ANSYS negative pivot term

[Leibi]

Hello

now the VGLUE (my last post)command works. I mesh after gluing the volumes together.
By solving my simulation i have another problem.
In the output window are the following letters written (see attachement).
I do not understand this becaus i constrained my model in all degrees of freedom.

Has someone an idea where the mistake is?

Thanks a lot!!!

 

[Leibi]

Here is the Output Window:

 

[Leibi]

One thing i have forgotten: if i execute the List-> Nodes command only nodes until Numer 20216 are shown but the negative pivot term is at node 30338.

I don't understand that.

Anybody?

Thanks a lot!

 

[Drej]

It could well be that your VGLUE command didn't work and/or that you have unconnected regions in your model. What type of analysis are you running? If your mesh is similar between connecting parts/regions where you performed the VGLUE (ie the node positions match) then you may be able to use the NUMMRG command to merge the nodes.

Personally, I would first perform a modal analysis on the model (without any loads) using the boundary conditions as shown in your picture and check for rigid body modes (check for very small or zero frequencies). Check/animate the modes also. If you see very small frequencies then the parts are likely unconnected (check the modes to confirm).


------------
See faq569-1083 for details on how to make best use of Eng-Tips.com

 

[Leibi]

Hi

i'm running a static strucutral analysis.
Unfortunately i have no experience with modal analysis and my benches fail.

 

[Drej]

Have you checked whether the VGLUE command actually worked i.e. is the mesh continuous and are the nodes at the VGLUEd interface shared or are they separate? Plot the node numbers at this region (and any others) and check for duplicates.


------------
See faq569-1083 for details on how to make best use of Eng-Tips.com

 

[Leibi]

Hi,
here's a picture of the mesh with the three glued volumes in the background. I glued before meshing.
By meshing with MeshTool i use the following settings:
Element Atrributes: Global
Smart Size: On
Mesh: Areas
Shape: Tri
Free: On

Then i pick one area and go to PickAll an Mesh the Volume. I use the Shell 99 Element for Composites (CFRP-Tube)

If i check the mesh, ANSYS tells me that there are no error elements to plot.

Thanks

 

[Leibi]

To the nodes: here i plotted all nodes first sorted by the z- coordinate (at the interfaces at z = 30 and z = 230) and then sorted second by x and third by y.

In my opinion there are no double elements. What do you think?

 

[Drej]

Output the model by issuing the CDWRITE command. Type CDWRITE,COMB,output,inp in ANSYS or go to Main Menu>Preprocessor>Archive Model>Write and choose COMB and then a filename such as "output.inp". This will output the model in ASCII format. Then upload this ("output.inp") file.


------------
See faq569-1083 for details on how to make best use of Eng-Tips.com

 

[Leibi]

Here is the file.

Thanks!

 

[Drej]

I ran your model and it solved without any problems in v8.1. However, there were warnings for every element stating "possibly inaccurate interlaminar shear stresses. Consider using more elements in this region."

Your model has a strange cross section considering you are using shells:

|-----------|
|-----------|
| |
| |
|-----------|
|-----------|

Should this be:

|-----------|
| |
| |
| |
| |
|-----------|

i.e. a hollow thin-walled cylinder?


------------
See faq569-1083 for details on how to make best use of Eng-Tips.com

 

[Leibi]

Thanks

that means it would be better if i create not the whole volume. I should better create a a hollow cyclinder with a wall thickness of about 0, so that it's like an curved area and mesh this area with the shell element.

Thanks

 

[Leibi]

Is that right what I meant?

Thanks

 

[Leibi]

Hi
now my simulation is running. I was on the right way but my ANSYS version has a mistake; on another PC it works!

Thanks for your help!