random vibration environment analysis

[cavva78]

Hello everyone,
could anyone plase tell me what is the correct procedure to perform a structural analysis by FEA on an item subjucted to random vibration? I work in aerospace field and use NX.Nastran with Femap. I know the input PSD for this item and I would like to verify its strength. It includes some welded connections also: how are they usually modelled?

Thanks and regards

 

[GBor]

Random vibe is based on a natural frequency analysis, so you will need to run that first. Set your model up with your realistic boundary conditions. Input your PSD. There are some other considerations, but I'm not familiar enough with NX to know what to tell you...

Garland E. Borowski, PE
Engineering Manager
Star Aviation

 

[spongebob007]

You have to do a response simulation in NX. In order to do this, you need to have the NASTRAN response simulation license. Otherwise you are up the creek. NX also has the ability to perform the fatigue analysis.


The NX documentation should be able to give you all the information you need about setting up and running the models?

 

[cavva78]

Hello,

I don't use NX but Nastran with Femap as pre-post, with the licence for dynamic analysis. I am able to perform freq response and to apply PSD, but I would like to know how to consider the results, since I can have PSD of component stresses, and not Von Mises. Is there any standard procedure?

 

[spongebob007]

I don't know a whole lot about random vibration analysis, but don't you want to use RMS stress for fatigue evaluation? The reason I ask is that I have a textbook that has an example of determining the fatigue life of a cantilever beam subject to random vibration. The example uses the RMS bending stress to determine the fatigue life.

I know NASTRAN is capable of calculating the RMS VonMises Stress. I can also get X-Y plots of the PSD of VonMises stresses for an individual element.