[Solved] two-body part (mesh seperately) FEA error

About the development of the FEM module/workbench.

Moderator: bernd

john huang
Posts: 20
Joined: Tue Sep 27, 2022 9:44 am

[Solved] two-body part (mesh seperately) FEA error

Post by john huang »

Hi,

First thank you so much for your attention and help!

I am novice with limited FreeCAD FEA experience. I just finished a single body (a block only) FEA and saw the results. Soon I got stuck when I went a little further - to do FEA on a part of two bodies.

Here is my case:
A part is consisted of a block and a cylinder with a ConstraintTie set between the contacted faces of them. A ConstraintFixed is set on the left face of the block, and a ConstraintForce is set on the top face of the cylinder. Material is set as steel (not important). Both of the bodies have been meshed by Gmesh seperately. Now when I double-click the 'SolverCcxTools001' and click the button 'Write .inp files' as usual, pops up this error: no mesh object defined in the analysis, while we obviously have had two mesh objects defined.

Please let me know where I did wrong and how I could do this analysis properly.

BTW, is it possible to mesh these two bodies at the same time? If not, then what difference it makes that the two bodies belong to same one part? Or maybe should I merge the two bodies into one single body and thus reducing the two-body analysis to a single-body analysis?

regards
Attachments
0_description.PNG
0_description.PNG (265.66 KiB) Viewed 976 times
1_error_popup.PNG
1_error_popup.PNG (253.8 KiB) Viewed 976 times
Last edited by john huang on Wed Oct 19, 2022 3:07 am, edited 1 time in total.
User avatar
NewJoker
Veteran
Posts: 1490
Joined: Sun Oct 11, 2020 7:49 pm

Re: two-body part (mesh seperately) FEA error

Post by NewJoker »

You have to apply Part Fuse, Part Compound or similar boolean tool to this model and then create the mesh. Unfortunately, multiple meshes are not supported at the moment.
john huang
Posts: 20
Joined: Tue Sep 27, 2022 9:44 am

Re: two-body part (mesh seperately) FEA error

Post by john huang »

NewJoker wrote: Wed Sep 28, 2022 10:50 am You have to apply Part Fuse, Part Compound or similar boolean tool to this model and then create the mesh. Unfortunately, multiple meshes are not supported at the moment.
Thank you very much @NewJoker !
I tried Part_Fuse and Part_Compound, and here are what I learned:

1. for Part_Fuse, it works! It helped to reduce it to the single-body analysis. HOWEVER, it also means we could ONLY do this analysis with the whole fusion being assigned SAME material. In order to assign different materials to them, I did another test. I tried assigning different materials in FEM module BEFORE fusion, and then back to Part module and do Part_Fuse, making a fusion with different materials. Then when I went to FEM module again and tried meshing, an error occurred: The shapes for the mesh group for the reference shapes of analysis member: MaterialSolid001 could not be found , which implied that we had lost the second material after Part_Fuse, though that material was still shown in the model tree.

2. for Part_Compound, I encountered another error after I prepared well every step (including defining the ConstraintTie with positive/zero/negtive different tolerances) and clicked 'run CalculiX': no nodes found in Frd file. I feel that it's possible to complete this different-material analysis by using Part_Compound, please let me know where I probably made mistakes. BTW what is this 'Frd file'?

regards
Attachments
3_error_index_out_of_range.PNG
3_error_index_out_of_range.PNG (191.34 KiB) Viewed 848 times
4_CalculiX_execute_error.PNG
4_CalculiX_execute_error.PNG (287.78 KiB) Viewed 848 times
User avatar
NewJoker
Veteran
Posts: 1490
Joined: Sun Oct 11, 2020 7:49 pm

Re: two-body part (mesh seperately) FEA error

Post by NewJoker »

Indeed, Part Compound can be used for analysis involving multiple materials. When it comes to the error message, can you share the file ?

Frd is CalculiX's field results file.
john huang
Posts: 20
Joined: Tue Sep 27, 2022 9:44 am

Re: two-body part (mesh seperately) FEA error

Post by john huang »

NewJoker wrote: Thu Sep 29, 2022 10:51 am Indeed, Part Compound can be used for analysis involving multiple materials. When it comes to the error message, can you share the file ?
Of course here are the files that analysis generated. I guess you want the .inp file, but this extension file couldn't be attached, how should I send it to you?

As for the nodes missing tied contact, was it because that the tolerance value not appropriate?
Attachments
5_error_files.PNG
5_error_files.PNG (18.79 KiB) Viewed 786 times
User avatar
NewJoker
Veteran
Posts: 1490
Joined: Sun Oct 11, 2020 7:49 pm

Re: two-body part (mesh seperately) FEA error

Post by NewJoker »

john huang wrote: Fri Sep 30, 2022 3:48 am I guess you want the .inp file, but this extension file couldn't be attached, how should I send it to you?
Actually, FreeCAD’s project file (.FCStd) would be sufficient.

john huang wrote: Fri Sep 30, 2022 3:48 am As for the nodes missing tied contact, was it because that the tolerance value not appropriate?
Yes, it can be the case. But too high tolerance may stretch the elements too much and distort the mesh even causing errors.
john huang
Posts: 20
Joined: Tue Sep 27, 2022 9:44 am

Re: two-body part (mesh seperately) FEA error

Post by john huang »

NewJoker wrote: Fri Sep 30, 2022 7:22 am Actually, FreeCAD’s project file (.FCStd) would be sufficient.
Here is a sample project file. A rough mesh of this compound was used to make the project file as slim. And here I just set the tolerance 1mm, but some other values I tried also resulted the same error. I guess the appropriate value should be positive, right?
Thank you for your time again! @NewJoker
Attachments
two-material compound.FCStd
(149.52 KiB) Downloaded 7 times
User avatar
NewJoker
Veteran
Posts: 1490
Joined: Sun Oct 11, 2020 7:49 pm

Re: two-body part (mesh seperately) FEA error

Post by NewJoker »

It will work if you use Part BooleanFragments. Then you don't need tie constraint because the mesh is continuous (you can notice it if you look at its connectivity between both solids). In FreeCAD tie (or contact) is only necessary when there's a gap between the parts.

two-material compound - working.FCStd
(22.88 KiB) Downloaded 7 times

P.S. To reduce the size of the file you can clear the mesh but leave its settings so that it can be easily regenerated.
john huang
Posts: 20
Joined: Tue Sep 27, 2022 9:44 am

Re: two-body part (mesh seperately) FEA error

Post by john huang »

NewJoker wrote: Sun Oct 02, 2022 11:09 am It will work if you use Part BooleanFragments. Then you don't need tie constraint because the mesh is continuous (you can notice it if you look at its connectivity between both solids). In FreeCAD tie (or contact) is only necessary when there's a gap between the parts.


two-material compound - working.FCStd


P.S. To reduce the size of the file you can clear the mesh but leave its settings so that it can be easily regenerated.
It works! Thank you very much! @NewJoker
john huang
Posts: 20
Joined: Tue Sep 27, 2022 9:44 am

Re: two-body part (mesh seperately) FEA error

Post by john huang »

hi @NewJoker, I want to set this topic as 'solved', adding a [solved] before the headline, could you please show me the button?
Post Reply