained vibration but tracing and editing manually solved the problems.
*Step
*Frequency, Solver=Pardiso
10
*Boundary
*Node file
U
*El file
S, E, NOE
*End step
So basically two key changes are needed to actually run FEMaster from PrePoMax - switching to Abaqus format (I guess that it could be the hard-coded setting in PrePoMax) and removing the keywords with OP=NEW. Or we can ask the FEMaster dev to support them.
1 Like
I think I can implement it in PrePoMax.
4 Likes