To all ,
to anyone using the NX 'Response Simulation' Module
Dose anyone know the type of damping when onn specifies the 'Damping Factor" in the NX 'Response Simulation' Module (% viscous) in relation to the NASTRAN TABDMP1 card which as the option "G", "CRIT" or "Q" ?
Solved! Go to Solution.
Are you comparing results or performance? I'd be curious to hear your findings... If it's performance, make sure to include stress in solid elements, that's a huge bottleneck...
Not reached the "real" performance test yet. The test model I created is simple and made of shell elements. Running the nmodes is trivial so the overall analysis time is short. I am currently having a unit issue so the answer at a couple of nodes are different (between sol111 & Resp. sim).
Going through the .bdf to understand the nastran cards created from the GUI is a "pain" because it's seem so convoluted. Will need to dig out the random template I created 15years ago!!
I have noticed however that data is written to the .f06 file regradless of the PRINT or PLOT option (on the stress). Something I am doing wrong? I have noticed that on my test case the .f06 is large (~1.6Gb) when I requested STRESS=ALL. A bit silly I know but I wanted to know. If I recall there was nothin in the .op2 file
I have a real model for which I'd like to run a SOL111 with STRESS=ALL (or nearly ALL) just to see how bad it is
SOL111 (and SOL103 for SRS) is a convoluted mess in NASTRAN in general :-)
To limit the output size, I would not do the normal f06 output (although NXN introduced op2 output of "VM" RMS stresses), but rather XYOUT entries... It's not pretty, but works a bit better...
Couldn't agree more with you. but then MAYA never replied to my request for info on their tool
There is a way of not writing anything to the .f06 file and write all the rms stress to the .op2 (No XYOUT card here). Indeed from v10 one can get the random VM stress too using FREQVM (based on early-1990 method!)
performance test is running - a real model with stress=all (~200,000 elms). Getting the nmodes took about 25 minutes but Nastran is still computing the rms values - not looking too good
comparing results- - Based on my dummy/totally-made-up test case,
exiting in z direction, held in XY
damping (CRIT) 0.04 for all modes up to 2000Hz
Accel response of a node, see attached, is not too bad for Z response but not quite the same for X&Y.
Strange about MAYA, they are usually great with following up... We've been using SATK for several years and have not looked back (Link to our SATK success story).
I prefered XYOUT to STRESS(RMS) because the post-processors we used weren't supporting it or it wasn't clear what we were displaying... For instance, there still was a "divided by 2pi" issue in NX on importing zero crossings from op2...
Not bad! The difference might be the number of spectral lines in Response Sim... Now go ahead and take a big fat solid mesh and get stresses, I dare you :-)