They're Soliciting Tail-strike Solutions
Thread Starter
Join Date: Dec 1998
Location: England
Posts: 242
Likes: 0
Received 0 Likes
on
0 Posts
They're Soliciting Tail-strike Solutions
An airframe mechanic mate of mine and I were discussing the CI-611 accident and he sent me this:
The FAA a few years ago prohibited using dissimilar metals (such as stainless steel) in airframe doubler repairs because over each cycle of temperature and pressure they'd "work" and even the area being repaired (doublered) would be "worked" by the vastly different coefficients of expansion and ductility etc of a dissimilar metal (and disregarding any possible electrolytic interaction). However what they did NOT do was insist that previous repairs (the ones most at risk due to the nature of the glitch) be inspected or re-done. CI-611 was probably the legacy of that policy. Now that they've established that it's a definite problem, they are trying to head it off at the pass. Makes you wonder why their aeronautical engineers couldn't work out that any such dissimilar metal repair would work itself into a failure eventually. Think of it as the repaired section and the doubler both working on each other cyclically (and more importantly, on the riveting securing the repair).
Perhaps if they made manufacturers build longer undercarriage legs (especially for stretched models) there'd be less tail-scrapes (or the more injurious tail-hits on over-rotated landings). But that would mean more weight and less performance (and less carry-on future repair business). That engineering order we received at Eastern to get those stainless doublers off our DC-9s came from Douglas back in the late
70s when I was then assigned to structural repairs.
So some 30 years later, the FAA 'sees a problem'....
TO WHICH I REPLIED WITH THIS
Actually another way of avoiding tail-scrapes on landing would be to arrange the trailing-edge flaps such that they auto-deployed a little more during the landing flare (on a RADALT or body-angle cue perhaps) - and thus kept the body angle less pitched at oleo compression. That should be able to be contrived by the flight-control designers. In my experience, very few airframes on stretched models wouldn't have taken a hit or two on landing over the years (but admittedly, some worse than others). Having the standard fowler flap arrangement deploy a few extra degrees as the nosewheel comes off would also probably avoid the classic tailstrike because of overrotation on take-off. At take-off settings the fowler flap extends initially(without drooping much more) and at land-flap type settings it just droops, and although that creates instant lift there is much greater drag (which helps in particular for an early planting and stopping on contaminated runways). The pilot wouldn't notice the effect on take-off (except perhaps that the aircraft would seem to leap into the air). On landings he should (if it's done right by the flight-control designers) just find that he won't have to flare as much to achieve an acceptable descent rate.
I suppose you could achieve much the same thing (tail-clearance on stretched jets) on take-off by having a mechanism to fully extend the oleos at rotate. Not sure if that would/could be used to work for a late over-flared landing "save" though.
Each fuselage repair costs many millions and constitutes a future potential failure area, so the designer who comes up with a workable and patentable solution for it that doesn't lead to weight growth - well he would/should make millions (maybe).
Anybody agree (or have similar views but an alternate solution)? I hasten to add that I have had a landing tail-strike - but that was more related to idiocy/fatigue than anything else.
I'll post this in Rumours and News - and realize that it may be moved (but it's really a matter of soliciting wide-spread opinions and that may happen more readily through being initially there). Those who feel the need to suggest non-technical solutions should probably just save their verbiage. The requirement here is to eliminate as much as possible the human element (or rather, accommodate it).
The FAA a few years ago prohibited using dissimilar metals (such as stainless steel) in airframe doubler repairs because over each cycle of temperature and pressure they'd "work" and even the area being repaired (doublered) would be "worked" by the vastly different coefficients of expansion and ductility etc of a dissimilar metal (and disregarding any possible electrolytic interaction). However what they did NOT do was insist that previous repairs (the ones most at risk due to the nature of the glitch) be inspected or re-done. CI-611 was probably the legacy of that policy. Now that they've established that it's a definite problem, they are trying to head it off at the pass. Makes you wonder why their aeronautical engineers couldn't work out that any such dissimilar metal repair would work itself into a failure eventually. Think of it as the repaired section and the doubler both working on each other cyclically (and more importantly, on the riveting securing the repair).
Perhaps if they made manufacturers build longer undercarriage legs (especially for stretched models) there'd be less tail-scrapes (or the more injurious tail-hits on over-rotated landings). But that would mean more weight and less performance (and less carry-on future repair business). That engineering order we received at Eastern to get those stainless doublers off our DC-9s came from Douglas back in the late
70s when I was then assigned to structural repairs.
So some 30 years later, the FAA 'sees a problem'....
TO WHICH I REPLIED WITH THIS
Actually another way of avoiding tail-scrapes on landing would be to arrange the trailing-edge flaps such that they auto-deployed a little more during the landing flare (on a RADALT or body-angle cue perhaps) - and thus kept the body angle less pitched at oleo compression. That should be able to be contrived by the flight-control designers. In my experience, very few airframes on stretched models wouldn't have taken a hit or two on landing over the years (but admittedly, some worse than others). Having the standard fowler flap arrangement deploy a few extra degrees as the nosewheel comes off would also probably avoid the classic tailstrike because of overrotation on take-off. At take-off settings the fowler flap extends initially(without drooping much more) and at land-flap type settings it just droops, and although that creates instant lift there is much greater drag (which helps in particular for an early planting and stopping on contaminated runways). The pilot wouldn't notice the effect on take-off (except perhaps that the aircraft would seem to leap into the air). On landings he should (if it's done right by the flight-control designers) just find that he won't have to flare as much to achieve an acceptable descent rate.
I suppose you could achieve much the same thing (tail-clearance on stretched jets) on take-off by having a mechanism to fully extend the oleos at rotate. Not sure if that would/could be used to work for a late over-flared landing "save" though.
Each fuselage repair costs many millions and constitutes a future potential failure area, so the designer who comes up with a workable and patentable solution for it that doesn't lead to weight growth - well he would/should make millions (maybe).
Anybody agree (or have similar views but an alternate solution)? I hasten to add that I have had a landing tail-strike - but that was more related to idiocy/fatigue than anything else.
I'll post this in Rumours and News - and realize that it may be moved (but it's really a matter of soliciting wide-spread opinions and that may happen more readily through being initially there). Those who feel the need to suggest non-technical solutions should probably just save their verbiage. The requirement here is to eliminate as much as possible the human element (or rather, accommodate it).
Join Date: Sep 2002
Location: schiphol
Posts: 11
Likes: 0
Received 0 Likes
on
0 Posts
Maybe a tail-ground proximity warning (horn and/or light) to warn for overrotation on T/O or landing?
Or something on the PFD like the max pitch indicators when flaps are extended?
rgrds
Or something on the PFD like the max pitch indicators when flaps are extended?
rgrds
Join Date: Aug 2000
Location: British Isles
Posts: 120
Likes: 0
Received 0 Likes
on
0 Posts
Dear OVERTALK,
my 10 pennyworth
I wouldn't like any flying control doing something I hadn't told it too, that's part of the reason I don't fly Airbuses!
Tailstrikes are usually caused by incorrect technique, so the best cure I believe is through training and monitoring of trends such as the SESMA program in British Airways.
This is better than designing out the problem and encouraging bad habits which could be transferred to other types.
Best regards, PTC
my 10 pennyworth
I wouldn't like any flying control doing something I hadn't told it too, that's part of the reason I don't fly Airbuses!
Tailstrikes are usually caused by incorrect technique, so the best cure I believe is through training and monitoring of trends such as the SESMA program in British Airways.
This is better than designing out the problem and encouraging bad habits which could be transferred to other types.
Best regards, PTC
Join Date: Jan 2003
Location: Bakewell, Derbyshire, UK
Posts: 7
Likes: 0
Received 0 Likes
on
0 Posts
what about replacing the nose wheel with a wider, dual-wheel assembly, and simply digging a trench down the centre line so that you couldnt slam your ass into the ground on T/O or flaring
Join Date: Nov 1999
Posts: 324
Likes: 0
Received 0 Likes
on
0 Posts
for Peeteechase - It's not just the Airbus
We "solved" this problem on the MD-11 (finally). Some of you probably remember the AD on the MD-11 that lead to the mandatory installation of the "-908 FCC software upgrade." That product was my swan-song for MD-11 flight control development. The major addition in this software upgrade was the extension of the Pitch Rate Damping (PRD) function to encompass pressure altitudes all the way down to field elevation. There were two other additions to the LSAS (Longitudinal Stability Augmentation System) control laws that can command up to 5 Deg of elevator authority:
"Positive Nose Lowering" (PNL) - This is a subfunction that is armed for activation on landing when the automatic ground spoilers are armed. It is a two-stage elevator input that begins by adding about 2 Deg of nose-down input when the main wheels first spin up. Once the FCC "sees" the ground spoilers deploy past 10 degrees, it adds approximately another 2 degrees of nose down. One only need perform a Max braking+AGS landing both with and without the PNL function to know how far it can go to avoiding tailstrikes due to nose rise after touchdown AGS deployment.
"Pitch Attitude Protection" (PAP) - This is just a "politically correct" name for "Tailstrike Deterrence". This function is available on both takeoff and landing, as it is scheduled with respect to Radalt. The R/A schedules a maximum pitch attitude limit that, if exceeded, will cause LSAS to put in some nose-down elevator to counter the nose rise. The key attitude is 9.5 degrees at 0 feet R/A with the main gear compressed. As R/A increases, that pitch limit increases. If you are flying "hands off" (not sure who would be doing this that close to the ground, but I did a lot of it in the simulator testing this feature) LSAS will smartly bring the nose down for you. With "hands-on" it provides a nice little force deterrent to the pilot as if to say "I don't think you want to go there, buddy!" It is not so much of a force deterrent that it would be a concern of limiting the pilot's actions.... With 15-20 pounds of force you could easily pull right thru it and smack the tail.. but one hopes that below 10 feet R/A that any pilot worth his salt is trimmed-up well enough, and aware enough of his pitch attitude that this would never happen....yet it does happen on occasion!
All of this "magic" is made possible by the "hybrid" elevator control architecture of the MD-11. By "hybrid" I am referring to a mixture of conventional mechanical actuation and electronic fly-by-wire actuation. The LSAS is the software part, and it can use up to its 5 degree elevator limit to perform these functions while the pilot is flying. We refer to this as "series control" (i.e. the LSAS operates in series with the pilot's inputs).
Doing it with flaps, while it may be technically possible, would likely be quite messy and expensive, mostly due to the nature of how flap control systems are designed to be slow, methodical, and structured in such a way as to avoid high-frequency transients on the airframe response (i.e typically driven by ball-screw actuators, not the snappy response of the hydraulic actuators used for primary surfaces). I'd take a wild guess that failure modes would play a part in whether or not such a solution would ever meet safety requirements for "improbability" of a failure that could spell trouble.
I believe the LSAS is the best solution (and not just because I designed it!), because it provides the pilot with that immediate force-feedback when you are in the tailstrike neighborhood. It is tantamount to a high-time captain applying a bit of nose-down force on his column while he allows a relatively "green" F/O to complete a landing without risking the tailstrike.
"Positive Nose Lowering" (PNL) - This is a subfunction that is armed for activation on landing when the automatic ground spoilers are armed. It is a two-stage elevator input that begins by adding about 2 Deg of nose-down input when the main wheels first spin up. Once the FCC "sees" the ground spoilers deploy past 10 degrees, it adds approximately another 2 degrees of nose down. One only need perform a Max braking+AGS landing both with and without the PNL function to know how far it can go to avoiding tailstrikes due to nose rise after touchdown AGS deployment.
"Pitch Attitude Protection" (PAP) - This is just a "politically correct" name for "Tailstrike Deterrence". This function is available on both takeoff and landing, as it is scheduled with respect to Radalt. The R/A schedules a maximum pitch attitude limit that, if exceeded, will cause LSAS to put in some nose-down elevator to counter the nose rise. The key attitude is 9.5 degrees at 0 feet R/A with the main gear compressed. As R/A increases, that pitch limit increases. If you are flying "hands off" (not sure who would be doing this that close to the ground, but I did a lot of it in the simulator testing this feature) LSAS will smartly bring the nose down for you. With "hands-on" it provides a nice little force deterrent to the pilot as if to say "I don't think you want to go there, buddy!" It is not so much of a force deterrent that it would be a concern of limiting the pilot's actions.... With 15-20 pounds of force you could easily pull right thru it and smack the tail.. but one hopes that below 10 feet R/A that any pilot worth his salt is trimmed-up well enough, and aware enough of his pitch attitude that this would never happen....yet it does happen on occasion!
All of this "magic" is made possible by the "hybrid" elevator control architecture of the MD-11. By "hybrid" I am referring to a mixture of conventional mechanical actuation and electronic fly-by-wire actuation. The LSAS is the software part, and it can use up to its 5 degree elevator limit to perform these functions while the pilot is flying. We refer to this as "series control" (i.e. the LSAS operates in series with the pilot's inputs).
Doing it with flaps, while it may be technically possible, would likely be quite messy and expensive, mostly due to the nature of how flap control systems are designed to be slow, methodical, and structured in such a way as to avoid high-frequency transients on the airframe response (i.e typically driven by ball-screw actuators, not the snappy response of the hydraulic actuators used for primary surfaces). I'd take a wild guess that failure modes would play a part in whether or not such a solution would ever meet safety requirements for "improbability" of a failure that could spell trouble.
I believe the LSAS is the best solution (and not just because I designed it!), because it provides the pilot with that immediate force-feedback when you are in the tailstrike neighborhood. It is tantamount to a high-time captain applying a bit of nose-down force on his column while he allows a relatively "green" F/O to complete a landing without risking the tailstrike.
Join Date: Mar 2000
Location: Arizona USA
Posts: 8,571
Likes: 0
Received 0 Likes
on
0 Posts
Hmmm, positive nose lowering (PNL),
That's what the Captain is for when the First Officer pulls too hard on the pole during landing...or takeoff for that matter.
Slower rotation works better...and saves the a@@ end of the aeroplane from being bent.
Some will disagree...of course.
For those I can only say...watch the right seat guy, some of 'em get mighty frisky, from time to time.
That's what the Captain is for when the First Officer pulls too hard on the pole during landing...or takeoff for that matter.
Slower rotation works better...and saves the a@@ end of the aeroplane from being bent.
Some will disagree...of course.
For those I can only say...watch the right seat guy, some of 'em get mighty frisky, from time to time.
A simplistic view from somebody who does little aeroplanes for a living.
We routinely on light aircraft designs fit the back ends with sacrificial skids or in extremis wheels that aren't intended normally to be used. Whilst without a doubt it's better that the handling qualiteis prevent a tailstrike, if particular airliners are known to have a finite risk of this, I can't see why such a feature need be particularly heavy or expensive? Hey, you can probably persuade the airline beancounters that it'll pay for itself by shifting the CG aft.
I think perhaps though this should be an "as well" not an "instead" precaution.
G
We routinely on light aircraft designs fit the back ends with sacrificial skids or in extremis wheels that aren't intended normally to be used. Whilst without a doubt it's better that the handling qualiteis prevent a tailstrike, if particular airliners are known to have a finite risk of this, I can't see why such a feature need be particularly heavy or expensive? Hey, you can probably persuade the airline beancounters that it'll pay for itself by shifting the CG aft.
I think perhaps though this should be an "as well" not an "instead" precaution.
G
Join Date: Jul 2001
Location: Canada
Age: 67
Posts: 53
Likes: 0
Received 0 Likes
on
0 Posts
Genghis the Engineer, large aeroplanes do have tail skids, the 707 I think had a sacraficial wood skid. this was a rigid device with no "give". Newer and longer aircraft have "Donkeys D1cks"
which are basicaly a hydraulic or pneumatic ram with some give. They protect against a gentle tail scrape, however they will not protect against any real bang.
which are basicaly a hydraulic or pneumatic ram with some give. They protect against a gentle tail scrape, however they will not protect against any real bang.
Thank you, I am duly enlightened. I do wonder sometimes why in 5 years professional training in my youth, nobody felt it appropriate to drop some of the maths and include more useful stuff like that.
G
G
Join Date: Dec 1999
Location: LTN
Posts: 216
Likes: 0
Received 0 Likes
on
0 Posts
How about implementing a decent Flight Data Monitoring system to pick up these problems early before they turn into an incident/accident.
IMHO this may have prevent the Monarch incident in gib.
IMHO this may have prevent the Monarch incident in gib.
Or a non-linear LSS curve combined with heavily damped SPO to make elevator power reduce with increasing AoA. Would certainly be easy to include into a FBW aircraft and not that hard to include in a new conventionally controlled design.
G
G
Overrotation or Too High Deck Angle During Flare.
I think that is a simplistic treatment.
Most accidents can be blamed on the pilot. But in a very large number of cases, different aircraft characteristics would have prevented or alleviated the accident irrespective of the pilot's ineptitude. When we certify an aeroplane we are not actually trying to produce something which can't be crashed, or something that can be flown well. What we are trying to produce is something that should not, under normal circumstances (including moderate pilot ineptitude) won't have an accident.
After that we have twin objectives of something that'll do the job, and making sure that when it does all go utterly pear shaped it'll be survivable.
G
Most accidents can be blamed on the pilot. But in a very large number of cases, different aircraft characteristics would have prevented or alleviated the accident irrespective of the pilot's ineptitude. When we certify an aeroplane we are not actually trying to produce something which can't be crashed, or something that can be flown well. What we are trying to produce is something that should not, under normal circumstances (including moderate pilot ineptitude) won't have an accident.
After that we have twin objectives of something that'll do the job, and making sure that when it does all go utterly pear shaped it'll be survivable.
G
ex-Tanker
Join Date: Apr 2000
Location: Luton Beds UK
Posts: 907
Likes: 0
Received 0 Likes
on
0 Posts
A Pilot - or Engineering problem?
The pilot who started the thread has honestly stated that it happened to him, when fatigued. It could also have happened in turbulent conditions or wind shear. The only saviour here is the go around if the condition is recognised. As most of us haven't had a tail strike we aren't too cognisant of its likelihood.
Engineering tricks which reduce elevator effectiveness at high AOA may well cause embarrasment in other critical flight regimes - terrain avoidance in extremis - stall recovery etc. Flap movement during landing moves the whole ballpark as the pilot is in a deeply eyeball and hand environment - an anologue human mode, based upon short term learned tactile values. And by the way Unctious - if you had a connection with the DC-10 I don't know - the DC-10 CWS on landing would continue to trim if the pilot held off too long - and that was just as bad - causing a very long landing.
There are two areas where improvement can be made:
One is in the initial design - especially on stretched models.
The other is in training of recognition of impending critical angles - best done in a sim. and with the Go Around as correct handling.
Engineering tricks which reduce elevator effectiveness at high AOA may well cause embarrasment in other critical flight regimes - terrain avoidance in extremis - stall recovery etc. Flap movement during landing moves the whole ballpark as the pilot is in a deeply eyeball and hand environment - an anologue human mode, based upon short term learned tactile values. And by the way Unctious - if you had a connection with the DC-10 I don't know - the DC-10 CWS on landing would continue to trim if the pilot held off too long - and that was just as bad - causing a very long landing.
There are two areas where improvement can be made:
One is in the initial design - especially on stretched models.
The other is in training of recognition of impending critical angles - best done in a sim. and with the Go Around as correct handling.
Join Date: Jul 2000
Location: U.S.A.
Posts: 474
Likes: 0
Received 0 Likes
on
0 Posts
UNCTUOUS - great post. A question on the MD11 tail. It is my understanding that two MD11 aircraft suffered elevator/stabilizer damage during (false) GPWS escape manuevers in the Phillipines.
Does/can the software limit the loads the aircraft/tail will encounter with max deflection?
Does/can the software limit the loads the aircraft/tail will encounter with max deflection?
Join Date: Mar 2002
Location: N/A
Posts: 43
Likes: 0
Received 0 Likes
on
0 Posts
Capt CB's idea of a tail-ground proximity warning to warn for overrotation on T/O or landing is intriguing. I'd imagine such devices would be relatively inexpensive and simple to install...after all, even minivans are coming equipped with sonar to prevent bumping into obstacles while backing up.
An extremely simple computer could issue an aural advisory based on tail clearance and closure rate. Properly calibrated, these aural indications could provide the pilots with an accurate mental picture of tail clearance during rotation and flare.
Certainly, a proficient pilot should already have a rather accurate idea of tail clearance based on other cues. It seems to me, however, that this idea could be beneficial to pilots who are less experienced or task-saturated.
-WN
An extremely simple computer could issue an aural advisory based on tail clearance and closure rate. Properly calibrated, these aural indications could provide the pilots with an accurate mental picture of tail clearance during rotation and flare.
Certainly, a proficient pilot should already have a rather accurate idea of tail clearance based on other cues. It seems to me, however, that this idea could be beneficial to pilots who are less experienced or task-saturated.
-WN