Possible cause of AA flight 587 crash...a new thought

Vanity | 11/15/01 | Agent Smith

[Agent Smith]

Up until now, my best guess as to the cause of the 587 crash was defective/substandard bolts used in attaching the vertical stabilizer to the tail and a failure to detect the problem through inspections.

However, I heard on the news last night that the vertical stabilizer was not fabricated from aluminium, but from a carbon fiber composite. This material is very strong and light but can fail catastrophically if a stress fracture/crack develops. Based on the photos of the recovered stabilizer showing that it was cleanly severed from the tail, I now believe that this is the most likely cause of the accident.

The turbulence from the JAL 747 was the straw that broke the camel's back.

[Blueflag]

Thanks for the pics. Actually, I don't see the VS missing any pieces of import except the rudder and the little toe at the bottom front. That's my opinion.

To me the radar plot is VERY revealing. (1) the VS landed less than one mile from impact. The engines are very close to impact. Total radius of all parts looks to be 1/2 to 1/4 mile. This incident, at its end happened very quickly. Why did that plane fall apart so rapidly? Blows me away. Would like now to see the voice transmissions plotted along the radar track. When did the pilots report the shaking, ask for max power, report wake turbulence., etc., etc.

This just got way more perplexing for me.

BTW, go here for more info good thread

[Blueflag]

ONe more time on this wake turbulence theory. The time stamps on the radar plot demonstrate the A300 was :90 sec behind the 747. IT was about 1/2 - 2/3 mile to the left and about (near as I can make out) 300 - 400 feet lower. For folks who have flown or flown behind heavies, and taking into account there was an 11KT wind blowing to the SE, could the wake of the 747 have made it into the path of the A300?

.

Now, let me re-state my position that I do not believe wake turbulence could cause an A300 to shed parts like this.

[discostu]

I think I was fooled by the angle in the other pic I saw, that's the problem with 4 square inch 2D pictures of 27 foot tall 3D objects, you lose some details.

Definitely, this gets more perplexing with each answer not less. Crowley did once say that one of the things that made life interesting was that every "soldier" (exclamation mark) was followed by a "hunchback" (question mark"), but not every hunchback has a soldier (ie, not all questions have answers, but every answer makes more questions). This is definitely one of those cases.

[PaulKersey]

>Well, some reports and data indicate that in the air, their paths were as little as 45 seconds apart.

Even if that close, wake turbulence won't rip the tail off a normal aircraft. The wings absorb almost all of the vortex torque.

Again, the relative size of the 2 aircraft is important, and these were both jumbos.

[Blueflag]

Also, take a look at this (will take a while to load.) It is a photo of the VS assembly at the factory. The VS has the 'flanges' mounted ventrally, ready to attach to the fuselage. BAsed on this photo, I would say that the failure point on THIS A300 was therefore in the VS subassembly. HEre 'tis

A300 VS assembly photo

[GodBlessUSA]

I use to fly cessna. I was told to land following a jet. I hit the wake turbulance as I hit the runway. What I went through to keep that plane's wings from hitting the ground and keeping from wiping out was unbelivable. I was in a small cessna and it certainly didn't damage that plane at all. That was a long time ago. I know it's a totally different situation. I have also sat at Reagan Airport at the end of the runway and watch one jet take off and land right after another. I can't help the doubts.

[Blueflag]

Agreed -- BTW, radar data plotted above demonstrates the two aircraft were NEVER that close. Nice to get facts instead of reports.

I agree with you that WT could never tear apart a healthy A300 airframe. sumthin's up here.

[Blueflag]

AAAAAAAAAAGH. I stink at HTML. After the PDF loads, (you may need to hit re-load) go to page 14. Sorry.

[discostu]

Found it, does look like the leading part of the attachment curve is what failed first. I also note from the plots that the plane was in a fairly tight (for a jumbo jet) turn, I'm presuming that before cruising altitude and trimming most of the turn is done with the rudder (any big iron boys that can talk about that?), which as you pointed out earlier is kind of like skidding a plane, wonder if being in the middle of a rudder turn and losing the tail would accentuate lateral movement and therefore lateral stresses on other parts (like the engines). I'm in serious WAG territory now.

[Blueflag]

Actually I see nothing odd with the departure route. THe A300 went wheels up faster than the 747, as expected, then flew a similar (if tighter) radius, consistent with a lighter smaller plane.

Enjoyed the discourse. I have to go fly from LGA to ATL now.

Later.

[Tony in Hawaii]

There are six tabs which retain the vertical stab. Only four of these are visible in the pictures of the chunk of aft fuselage in this picture. Very interesting.

These are very interesting pictures, thanks for posting them, I hadn't seen them before.

[bootless]

Hmm, thanks for that. I'm starting to get information overload on this. I'll check back for that message, and see how that conversation is developing.

What do/did you fly?

[David]

"Some American production airplanes do not even have a vertical stabilizer."

Which ones?

Beechcraft V34 etc. Bonanza

[discostu]

Good luck. I'll throw this out incase others pick up on it. I don't see a problem with the departure route under normal conditions. Basically I'm wondering if losing the tail in the middle of that turn contributed to the secondary damage, could the combination of a tight turn and loss of rudder sent the plane into a flat spin, which then put serious lateral stress on the engines causing them to be torn from the plane?

[mad_as_he$$]

Nice collection of FACTS guys great job!!!

[David]

Other discussion threads have introduced a different plausible scenario. Indeed several aircraft have been successfully flown without a VS in place. But ... Suppose the VS detached from the airframe while the first officer has stepped on the left rudder (appropriately, no pilot error implied) to in part correct for WT. The VS detaches, the yaw moment is to port, and the right wing is generating more lift and roll moment than the left. The WT that had been trying to roll the aircaft to the right now pushes on the wings to roll it to the left (other side of vortex) and encounters an aircraft trying its best already to roll to the left. Without a VS in place, (because it failed, not the pilot) the plane departed controlled flight with no means of recovery.

Maybe. I tend to be really skeptical of fairy tales which rely on a series of coincidental events, particularly where the fairy telling the story has a history of lieing under similiar circumstances (TWA 800 for example). The government and the industry are determined to sell the line that this did not result from an enemy attack they could do nothing about and so have a strong motivation not to tell the truth.

Further, your hypothetical relies on having gotten to the wake turbulance in the first place. The odds against that are so high as to render it almost unbelievable. With the wind conditions in place, the wake vortex would have disipated almost immediately. And, whatever wake vortex there was would not have been in this flightpath because the wind would not only have disipated it, it would have also moved it a long way from the track of the 747. The TV depictions show the Airbus in the wake event because it was flying through the airspace where the 747 had been (a minute and a half earlier)--one of the many problems with that analysis is that the wake would no longer have been there by the time the Airbus got there because the wind tends to move whatever vortex it does not disipate. Wake turbulance just didn't have anything to do with this event.

Maybe there is some other explanation for why it was an accident--the bolts on the stab wore out; the stab bolts wore out at the very exact same time the bolts that held both engines on, both picked the exact same time to separate; it happened at the very same exact time whatever spark ignited TWA 800's fuel tank struck again generating the fire at the wing root testified to by a number of witnesses. Whatever you might think of that fairy tale, it is a lot more likely than the wake turbulance explaination.

[okie01]

"The TV depictions show the Airbus in the wake event because it was flying through the airspace where the 747 had been (a minute and a half earlier)--one of the many problems with that analysis is that the wake would no longer have been there by the time the Airbus got there because the wind tends to move whatever vortex it does not disipate. Wake turbulance just didn't have anything to do with this event."

However, wake turbulence sinks. And it is blown by the wind. The wind was from 320 at 11 knots. Thus, by turning inside the JAL 747 and by flying 400-800' below the track of the 747, AA 587 would have entered the area where the 747's wake was sinking/drifting.

Encountering wake turbulence would not be unexpected. That it would be sufficient to separate the VS from the A300 is what would not be expected. Unless, the VS was already doomed by a structural flaw...

[RaceBannon]

Vertical sections in the tail I believe would be called SPARS, (The equivilent of a rotor blade SPAR),

sections that were aligned horizontal (parallel to the ground) would be called Ribs.Think of your ribs, sibs on an aircraft would be aligned in rows, like your ribs, but on a fusalage they would appear to be verticel, going from the front of the aircraft to the rear.

Any relatively flat, I-Beam like piece that connects the Ribs would be BEAMS. Beams would make floor pieces or ceiling pieces within the fusalage.

Right angel pieces that ran perpendicular to the RIBS would be stringers.(And for the life of me, I forgot how they would be attached! I think riveted from RIB to RIB as individual pieces)They are basically a lightweight aluminum angle that provide stiffening for the skin bewteen each rib.

RIBS and BEAMS would be riveted to the outer skin, the SPAR would be possibly bolted to the RIBS or Riveted.<P. From the photo I see here, it looks like the vertical spar was sheared somehow. However, I am used to rotary craft, so this is a logical guess for me, not a declaration.

[eno_]

Excellent post, Race. It had to be a big sideways load, doncha think? All six mounts, all at once. Snap! 600 sq. feet sideways into 300kts sounds a lot more plausible than wake turbulence. Plus the witnesses saw some violent event up front before the tail came off. Bang! Yaw. Snap (engines too, moments later). Crash. Plus that makes the pristine condition of the vert stab surfaces more plausible.

[eno_]

On the FAA Web site, it looks as if the front four attachment points are on one piece of the tail cone, and the back two on another. They are all there. Intact. With fasteners in place.