NIC Beam Elements Support Torsional Modes?

[markdeckerZBQL7]

Running a NAFEMS benchmark for beam element modal simulation.  The torsional modes (#3 and #7) are skipped in NIC.  Running the same setup in Sim Mechanical and Abaqus produces the correct results.

[Andrew.Sartorelli]

Hi @markdeckerZBQL7,

With default settings I was seeing the same behavior as you. On the Modal Setup I changed the Mass Representation to Coupled, after which I did see the torsional modes. Although the torsional frequencies were differing by about 10%. I'll see if we can get that even closer.

Best Regards,

Andrew

[Andrew.Sartorelli]

Hi @markdeckerZBQL7,

Can you share which NAFEMS benchmark set you are using and the number of this specific benchmark problem? I know there are a number of different NAFEMS benchmark sets that have been published and I'd like to have the QA team investigate this a bit further. None of my models were able to get the torsional frequencies to match with what you've shown, regardless of the additional parameters I've used.

Best Regards,

Andrew

[markdeckerZBQL7]

Interesting.  Sim Mechanical produces a value of 71.19 (for a target of 71.2) for that first torsional mode.

[markdeckerZBQL7]

NAFEMS "Selected Bencmarks for Forced Vibration" R0016 March 1993 Test #5

[Andrew.Sartorelli]

@markdeckerZBQL7

What were the results with the Standard solver versus the Nastran solver in Simulation Mechanical? If those match up, then it may be something with the In-CAD pre-processor rather than Nastran.

Andrew

[markdeckerZBQL7]

I used the standard Algor solver in Sim Mech.  I went back and tried the Nastran solver but the default settings also skip the torsional modes.  I did not see any options in the dialog so I guess one would have to edit the Nastran input deck to throw that option for mass coupling?

[Andrew.Sartorelli]

Hi @markdeckerZBQL7,

If you pop it open in the Editor, try adding PARAM,COUPMASS,ON after the BEGINBULK statement and then running it.

Andrew

[Andrew.Sartorelli]

Hi @markdeckerZBQL7,

I've found that we are using what appears to be the same problem as a verification problem, 5.1.4, in the Autodesk Nastran Verification Manual. However, the results provided in "Selected Benchmarks for Natural Frequency Analysis,1987" are different from those provided in your version of the benchmark. Siemens is also using the same problem in the NX verification manual and has the same answers that we are providing.

Regardless of the differences in frequencies provided by NAFEMS for the torsional modes, we are not seeing those modes with some solver settings. This issue has been reported to the QA team for further investigation.

Regards,

Andrew

______________________________________________________________

If my post answers your question, please click the "Accept as Solution" button. This helps everyone find answers more quickly!