Discrepancy in Results between ccx 2.13 and ccx 2.17 Versions

1

Hello,

When I run cube inp file with ccx 2.13, I get the correct results. However, when I try the same file with ccx 2.17, the results differ. What might be causing this discrepancy? Additionally, I’m unable to download ccx 2.13 on Ubuntu. How can I resolve this issue?

ccx2.13

MicrosoftTeams-image (3)
MicrosoftTeams-image (3)901×732 83.2 KB

ccx2.17

MicrosoftTeams-image (5)
MicrosoftTeams-image (5)802×479 59.3 KB

Thanks,
Bhavita

2

Is it some publicly available file or just your own model ? In the latter case, you should share it so that we can have a look at the issue.

3

Hello @Calc_em
Here is link for inp file.
https://drive.google.com/drive/folders/1vbtk4_fmeJjwLdYtRoyHgMhkt0PrSXmH?usp=sharing

4

When I submit it using the newest CalculiX version (2.21), I get the following error:

 *ERROR in checktemp: no final temperature
        defined in node           15

meaning that you should add the temperature values for all the remaining nodes in the model under *TEMPERATURE.

5

I have already used temperature card for all nodes is model like this
*TEMPERATURE
Part-2, 201

*MODEL CHANGE, TYPE=ELEMENT, ADD
Part-2

** Outputs → frd file
*NODE FILE
U
*EL FILE
S, E, PEEQ
*NODE PRINT, NSET=ConstraintFixed, TOTALS=ONLY
RF

*END STEP
**----------------------------------------------------

**------------------------------------------------------------
*STEP, NLGEOM
*STATIC

*BOUNDARY, OP=NEW
Set-Ground, 1
Set-Ground, 2
Set-Ground, 3
Set-TopLayer-3, 1
Set-TopLayer-3, 2
Set-TopLayer-3, 3

*TEMPERATURE
Part-3, 201

6

The error points to the first step - you should define temperatures for all nodes already there. They can be just equal to the initial temperature in some regions if you don’t want them to undergo thermal straining.

7

@Calc_em Are you saying like this

*STEP
*STATIC

*TEMPERATURE
NALL,200

*NODE PRINT, NSET=Nall
U
*EL PRINT, ELSET=Set-all
S
*END STEP

**------------------------------------------------------------

**------------------------------------------------------------
*STEP, NLGEOM
*STATIC

*BOUNDARY, OP=NEW
Set-Ground, 1
Set-Ground, 2
Set-Ground, 3

*TEMPERATURE
Part-1, 201

*MODEL CHANGE, TYPE=ELEMENT, REMOVE
Part-2
Part-3
Part-4
Part-5
Part-6
Part-7
Part-8
Part-9
Part-10
Part-11

8

Yes, it can be done this way.

9

still incorrect results like earlier using ccx2.17

10

Try without *MODEL CHANGE first. If it works as expected then introduce model change to the model. It’s best to build complexity gradually.

11

Thank you @Calc_em, It’s working in ccx2.17 after removing model change card