Converting Sine on Rendom to Random only profile

[esector]

I have the following random profile:
4Hz – .0008 G^2/Hz
100Hz – 0.02 G^2/Hz
500Hz – 0.02 G^2/Hz
2000Hz – 0.0013 G^2/Hz
Sine tones at
F1=1.3G at 17Hz
F2=1.1G at 34Hz
F3=1.3G at 68Hz
F4=1.0G at 340Hz

I need to convert this sine on random profile to random only profile so that it can be compared against other random profiles.
Any help would be greatly appreciated.
Alex

 

[CESSNA1]

ESECTOR: I am not sure what you are asking. The random profile is the first set of data you have at 4,100,500, and 2000 Hz. Just remove the sine component. I assume this is for a shaker control system. A word of caution. The 4 Hz is on the ragged edge of capability for piezoelectric accelerometers. I have never seen a random profile go below 5 Hz. What do you mean by "compared against other random profiles"? What are you going to compare? This looks like a slightly modified NAVMAT 9492 Spectrum. The NAVMAT 94932 Spectrum is at 5, 80, 350, and 2000 Hz. The PSD for the 80 to 350 Hz ranges is
.04 G^2 rms/Hz. The slopes at each end of the spectrum are 3 Db/Oct.

Regards
Dave

 

[esector]

I have a system that was tested and approved under the following profile:
10Hz - 0.02 G^2/Hz
28Hz - 0.02 G^2/Hz
40Hz - 0.04 G^2/Hz
2000Hz - 0.04 G^2/Hz

I was asked to determine if this system would survive under “Sine on Random” profile listed in the first thread. So what I would like to do is to see which of the two profiles is more severe.
From my understanding I first have to make two profiles equivalent. To do that I have to convert “Sine on Random” to “Random” profile by converting Sine tones to equivalent [G^2/Hz] and then add it to existing PSD profile. Then I’ll be able to compare PSD levels between two profiles and see which one is higher in magnitude at a specific frequency.
Regards,
Alex

 

[tomirvine]

I would not convert the sine-on-random to an "equivalent" random.

Rather I would use a vibration response spectrum approach, with some thought given to histograms.

Tom Irvine

http://www.vibrationdata.com