Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

Register Log in
  • Congratulations cowski on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Proving hazardous relative motion <1E-9 3

Status
Not open for further replies.
tb2944

tb2944

Aerospace
Sep 29, 2016
5
Hi All,

This concerns relative motion between an aircraft system and the local structure, during a crash, which could cause a hazard.

This requires an explicit demonstration by a combination of test, analysis and experience that the probability of the hazard occurring is < 1E-9. I will therefore need to attribute a probability to the chance that there is hazardous relative motion.

Any ideas how a test could provide that level of confidence? A simple drop test using critical loading conditions? Or apply a factor on the loading to take account of scatter, providing a confidence level?

Alternatively, is there a suitable analysis method?

Thank you
 
Replies continue below

Recommended for you

Sort by date Sort by votes
tb2944,

What could cause the relative motion? Would this be your structure tearing loose during an otherwise survivable crash?

Probabilities sound like a challenge. I am aware of all sorts of yield and ultimate stresses, but I do not have Gaussian curves of them, showing me [&sigma;] and all that. I would also need Gaussian crash decelerations, in all possible directions.

If you not trying desperately to save every last gram of mass, why not just pick a survivable deceleration and a fairly generous safety factor. [pre][/pre]

--
JHG
 
Upvote 0 Downvote
Disclaimer: This isn't my area.

"...during a crash, which could cause a hazard."

I've seen information that states that there's a maximum g-load assumption that's considered survivable. Beyond that point, there's no concern about hazards because it won't bother anyone anyway.

There should be a formal requirement written down which states what the maximum g-loads in various conditions and directions. Then you can work from the one most-applicable g-loading number.

Maybe. Wait for others to confirm. Not my area.

 
Upvote 0 Downvote
Typically, there is a crash shock specification that allows you to do this sort of calculation. You can then assume a 0.1-s duration, and the safety factor will determine the probability. There are articles on the web that discuss the statistical nature of the yield strengths, etc.:


Your requirement seems ludicrous, on the face of it, given that whatever crash shock value you use is probabilistic uncertain already. Likewise, the crash duration is rather arbitrary as well. Crashing into a mountain is different than crashing in a forest, or crashing into the ocean.

In any case, your 10^-9 probability requires at least 6-σ above mean

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert! faq731-376 forum1529
 
Upvote 0 Downvote
tb2944...

Exactly what specification, regulation, etc are You talking about?

FAR? MIL-STD? JAR?

Regards, Wil Taylor

o Trust - But Verify!
o We believe to be true what we prefer to be true. [Unknown]
o For those who believe, no proof is required; for those who cannot believe, no proof is possible. [variation,Stuart Chase]
o Unfortunately, in science what You 'believe' is irrelevant. ["Orion", Homebuiltairplanes.com forum]
 
Upvote 0 Downvote
Thanks for the suggestions.

I will take survivable crash load factors from the FAR.

The difficult bit is to argue a probability of failure based on (maybe) crash load distribution and material allowables. Clearly the published ultimate crash load factors already contain a safety factor (1.5?) but I'd need more info on how this was derived to build a case that I achieve 1E-9 overall. I dimly remember that 1.5 is supposed to broadly relate to 1E-3 or 3sigma, so that could be a start.
@IRStuff - I couldn't find any more on the crash shock specification.

Regarding the next step, we plan to use a drop test to show that the relative movement of structure and systems in a particular area - following application of crash loading - does not exceed a given level.

Any ideas on an appropriate test factor to use? ie, to help argue (to 1E-9) that this level of movement is not exceeded?

 
Upvote 0 Downvote
Generally, crash shock is usually done by analysis, at least for most military systems, since test is likely to damage the unit under test beyond repairability and there's usually no requirement that the system operates after such an event. Nevertheless, you should still be able to determine some value for the variance of the critical strengths involved and by applying a 6-σ bound, you should be able to convince someone that you are going to meet the requirement.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert! faq731-376 forum1529
 
Upvote 0 Downvote
Where did this requirement come from? They will be able to provide a procedure for figuring this out. Or they have no idea what they are asking for.
 
Upvote 0 Downvote
There are probabilities and statistics involved in setting the airworthiness standards, but I don't believe that one drags the probabilities along into each structural calculation.

I had to review a 'telephone book thick' structural analysis report, and they didn't incorporate probabilities into each calculation. They simply chooses the worst case at every step of the calculation. This stacks up additional safety factors along the way. At the end, they show that the structural failure in question can't happen, at all, in the case studied.

If the failure ever actually did happen, then it would be due to some other external factor. Not because of some statistical variation.



 
Upvote 0 Downvote
TB2944,
You may be making your problem more difficult than it needs to be. You are being too vague for us to help you.
If you refer to "FAR" requirements you may be talking about reliability requirements such as FAR 25.1309 which have NOTHING to do with structural requirements.
If you mean dynamic loading such as seat structures or maybe landing gear impact, you are dealing with energy absorption, not reliability or statistics.

Before embarking on any new design, it is crucial to understand the requirements to be met, otherwise a lot of time and effort will be wasted proving things that don't matter.

Since there is nothing secret about the Federal Aviation Regulations (all published online) or the guidance documents which support the testing and analysis needed to show compliance with these regulations (also available online), why not be more specific and tell us what you're doing, and what regulation you need to satisfy?

STF
 
Upvote 0 Downvote
The numerical value presented is inconsistent with typical reliability requirements, which are usually expressed as mean time to (or between) failure with values in the thousands of hours. Failure rates for most aviation equipment would be expressed as failures per millions of hours. Only space equipment would have failure rates in the per billions of hours, and even then, the "per billion hours" is abstracted to FITs of "failures in time," so a satellite rated processor might have a failure rate of 250 FITs or 250 failures per billion hours.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert! faq731-376 forum1529
 
Upvote 0 Downvote
can we start at the beginning ... where is this requirement coming from ?

The only thing I can remember about moving bits and stationary bits, like flight controls passing through structure, where there is a nominal separation and you need to consider the effect of structural deformation (though i'd've thought that was for limit loads).

another day in paradise, or is paradise one day closer ?
 
Upvote 0 Downvote
"Extremely Improbable" is defined as 1 x 10-9.

Start with Wiki: "AC 25.1309–1 is an FAA Advisory Circular (AC) (Subject: System Design and Analysis) that describes acceptable means for showing compliance with the airworthiness requirements of § 25.1309 of the Federal Aviation Regulations..."

The actual ref (PDF): FAA AC 25.1309-1.PDF. Middle of Page 5.

As far as I know, one doesn't drag the probabilities along into structural calculations. It's Newtonian physics, not quantum mechanics. :)


 
Upvote 0 Downvote
as a famous "engineer" once said "ya cannay change the laws of physics"

another day in paradise, or is paradise one day closer ?
 
Upvote 0 Downvote
I think the OP is over-reading the requirement. The usage of "extremely improbable" is directed to critical or essential systems required for safe flight or safe landing. Therefore "during crash" the plane cannot possibly be doing safe flight, nor is it capable of a safe landing, so the "extremely improbable" requirement cannot possibly hold sway.

This would therefore mean that the equipment in question must simply be designed with normal safety margins against a normal crash shock scenario, so that it's unlikely to break loose during a crash that might be otherwise survivable. The rationale is to ensure that a survivor of the crash itself doesn't get killed by equipment that's broken loose.

This is why we're told to have our stuff either in overhead compartments or under our seats in a plane. But, the constraints on our carry-ons are fairly loose, so they simply want them to be unlikely to kill someone that might have survived the crash.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert! faq731-376 forum1529
 
Upvote 0 Downvote
IRstuff,
I think it's fun to guess at the OP's intentions, too, but until the OP clarifies his/her problem we'll still just be guessing.

STF
 
Upvote 0 Downvote
Not necessarily directly applicable but, there are 'standard failure rates' for certain common components that get used in FMEA/FMECA etc. I used to work on stuff that hang off aircraft and the detail FMECA that was outsourced to safety consultant had probabilities assigned for most components etc.

Posting guidelines faq731-376 (probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​
 
Upvote 0 Downvote
True, but "failure rate" by definition, implies a classical exponential, constant failure rate, distribution, which is not applicable for crash shock situation.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert! faq731-376 forum1529
 
Upvote 0 Downvote
It was 12 years+ ago and I forget details but I thought one of the failure modes allowed for over stress during crash and that there were applicable failure rates.

Posting guidelines faq731-376 (probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​
 
Upvote 0 Downvote
Anything's possible, I suppose, but typically, reliability calculations deal strictly with random failures, and they can be accelerated with temperature or vibration, but that's pretty much it.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert! faq731-376 forum1529
 
Upvote 0 Downvote
Status
Not open for further replies.

Similar threads

manees90
  • Locked
  • Question
How to Mesh Surfaces with 90-Degree Bends Using Shell Elements to Minimize Artificial Stresses
Replies
1
Views
5K
SWComposites
SWComposites
RogerPink
  • Locked
  • Question
SpaceX Falcon 9
Replies
5
Views
447
RogerPink
RogerPink
JBlack68
  • Locked
  • Question
Is equation 9.2.6 in ECSS-E-HB-32-23A correct?
Replies
0
Views
112
JBlack68
JBlack68
geeked
  • Locked
  • Question
Lockheed Stress Memo No. 1e
Replies
3
Views
211
40818
40818
jetmaker
  • Locked
  • Question
Static and Fatigue properties of 21-6-9 Cres Tubing
Replies
2
Views
204
RPstress
RPstress

Part and Inventory Search

Sponsor

Back
Top Bottom