Sine on Random Vibration Test Requirements Interpretation

[jimbod20]

I’m hoping contributors to this site can help me understand how to interpret sine on random test requirements.

I’m working design and qualification of a part which must be subject to vibration test per RTCA DO-160G section 8 category R (helicopter) ‘multi sine on noise test’. ‘Multi sine on noise’ is specific nomenclature from my customer specification. An outline of the test requirements are provided in the attachment.

DO-160 defines discrete test spectrum sinusoidal frequencies for sine on random test. The sinusoidal rays I have been given for test appear to be discrete “frequency windows” with different beginning and ending amplitudes. Ray 4 and 5 frequency windows overlap with a different initial direction. I cannot find examples of sinusoidal rays similar to defined above.

DO-160 Comparison:
• R2 is 2x frequency of R1 which is consistent with table 8-2a in DO-160G.
• The noise level of .02 g^2/Hz and .002 g^2/Hz appears to match endurance levels defined in table 8-2B and corresponding test data plot figure 8-6 in DO-160G.
• Test procedure denotes performance test for 30 minutes, endurance test minimum of 2 hours then 30 minutes of performance again at levels specified.
o Performance/endurance/performance test sequence is consistent with DO-160G. Random test levels appear to be endurance levels for both performance and endurance test phase.

Questions:
• Does Fmin and Fmax for each ray define a frequency range within which the magnitude of the sine amplitude varies from Fmin level to Fmax level?
o For example, Sinusoidal vibration is 3.2g’s at X Hz and ramps linearly to 4 g’s at 1.83*X hz? Rate at .4 octave/minute.
o The shake table would then transition to Ray 2. Sinusoidal vibration is 4g’s at 2X Hz and ramps linearly to 5 g’s at 3.66*X Hz? Rate at .4 octave/minute.
o And so on through Rays 3-5.


• Why does the spec provide 5 rays when DO-160 talks about 4 sinusoidal rays? Rays 4 and 5 have a similar frequency range and amplitude is identical. The distinction is the difference in initial direction. I don’t know what this means?
o Is the initial direction somehow telling me there is a phase relationship requirement for test.

 

[IRstuff]

"• Does Fmin and Fmax for each ray define a frequency range within which the magnitude of the sine amplitude varies from Fmin level to Fmax level? "
"Rate at .4 octave/minute." Doesn't that answer the first question?

However, I don't see anything you've written that implies that the "rays" are sequential in time. In general, sine tones are all there at the same time, particularly since some of the rays are clearly harmonically related.

"• Why does the spec provide 5 rays when DO-160 talks about 4 sinusoidal rays? Rays 4 and 5 have a similar frequency range and amplitude is identical."
The customer's spec supersedes DO-160. They may have other things that are vibrating. DO-160 is a general specification, and cannot possibly account for all variations of all aircraft.

"The distinction is the difference in initial direction. I don’t know what this means?"

The direction, I presume, is telling you which way to ramp the frequency.



TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

Of course I can. I can do anything. I can do absolutely anything. I'm an expert!

 

[jimbod20]

Thanks IRstuff.
I hope to gain knowledge/insight here before I go back to discuss/confirm with my customer. Please note, I have limited experience with vibration test.

The response regarding direction seems reasonable. Maybe direction and speed dictate 'sweeps' through the rays vs four discrete sinusoidal vibrations for duration of test as outlined in DO-160. As noted, a custom set of requirements.

I have come to realize that I am struggling with what the test process might look like. Maybe others have seen something similar to what I outline here?

If the speculation of direction and speed is correct, does the following seem reasonable or make sense?

Test notes/outline:
"Initial Direction" implies Rays 1, 2, 3 and 4 ramp up in frequency? Ray 5 ramps down in frequency at the same time rays 1-4 ramp up.
"speed" implies/ramp rate for all five rays? The result is random plus 5 discrete sinusoidal vibrations at a point in time? This might be similar to DO-160.

If the notes/outline seem reasonable, Is there an industry practice/guideline regarding how many discrete frequencies are selected for each ray? Maybe I am stuck on discrete frequencies?

 

[IRstuff]

That would be my interpretation. "Discrete frequency" is a mythical thing. Most tones assume a VERY specific operating condition of the engine, i.e., a very specific RPM. However, most aircraft fly only a small portion of their flights at a fixed RPM, and with modern autopilots, the flight computer may be adjusting the RPMs as required.

As for your specific question, the general practice is to use the DO-160 or MIL-STD-810 frequencies, if no measured data is available, or, real data is collected and some sort of envelope is developed for both the random and sine tones. If one considers the entire flight envelope from taxi to takeoff to flight to landing to taxi, we should get all sorts of tones, depending on the RPM of the engine, and how the airframe is stressed and what vibration modes there might be.

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

Of course I can. I can do anything. I can do absolutely anything. I'm an expert!

 

[jimbod20]

I have studied MIL-STD-810/DO-160 and recognize the standards practice is to identify 'discrete frequency' sinusoidal rays which coincide with flight vehicle excitation sources. That being said, the interpretation of my customers specification seems reasonable. Flight vehicle surveyed sine tones are beyond 4 discrete frequencies

I have searched for and studied vibration test controller specifications. I see reference to both 'swept' and 'fixed tone' sine control in sine on random control modes. Swept sine control seems consistent with interpretation of my customers spec.

If the controller produces a swept sine ray, some finite time is required for the ray to occur. For example, the period of a 25 hz sinusoidal ray (toward the lower end of the test spectrum) would be .040 seconds. Time is required to produce and maintain each frequency. If my speculation is correct, the controller would need to establish some number of discrete rays/frequencies within the test defined 'frequency ray window'. Maybe a function of the frequency window size and speed sweep rate. Do you know if this outline is reasonable and consistent with test controllers capability?

 

[IRstuff]

They gave you the sweep time: 0.4 octave/min You just need to give that to the operator, and the controller will take care of all of that.

So, no, there is no discreteness; it's a frequency chirp

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

Of course I can. I can do anything. I can do absolutely anything. I'm an expert!