Page 5 of 6 FirstFirst ... 3456 LastLast
Results 81 to 100 of 119

Thread: Cargo airplane collapsed in Kyrgyzstan

  1. #81
    Senior Member Evan's Avatar
    Join Date
    Jan 2008
    Posts
    5,361

    Default

    Quote Originally Posted by BoeingBobby View Post
    Unlike ATL, Atlas although not the best, buys you a hotel at every overnight, we have gateway basing so no crash pads, buys you a ticket to and from your gateway, business class tickets on all international flights, and we are catered on every flight.
    BB, Can you tell us what happens on the 744 if the vertical mode goes to GS CAP and then loses the glideslope signal? Does it then transition to V/S by default? This flight transitioned to ALT HOLD when it reached 3400 with no glideslope capture, then flew level until it traversed the mirror glideslope at 9°, which activated GS CAP but of course it flew right through that. Yet it descended on autopilot. Why does it not remain in ALT HOLD in that scenario?

  2. #82
    Member BoeingBobby's Avatar
    Join Date
    Jun 2009
    Location
    MIA
    Posts
    916

    Default

    Quote Originally Posted by ATLcrew View Post
    I wasn't aware there were still airlines who DIDN'T pay for hotels on overnights. Apart from an occasional late-running van, I've had almost no issues at all with our hotels. If anything, some of them are, in my opinion, a little too nice.
    I meant at base, not out on a trip.

  3. #83
    Member BoeingBobby's Avatar
    Join Date
    Jun 2009
    Location
    MIA
    Posts
    916

    Default

    Quote Originally Posted by Evan View Post
    BB, Can you tell us what happens on the 744 if the vertical mode goes to GS CAP and then loses the glideslope signal? Does it then transition to V/S by default? This flight transitioned to ALT HOLD when it reached 3400 with no glideslope capture, then flew level until it traversed the mirror glideslope at 9°, which activated GS CAP but of course it flew right through that. Yet it descended on autopilot. Why does it not remain in ALT HOLD in that scenario?
    That is a damn good question Evan. However I do not honestly know the answer. I fly 10 Red Bull aircraft with the wings off in the cargo compartment from Hahn Germany to Los Angles tomorrow. I will see if any of the other crew has the answer and let you know.

  4. #84
    Senior Member 3WE's Avatar
    Join Date
    Jan 2008
    Posts
    4,016

    Default

    Quote Originally Posted by Evan View Post
    BB, Can you tell us what happens on the 744 if the vertical mode goes to GS CAP and then loses the glideslope signal? Does it then transition to V/S by default? This flight transitioned to ALT HOLD when it reached 3400 with no glideslope capture, then flew level until it traversed the mirror glideslope at 9°, which activated GS CAP but of course it flew right through that. Yet it descended on autopilot. Why does it not remain in ALT HOLD in that scenario?
    Holy crap! I don't have a clue what you said.

    I take comfort in two things...a. 1. A 49 year, 40,000 hour veteran does not know the answer either and 2. I do know that NONE of that excuses the need to monitor the instruments and reconcile things to the approach plate...

    If they have to know all that crap, no wonder they forget the basics occasionally.
    Les règles de l'aviation de base découragent de longues périodes de dur tirer vers le haut.

  5. #85
    Senior Member Evan's Avatar
    Join Date
    Jan 2008
    Posts
    5,361

    Default

    Quote Originally Posted by 3WE View Post
    Holy crap! I don't have a clue what you said.

    I take comfort in two things...a. 1. A 49 year, 40,000 hour veteran does not know the answer either and 2. I do know that NONE of that excuses the need to monitor the instruments and reconcile things to the approach plate...

    If they have to know all that crap, no wonder they forget the basics occasionally.
    Put simply, It appears that in this case they were above and beyond the glideslope, so after leveling off at 3400' and 'hunting', the autopilot found the false glideslope signal instead, causing the mode to become GS CAP (glideslope capture). GS CAP is supposed to then transition to GS and ride the ILS down, but that's not going to happen with a 9° false glideslope, so instead it appears to have flown through the false glideslope, whereupon GS CAP changed to... WHAT? The a/c flew a nice 3° glidepath angle down to DH, I'm guessing in V/S mode but maybe not. But why, if the glideslope is lost (or full downward deflection is beyond the autopilot capabilities) does it not do the safe thing and revert to ALT HOLD again? If it had, this accident wouldn't have occurred.

    Or maybe that final 3° glidepath was manually commanded by the crew...

    Yes, it's no excuse for not piloting, but it should provide a defense against not piloting.
    Attached Thumbnails Attached Thumbnails Click image for larger version. 

Name:	mycargo_b744_tc-mcl_bishkek_170116_7.jpeg 
Views:	21 
Size:	35.3 KB 
ID:	7218  

  6. #86
    Senior Member Gabriel's Avatar
    Join Date
    Jan 2008
    Location
    Buenos Aires - Argentina
    Posts
    5,852

    Default

    Evan, note the the glideslope signal was not lost in this case.

    If the glide slope signal was lost, I would expect the AP to default to VS hold (and the VS flag to show up). That would be sort of "fail passive" (but with the AP still engaged). The alternative would be to disconnect the AP with the plane left in trim (that would be truly fail passive), but this seems more dangerous and unnecessary. Defaulting to ALT HLD is in my opinion even more dangerous. The AP would be not only making significant pitch inputs to transition from the descent to level flight (a disconnect at this moment could leave the plane out of trim), but also the AT may be disconnected which risks a stall if the pilot doesn't apply throttle in time.

    However, in this case what happened is that the GS needle was low (indicating the plane was above the glideslope), then (when they crossed the false 9 deg glide slope) the needle became alive, probably deflected up for a moment, then started to move down (as the plane approached the false glide slope from below), then approached the center (and the AP changed from ALT HLD to APP) and then keep moving down and deflected fully down again (as the plane was now above both the real and the false glide slope).

    The following applies for the MD-80 AP, but I guess that all or most APs have something similar (even if only internally, transparent to the pilot). When the AP goes to APP it can be in 2 different status in the FMA (Flight Mode Annunciator): GS CAP and GS TRK. GS CAP (CAP for Capture) is a mode where the AP will try to capture the GS, that is, to maneuver the plane towards the GS deflection, and it happens in the beginning when the AP transitions to APP and at any time where the GS goes anywhere off center by more than a hair. GS TRK (TRK for Track) is the status where the GS is centered (or barely off center) and the APP is basically keeping it centered. I am not sure what is the difference between the 2 in practical terms, but I guess that GS CAP is a bit more aggressive with the interception.

    My guess is that in this case the AP transitioned from ALT HLD to APP and remained in APP (in GS CAP) until the GA was initiated. If I am correct, we should see that the approach was not really parallel to the stipulated one, but slightly steeper (although nowhere close to the 9 deg).

    --- Judge what is said by the merits of what is said, not by the credentials of who said it. ---
    --- Defend what you say with arguments, not by imposing your credentials ---

  7. #87
    Senior Member Evan's Avatar
    Join Date
    Jan 2008
    Posts
    5,361

    Default

    Quote Originally Posted by Gabriel View Post
    Evan, note the the glideslope signal was not lost in this case.

    If the glide slope signal was lost, I would expect the AP to default to VS hold (and the VS flag to show up). That would be sort of "fail passive" (but with the AP still engaged). The alternative would be to disconnect the AP with the plane left in trim (that would be truly fail passive), but this seems more dangerous and unnecessary. Defaulting to ALT HLD is in my opinion even more dangerous. The AP would be not only making significant pitch inputs to transition from the descent to level flight (a disconnect at this moment could leave the plane out of trim), but also the AT may be disconnected which risks a stall if the pilot doesn't apply throttle in time.

    However, in this case what happened is that the GS needle was low (indicating the plane was above the glideslope), then (when they crossed the false 9 deg glide slope) the needle became alive, probably deflected up for a moment, then started to move down (as the plane approached the false glide slope from below), then approached the center (and the AP changed from ALT HLD to APP) and then keep moving down and deflected fully down again (as the plane was now above both the real and the false glide slope).

    The following applies for the MD-80 AP, but I guess that all or most APs have something similar (even if only internally, transparent to the pilot). When the AP goes to APP it can be in 2 different status in the FMA (Flight Mode Annunciator): GS CAP and GS TRK. GS CAP (CAP for Capture) is a mode where the AP will try to capture the GS, that is, to maneuver the plane towards the GS deflection, and it happens in the beginning when the AP transitions to APP and at any time where the GS goes anywhere off center by more than a hair. GS TRK (TRK for Track) is the status where the GS is centered (or barely off center) and the APP is basically keeping it centered. I am not sure what is the difference between the 2 in practical terms, but I guess that GS CAP is a bit more aggressive with the interception.

    My guess is that in this case the AP transitioned from ALT HLD to APP and remained in APP (in GS CAP) until the GA was initiated. If I am correct, we should see that the approach was not really parallel to the stipulated one, but slightly steeper (although nowhere close to the 9 deg).
    Well, the thing is, the autopilot never transitioned to G/S mode, yet it flew a steady 3° FPA, parallel to the actual G/S, down to DH. I'd like to know exactly how it did that.

    The AP was certainly in APP mode, LOC in lateral mode, when it passed through the false glideslope. That caused it to transition to G/S CAP for about one second (and the Russian plot I attached above suggests it did briefly try to align with the steep 9° convergence of the two G/S signals, but by then it was not possible to maintain the false glideslope and the PFD went to full downward deflection again. Immediately thereafter, the a/c entered a steady 3° descent. Was this selected guidance by the crew or a default mode transition (I doubt it was hand-flown), and if so, what was the mode in descent?

    I don't see how reverting from a failed glideslope capture back to ALT mode would be particularly dangerous, as it would be the same as pressing ALT HOLD at that moment.

    The G/S issue is simple, and when I say "lost the glideslope", I mean lost the convergence of the two signals that creates it. Attempting to catch the G/S from above, they would have been in the predominance of the 90 HZ signal (full downward deflection) until encountering the false 150 HZ signal (causing the upwards deflection and momentarilly triggering G/S CAP). But a moment later they would be back in the predominannce of the 90HZ signal (back to full downward deflection) and thus no transition to G/S mode. **

    I'm sure the reason BoeingBobby can't answer this off the top of his head is because this is not supposed to happen (the plane is not designed to capture a 9° glideslope) and the autopilot behavior in this scenario (G/S CAP followed by a failure to capture) is almost certainly not documented or instructed.

    Anyway, hoping he can shed some light on this part of the mystery...

    **EDIT - some G/S systems reverse the phase of these side lobe signals at the 9° angle, but the Russian reports and plot seem to indicate that this was not the case. Need to learn more about that...

  8. #88
    Senior Member Evan's Avatar
    Join Date
    Jan 2008
    Posts
    5,361

    Default

    Quote Originally Posted by BoeingBobby View Post
    That is a damn good question Evan. However I do not honestly know the answer. I fly 10 Red Bull aircraft with the wings off in the cargo compartment from Hahn Germany to Los Angles tomorrow. I will see if any of the other crew has the answer and let you know.
    One thing I've discovered is that—apparently—on G/S CAP, the FMS commands a -.5G pitch manuever to establish the desired 3° glidepath. If the automation does not then transition to G/S, perhaps it just remains on that path... in V/S? or...?

  9. #89
    Senior Member 3WE's Avatar
    Join Date
    Jan 2008
    Posts
    4,016

    Default

    Quote Originally Posted by Evan View Post
    ...it's no excuse for not piloting, but it should provide a defense against not piloting...
    The magenta line needs to start working vertically, and be combined with synthetic vision and maybe a good speed control and stall avoidance... cue the app joke which really means it wouldn't be too hard.
    Les règles de l'aviation de base découragent de longues périodes de dur tirer vers le haut.

  10. #90
    Senior Member Gabriel's Avatar
    Join Date
    Jan 2008
    Location
    Buenos Aires - Argentina
    Posts
    5,852

    Default

    Quote Originally Posted by Evan View Post
    **EDIT - some G/S systems reverse the phase of these side lobe signals at the 9° angle, but the Russian reports and plot seem to indicate that this was not the case. Need to learn more about that...
    Due to the nature of the lobes and their reflections, there are different slopes where the 90 hz signal has the same amplitude than the 150 hz signal.

    The 1stt one (the original one, the "real" one) is usually at about 3 deg slope.

    The remaining ones happen and multiples of the original one: the 2nd at 6 deg, the 3rd at 9 deg, the 4th at 12 deg, and so on.

    The odd ones (1st, 3rd...) are "upside up", i.e. the 90 hz signal is stronger than the 150 hz signal above the slope (and vice versa), and the ILS instrument reacts normally (the needle goes down when the plane is above the slope and vice versa).

    The even ones (2nd, 4th...) are "upside down" (inverted), with the 150 hz being dominant above the slope, so the indication in the plane is reversed.

    This 747 intercepted the 3rd slope (9 deg), which is NOT inverted.

    --- Judge what is said by the merits of what is said, not by the credentials of who said it. ---
    --- Defend what you say with arguments, not by imposing your credentials ---

  11. #91
    Senior Member Evan's Avatar
    Join Date
    Jan 2008
    Posts
    5,361

    Default

    Quote Originally Posted by Gabriel View Post

    The odd ones (1st, 3rd...) are "upside up", i.e. the 90 hz signal is stronger than the 150 hz signal above the slope (and vice versa), and the ILS instrument reacts normally (the needle goes down when the plane is above the slope and vice versa).

    The even ones (2nd, 4th...) are "upside down" (inverted), with the 150 hz being dominant above the slope, so the indication in the plane is reversed.
    Uh yeah... this is where it gets tricky. There are five different types of glideslope antenna systems in use today. Three of them are imaging-type antennas. With those systems, the 6° and 12° false glideslopes are 'null' and the 9° one is always INVERTED. The other two operate as you described. I think you are correct that in this accident, we are dealing with non-imaging type antennas, so the 9° would NOT be inverted and thus the command after passing through would be 'fly down'.

  12. #92
    Senior Member Evan's Avatar
    Join Date
    Jan 2008
    Posts
    5,361

    Default

    Quote Originally Posted by 3WE View Post
    The magenta line needs to start working vertically...
    It already does on the newest displays
    Attached Thumbnails Attached Thumbnails Click image for larger version. 

Name:	AERO_2012q1-3_fig3.jpeg 
Views:	29 
Size:	45.6 KB 
ID:	7232  

  13. #93
    Senior Member 3WE's Avatar
    Join Date
    Jan 2008
    Posts
    4,016

    Default

    Quote Originally Posted by Evan View Post
    It already does on the newest displays
    I was meaning on 'all aircraft', not just 'the newest displays'...and synthetic vision where you simply fly 'visually' through boxes... effortless SA and confirmation for the approach profile, unlike this AND the NZ bunch on your other thread.
    Les règles de l'aviation de base découragent de longues périodes de dur tirer vers le haut.

  14. #94
    Member
    Join Date
    Jan 2008
    Location
    MA, USA
    Posts
    684

    Default

    It seems like with a system that has GPS and an airport database with the right info, this should be almost a non-issue.

    The GPS knows where the plane is in 3 dimensions, and the airport database can know where the runway threshold is, including its elevation. That's assuming of course you've correctly programmed it with the destination airport and runway.

    With that info, it's easy to compute the descent angle required to reach the threshold from the aircraft's current position. If the GS indication is centered and the descent angle isn't 3 degrees (or whatever's correct for that runway), it could display a warning. Something similar could be done with the localizer if localizers are prone to similar issues.

    I'd certainly expect any system capable of displaying an image like Evan posted to be able to do that.
    Be alert! America needs more lerts.

    Eric Law

  15. #95
    Senior Member Gabriel's Avatar
    Join Date
    Jan 2008
    Location
    Buenos Aires - Argentina
    Posts
    5,852

    Default

    An EGPWS should have been able to prevent this accident too. After all it was a CFIT.

    --- Judge what is said by the merits of what is said, not by the credentials of who said it. ---
    --- Defend what you say with arguments, not by imposing your credentials ---

  16. #96
    Senior Member 3WE's Avatar
    Join Date
    Jan 2008
    Posts
    4,016

    Default

    Quote Originally Posted by elaw View Post
    ...It seems like with a system that has GPS and an airport database with the right info, this should be almost a non-issue...
    Brief predominately serious moment for 3BS:

    Two primary issues stand in the way:

    1) Certification- yeah sure, it would be 'nothing' to add this sort of technology- and that's what I'm talking about...10-year old MSFS could provide a half decent, totally fake picture of reality WITH BOXES FOR YOU TO FLY THROUGH!...but what if the computer system hiccups? Liability, cumbersome government regulators, the public's disdain for mistakes and 1a) Unforseen circumstances???

    2) $. Because of #1, it would cost $...lots of $...and, like it or not, old, used, lower-tech aircraft are pretty darn important to aviation. I was surprised to see 'old fashioned' instruments on the MD-88 I rode this week...I sort of assumed (as is a major fault of mine) that almost all passenger jets in the US had gone glass- The cargo industry REALLY likes old planes...I still see ITS-Mobiles flying about Graceland International Airport.

    ...and I haven't become Evan...I still think Airmanship and fundamentals are important...but the old fashioned cryptic ness of ADF needles and ILS needles and doing some quick height-distance math when an I-pad can show you what you'd see in CAVU conditions...

    Which is better for a quick check of all-important SA?

    As Evan will rant (with some validity)...it's all about the $$

    And ATL, et al. will counter rant with equally valid: Stats say we are really really really really really really really safe with our systems as they are now, don't fix what isn't broke...(Gabe's EGPWS reference noted.)

    [/serious]I restate from another thread, I hope that safety officials are taking note!
    Les règles de l'aviation de base découragent de longues périodes de dur tirer vers le haut.

  17. #97
    Senior Member Evan's Avatar
    Join Date
    Jan 2008
    Posts
    5,361

    Default

    Quote Originally Posted by Gabriel View Post
    An EGPWS should have been able to prevent this accident too. After all it was a CFIT.
    I guess, since they were always over the field when in ground proximity, and TAWS doesn't know which end of the runway you are landing on, the alerts would be inhibited in landing configuration.

    ILS is a pretty antiquated system but also probably proven reliable because of its analog simplicity. There are layers of protection against failure, including back-up amplifiers and nearby sensors that detect any signal anomalies. When those sensors detect a problem, they alert ATC and switch to the back-up amplifiers, and if the back-ups fail they shut the system down. The only way you are going to get into these erroneous situations is if maintenance or testing work is being done, maintenance procedures are not followed and the flight crew is not monitoring altitude and distance. That's a lot of cheese to pass through. You can avoid false glideslope by always catching the glideslope from below and remaining vigilant about altitude and distance. What's that thing that comes between AVIATE and COMMUNICATE again?

    I favor an internal guidance safeguard like EGWPS for 3D positional awareness. Maybe in the future we will have an infrared terrain-mapping version of radalt comparing itself to a 3D terrain database.

  18. #98
    Senior Member 3WE's Avatar
    Join Date
    Jan 2008
    Posts
    4,016

    Default

    Quote Originally Posted by Evan View Post
    Today, we have radar mapping synthetic vision and 3D terrain databases.
    Fixed.

    ...just not in all planes, especially older cargo planes.

    ...I still like the idea of checking your altitude at "the outer marker" and VSI afterwards...maybe even the old ILS signals are a good confirmation of what the synthetic glideslope is showing you. (Primary secondary inversion kind of thing)...
    Les règles de l'aviation de base découragent de longues périodes de dur tirer vers le haut.

  19. #99
    Senior Member Gabriel's Avatar
    Join Date
    Jan 2008
    Location
    Buenos Aires - Argentina
    Posts
    5,852

    Default

    Quote Originally Posted by Evan View Post
    I guess, since they were always over the field when in ground proximity, and TAWS doesn't know which end of the runway you are landing on, the alerts would be inhibited in landing configuration.
    I would be very disappointed if that was the case. I don't understand the workings of the EGPWS as well as the GPWS, but one of the main difference between the them is the "look ahead" feature. And "ahead" requires heading information. I would expect, for example, that while descending across certain altitude at one specific coordinate and one specific descent rate, whether the EGPWS triggers a "terrain ahead" warning will depend on whether I am heading towards the mountain or away of it

    I favor an internal guidance safeguard like EGWPS for 3D positional awareness. Maybe in the future we will have an infrared terrain-mapping version of radalt comparing itself to a 3D terrain database.
    Almost every new Cessna C-172 comes out of the factory with Synthetic Vision (SVT) where you can actually see ahead the terrain, runway and obstructions (except inconsistent obstructions like deer or trucks). EVS (enhanced vision system, using an IR camera that can look through the fog and clouds and smoke and depicting the image directly on the PFD, together with the pitch information and sometimes together with the SVT) is also already available for small GA airplanes. To be fair, at least SVT is available on most new airliners coming out of the factory too.

    https://www.google.com.mx/search?q=S...nthetic+vision
    https://www.google.com.mx/search?q=S...n+system+(evs)



    What's that thing that comes between AVIATE and COMMUNICATE again?
    NAVIGATE, but that's an emergency procedure

    --- Judge what is said by the merits of what is said, not by the credentials of who said it. ---
    --- Defend what you say with arguments, not by imposing your credentials ---

  20. #100
    Senior Member Evan's Avatar
    Join Date
    Jan 2008
    Posts
    5,361

    Default

    Quote Originally Posted by Gabriel View Post
    I would be very disappointed if that was the case. I don't understand the workings of the EGPWS as well as the GPWS, but one of the main difference between the them is the "look ahead" feature. And "ahead" requires heading information. I would expect, for example, that while descending across certain altitude at one specific coordinate and one specific descent rate, whether the EGPWS triggers a "terrain ahead" warning will depend on whether I am heading towards the mountain or away of it.
    Fair, but there are certain areas of that terrain database, namely in direct proximity to airfields, on both sides of the extended centreline where very low altitude flight will naturally occur, where EGPWS (and TAWS) MUST be inhibited. It you are outside of this area and approaching terrain threats, the system can of course sense the course you are on, but I doubt that it sounds a terrain warning in the area just beyond the overrun where this plane came down. I expect that this area is designated as an 'inhibit' zone. Might be wrong about that...

    Anyway, in this case the warnings would have needed to occur over the runway. I highly doubt EGPWS gives altitude alerts because you are too far past the TDZ. That would be cool if it did though.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •