Vibration and Shock Analysis

[Tegguy]

First I will appologize I'm far more a mechanical design engineer than a stress analysist at this point in my career so I'm probably going to phrase stuff wrong. I'm also not completing the analysis but am helping on the design side.

When a model analysis is competed you end up with a list of resonance frequencies. When you take that and compare it to the flown down Random vibe (Frequence and G^2/Hz) and shock requirements (G and frequency). I can understand wanting to try to keep your first few frequencies outside the peaks in the random vibration profile but I have a few questions

1) In random vibe is this correct thinking if I have a break point at 250hz where my PSD decreases should I push the structure/design to be above this frequency at the first mode?
2) In random vibe if I have a breakpoint at say 700hz but there isn't a huge change in PSD is there a reason not to have a structural mode at that point?
3) How many modes matter when looking at the modal analysis?
4) Does shock behave the same as random vibe for questions above?

if I have a 1500g shock @ 1800hz but also a 1500g shock @ 10000hz is it only those two points that matter or everything in between those as well?​

Sorry if this is a fundamental question just working with a newer team and trying to make sure we're making the right choices.

 

[FEAster_man]

1) For SDOF this is true. For more complex structures it depends. If that's a primary mode, it's probably true in vast majority of cases. But for example if your local mode that you pushed above 250Hz starts to couple with some other modes, now the overall response can be more severe (at least from the perspective of a secondary structure).
2) Not sure I understand the question.
3) Since we are in spacecraft forum the standard frequency range that we are interested in is 20-2000Hz. In this frequency range you'd like to have the total mass participation of at least 80%. If that's not the case you can use residual vectors to improve the accuracy of the dynamic solution. To improve the solution efficiency modes with modal participation factors lower than some threshold value can be omitted. They do not play a significant role in the overall results.
4) Shock is also a load case defined by a spectrum, and the response will depend on the natural frequencies of your component. So if the SRS is defined as
100 Hz - 15g
1800 Hz - 100g
10000 Hz - 100g
you are interested in the response from the frequency range 100-10000 Hz.

 

[GregLocock]

I suggest you head over to Tom Irvine's website since he almost literally wrote the book on this stuff.

https://www.vibrationdata.com

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?