FEM analysis : ccx_dat_file empty

About the development of the FEM module/workbench.

Moderator: bernd

User avatar
vkt
Posts: 5
Joined: Sun Dec 18, 2022 5:48 pm

Re: FEM analysis : ccx_dat_file empty

Post by vkt »

I have been experimenting and found a method to use element-based rather than node-based surfaces to get heat flux output to the .dat file. In my thermal problem, I have convection applied to surfaces selected in FEM. I then edit the .inp file (remembering to use control+S to save the file after editing in the SolverCcx tools) to copy the *FILM section from the history data section into the model data section. The *FILM defines the surface by element numbers, faces, sink temperature and HTC on separate lines so this can be edited to give an element-based surface definition by stripping off the last entries on each line and changing F to S:

*FILM
** Heat flux on face Pad:Face5
450,F4,303,0.005
490,F2,303,0.005 …….

becomes:

*SURFACE,NAME=Face1,TYPE=ELEMENT
450,S4
490,S2 …….

And this surafce can be used with *Section Print to get the total flux on that surface:

*SECTION PRINT,NAME=Face1,SURFACE=Face1
FLUX

This seems to work correctly and I should be able to write a script to do the edit for surfaces with lots of element faces.

I am still a bit puzzled by the units used in Calculix. From the comments (listed below) at the end of my .inp file, I see that the thermal analysis is using t (tonne) mm units which I have converted to SI to do a benchmark check (which worked OK). But the mechanical material properties appear to be in kg mm units (32 GPa became 32000 in the .inp which must be in kg/(s^2.mm)).

** The unit system which is used at Guido Dhondt's company: mm, N, s, K
** Since Length and Mass are connected by Force, if Length is mm the Mass is in t to get N

I would appreciate any suggestion on how to use SI units throughout.
Many thanks.
Post Reply