Siemens NX Nastran - Random analysis (PSD curve)

[filipadamek]

Hi all,

I would like to ask something about random analysis. I´m solving a random analysis problem in NX Nastran. It's going about PSD units. I went through these posts:

http://www.eng-tips.com/viewthread.cfm?qid=365454

http://www.eng-tips.com/viewthread.cfm?qid=345410

https://community.plm.automation.si...andom-Response-Analysis-ASD-Units/td-p/407823

and many more...

But just to clarify. In my results I still get quite small RMS stress and I don't believe it... If I want the results in G, I will apply the PSD loading curve in G^2/Hz. Then similiarly, to get the results in Pa, PSD in (m/s^2)^2/Hz; to get the results in MPa, PSD in (mm/s^2)^2/Hz. Am I right? So If I have for instance PSD value 0.1 G^2/Hz and I want the results in MPa, I will simply convert it to (mm/s^2)^2/Hz => I will get 0.1*(9810*9810)=9623610 (mm/s^2)^2/Hz.

Also I want to ask you one more thing. The pre calculation for the random analysis is the harmonic analysis with the unit loading. Does it mean that If I want it in: G units => applying the unit loading of 1 G
Pascals => applying the unit loading of 1 m/s^2
Megapascals => applying the unit loading of 1 mm/s^2

Is it OK or Am I having something wrong?

Thank you for your answers.

Filip

 

[JBlack68]

as long as you work in consistent unit there is no issue
Are you using NX sol111 or the Response Simulation module?
If you use mm and mass = tonnes (Mg), a psd in g^2/Hz then you stress should be in MPa^2/Hz, disp in mm^2/Hz, accel in (mm/s^2)/Hz etc

 

[filipadamek]

I use SOL111. I can post here the input *dat file after weekend... First time I have used: tonnes,mm, unit loading SPCD = 9810 mm/s^2, PSD in G^2/Hz => but I got very small RMS stress (something like 0.0001... MPa). Then I've tried: tonnes,mm, unit loading SPCD = 1 mm/s^2, PSD in (mm/s^2)^2/Hz => higher RMS stress values (0.01... MPa). But still I think there is an issue with units...

Filip