What is the material?
What is keeping you from using a more conservative strength?
I seem to recall NASA using a 1.4 ultimate factor. But you should verify this for your component and customer. Surely you are working to a design spec?

Maybe there're two ways at looking at random vibration ...

can the material withstand the load as a static load, including dynamic factors ?
then can the structure withstand the necessary time interval (of random loading, to some PSD) ?

Admittedly the first seems practically useless, but I can see some point to the question.

ultimate factors ... depends on how much you tet and analyze and know the launch vibration loads. 1.15 (geeze, a fitting factor) seems low ... I've heard military will use 1.33. 1.50 is typical of commercial aviation, but may be too expensive for space

another day in paradise, or is paradise one day closer ?

Random vibe isn't typically an issue with strength of material, per se, isn't it? Isn't it more of an issue of modal excitations? The modal excitations loosen screws and bolts and deteriorate electronics.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert! https://www.youtube.com/watch?v=BKorP55Aqvg
FAQ731-376: Eng-Tips.com Forum Policies forum1529: Translation Assistance for Engineers Entire Forum list http://www.eng-tips.com/forumlist.cfm

satellites are often designed by the launch vibration ... why they test on a shaker table

another day in paradise, or is paradise one day closer ?

Thanks a lot for the overwhelming response!

The material is aluminum. The safety factor looks reasonable after checking with the launcher specification and standards.

Now I wonder should I use yield strength for random vibration if anyone has the experience to share? :)

guessing you don't want plasticity at those load levels ?

another day in paradise, or is paradise one day closer ?

@rb1957, yes, no plasticity

and hence yield strength as an allowable

another day in paradise, or is paradise one day closer ?

As I see random vibration is more like a durability than a strength check, using yield strength seems anti conservative, right?

see my first post

another day in paradise, or is paradise one day closer ?

Dear Sarclee,
in random vibration you get RMS stress plot from analysis (e.g. by Patran Random utility, that postprocesses Nastran SOL111 FRF).
Maximum RMS stress value has to be multiplied by 3 to get the so called 3 sigma value.
In this way you are confident that in 99.7 % of cases that stress value shall not be exceeded.
Then your customer has surely specified a yield margin of safety defined as (material yield stress)/(SFy x 3 RMS stress)-1.
SFy is the product of includes various factors that depend from project specifications or general regulations, for example:
SF qualification_yield: 1.1
SF modelling: 1.1
Analysis SFs: 1
Overall SF Yield: 1.21

I remark it again: the above factors depend on project specifications or applicable regulations (NASA? ECSS? It depends on the project).

Regards,
Stefanix1972.

@rb1957, I am not using static load...

Thanks stefanix1972! Just to clarify if the rms stress I obtained is at 3 sigma, I don't have to multiply by three?

MoS = Yield/ (SF*3*RMS@3sigma) - 1

Is it correct?

you're using a static allowable, so you're using a quasi-static load (like 3 sigma) to represent the loading spectrum.

there is another calc to be done ... to compare the spectrum loading with the fatigue allowable (s/N curve).

another day in paradise, or is paradise one day closer ?

Dear Sarclee,
by random analysis, if you input nominal PSD (e.g. in acceleration expressed in m/sec2) you get the RMSstress value in N/m2.

This RMSstress value has to be multiplied by three:

MoS = Yield/ (SF*3*RMSstress) - 1

In synthesis, be careful not to multiply twice by three.
You also need to pay attention to PSD curve units: if it is in g-units, RMS stress value has to be multiplied by 9.81 m/sec2 to get N/m2.

Regards,
Stefanix1972.

@Stefanix1972

Thanks. That is what I concerned if I take 3 sigma stress, should I still take the multiplier of 3?

MoS = Yield/ (SF*3*RMSstress) - 1

I was told, not space industry, that random vibration is a durability event, either we use fatigue strength in calculating MoS or we use Miner's rule to do cumulative damage ...

But based on the comments above, it seems the space industry is using yield strength in calculating MoS for Random Vib...

Dear Sarclee,
you are right: in space industry random vibration is relevant in the 120 seconds from launch onwards, so there is no real concern for fatigue due to random vibration. There is a minor concern for fatigue for the effects of ground tests, transportation loads, launch loads.
Thermal loads can be assessed for fatigue effects if you are dealing with a spacecraft in orbit around a planet.
In aeronautics, automotive, etc, of course random fatigue has to be taken into account.
I have not a specific background on this, but I would check on Steinberg and MIL-STD-810 last versions.

Regards,
Stefanix1972.

Thanks Stefanix1972 for clearing my doubt!