KE 773 landing incident at NRT
Longtimelurker
Join Date: Nov 1998
Location: killington Vt
Posts: 391
Likes: 0
Received 0 Likes
on
0 Posts
And if you are on the money you don't have to fly the other wing down..LOL.... this is really the best way to land a large jet.
Tote ..You may get better results by turning off the A/T in gusty xross wind.
Tote ..You may get better results by turning off the A/T in gusty xross wind.
Join Date: Nov 1999
Location: hoschton, GA, USA
Posts: 147
Likes: 0
Received 0 Likes
on
0 Posts
So right you are, Filew....
Can't count the number of times I've climbed the stairs at the Narita View with a bottle of Jack Daniels to watch people try to land in the fall and spring winds....
Autothrottles really hinder in gusty conditions. Speed goes low, Autothrottles spool up and nose pitches up, speed goes low, nose pitches down.
Better to turn them off and maintain one throttle setting that works, let the gusts average out.
Can't count the number of times I've climbed the stairs at the Narita View with a bottle of Jack Daniels to watch people try to land in the fall and spring winds....
Autothrottles really hinder in gusty conditions. Speed goes low, Autothrottles spool up and nose pitches up, speed goes low, nose pitches down.
Better to turn them off and maintain one throttle setting that works, let the gusts average out.
411a;
The L1011-500 was particularly susceptible to this if you grabbed a mit-full of reverse, especially as the speed and stabilizer effectiveness reduced. The solution was of course, as you say, obviously done with care. The FCOMs had very clear cautions on reverse thrust selection as well as the effect of spoiler deployment but seasoned crews understood the dynamics.
One had/has to be careful with the B767-300 to ensure the nosewheel didn't slam down in an over-aggressive response to spoiler deployment followed by a handful of reverse, however. The phenomena is known and has damaged more than one airplane.
There's nothing sweeter in flying than a well-executed, pretty crosswind landing.
aquadalte, great photos.
Quote:
Surely reverse thrust should tend to pull the nose down, and so the tail up.
LOL....you might think so, however on some types , selecting agressive reverse definitely does result in a nose-up moment....easily corrected by straightarming the column.
Surely reverse thrust should tend to pull the nose down, and so the tail up.
LOL....you might think so, however on some types , selecting agressive reverse definitely does result in a nose-up moment....easily corrected by straightarming the column.
One had/has to be careful with the B767-300 to ensure the nosewheel didn't slam down in an over-aggressive response to spoiler deployment followed by a handful of reverse, however. The phenomena is known and has damaged more than one airplane.
There's nothing sweeter in flying than a well-executed, pretty crosswind landing.
aquadalte, great photos.
Join Date: Apr 2008
Location: nowhere
Posts: 1
Likes: 0
Received 0 Likes
on
0 Posts
Totempole - are you saying that the Alteon instructors should have been teaching that awful crosswind method known as wing down?? Been on the 767 for 10 years, the 777 for nigh on 8 and have only ever used the decrab method, and have never had what you described. This method is FAR better than the wing down crosscontrolled method.
Our ops manual for the aircraft has a table which I assume comes from Boeing that lists Max crosswind for no crab. It is 24 knots at max landing weight of 164,000lbs. 20 knots at 120,000 lbs.
Last edited by JammedStab; 12th Aug 2009 at 03:40.
Technique
As always there are various methods to skin a cat. While wing down/cross controlled works fine in a Cessna/Piper or similar, and various straight winged large planes, it has issues when applied to a swept wing aircraft.
Wing down can be flown by a competent pilot on most aircraft. On the A320/330/340 etc it is not the recommended method due to the interaction of the roll control laws. (AI test pilot comment on matter: "Why would you want to do that?" (in French accent).
The application of cross control on the swept wing jet, based on flight data, shows positive correlation with Euler angle instability in roll/pitch/yaw, randomised rudder inputs resulting in roll deviations, and variations in total drag, causing airspeed variation, requiring compensation in thrust, causing changing thrust/pitch couples and flight path instability from the trim change and speed stability for the off speed conditions. The most significant factor isolated in analysis if stability on approach was the activity of the rudder; the tracking task requires changes in the rudder angle, which results in a lag before roll occurs (which is a variable lag dependent on aoa amongst other issues). In almost 50% of the cases examined where cross control was applied, the aircraft resulted in a forward slip that was pilot induced due to excessive rudder application, and resulted in high lateral loads on the gear axles.
In comparison, an approach flown does into the flare in crab has effectively no crosswind issue to deal with until the decrab is initiated, which can be after the flare has been initiated and developed. The tracking task is simplified to roll control only to adjust the track to maintain the centreline, which is an everyday tracking task. Once in the flare, the nose can be accurately yawed to align to the centreline or runway QDM visually, using very large visual cues. The associated roll lags slightly and is offset by aileron as required.
2 cases may occur to confound the above; floating or wind change around the flare changing drift. In both cases the aircraft can either be landed with drift on, (all planes can do that) if required to maintain the lateral landing zone of the runway, or can have slight into wind roll applied to minimise downwind drift.
Horses for courses; every pilot I know is the worlds best, (with every fighter pilot being the best of the best of the best) so this doesn't apply to anyone other than.... me, naturally. Unfortunately, from a systems reliability view, we work with the lowest competency that is acceptable to the system on the worst day, which is obviously no one who is reading this post.
As far as this threads topic of tail skid adjustment is concerned, as has been mentioned previously the crosswind was hardly a big deal, so any issues that are apparent have little to do with crosswind. With respect, this is not the first tail strike of a geometry limited design; I have personally watched 4 in the last 30 years (none were Asian carriers) while waiting at the holding point. The neatest was a US carrier B757-200 where the APU touched down before the main gear on a calm day.
The majority of pod scrapes, and high lateral g landings I have had to look at in the past have been related to poorly conducted cross control approaches. This doesn't mean you cannot do it, it just intimates that occasionally it is not done very well.
As far as landing on a single wheel, while the CS25 requires a single wheel touchdown load analysis to be conducted by the manufacturer, 14 CFR Part 25 does not. Boeing however does assess the structural load as such. ( At least one other manufacturer didn't...)
Alteon (Boeing) generally teaches what the customer has requested, if it is within the discretionary bounds of operational technique.
Final thought: if you really wish to persist with crossing controls on a swept wing jet, for those who have spoilers incorporated for supplementary or primary roll control, have you thought of what your stall speed becomes when you are in unbalanced flight, with spoilers up on a single wing? I lost a friend in such a case, along with the whole crew when they unfortunately found the answer at 5000' agl.
Warm Springs, GA
Wing down can be flown by a competent pilot on most aircraft. On the A320/330/340 etc it is not the recommended method due to the interaction of the roll control laws. (AI test pilot comment on matter: "Why would you want to do that?" (in French accent).
The application of cross control on the swept wing jet, based on flight data, shows positive correlation with Euler angle instability in roll/pitch/yaw, randomised rudder inputs resulting in roll deviations, and variations in total drag, causing airspeed variation, requiring compensation in thrust, causing changing thrust/pitch couples and flight path instability from the trim change and speed stability for the off speed conditions. The most significant factor isolated in analysis if stability on approach was the activity of the rudder; the tracking task requires changes in the rudder angle, which results in a lag before roll occurs (which is a variable lag dependent on aoa amongst other issues). In almost 50% of the cases examined where cross control was applied, the aircraft resulted in a forward slip that was pilot induced due to excessive rudder application, and resulted in high lateral loads on the gear axles.
In comparison, an approach flown does into the flare in crab has effectively no crosswind issue to deal with until the decrab is initiated, which can be after the flare has been initiated and developed. The tracking task is simplified to roll control only to adjust the track to maintain the centreline, which is an everyday tracking task. Once in the flare, the nose can be accurately yawed to align to the centreline or runway QDM visually, using very large visual cues. The associated roll lags slightly and is offset by aileron as required.
2 cases may occur to confound the above; floating or wind change around the flare changing drift. In both cases the aircraft can either be landed with drift on, (all planes can do that) if required to maintain the lateral landing zone of the runway, or can have slight into wind roll applied to minimise downwind drift.
Horses for courses; every pilot I know is the worlds best, (with every fighter pilot being the best of the best of the best) so this doesn't apply to anyone other than.... me, naturally. Unfortunately, from a systems reliability view, we work with the lowest competency that is acceptable to the system on the worst day, which is obviously no one who is reading this post.
As far as this threads topic of tail skid adjustment is concerned, as has been mentioned previously the crosswind was hardly a big deal, so any issues that are apparent have little to do with crosswind. With respect, this is not the first tail strike of a geometry limited design; I have personally watched 4 in the last 30 years (none were Asian carriers) while waiting at the holding point. The neatest was a US carrier B757-200 where the APU touched down before the main gear on a calm day.
The majority of pod scrapes, and high lateral g landings I have had to look at in the past have been related to poorly conducted cross control approaches. This doesn't mean you cannot do it, it just intimates that occasionally it is not done very well.
As far as landing on a single wheel, while the CS25 requires a single wheel touchdown load analysis to be conducted by the manufacturer, 14 CFR Part 25 does not. Boeing however does assess the structural load as such. ( At least one other manufacturer didn't...)
Alteon (Boeing) generally teaches what the customer has requested, if it is within the discretionary bounds of operational technique.
Final thought: if you really wish to persist with crossing controls on a swept wing jet, for those who have spoilers incorporated for supplementary or primary roll control, have you thought of what your stall speed becomes when you are in unbalanced flight, with spoilers up on a single wing? I lost a friend in such a case, along with the whole crew when they unfortunately found the answer at 5000' agl.
Warm Springs, GA
Join Date: Jul 2009
Location: Nanaimo
Age: 75
Posts: 34
Likes: 0
Received 0 Likes
on
0 Posts
Typical quick draw Mcgraws........I have not advocated cross controls all the way to touchdowns! By all means crab all the way ( that's what I do ) to about 100 feet, then ease on the downwind rudder to decrab but ease the upwind down. In very strong gusty/variable crosswinds, it would be prudent to keep some crab and straighten on the runway....you will get some uncomfortable " twisting sensation " but no harm. You have to understand the B777 have a long wheel truck but aft steerable wheels ( which touches downfirst ), so this " twisting sensation " is just the mechanics of the aircraft forward inertia/momentum at work. Non B777 qualified instructors misinterpret this as hard landing with residual drift. The autopilot touches down in crosswinds in this fashion.
I have seen people decrabbing fully on flare only to overcontrol with upwind wings lifted up or have the aircraft with one wheel truck bearing the full weight for extended periods due to too much into wind control inputs due to abrupt last minute decrab at flare.
I have seen people decrabbing fully on flare only to overcontrol with upwind wings lifted up or have the aircraft with one wheel truck bearing the full weight for extended periods due to too much into wind control inputs due to abrupt last minute decrab at flare.
Last edited by totempole; 12th Aug 2009 at 03:57.
B777 X wind oddities
the B777 has a couple of very minor unusual crosswind characteristics:
Both 200 & 300 in a strong crosswind will (without rudder input) will initially diverge track towards the downwind side... up to approximately 60kts, thereafter the aircraft will start to diverge track towards the windward side of the runway. The B747/744/757/767, MD11, A300, A320, A330 and A340 don't do this... The -200 simulator does replicate this characteristic.
The B777-300 in a strong crosswind, above 35kts component, when doing a decrab, the cockpit (pilot eye) lateral position relative to the runway does not move downwind, the gear position shifts upwind. A B747/744 or B767 does the opposite, as do long body Airbus', the pilot eye position moves downwind, and the gear position remains fairly constant in relationship to the runway centerline. -200 simulator doesn't do this, nor does the plane, -300 simulator unknown.
In respect to the lateral forces on touchdown, the aircraft acts as a normal plane with the cg forward of the gear, and will align fairly tidily with the aircraft track as a natural dynamic. The lateral load at touchdown is quite high in such a case, landing with full decrab, but if the pilot has applied a yaw input prior to the touch, it makes a significant reduction to the peak lateral g recorded and felt in the aircraft. As long as the yaw rate has been initiated, there is a lessening of the peak acceleration going from no yaw* to the inertially developed rate.
A similar issue occurs in the measurement of landing g, if the data analysis is merely based on peak recorded values; the accelerometer is not at the cg, or at the centroid of all forces, and records both vertical acceleration against the fixed body axis, and also the rotation of the body. So a late flare will result in additive accelerations being measured, that of the pitch rate and the gear rebound. (simplistically). The -300 accelerometer is... almost exactly the fuselage plug size forward of the -200's (surprising as the location is determined normally by analysis of the natural fuselage harmonic node locations, and is located at or near a node...) and the screening value for hard landings where applied is about 0.2g higher as a result to give rational data. Same issue apples for the lateral g sensing.
In relation to the simulator fidelity in general, the landing phase is not a bad representation at all, and is required by existing MOS to be valid. The values of almost all parameters of static and dynamics is in the the region of no more than 10% (very simplistically) from flight data, and the time transport delays are pretty small. The simulator problems of fidelity should (if the sim is being maintained with any semblance of quality assurance) only be significant outside of the normal operational envelope. Within the envelope the data has to have been either recorded or modeled for validation. The biggest problem is wherever the operation encounters non linear effects such as cross coupling in a stall condition.
Fidelity issues of note were a 3 holer (XXL size) fully stalled at 630,000 climbing at 6000 FPM, with 186,000 lbs of thrust installed, small twin jet stalled with full rudder applied, and ailerons controlling the roll comfortably (accident data shows same aircraft condition cross coupling with a roll rate in excess of 180 degrees a second). Latter case dynamics are now updated, finally, and may be more representative of reality, an off center axis entry into the microburst model will result in departure if the plane is stalled while at a high yaw rate due to the microburst. (now crew occasionally complain about the "unreasonable" handling of the simulator in this situation). A related issue that was noted on a series of sims evaluated a few years ago was that out of wind aileron did non give any roll effect in crosswinds up to 50kts, until the nose wheel was lifted off the ground, then it made quite a difference. In this case there are aerodynamic and geometry issues that effect the roll authority at low angels of attack, but there was no discernible effect identified by the meta centric height at any of the gear. Flight data showed that real planes do notice aileron input and crosswind effects, particularly swept wing aircraft.
The NRT thread says more about things unrelated to crosswinds, and if recent mutterings from Randy Babbitt are valid, are not isolated geograpically, but are an indication of global systemic malaise.
Warm Springs, GA
Both 200 & 300 in a strong crosswind will (without rudder input) will initially diverge track towards the downwind side... up to approximately 60kts, thereafter the aircraft will start to diverge track towards the windward side of the runway. The B747/744/757/767, MD11, A300, A320, A330 and A340 don't do this... The -200 simulator does replicate this characteristic.
The B777-300 in a strong crosswind, above 35kts component, when doing a decrab, the cockpit (pilot eye) lateral position relative to the runway does not move downwind, the gear position shifts upwind. A B747/744 or B767 does the opposite, as do long body Airbus', the pilot eye position moves downwind, and the gear position remains fairly constant in relationship to the runway centerline. -200 simulator doesn't do this, nor does the plane, -300 simulator unknown.
In respect to the lateral forces on touchdown, the aircraft acts as a normal plane with the cg forward of the gear, and will align fairly tidily with the aircraft track as a natural dynamic. The lateral load at touchdown is quite high in such a case, landing with full decrab, but if the pilot has applied a yaw input prior to the touch, it makes a significant reduction to the peak lateral g recorded and felt in the aircraft. As long as the yaw rate has been initiated, there is a lessening of the peak acceleration going from no yaw* to the inertially developed rate.
A similar issue occurs in the measurement of landing g, if the data analysis is merely based on peak recorded values; the accelerometer is not at the cg, or at the centroid of all forces, and records both vertical acceleration against the fixed body axis, and also the rotation of the body. So a late flare will result in additive accelerations being measured, that of the pitch rate and the gear rebound. (simplistically). The -300 accelerometer is... almost exactly the fuselage plug size forward of the -200's (surprising as the location is determined normally by analysis of the natural fuselage harmonic node locations, and is located at or near a node...) and the screening value for hard landings where applied is about 0.2g higher as a result to give rational data. Same issue apples for the lateral g sensing.
In relation to the simulator fidelity in general, the landing phase is not a bad representation at all, and is required by existing MOS to be valid. The values of almost all parameters of static and dynamics is in the the region of no more than 10% (very simplistically) from flight data, and the time transport delays are pretty small. The simulator problems of fidelity should (if the sim is being maintained with any semblance of quality assurance) only be significant outside of the normal operational envelope. Within the envelope the data has to have been either recorded or modeled for validation. The biggest problem is wherever the operation encounters non linear effects such as cross coupling in a stall condition.
Fidelity issues of note were a 3 holer (XXL size) fully stalled at 630,000 climbing at 6000 FPM, with 186,000 lbs of thrust installed, small twin jet stalled with full rudder applied, and ailerons controlling the roll comfortably (accident data shows same aircraft condition cross coupling with a roll rate in excess of 180 degrees a second). Latter case dynamics are now updated, finally, and may be more representative of reality, an off center axis entry into the microburst model will result in departure if the plane is stalled while at a high yaw rate due to the microburst. (now crew occasionally complain about the "unreasonable" handling of the simulator in this situation). A related issue that was noted on a series of sims evaluated a few years ago was that out of wind aileron did non give any roll effect in crosswinds up to 50kts, until the nose wheel was lifted off the ground, then it made quite a difference. In this case there are aerodynamic and geometry issues that effect the roll authority at low angels of attack, but there was no discernible effect identified by the meta centric height at any of the gear. Flight data showed that real planes do notice aileron input and crosswind effects, particularly swept wing aircraft.
The NRT thread says more about things unrelated to crosswinds, and if recent mutterings from Randy Babbitt are valid, are not isolated geograpically, but are an indication of global systemic malaise.
Warm Springs, GA
Join Date: Jul 2003
Location: Canada
Posts: 257
Likes: 0
Received 0 Likes
on
0 Posts
As always there are various methods to skin a cat. While wing down/cross controlled works fine in a Cessna/Piper or similar, and various straight winged large planes, it has issues when applied to a swept wing aircraft.
I currently own an (aerobatic) Beech Sundowner 180. The Cessna (a 150) worked just fine with the wing-low method. My Cherokee managed it OK as well. But my current Beech doesn't do so well with that method. The cross-control creates a lot of drag, and in gusty conditions juggling the power, amount of rudder, and aileron makes for an unstabilized approach. Like you big boys, this is a plane that does best with a stabilized approach with speed nailed right on 80 knots (flapless or 1 notch), 75 knots (2nd notch of flaps) or 70 knots (full flaps). It's a plane that has a "reputation" of biting on landing with porpoising, and tight speed control is the antidote to this. The one-wing-low method makes this very difficult to achieve and results in very sloppy x-wind landings.
So my x-wind method is to crab it down all the way into the flare and then straighten with rudder, adding a touch of in-wind aileron and downwind rudder to land on the upwind main first.
It gives a very nice chirp-chirp-chirp every time, very elegant and using this method it is very easy to control even at the max demonstrated 17 kt crosswind limit (and beyond if you get it right). I've impressed many a flying buddy with this method, but the reality is that the very responsive roll rate of the Sundowner deserves most of the credit! Working in this manner, I have never been "bitten" by a impending series of pilot-induced oscillations.
Beech
Join Date: Aug 2008
Location: Falls Church
Age: 78
Posts: 21
Likes: 0
Received 0 Likes
on
0 Posts
Wow, fdr....impressive but my gyros toppled in the midst of reading it. What totempole had described, I have seen it happened. Me thinks the easiest approach would be to do like what the autopilot do in a crosswind approach and landing to touchdown.
The KAL B773 NRT incident was exacerbated by the airline's recent emphasis on " reactive " simulator training requiring bounced landing recovery. Just because other fleets ( A330, B737, B744 and A300-600 ) had many incidences of botched bounced landings, the smart asses made it compulsory for all fleets to practise bounced landing recovery......sometimes with the unfortunate overemphasis. As far as I have experienced on the B777 for the last decade, it is extremely hard to bounce. The recent sim training must have contributed to the PF's misdiagnosing gear strut compression as a bounce.
The KAL B773 NRT incident was exacerbated by the airline's recent emphasis on " reactive " simulator training requiring bounced landing recovery. Just because other fleets ( A330, B737, B744 and A300-600 ) had many incidences of botched bounced landings, the smart asses made it compulsory for all fleets to practise bounced landing recovery......sometimes with the unfortunate overemphasis. As far as I have experienced on the B777 for the last decade, it is extremely hard to bounce. The recent sim training must have contributed to the PF's misdiagnosing gear strut compression as a bounce.
Join Date: Apr 2008
Location: nowhere
Posts: 1
Likes: 0
Received 0 Likes
on
0 Posts
Final thought: if you really wish to persist with crossing controls on a swept wing jet, for those who have spoilers incorporated for supplementary or primary roll control, have you thought of what your stall speed becomes when you are in unbalanced flight, with spoilers up on a single wing? I lost a friend in such a case, along with the whole crew when they unfortunately found the answer at 5000' agl.
What happened to the crew at 5,000 feet?
the aircraft stalled approximately 50kts above Vs1.0g speed. The resulting rotation rate was sufficient to throw an outboard engine off the pylon.
Splash,
grief,
flowers,
eulogies etc.
This was an extreme case, as it occurred in a VMCA demonstration with rudder boost off, but the physics remains, high beta angles, and spoiler rise results in an elevated stall speedparticularly on a swept wing. If trying to maintain a total lift evel such as in level flight, or a constant flight path, then the aoa is increased. Recently listened to a qualified & experienced group of crew discussing the IAS bands on a B744 with speedbrake raised, comment was that the increased indication is "just a glitch in the display...". Really?
Being out of balance wil increase the stall speed where spoilers are used for roll control. Using speedbrake will for a 1.0g condition increase stall speed. The amounts are dependent on the aircraft design.
Splash,
grief,
flowers,
eulogies etc.
This was an extreme case, as it occurred in a VMCA demonstration with rudder boost off, but the physics remains, high beta angles, and spoiler rise results in an elevated stall speedparticularly on a swept wing. If trying to maintain a total lift evel such as in level flight, or a constant flight path, then the aoa is increased. Recently listened to a qualified & experienced group of crew discussing the IAS bands on a B744 with speedbrake raised, comment was that the increased indication is "just a glitch in the display...". Really?
Being out of balance wil increase the stall speed where spoilers are used for roll control. Using speedbrake will for a 1.0g condition increase stall speed. The amounts are dependent on the aircraft design.
Last edited by fdr; 13th Aug 2009 at 03:12.
Join Date: Jan 2008
Location: Australia
Posts: 125
Likes: 0
Received 0 Likes
on
0 Posts
A few examples of Croswind landings at NRT...all footage taken on the SAME DAY allowing for comparison of technique and aircraft performance...
An ANA 777
YouTube - Crosswind Landing - by All Nippon Airways (ANA) Boeing 777-200ER ?JA717A?
An MAS 777
YouTube - Crosswind Landing - by Malaysia Airlines Boeing 777-200ER ?9M-MRJ?
A QF A330
YouTube - Crosswind Landing - by Qantas Airbus A330-300 ?VH-QPG?
A FEDEX MD11
YouTube - Crosswind Landing - by FedEx Express McDonnell Douglas MD-11(F) ?N587FE?
An SIA A380
YouTube - [720p] Crosswind Landing - by Singapore Airlines Airbus A380-800 ?9V-SKD?
An AF 747
YouTube - [720p] Hard Landing!! @ Narita - by Air France Cargo Boeing 747-400ERF ?F-GIUC?
An ANA 777
YouTube - Crosswind Landing - by All Nippon Airways (ANA) Boeing 777-200ER ?JA717A?
An MAS 777
YouTube - Crosswind Landing - by Malaysia Airlines Boeing 777-200ER ?9M-MRJ?
A QF A330
YouTube - Crosswind Landing - by Qantas Airbus A330-300 ?VH-QPG?
A FEDEX MD11
YouTube - Crosswind Landing - by FedEx Express McDonnell Douglas MD-11(F) ?N587FE?
An SIA A380
YouTube - [720p] Crosswind Landing - by Singapore Airlines Airbus A380-800 ?9V-SKD?
An AF 747
YouTube - [720p] Hard Landing!! @ Narita - by Air France Cargo Boeing 747-400ERF ?F-GIUC?
fdr
Having only flown the -200, I'll take your word for it but I'm interested in the physics of it... From the description, it sounds like the CG rotates around the nose of the 'plane rather than the other way round as per normal aircraft. Are Boeing using some sort of exotic matter to build the -300? 
I'm approaching 5 digits of flying time on the 777 and, like gleneagles, would say that aping what the autopilot does is not a bad place to start. I've witnessed many different approaches to the problem, some of which resulted in valid solutions!
As an aside, the proponents of decrab vs. wing down usually demonstrate what I call the "very late wing down" technique...
The B777-300 in a strong crosswind, above 35kts component, when doing a decrab, the cockpit (pilot eye) lateral position relative to the runway does not move downwind, the gear position shifts upwind.
I'm approaching 5 digits of flying time on the 777 and, like gleneagles, would say that aping what the autopilot does is not a bad place to start. I've witnessed many different approaches to the problem, some of which resulted in valid solutions!
As an aside, the proponents of decrab vs. wing down usually demonstrate what I call the "very late wing down" technique...
Join Date: Jan 2008
Location: Herts, UK
Posts: 748
Likes: 0
Received 0 Likes
on
0 Posts
From the description, it sounds like the CG rotates around the nose of the 'plane rather than the other way round as per normal aircraft. Are Boeing using some sort of exotic matter to build the -300?
I imagine it's a case of inertial Vs aero forces dominating the movement... if the resistive forces to slow(er) movements are dominated by aero forces the overall aero centres move the effective 'hinge' - in this case the fuselage yaw centre, fwd of the c.g.
If only a nosewheel is on the ground then it's not hard to picture this as the pont where rotation might be centred due to it's lateral friction, regardless of the c.g.
Join Date: Mar 2000
Location: Arizona USA
Posts: 8,571
Likes: 0
Received 0 Likes
on
0 Posts
This was an extreme case, as it occurred in a VMCA demonstration with rudder boost off, but the physics remains, high beta angles, and spoiler rise results in an elevated stall speedparticularly on a swept wing.
HarryMann;
Regarding "inertial" vs "aero" forces, were you thinking solely of the above case, "only a nosewheel on the ground"? While I'm trying to picture the dynamics you describe for an aircraft in the air, I can't imagine where any transport would be only on the nosewheel, "about which the airplane would pivot", so I'm sure I'm misunderstanding what you mean.
fdr;
This may be "off-thread" but the thread has wandered anyway so I will risk it in favour of understanding.
While your comments were a good read, I'm still puzzling over the dynamics that you describe thus:
Landing the DC8-61 and 63 series aircraft on US runways (150' wide, Canada's and the ones used in Europe were 200' wide), in a heavy crosswind meant that one placed the cockpit slightly on the upwind side of the centerline so that the gear was on the centerline at touchdown.
A de-crabbing of a long fuselage meant either the cockpit was on centerline and the gear was off the centerline on the downwind side, or the cockpit was flown slightly upwind of the extended centerline to place the gear on the centerline. I would have thought this was well understood on even longer aircraft such as the B777-300 and A340-600 but you say these two aircraft, for example, behave exactly opposite to one another in very high crosswinds. One decrabs "at the cockpit" and the other "at the center of gravity", roughly. I'm trying to understand how.
I can see, given the mass involved, that the pivot may shift slightly due to the effects of the crosswind upon the fuselage's length but it seems that this would apply uniformly over all types.
I'm also still trying to understand the following comment from fdr and hope perhaps a more detailed explanation might be offered. Not doubting, yet, but trying to imagine the dynamics and "why?" For me, it makes no sense at all but I'm suspending judgement in favour of curiosity:
Regarding the vertical 'g' parameter for the B777 and your comment:
A UDRI, (University of Dayton Research Institute) paper for the FAA entitled, Statistical Loads Data for the Boeing 777-200ER Aircraft in Commercial Operations, discusses these factors at length. The paper also discusses, as do your comments, the variability of the vertical loads accelerometer. In fact we noticed early on that the B777-300 and -200 series vertical acceleration parameter was unreliable, tending to be "spikey", triggering many false "hard landing" events. Here are the comments from the FAA document.
The B777 AMM states quite clearly at the beginning of the hard landing check, that unless the pilot reports the landing as hard, it didn't happen and no maintenance inspection will take place. The inspection procedure then goes on to outline the 'g' limits requiring different inspection levels. The 'g' limits change according to roll angle and samples per second, (8x or 16x - ours were 10x).
My main question is about the de-crabbing phenomenon you describe and how it happens. Thanks fdr.
If only a nosewheel is on the ground then it's not hard to picture this as the pont where rotation might be centred due to it's lateral friction, regardless of the c.g.
fdr;
This may be "off-thread" but the thread has wandered anyway so I will risk it in favour of understanding.
While your comments were a good read, I'm still puzzling over the dynamics that you describe thus:
Originally Posted by fdr
The B777-300 in a strong crosswind, above 35kts component, when doing a decrab, the cockpit (pilot eye) lateral position relative to the runway does not move downwind, the gear position shifts upwind. A B747/744 or B767 does the opposite, as do long body Airbus', the pilot eye position moves downwind, and the gear position remains fairly constant in relationship to the runway centerline. -200 simulator doesn't do this, nor does the plane, -300 simulator unknown.
A de-crabbing of a long fuselage meant either the cockpit was on centerline and the gear was off the centerline on the downwind side, or the cockpit was flown slightly upwind of the extended centerline to place the gear on the centerline. I would have thought this was well understood on even longer aircraft such as the B777-300 and A340-600 but you say these two aircraft, for example, behave exactly opposite to one another in very high crosswinds. One decrabs "at the cockpit" and the other "at the center of gravity", roughly. I'm trying to understand how.
I can see, given the mass involved, that the pivot may shift slightly due to the effects of the crosswind upon the fuselage's length but it seems that this would apply uniformly over all types.
I'm also still trying to understand the following comment from fdr and hope perhaps a more detailed explanation might be offered. Not doubting, yet, but trying to imagine the dynamics and "why?" For me, it makes no sense at all but I'm suspending judgement in favour of curiosity:
Both 200 & 300 in a strong crosswind will (without rudder input) will initially diverge track towards the downwind side... up to approximately 60kts, thereafter the aircraft will start to diverge track towards the windward side of the runway. The B747/744/757/767, MD11, A300, A320, A330 and A340 don't do this... The -200 simulator does replicate this characteristic.
A similar issue occurs in the measurement of landing g, if the data analysis is merely based on peak recorded values; the accelerometer is not at the cg, or at the centroid of all forces, and records both vertical acceleration against the fixed body axis, and also the rotation of the body. So a late flare will result in additive accelerations being measured, that of the pitch rate and the gear rebound. (simplistically). The -300 accelerometer is... almost exactly the fuselage plug size forward of the -200's (surprising as the location is determined normally by analysis of the natural fuselage harmonic node locations, and is located at or near a node...) and the screening value for hard landings where applied is about 0.2g higher as a result to give rational data. Same issue apples for the lateral g sensing.
The installation of acceleration sensors in the cockpit rather than near the center of gravity is a deviation, albeit approved by the appropriate airworthiness authorities, from the Federal Aviation Regulation specifications. The transfer functions and filtering techniques used in the transfer of the sensed accelerations at the cockpit to those expected at the center of gravity are oriented towards flight conditions. As a result, the aircraft dynamics experienced during landing conditions are not properly accounted for and the recorded accelerations for these conditions will include unknown errors. These errors were clearly manifested in the vertical and lateral load factors recorded for the touchdown condition when comparing the B-777-200ER load factors with those of other aircraft. This is of particular concern if vertical load factors recorded during touchdown were to be used by the airlines for routine hard landings identification or by the National Transportation Safety Board in landing accident investigations. In such cases, the use of the recorded values would be very questionable. It is recommended that a task be considered that will install a center of gravity accelerometer in one airplane and record both the converted and direct measured accelerations to determine the extent of the differences that may exist. Perhaps some airline might be willing to consider such a task on one airplane for a few hundred hours to at the very least resolve any hard landing concerns.
My main question is about the de-crabbing phenomenon you describe and how it happens. Thanks fdr.
Join Date: Nov 2004
Location: Here, there, and everywhere
Posts: 1,122
Likes: 0
Received 12 Likes
on
7 Posts
Join Date: Oct 2006
Location: Gone Flying...
Age: 63
Posts: 270
Likes: 0
Received 0 Likes
on
0 Posts
Looks to me like most of the spoilers on the left wing didn't deploy after touchdown(anyone know why) and he might have come close to scraping an engine pod.
aquadalte;
That is precisely why all the spoilers except for the one ground spoiler on the left wing are down; they are responding as ordered, with right full control wheel. Notice how deeply the right wing dips at touchdown as the right wing spoilers come up - the left wing is still flying - exactly the opposite outcome intended by the use of full-right control wheel, (which is, "keep the upwind wing from rising"). The engine comes very close to the runway. The left spoilers slowly come up near the end of the landing roll as the wheel is neutralized.
Could it be because the pilot applied full right yoke? (during the roll out, after touch down.)
Join Date: Nov 1999
Location: hoschton, GA, USA
Posts: 147
Likes: 0
Received 0 Likes
on
0 Posts
Can't speak for other airlines, but my flight manual (B-777) says to to forward slip up to 31 knots crosswind component, and use a combination of crab/slip up to 45 knots crosswind componet, lining up on the upwind runway edge lights.
Always fun to watch the First Officers face.
When you kick out of the crab/slip, the airplane lands on the centerline.
Be that as it may, I have been flying in and out of Narita since 1988. Narita can have the worse oreographic turbulence landing on 16 that I have ever encountered.
Usually approaches to NRT are smooth (laminar air) until the last 300 feet if there is a strong crosswind. Then all hell breaks loose.
Never did like golf or golf courses, even though I live on one. For those of you that don't know, there is an undulating golf course right next to the NRT airport. Source of oreographic turbulence.
Always fun to watch the First Officers face.
When you kick out of the crab/slip, the airplane lands on the centerline.
Be that as it may, I have been flying in and out of Narita since 1988. Narita can have the worse oreographic turbulence landing on 16 that I have ever encountered.
Usually approaches to NRT are smooth (laminar air) until the last 300 feet if there is a strong crosswind. Then all hell breaks loose.
Never did like golf or golf courses, even though I live on one. For those of you that don't know, there is an undulating golf course right next to the NRT airport. Source of oreographic turbulence.