Rigid Element Passing Through Solid doesn't solve

[Inv_kaos]

Say I create a new part and join a series of nodes on bolt holes to a centroid with rigid elements. I then apply a torsional moment to this central node. These rigid elements usually pass through solid brick elements and has always worked in the 2012 version but now my analysis doesn't finish with the same setup in 2013. I don't know if it is because the elements are passing through the other bricks or not but if I suppress the part with the rigid elements it solves. In 2013 when it stops the log file gives no error messages. I have just tested the exact same part in both versions and can confirm it only happens in 2013. Is this a known issue or a change in functionality?

Cheers.

Stew

[John_Holtz]

Hi Stew,

As far as I know, it is neither a known issue or a change in functionality. Can you provide an archive of your model? See "Create, Post, or Provide an Archive of your model".

In the meantime, you can probably change the element type from rigid to beam and proceed with the analysis.

[Inv_kaos]

Hi John,

The beam option worked on this file but still failed on a previous file. I also tried local centroids with spokes at each bolt hole then used 8 rigid spokes to the main centroid in another part - still didn't work, failed on the beam elements at the local centroids. I was hoping the reboot this morning would have fixed it but get the same issue.

I would think Autodesk would allow *.ach file to be uploaded without having to be ziped first, I just waited for a 95MB and a 35MB file to upload then be told it is not a supported format. Anyway I will try to make up a smaller sample file that has the same issues.

Cheers

[Inv_kaos]

Clearly I didn't ready you tip on posting archive files before I posted the last message - good tip.

I can't create a suitable example so I have uploaded the original file in the 2013 format.

http://www.sendspace.com/file/sb08kz

Cheers.

[Inv_kaos]

Hi John, Did you manage to have a look at this file? I have had this happen again in a similar set-up but this time the part is more critical. I have had to redo it all in the 2012 release costing significant delays in an urgent project. Can you please advise if this has been logged and if you know when an update or service pack will be released? Cheers. Stew

[John_Holtz]

Hi Stew,

I just took a look at your model. Strangely, I made a coarser mesh and was able to run it in 2013. But I also tried your original mesh and it failed. I presume it is the same type of failure that you encountered. My guess is that it is related to the amount of memory available.

• So a coarser mesh and setting the "Setup > Model Setup > Parameters > Solution > Percent memory allocation" to a larger amount may get the model to run for you.
• You could also try to schedule an analysis so that you can close the interface and free-up additional memory before the analysis runs.

I also tested your original fine-mesh model with the upcoming service pack, and it ran. No one at Autodesk, including me,  can say for sure that the service pack will be released next week as currently scheduled. Also, there is always a (slim?) chance that "the fix" will be removed from the release if testing discovers something. I will keep my fingers crossed. (Does that mean good luck around the world like it does here in the United States?)

P.S. I assume that you saw the warning from the analysis that the moment was dropped because the rigid elements were not setup to include the rotation about X degree of freedom (DOF), and you have corrected it in your model.

[Inv_kaos]

Thanks John, Since I can run the same mesh size in 2012 I wouldn't have thought it is purely a memory issue unless it only affects 2013. However it would make sense that I could not get it to repeatedly fail with every model, I will check the memory allocation and let you know if it fixes it. Hopefully the SP will inadvertently fix the issue. Yeah I noticed the moment wasn't applied when I checked my results, changed the definition and re-ran. Cheers.