CalculiX 2.12 Available

[makkemal]

http://www.dhondt.de/new_calc.htm

New features in Version 2.12 of CalculiX (check the documentation!)

Procedures

Specific gas network elements (such as GAS PIPE FANNO) can be used in a purely thermal calculation (parameter THERMAL NETWORK on the *STEP card); in combination with the user subroutine film.f the heat transfer coefficient is determined by use of the Glielinski equation (cf. example beamhtfcnu.inp).
Improved the convergence of gas networks.
Updated the calculation of sensitivities (cf. *SENSITIVITY)
Coded *GREEN function calculations (response to unit forces at certain locations of the structure)
accelerated the reading of the global model in *SUBMODEL calculations.

Coupling

Implemented the BEAM nonlinear MPC.
Coded *COUPLING followed by *KINEMATIC.
Coded *COUPLING followed by *DISTRIBUTING.

Elements

The GASPIPE network element was removed. Instead, the GASPIPE FANNO element should be used.
Coded the SPRING1 and SPRING2 element.
Coded the MASS element.
For shells a local coordinate system is automatically established by projecting the global or local (if defined by the user) x-axis on the shell surface.
Coded 6-node composite elements.

Documentation

Updated the theoretical description of gas network elements.

Output

Added the orientation in the dat-file for the quantities stored in local coordinates.
Coded *SECTION PRINT for the output of heat flux, section forces, section moments and section areas.
Included the relative stress error (heuristic relations) in the output of the error estimator ERR

[easyw-fc]

http://www.dhondt.de/new_calc.htm

New features in Version 2.12 of CalculiX (check the documentation!)

Hi Michael
I have built a 64bit MT of the ccx 2.12 version for windows 64bit here
https://github.com/easyw/ccx_prool/tree ... lder/win64
win64/ccx2_12_MT-64.exe
would you mind to test it?

I've done some little testing and it seems fine to me

Thx in advance
Maurice

[UR_]

Hello,
this is result of test suite from here
http://www.dhondt.de/ccx_2.12.test.tar.bz2

error.1180.txt

(3.58 KiB) Downloaded 110 times

done on
win 10 x64 14393.969
OMP_NUM_THREADS=1

Has anybody a result file for native (linux) build?
Divergences are a little bit frightening

Tomorrow i will do a test run on this build
https://github.com/ddfem/ccx_mingw64

[UR_]

Now tested this build:

https://github.com/ddfem/ccx_mingw64/bl ... 2_MTNO.exe

error.1480.txt

(640 Bytes) Downloaded 99 times

also done with:
win 10 x64 14393.969
OMP_NUM_THREADS=1

See some informations here:
http://mecway.com/forum/discussion/405/ccx-2-12

reducing compiler optimizations seems to work.

[makkemal]

@easyw-fc
Thanks
Caluclix launcher also updated
https://yadi.sk/d/UB72nDqrtw3aZ

[easyw-fc]

Caluclix launcher also updated
https://yadi.sk/d/UB72nDqrtw3aZ

I added also 32b MT build
ccx2_12_MT-32.exe

Maurice

[looo]

do you have an instruction (script) how to build calculix. I would like to create packages for conda.
are you using spooles and arpack?

[easyw-fc]

do you have an instruction (script) how to build calculix. I would like to create packages for conda.
are you using spooles and arpack?

Hi @looo
if you were meaning for windows, you can look at my repo which is a slight modified of the ccx_prool repo
https://github.com/easyw/ccx_prool
there you can find a makeMT212.sh file to build it...
I'm using Cygwin64 with spooles and arpack

[looo]

Thank, nice you have set up a repo for calculix. This will definitely help a lot.
linux is first, if this works I may try to build with windows. But I have no idea if there is a fortran compiler in visual studio that works for calculix.

[Jee-Bee]

Finaly *MASS is added