airtraffic

Author Topic: N121JM GLF4 Crash at KBED  (Read 36919 times)

Offline RonR

  • Moderator
  • Hero Member
  • *****
  • Posts: 1189
N121JM GLF4 Crash at KBED
« on: June 01, 2014, 09:31:13 AM »
Here is what little audio I could find on the GLF4 accident at KBED.  Unfortunately, it doesn't reveal much at all...

Ron




Offline dave

  • Site Founder
  • Administrator
  • Hero Member
  • *****
  • Posts: 4597
    • LiveATC.net
Re: N121JM GLF4 Crash at KBED
« Reply #1 on: June 01, 2014, 10:36:57 AM »
There was some sort of buffering issue or network issue at the receiver site.  I doubt a takeoff clearance would have revealed much of anything anyway.  What is in those engines will be the key to this investigation.  FYI a Gulfstream 4 can take off and fly on a single engine.  That they couldn't take off likely means both engines lost thrust.  And if one exploded it could have taken the other one out, too.  Still, a very sad day here.

Offline FL510

  • Newbie
  • *
  • Posts: 1
Re: N121JM GLF4 Crash at KBED
« Reply #2 on: June 01, 2014, 02:14:53 PM »
From what I've read so far, this was most likely a high speed abort due to an on board fire.  One witness heard the engines spooling down as if being shut down which is normal procedure for an emergency evacuation due to a fire.

I seriously doubt the engines had anything to do with this accident. 

Offline mielsonwheals

  • Newbie
  • *
  • Posts: 34
Re: N121JM GLF4 Crash at KBED
« Reply #3 on: June 03, 2014, 12:28:49 PM »
Quote
From what I've read so far, this was most likely a high speed abort due to an on board fire.  One witness heard the engines spooling down as if being shut down which is normal procedure for an emergency evacuation due to a fire.

I seriously doubt the engines had anything to do with this accident.

I am going to have to disagree here. If you look at the recent post crash pictures, the thrust reverses are in the stowed position, which would lead me to believe they did not attempt a high speed abort.

You can't take witnesses accounts too seriously, especially ones that are not familiar with aviation. I work at the airport, and was unfortunately there the night it happened. Too early to speculate, and most of all very sad.

Offline InterpreDemon

  • Hero Member
  • *****
  • Posts: 514
Re: N121JM GLF4 Crash at KBED
« Reply #4 on: June 03, 2014, 01:14:02 PM »
It is all wild speculation, of course, but after reviewing the photos I also believe that fire was the result and not the cause, that it was more likely a problem with indicators or controls, top of the list for me being the trim of the stabilizer. I think the G4 couples flap and horizontal stabilizer motion, the two gearboxes being connected, and something may have happened to get them out of sync. If the stab setting was out of whack, for example still in nose-down trim for full flaps during the prior landing, unless they noticed that the stab indicator was not in sync with the flap indicator the pilots would not have realized there was a severe nose down trim issue until they attempted to rotate, which would be confusing if they had pre-set the normal (elevator tab) trim per takeoff checklist.

Unfortunately, confusion delays decisions, and later at night at sea level in cool air with just a handful of big-shots on board and plenty of runway in front of them, they may have been operating in "chauffeur mode", planning a gradual acceleration, long takeoff run and gentle rotation... basically throwing any maximum performance and/or balanced field calculations out the window and leaving them with too much useless runway behind them by the time they realized they had a problem.

That's just my instinctive first impression, but I'm almost always wrong at least 10% of the time.
« Last Edit: June 03, 2014, 06:46:28 PM by InterpreDemon »

Offline Eric M

  • Full Member
  • ***
  • Posts: 114
Re: N121JM GLF4 Crash at KBED
« Reply #5 on: June 03, 2014, 10:32:36 PM »
I've been wondering for a couple days how a G4 could manage to use up 7,000' of runway and still need more. Chauffeur mode would certainly explain it.

Offline InterpreDemon

  • Hero Member
  • *****
  • Posts: 514
Re: N121JM GLF4 Crash at KBED
« Reply #6 on: June 04, 2014, 01:10:47 AM »
A normal takeoff with full power from brake release takes about 5200', so when they reach the VR of 135kts you might figure they had used up at least 3500' and, if just "taking it easy" perhaps 4500' or more... and they have full or METO power on at the time of rotation, unlike aborting a takeoff at V1 ( at 110kts and at least 1000' earlier) due to faulty power or an engine failure. Assuming it takes at least five seconds (1000') for them to understand they must abort... they realize they can't rotate, pull the power and slam the brakes, they could easily be hurtling along at 150kts with full power and only three or four football fields, perhaps five seconds at 173mph, in front of them before they hit the localizer and approach lights, rupture a tank and roll into the ditch. The reason the reverse thrusters were not deployed is because they cannot be until the engines spool down sufficiently, by that time they were already hitting ground equipment. Thus when you look at the total stop distance off the runway (another 2000') and add that to the estimated runway remaining at the time of rotation it feels well within the range of plausibility.

We'll know in about 18 months, I suppose, perhaps earlier if they release the CVR and FDR transcripts.

Offline MatD

  • Newbie
  • *
  • Posts: 1
Re: N121JM GLF4 Crash at KBED
« Reply #7 on: June 04, 2014, 10:00:07 AM »
As a pilot and aviation enthusiast with an undergraduate degree in mathematics and a master's in computer science I felt the need to crunch some number prior to the final NTSB report that should be due out in 12-18 months. The information that follows is based on the facts available today along with many (but reasonable) assumptions. The facts are:
1. 49 seconds from roll start to final stop
2. 165 knots top speed attained
3. 100 knots at impact
4. 7,762' of pavement (including overrun at departure end)
5. 854' of terrain after end of pavement to stop location

My assumptions are primarily acceleration and deceleration rates associated with takeoff thrust, breaking, reverse thrust, impact with objects, and types of surfaces. For my calculations I used V1=117, VR=127, and V2=137 (although V2 isn't really a factor) all speeds in knots.

Based on those facts and assumptions, I came up with this timeline (all times indicate elapsed SECONDS from start of roll):
15 secs. - 80 knots announcement, 1,080' of runway used, 6,682' of pavement remaining
22 secs. - V1 (117 knots), 2,277' of runway used, 5,485' of pavement remaining
24 secs. - VR (127 knots), 2,700' of runway used, 5,062' of pavement remaining
26 secs. - V2 (137 knots), 3,159' of runway used, 4,603' of pavement remaining
31 secs. - 165 knots (max speed), 4,464' of runway used, 3,298' of pavement remaining
40 secs. - cross over departure threshold into overrun - estimated speed: 146 knots
44 secs. - end of paved surface - estimated speed: 126 knots
48 secs. - impact with light system before ditch - estimated speed: 95 knots
49 secs. - impact with ditch and full stop - estimated entry speed: 89 knots (11 knots slower than what NTSB stated)

I see two critical time periods, both of which are just seven seconds each. First, from 15-22 seconds between 80kts and V1, did the pilots put back pressure on the stick to confirm the elevator was functional? This could have provided the first indication of a problem. The second critical period was from 24-31 seconds after the VR announcement to max speed (at which point I assume they stopped accelerating). They covered over 1,300' during that span at which point they did not have the necessary stopping distance. I don't say this to second-guess the pilots, I'm just pointing out where critical runway distance was consumed and the speeds they were dealing with. The final 18 seconds presumably was used to attempt a full stop.

It will be interesting to see how this matches the final NTSB analysis and report. The bottom line is even though the entire event lasted a mere 49 seconds the actual time that the pilots had to assess the problem and determine a course of action was very much shorter. As previous posts mentioned, pilots are trained not to abort a takeoff after V1, a fact which surely contributed to the second critical time period mentioned above.

Please let me know if you see any glaring mistakes or incorrect assumptions. I'm not always right, but I'm never intentionally wrong.

Offline InterpreDemon

  • Hero Member
  • *****
  • Posts: 514
Re: N121JM GLF4 Crash at KBED
« Reply #8 on: June 04, 2014, 12:49:04 PM »
It appears your mathematical analysis is in agreement with my guestimations (at least to the order of magnitude, which is close enough) which are what lead me to my suspicion that the stabilizer is the culprit, causing an abort at least five seconds after attaining VR instead of at or before V1. Actually, the "decision" to abort was never made; circumstances prevented take-off, doubtless there was confusion and delayed reaction and the pilots did what they could with the woefully insufficient runway and time remaining. For a event duration of only fifty seconds even five seconds of "Hey, what the f*** is going on here!" is an eternity.

It would be interesting for you to run the same analysis with an abort executed at V1 and see if they had properly balanced the field or would/could have stayed off the grass.

Regarding testing back-pressure, I'm not sure that would have revealed a nose-down horizontal stabilizer setting, possibly they might have felt a nose-heavy condition as the speed increased, perhaps a bit of wallowing, but I think what most likely happened is everything felt pretty much normal until they tried to rotate, pulled back to their bellies and nothing happened. I don't rule out elevator control failure... they do use cables with hydraulic assist, but it seems to me that would be felt before they even took the runway.
« Last Edit: June 05, 2014, 01:58:40 PM by InterpreDemon »

Offline Marty Becker

  • Full Member
  • ***
  • Posts: 208
Re: N121JM GLF4 Crash at KBED
« Reply #9 on: June 13, 2014, 09:41:02 PM »
NTSB Preliminary Report

http://t.co/tt6AC7iAfL

Offline bigj93702

  • Newbie
  • *
  • Posts: 13
Re: N121JM GLF4 Crash at KBED
« Reply #10 on: June 14, 2014, 07:03:06 PM »
I read the report (and I am just an Aviation enthusiast)...

My bottom line on this report is: High speed abort after Rotate due to control issues and possibly Flaps were not set as required?   IS this a correct interpretation? 

Offline Eric M

  • Full Member
  • ***
  • Posts: 114
Re: N121JM GLF4 Crash at KBED
« Reply #11 on: June 15, 2014, 12:38:13 AM »
I think more likely the control wheel gust lock wasn't removed.

Offline NEaviator

  • Newbie
  • *
  • Posts: 7
Re: N121JM GLF4 Crash at KBED
« Reply #12 on: June 15, 2014, 06:42:25 AM »
I think more likely the control wheel gust lock wasn't removed.

Eric,

The report says, "The gust lock handle, located on the right side of the control pedestal, was found in the forward (OFF) position, and the elevator gust lock latch was found not engaged."

Offline Eric M

  • Full Member
  • ***
  • Posts: 114
Re: N121JM GLF4 Crash at KBED
« Reply #13 on: June 15, 2014, 11:16:20 PM »
Oh! Scratch that, then.

Offline InterpreDemon

  • Hero Member
  • *****
  • Posts: 514
Re: N121JM GLF4 Crash at KBED
« Reply #14 on: June 15, 2014, 11:30:55 PM »
No, you're both right, and so was I. The report said that FDR data showed the elevator position during taxi and takeoff roll was consistent with the gust lock(s) being engaged, but the gust lock controls were found (afterward) to be in the disengaged position, which is entirely consistent with my earlier post about them only discovering they had a problem at rotation... that possibly the elevator failed or the stabilizer was in a nose-down state. What it was, was the elevator being held in a nose-down state by the gust lock, which they certainly would have realized when they attempted to rotate although it would have taken a few seconds at least for them to realize their mistake and disengage the gust locks... by which time it was too late. That's why you always check your control surfaces for proper freedom and motion before you take the runway.

Edit: What is strange is that they should not have been able to throttle up with the gust locks engaged. Perhaps the the process was more sinister in that they taxied with the locks on, at that time of night in "chauffeur mode" they rolled right onto the runway at a good clip and as they gently throttled up they got no power, realized the locks were on and turned them off... but by that time they were moving along at a rate where the forces on the elevator (having been held nose-down) were such that the latch could not disengage, but after the aircraft came to rest (or shortly beforehand) it did. Should be interesting to read the CVR transcripts when they come out.
« Last Edit: June 15, 2014, 11:46:50 PM by InterpreDemon »

Offline svoynick

  • Full Member
  • ***
  • Posts: 120
Re: N121JM GLF4 Crash at KBED
« Reply #15 on: June 19, 2014, 10:27:05 PM »
What is strange is that they should not have been able to throttle up with the gust locks engaged. Perhaps the the process was more sinister in that they taxied with the locks on, at that time of night in "chauffeur mode" they rolled right onto the runway at a good clip and as they gently throttled up they got no power, realized the locks were on and turned them off... but by that time they were moving along at a rate where the forces on the elevator (having been held nose-down) were such that the latch could not disengage, but after the aircraft came to rest (or shortly beforehand) it did. Should be interesting to read the CVR
transcripts when they come out.
Man, that would be sinister.  How fast can you get going from a taxi on only 6% thrust, though?  Are you hypothesizing that they disengaged the gust lock controls on the roll, but by that time - on only 6% thrust - the latch was held "captive" by the aerodynamic forces, but not knowing that, they then throttled up from there?

At first, I was wondering how a latch mechanism could be held by aerodynamic forces that couldn't be overcome by pilot inputs to the control itself - a little forward/back pressure applied to the control - but then I thought about it this way:  with the elevator gust-locked "down" (i.e. commanding a nose-down pitch), moderate airflow over the horizontal stabilizer would tend to apply an "upward" force on the elevator surface, toward its neutral, "in-trail" position. 

If it were such an upward force on the elevator that held the gust-lock latch in place and prevented its retraction (in spite of the deactivation of the gust lock handle in the cockpit), it would take a downward force on the elevator (e.g. a nose-down command from the pilot) to neutralize the forces on the latch to allow it to retract.

But one can imagine that a pilot who has just disengaged a gust lock would be wanting to get that elevator neutralized from its nose-down position, and ready to apply nose-up back pressure suitable for a takeoff.  So, in fact, you might imagine that the first instinct of a pilot, upon releasing the gust lock, would be to maintain back pressure to neutralize the elevator, which would only add to the force holding the gust lock latch from retracting.

If anything, it would take a conscious and counter-intuitive application of nose-down force to the control to neutralize the force on the elevator to allow the gust lock latch to release, and if you're thinking you've got to get the elevator freed up so you can pull the nose UP for takeoff, pushing it DOWN would be against all instinct.

Please note, these are ONLY speculations, both about any possible pilot actions and about the mechanics of this particular gust lock and latch mechanism.  I'm not knowledgeable about the G-IV systems, and welcome any clarifications or criticisms of my thoughts here.

( But it did make my heart heavy to read:  ...FDR data parameters...did not reveal any movement consistent with a flight control check prior to the commencement of the takeoff roll.  Dang. )