sine over random inputs

[mneal]

I am trying to figure out how to perform a sine over random vibration input to a assembly. It will be used on a MIL helicopter (MIL 810) and it has a profile containing both g^s/Hz and Sine inputs. Typically if I have just a g^2/Hz input I do a static of the 3 sigma GRMS value, but with both inputs I am unsure how to input /treat it. Any advice from the experts? Thanks in advance.

 

[mch]

Technically it is incorrect to apply a random and a sine input, since one is random (stoachistic) and the other is repeatic (harmonic).


When I've been asked to do this type of analysis in the past I suggsted converting the random part of the input into a uniform acceleration over the requred frequency range, with the value of 3xRMS value of the random input, as you mentioned. (This is because the RMS value of the random analysis is at the 1-sigma level, and you need to make sure that the sine input is at the 3-sigma level or higher.) The background level & the sine 'spikes' can be input using a table function, and a dynamic frequency analysis carried out in the normal way.


Regards,


Rod Giles

 

[Shawzie9]

If you convert the random levels to a static load with a value at the3SigmaGrms, you are not considering the dynamic effects of the whole assembly at all - this simulation is not very characteristic of your real environment. Ideally you need to calculate all the modes of your assembly within the frequency range specified, then perform a spectrum analysis with the PSD input. The sine spikes could be converted to a random level using Miles' Equation. The calcs are found here:


www.[B]cots[/B]journalonline.com/ pdfs/2003/04/COTS04_test&screen2.pdf


If you plan to compare your analysis to results from the vibe table, you need to interrogate the software you are using with the vibe table and find out how IT is doing the conversion. That way you are sure you simulate the same amount of energy being input to your system as you claim to be.