PPRuNe Forums

PPRuNe Forums (https://www.pprune.org/)
-   Rumours & News (https://www.pprune.org/rumours-news-13/)
-   -   AF 447 report out (https://www.pprune.org/rumours-news/489790-af-447-report-out.html)

DozyWannabe 31st Aug 2012 18:48


Originally Posted by roulishollandais (Post 7385472)
@TurbineD
Thank you for the AA587 NTSB report.Good to reread to see the Regulators ' immobility

I don't see immobility. If they found the composite/metal V/S attachment to be suspect they could have taken the Ultimate Design Load of the 767 V/S to compare - that would cover the theoretical angle. A practical demonstration would be simplicity itself - take a 767 up, fly it through the wake of a 747 and replicate the F/O's rudder inputs. Neither of these tests would require Airbus's input or co-operation.

That they did not perform the latter action (despite being willing in the past to risk wake turbulence encounters in a 737 during the UA535/USAir427 investigations and spray an ATR72 with frozen liquid in-flight) implies that they knew damn well that to replicate those rudder inputs would have been incredibly dangerous.

Lyman 31st Aug 2012 19:06

Dozy.

The photographic evidence of the VS/join FAIL tells the story. At the time, below 250 knots, the A300 Rudder was not limited. On the way down, the Rudder separated from the seven hinge points that attached to the Fin. It has always seemed shortsighted to spend so much time on a new system, and then forget to degrade one system, (e Rudder hinges) in favor of a far more critical one, the FIN JOIN. Any Trim device that can flat plate its mount to beyond critical load needs some serious reconsideration, which it got, the A350 Has seven (fourteen total) saddles/pins/tabs. The idea in engineering is to spread the load beyond failure territory. The Rudder hinges, the Saddle Pins, and the Steel saddles all survived the enormous Yaw 587 encountered. The Composite tabs broke like Ritz crackers. A simple solution would have been to "stack" two separate V/stabs, the upper one with weaker tabs, such that a portion of the V/stab remains if sideloads fail the top. It won't happen. Too unusual and expensive. No one would accept the "frangible" aircraft....

Cheers

Organfreak 31st Aug 2012 19:12

Dozy Wannabe said,

...implies that they knew damn well that to replicate those rudder inputs would have been incredibly dangerous.
Or not. Perhaps they could not take the risk, no matter how small. Perhaps there wasn't the funding. Your 'logic' is unimpressive.

Lyman 31st Aug 2012 19:19

No, they chose to fix it without testing other systems. The proof for any design is in vetting the design, not showing that others will "fail too".

DozyWannabe 31st Aug 2012 20:34


Originally Posted by Organfreak (Post 7389210)
Or not. Perhaps they could not take the risk, no matter how small. Perhaps there wasn't the funding. Your 'logic' is unimpressive.

I addressed that here:


Originally Posted by DW
That they did not perform the latter action (despite being willing in the past to risk wake turbulence encounters in a 737 during the UA535/USAir427 investigations and spray an ATR72 with frozen liquid in-flight)...

Bear in mind that those tests on the 737 were performed when there was a very real prospect that the rudder hardovers were caused by a wake turbulence encounter, and that the ATR wing's problems with ice were well-known. Those are some pretty big risks right there, and the NTSB performed them for real (with FAA/NASA assistance) without flinching.

As for funding, in 2001 the US government was still running Clinton's surplus.

@Lyman:

There was no need to test that the A300-600 V/S could fail if Ultimate design load was exceeded - that was a given. In order to practically test the theory that a metal-metal attachment could have survived those loads would have required testing on another type. My guess is that the NTSB crunched the numbers and found that it would have made no difference.

Regarding the B-52, Turbine D put it quite succinctly:


Originally Posted by Turbine D (Post 7383201)
This particular aircraft in the video was a test aircraft examining structural integrity limits. Other in service aircraft suffered from this problem from time to time and some did not fly, but crashed. Turbulence and buffeting were the causes, two aircraft in 1963 lost fins, one in 1964 lost the entire tail, all crashed.

The famous incident where the V/S was practically lost involved a crew of test pilots who were drilled to counteract a failure of that type and magnitude. Regular military pilots weren't so fortunate in the same situation. The JAL123 747SR had a similar vestigial part of the V/S remaining, but the loss of hydraulics still led the aircraft to crash. The B-52 is not an airliner.

I reckon you'd find that the airspeed-based rudder authority limits were very close when comparing the A300 and the B767. Variable-stop or not, I don't think it would have made a lot of difference whne putting bootfuls in like the AA587 F/O did.

Lyman 31st Aug 2012 20:44

The steel saddles survived, the FIN tabs, their corollary, snapped off. The mode of failure was exclusive to resin, its characteristics were not exonerated. They used a completely different material in a mirrored design application, and the material failed. Why test steel?

DozyWannabe 31st Aug 2012 20:51


Originally Posted by Lyman (Post 7389351)
The steel saddles survived, the FIN tabs, their corollary, snapped off.

Don't you mean Aluminium?

The attach points failed at a load considerably beyond that which was expected - if that's not enough of a clue then I don't know what would be. In terms of anecdotal evidence regarding strength of construction, the only airliner to land safely after a SAM strike was an A300.

In any case, the metal components were only handling a fraction of the load. I'm happy to be proven wrong - just take a B767 up and replicate the inputs - any takers?

Lyman 31st Aug 2012 21:40

I had assumed Steel, the NASA fracture analysis simply references "metallic" wrt the clevis, (fuselage portion) of the join. No matter, the pictures show the metallic structures to have survived, unscathed

.

The Composite failures were familiar to me, as to type and extent, in regard to tensile, compression, and torsional character.

The Graphite resin composite that made up the attach wings of the VS was lapped, and glued, to the load carrying member, that is a red flag, as was the fact that the holes for the Bushings and pins/bolts were drilled, (albeit with a core box bit). The center lug was 2 3/4 inches thick, and the fore and aft attach points were less, 1.78 and 2.

There was evidence of delamination, and matrix rich pockets that disintegrated, as would be expected, as well as the cross laminar fracturing that came with load limit exceedence.

The failure of each of the three points on the starboard, was consistent with diameter of material; the center point, the thickest, caused a cross laminar fracture above it, the lug hole was intact. Fore and aft, the lugs separated through the core box drilled portion, as would be expected, simply from a dimensional standpoint.

The importance of the sequential failures shows that the separation would not have occurred, should the fore and aft joins have exhibited the same (or at least increased) material present at the pin portion of the attachment.

From the improved design shown in later iterations of the 300 family, the conclusion is that the engineered design took insufficient consideration with regard to weight/strength of the six similar though variably constructed points of attachment.

Dozy, you want another system tested? They have been, and the fesults is consistent with a proper design and limit load calculatuons.

A simple look at the construction and mode of failure shows the weak spots.

It doesn't take a lot to conclude that for the application, insufficient design and material was used in a new and unique system....It failed, and 260 people died.

Would a different, improved design have weighed more? Would it be stronger?

Yes, and Yes....

DozyWannabe 31st Aug 2012 21:49


Originally Posted by Lyman (Post 7389427)
There was evidence of delamination, and matrix rich pockets that disintegrated, as would be expected, as well as the cross laminar fracturing that came with load limit exceedence.

That doesn't mean anything unless it was proven that metallic materials would not have failed at the same load level.


From the improved design shown in later iterations of the 300 family
There was no later iteration of the A300 family, the A300(B4)-600 was the last. There was a -600R, but the only change there was the addition of a trim tank in the V/S.

Lyman 31st Aug 2012 22:02

It does if you analyze the response of dissimilar materials, no test flight is needed.

DozyWannabe 31st Aug 2012 22:06

Sources please...

(Bearing in mind that the ultimate design load would have been calculated taking that into account...)

Brian Abraham 1st Sep 2012 07:05

I don't know what all Lymans chatter is about.

The vertical stabilizer was exposed to aerodynamic loads that were about twice the certified limit load design envelope and were more than the certified ultimate load design envelope. (From the NTSB report)

Lyman 1st Sep 2012 12:13

The trimming device that put the aircraft sideways survived.

The aircraft has a Rudder Limiter Control Unit.

"Rudder with Care" appears on the PFD when the Rudder is over deflected, but not at speeds below 250 knots.

The RTLU was capable of preventing this crash, but the system was not designed to.

Not in this case....

Two problems:

1. Not addressing a potentially lethal problem whilst equipped with a device that is supposed to save itself.

2. Ignorance of fail safe structural design in this aero domain, eg, the Rudder fasten/join is too robustly built at loads approaching VS separation.


It is not difficult to determine what load should fail the VS, NTSB did. Somewhat short of this value, the Rudder hinges should fail.

FullWings 1st Sep 2012 12:24


It is not difficult to determine what load should fail the VS, NTSB did. Somewhat short of this value, the Rudder hinges should fail.
Not necessarily. You can break the fin off by using quite moderate rudder inputs, timed to increase the amplitude of yaw oscillation until the forces on the vertical stabiliser become too great in part of the cycle. This is a dynamic situation...

Lyman 1st Sep 2012 12:31

That is true, fullwings, but in Pitch, the wings are protected by a 2.5 load factor.

Did someone forget to add load factor in Yaw? That is what the limiter is supposed to accomplish. The elevators cannot provide more than 2.5 in any aspect, what makes the VS chopped liver? How does the possibility of failure of VS escape the designers?

At accelerometer sensed limits, the Rudder should lose hydraulics, and trail in the airstream, unloaded...

Lyman 1st Sep 2012 12:36

Or, better still, the hydraulics can synch to opposite tail swing, attenuating the increasing Yaw oscillation..in any case, the rudder pedals should go off line, instant. Nobody dies from a rudder in free stream.

FullWings 1st Sep 2012 12:41

I suppose it's much more difficult to apply to a non-FBW aircraft. It would have to be quite a complex system to stop a damaging yaw excursion. There's always the possibility that the control input needed to stop swinging going out of limits would exceed structural limits for the fin/rudder anyway. The momentum for this event could be built up over several cycles and be hard to dissipate all at once.

A FBW system could/does model this and stop it getting out-of-hand in the first place.

Lyman 1st Sep 2012 12:48

The VS itself should damp the oscillation, without rudder, no fbw needed, and then there is the over built rudder problem. Losing rudders after a sensor picks up Yaw cycles is simply a solenoid and a spark away?

PIYO: pilot induced yaw oscillation??

Then again, it might need to be linked to synched N1s.

FullWings 1st Sep 2012 12:56

Great ideas which, unfortunately, have to be concisely specified in a mathematical way before they can be implemented in a control system. Maybe this process is slightly more difficult than it appears on the surface? Also, these systems have to function throughout the entire flight envelope, without bringing the law of unintended consequences into play...

Lyman 1st Sep 2012 15:04

FullWings Great ideas which, unfortunately, have to be concisely specified in a mathematical way before they can be implemented in a control system. Maybe this process is slightly more difficult than it appears on the surface? Also, these systems have to function throughout the entire flight envelope, without bringing the law of unintended consequences into play...


I'll wager the maths have been done to death, and the engineering modelled and rejected as too expensive. Also too heavy. It is frustrating that compromises are accepted, and that training is allowed to kill. That is, assuming the Rudder was the problem. Eleven years on, is it still too soon to discuss the mechanical possibilities of damage by foul play?


All times are GMT. The time now is 04:20.


Copyright © 2024 MH Sub I, LLC dba Internet Brands. All rights reserved. Use of this site indicates your consent to the Terms of Use.