AF 447 Thread No. 8
Join Date: Nov 2009
Location: Somewhere
Posts: 73
Likes: 0
Received 0 Likes
on
0 Posts
Just a minor observation on that Tarom incident, referring to the tendency of pilots to pull back the stick when confronted by a stall. The A310 has a yoke control. Where does that leave the "sidestick feedback" argument?
Join Date: Jul 2009
Location: France - mostly
Age: 84
Posts: 1,682
Likes: 0
Received 0 Likes
on
0 Posts
Roulis,
The C-star law does not depend on airspeed information. From Favre, C.(1994) 'Fly-by-wire for commercial aircraft: the Airbus experience', International Journal of Control, 59: 1, 139 — 157 :
The load factor comes from an accelerometer and the pitch rate from a gyro. I don't know where these are located, but my guess is in the ADIRU's under the cockpit floor.
The C-star law does not depend on airspeed information. From Favre, C.(1994) 'Fly-by-wire for commercial aircraft: the Airbus experience', International Journal of Control, 59: 1, 139 — 157 :
The load factor comes from an accelerometer and the pitch rate from a gyro. I don't know where these are located, but my guess is in the ADIRU's under the cockpit floor.
Last edited by HazelNuts39; 19th Jun 2012 at 18:46.
Join Date: Oct 2011
Location: Lower Skunk Cabbageland, WA
Age: 74
Posts: 354
Likes: 0
Received 0 Likes
on
0 Posts
@Jazz Hands,
Doesn't disqualify the SS issue from consideration at all. Other yokes have been pulled back to stall as well. Not logical, sorry, BZZZZT!
Doesn't disqualify the SS issue from consideration at all. Other yokes have been pulled back to stall as well. Not logical, sorry, BZZZZT!
Join Date: Dec 2010
Location: Middle America
Age: 84
Posts: 1,167
Likes: 0
Received 0 Likes
on
0 Posts
Regarding the occurrence of UAS at cruise and what information may have been available on June 1, 2009 refer to the BEA IR3, pages 63 and 64. It would appear to me that Air France provided (distributed) information to their crews regarding the ongoing investigation of six UAS incidents and made some recommendations. The 4th recommendation pointed out that corrections made should be "fabiles" or feeble in nature, meaning delicate or gentle. I assume this recommendation would apply to both roll and pitch corrections.
Join Date: Oct 2011
Location: Lower Skunk Cabbageland, WA
Age: 74
Posts: 354
Likes: 0
Received 0 Likes
on
0 Posts
I'm beginning to think that those pilots (447) never read anything, much less advisories from AB or AF. Ignorance is not, in this case, any kind of bliss.
Should it be mandatory for F/Os to take a quiz on everything that's handed out? Yes it should.
Should it be mandatory for F/Os to take a quiz on everything that's handed out? Yes it should.
Join Date: Jan 2005
Location: W of 30W
Posts: 1,916
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by Jazz Hands
Just a minor observation on that Tarom incident, referring to the tendency of pilots to pull back the stick when confronted by a stall. The A310 has a yoke control. Where does that leave the "sidestick feedback" argument?
The PF did not insist as he then fully pushed it forward.
Join Date: Jun 2011
Location: france
Posts: 760
Likes: 0
Received 0 Likes
on
0 Posts
Hi HazelNuts39,
Sorry to come back to C*,
the speeds which keep my interest are V and Vco, as I readed these two documents, in post #1 Thread8 AF447
Thank you for your help... Perhaps I misunderstand something, but were ? what ? why? ...
.
Sorry to come back to C*,
the speeds which keep my interest are V and Vco, as I readed these two documents, in post #1 Thread8 AF447 Thank you for your help... Perhaps I misunderstand something, but were ? what ? why? ...
Originally Posted by a) Alpa FBW Primer
C Star
C* (pronounced "C Star") is the popular name for a control law in which Nz (g) and pitch-rate feedback are blended. (In the late 60s and early 70s, Nz feedback was called the C law. NASA space shuttle approach studies added pitch-rate feedback, which was called C*.) At low speed in a C* airplane, pitch rate is primary;at higher speeds, g is primary. The changeover is transparent and occurs at about 210 knots in the A320 ("Fly-By-Wire for Commercial Aircraft: The Airbus Experience," C. Favre, 1991).
C*U ("U" represents forward velocity in flight equations) is a modified C* control law used in the B-777 to provide apparent speed stability. The trim switches set a reference speed that is summed with the actual speed in the feedback loop in such a way that the pilot feels conventional control force cues as speed changes. You "trim a speed," not the stabilizer (weight off wheels). Because the max trim reference speed is 330 knots, you would have to push on the control wheel to further increase speed toward Vmo. This provides a tactile high-speed cue.
Fly-by-wire allows designers to optimize the effective dynamics for different flight tasks--for example, an approach mode or a flare mode. This is called task tailoring and produces a multi-mode FCS.
In both the A320 and the B-777, the control laws are not fully active during takeoff until after liftoff because the sensors used for feedback would sense a lot of vibration and "noise" during the ground roll. Landing requires other transitions. Accident investigators should thoroughly understand mode transition points and effects
C* (pronounced "C Star") is the popular name for a control law in which Nz (g) and pitch-rate feedback are blended. (In the late 60s and early 70s, Nz feedback was called the C law. NASA space shuttle approach studies added pitch-rate feedback, which was called C*.) At low speed in a C* airplane, pitch rate is primary;at higher speeds, g is primary. The changeover is transparent and occurs at about 210 knots in the A320 ("Fly-By-Wire for Commercial Aircraft: The Airbus Experience," C. Favre, 1991).
C*U ("U" represents forward velocity in flight equations) is a modified C* control law used in the B-777 to provide apparent speed stability. The trim switches set a reference speed that is summed with the actual speed in the feedback loop in such a way that the pilot feels conventional control force cues as speed changes. You "trim a speed," not the stabilizer (weight off wheels). Because the max trim reference speed is 330 knots, you would have to push on the control wheel to further increase speed toward Vmo. This provides a tactile high-speed cue.
Fly-by-wire allows designers to optimize the effective dynamics for different flight tasks--for example, an approach mode or a flare mode. This is called task tailoring and produces a multi-mode FCS.
In both the A320 and the B-777, the control laws are not fully active during takeoff until after liftoff because the sensors used for feedback would sense a lot of vibration and "noise" during the ground roll. Landing requires other transitions. Accident investigators should thoroughly understand mode transition points and effects
Originally Posted by b) Cranfield, report n°9303
The application of a C* flight controllaw to large civil transport aircraft
The application of a C* flight controllaw to large civil transport aircraft
c*= Knz.nz + Kq.q (2)
q= (nz.g.Ttheta2/V) (s+1/Ttheta2) (3)
qSS=nz.g/V (4)
Kq=Vco/g (6)
q= (nz.g.Ttheta2/V) (s+1/Ttheta2) (3)
qSS=nz.g/V (4)
Kq=Vco/g (6)
Join Date: Nov 2010
Location: FR
Posts: 477
Likes: 0
Received 0 Likes
on
0 Posts
jcjeant, re: your #1339
I'm happy we agreed on the existence of the procedure.
Indeed, we can safely assume that all speed indications were false on AF447.
On the other hand, the condition for "doing nothing" is :
Si les informations erronées de vitesse ou d'altitude n'affectent pas la sécurité du vol (trajectoire stabilisée)
If the incorrect informations of speed or altitude do not affect the safe conduct of the flight (stabilized flight path)
One can argue that, coming from an A/P ON (in cruise) situation, the path was stable enough to skip the memory items. Apparently, that's the position of Mr Rosay (perhaps with his employer's interests in mind, too).
My position remains unchanged:
1/ I read the condition as "if you don't know what to set, then memory items for the time needed to pull the QRH"
2/ but even a crew which won't have the same understanding as Mr Rosay won't put its aircraft at risk by following the 5°/CLB memory items (re: excellent HN39's graphs
)
3/ given all the previous discussions, for a safer world, the "if" condition in the procedure should be clarified (perhaps it's already done, I hope so)
4/ all of this is barely related to AF447, as this crew didn't follow one (do nothing) or the other (5° and CLB) possibility.
I'm happy we agreed on the existence of the procedure.
Indeed, we can safely assume that all speed indications were false on AF447.
On the other hand, the condition for "doing nothing" is :
Si les informations erronées de vitesse ou d'altitude n'affectent pas la sécurité du vol (trajectoire stabilisée)
If the incorrect informations of speed or altitude do not affect the safe conduct of the flight (stabilized flight path)
One can argue that, coming from an A/P ON (in cruise) situation, the path was stable enough to skip the memory items. Apparently, that's the position of Mr Rosay (perhaps with his employer's interests in mind, too).
My position remains unchanged:
1/ I read the condition as "if you don't know what to set, then memory items for the time needed to pull the QRH"
2/ but even a crew which won't have the same understanding as Mr Rosay won't put its aircraft at risk by following the 5°/CLB memory items (re: excellent HN39's graphs
)3/ given all the previous discussions, for a safer world, the "if" condition in the procedure should be clarified (perhaps it's already done, I hope so)
4/ all of this is barely related to AF447, as this crew didn't follow one (do nothing) or the other (5° and CLB) possibility.
Join Date: Jul 2009
Location: France - mostly
Age: 84
Posts: 1,682
Likes: 0
Received 0 Likes
on
0 Posts
roulishollandais,
If Knz is not set equal to 1, equation (6) reads:
Kq/Knz=Vco/g
V in equation (4) and Vco in equation (6) are not airspeeds but are kinematic speeds. Equation (4) applies equally in space without air, usually written as an=ω*V, for any body moving along a curved path and maintaining its orientation relative to the direction of movement.
Originally Posted by Cranfield, report n°9303
The application of a C* flight controllaw to large civil transport aircraft
c*= Knz.nz + Kq.q (2)
qSS=nz.g/V (4)
if Knz is set equal to 1, (...):
Kq=Vco/g (6)
c*= Knz.nz + Kq.q (2)
qSS=nz.g/V (4)
if Knz is set equal to 1, (...):
Kq=Vco/g (6)
Kq/Knz=Vco/g
V in equation (4) and Vco in equation (6) are not airspeeds but are kinematic speeds. Equation (4) applies equally in space without air, usually written as an=ω*V, for any body moving along a curved path and maintaining its orientation relative to the direction of movement.
Last edited by HazelNuts39; 19th Jun 2012 at 20:28. Reason: an=ω*V added
Join Date: Jul 2009
Location: France - mostly
Age: 84
Posts: 1,682
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by roulishollandais
it means that kinetic speed information comes from inertial data after integration '?
Originally Posted by Favre (1994)
A homogeneous law, ensuring aircraft behaviour independent of the flight conditions and, in particular, independent of the centre of gravity location, is achieved by tabulating the gains as a function of the computed air speed, high-lift configuration and centre of gravity location.
Pitch = Pitch(0) + Ks/Kq * ∫ Sdt - Kn/Kq*(Vz - Vz(0))/g,
where S=side stick angle, Vz is vertical speed, and Pitch(0) and Vz(0) are the initial values of pitch and vertical speed.
P.S. About the cross-over velocity Vco.
The 'steady state' relationship between nz and q:
qSS = nz*g/V (4) may be written as:
Kn*nz = V/Vco * (Kq*q) and shown as:
Originally Posted by Cranfield report 9303
Section8.3: The cross-over velocity defines the ratio of pitch rate and normal acceleration feedback gains. (...) it had been argued that the cross-over velocity chosen had no apparent relation with the existing flying qualities specification (...) and so is open to question. By relaxing the definition of the cross-over velocity it becomes possible to specify feedback gains independently ...
Last edited by HazelNuts39; 22nd Jun 2012 at 14:43. Reason: Last line edited pending correction
Some basics of the control laws
Thanks to 'nuts and Rouli for bring up one aspect of the control laws that is independent of all the "autopilot" functions.
Part of the quote mentioned by Rouli strikes home, as it was inplemented operationally years before the shuttle entered drop tests and so forth:
The biggest thing we did not have was the airspeed feedback loop. The Nz and body rates ruled! Only thing dynamic pressure did was provide values for the "gains" that determined how much the control surfaces moved to produce the commanded Nz or roll rate ( gear up). AoA was much more important than speed, and it limited us from about 15 deg to 27 degrees with respect to gee. Figure about 15 degrees at 9 gees and 27 degrees at 1 gee ( draw the line yourself). Gear down blended AoA with the gee command and body rates. Gains were still a function of dynamic pressure. So we had the "feeling" that we were trimming for speed or AoA and not Nz as with gear up. Wasn't a real strong "feel" compared to conventional control systems, but better than a pure body rate or gee command.
Interestingly, if dynamic pressure was lost or deemed unreliable, the system used "standby gains". So gear up was 300 knots or so, and gear down was 180 knots. Simple deal, and landing gear handle switch provided the changeover. The biggest thing was the system used AoA intil the bitter end. The one example I provided was the guy who lost most of the radome due to a bird strike. This included the AoA probes. So he flew for 7 or 8 minutes using nothing but Nz and body rates and standby gains.
I am glad to see that some of our primitive control laws were improved for the Shuttle and later FBW planes.
Part of the quote mentioned by Rouli strikes home, as it was inplemented operationally years before the shuttle entered drop tests and so forth:
Originally Posted by a) Alpa FBW Primer
C Star
C* (pronounced "C Star") is the popular name for a control law in which Nz (g) and pitch-rate feedback are blended. (In the late 60s and early 70s, Nz feedback was called the C law. NASA space shuttle approach studies added pitch-rate feedback, which was called C*.) At low speed in a C* airplane, pitch rate is primary;at higher speeds, g is primary. The changeover is transparent and occurs at about 210 knots in the A320 ("Fly-By-Wire for Commercial Aircraft: The Airbus Experience," C. Favre, 1991).
C*U ("U" represents forward velocity in flight equations) is a modified C* control law used in the B-777 to provide apparent speed stability. The trim switches set a reference speed that is summed with the actual speed in the feedback loop in such a way that the pilot feels conventional control force cues as speed changes. You "trim a speed," not the stabilizer (weight off wheels). Because the max trim reference speed is 330 knots, you would have to push on the control wheel to further increase speed toward Vmo. This provides a tactile high-speed cue.
C Star
C* (pronounced "C Star") is the popular name for a control law in which Nz (g) and pitch-rate feedback are blended. (In the late 60s and early 70s, Nz feedback was called the C law. NASA space shuttle approach studies added pitch-rate feedback, which was called C*.) At low speed in a C* airplane, pitch rate is primary;at higher speeds, g is primary. The changeover is transparent and occurs at about 210 knots in the A320 ("Fly-By-Wire for Commercial Aircraft: The Airbus Experience," C. Favre, 1991).
C*U ("U" represents forward velocity in flight equations) is a modified C* control law used in the B-777 to provide apparent speed stability. The trim switches set a reference speed that is summed with the actual speed in the feedback loop in such a way that the pilot feels conventional control force cues as speed changes. You "trim a speed," not the stabilizer (weight off wheels). Because the max trim reference speed is 330 knots, you would have to push on the control wheel to further increase speed toward Vmo. This provides a tactile high-speed cue.
Interestingly, if dynamic pressure was lost or deemed unreliable, the system used "standby gains". So gear up was 300 knots or so, and gear down was 180 knots. Simple deal, and landing gear handle switch provided the changeover. The biggest thing was the system used AoA intil the bitter end. The one example I provided was the guy who lost most of the radome due to a bird strike. This included the AoA probes. So he flew for 7 or 8 minutes using nothing but Nz and body rates and standby gains.
I am glad to see that some of our primitive control laws were improved for the Shuttle and later FBW planes.
Join Date: Jun 2011
Location: france
Posts: 760
Likes: 0
Received 0 Likes
on
0 Posts
Thank you Gums and Hazelnuts39 : PPrune has the best specialists !
Reading you, Gums, i understand that it is airspeed (or an equivalent information as dynamic pressure) who determine the gains and not kinetic speed . It seems to me very logic.
Reading Hazzelnuts39 i don't understand that A330 (edit : add "does not use", del "used") does not use standby gains in case of loss of dynamic pressure, unlike the F-16. And that VCO=1200 KT.
but on the F-16 HUD (this of Gilbert Klopfstein) you have the speed vektor himself.
You are right to say and repeat that airspeed is not very important, much more is the Angle of Attack with adequate training.
But AF447 lost the PF (and the headdown Pilot managing !) without speed nor AoA. (Sure they still had pitch and power who were enough to save their lifes : but i let to other the Rosay discussion about the pitch, despite having had an ice-clogged pitot tube I guess for the "do-nothing" not written in the procedure).
It seems easy for engineers to explicite and display anywhere the inertial speed information in UAS case. K/s is not very expensive...
I will try to find the Favre's book and understand better why this C*,seems so unusefull and inadapted to transport aircraft.
Still a question, gums : why had the Pelican F-16 finally a PIO/APC
PPrune has the best specialists !Reading you, Gums, i understand that it is airspeed (or an equivalent information as dynamic pressure) who determine the gains and not kinetic speed . It seems to me very logic.
Reading Hazzelnuts39 i don't understand that A330 (edit : add "does not use", del "used") does not use standby gains in case of loss of dynamic pressure, unlike the F-16. And that VCO=1200 KT.
The biggest thing we did not have was the airspeed feedback loop. The Nz and body rates ruled! Only thing dynamic pressure did was provide values for the "gains" that determined how much the control surfaces moved to produce the commanded Nz or roll rate ( gear up). AoA was much more important than speed, and it limited us from about 15 deg to 27 degrees with respect to gee. Figure about 15 degrees at 9 gees and 27 degrees at 1 gee ( draw the line yourself).
You are right to say and repeat that airspeed is not very important, much more is the Angle of Attack with adequate training.
But AF447 lost the PF (and the headdown Pilot managing !) without speed nor AoA. (Sure they still had pitch and power who were enough to save their lifes : but i let to other the Rosay discussion about the pitch, despite having had an ice-clogged pitot tube I guess for the "do-nothing" not written in the procedure).
It seems easy for engineers to explicite and display anywhere the inertial speed information in UAS case. K/s is not very expensive...
I will try to find the Favre's book and understand better why this C*,seems so unusefull and inadapted to transport aircraft.
Still a question, gums : why had the Pelican F-16 finally a PIO/APC
Last edited by roulishollandais; 21st Jun 2012 at 16:12. Reason: grammatic : does not use standby gains
Join Date: Jun 2009
Location: NNW of Antipodes
Age: 81
Posts: 1,330
Received 0 Likes
on
0 Posts
HazelNuts39
Is that a typo? Was it meant to be 210 kt?
Thanks for all your good work on C*, and am slowly working my way through the references. I note that Boeings patent was dated 1989, which is a clear indication that "change" happens very slowly in the aviation industry.
... the A330 control law uses a cross-over velocity of 1200 kt (in alternate law)
Thanks for all your good work on C*, and am slowly working my way through the references. I note that Boeings patent was dated 1989, which is a clear indication that "change" happens very slowly in the aviation industry.
Join Date: Jul 2009
Location: France - mostly
Age: 84
Posts: 1,682
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by mm43
Is that a typo?
Vco = Kq/Kn * g = Kq/Kn * 32.17 ft/sec
Apologies for the confusion.
Last edited by HazelNuts39; 22nd Jun 2012 at 14:56. Reason: Erroneous value.
Good curve fit, 'nuts.
A year or more ago I raised the issue of "standby gains" to use a given value for dynamic pressure should the normal sensors go off to wonderland or the system deems the values FUBAR. Unfortunately, the 'bus laws are not designed to provide the AoA/speed "feeling"most pilots are used to. Only time we had that in the Viper was gear down, and even that mechanization was a kludge.
The 'bus laws are primarily a Nz command ( just as ours were, and best I can tell) and looks to me that AoA and other inputs only come into play when approaching a stall AoA. And then the system appears to ignore AoA if the dynamic/static pressure inputs are deemed to be unreliable ( and the infamous 60 knot deal), or we are in sub-mode "c", of mode "b", and the beat goes on.
As 'doze has pointed out over and over, and on end, our FLCS was designed for a different mission, but I can tell you that we would like to have had an airspeed/dynamic pressure feedback when gear up to "feel" more like a normal jet. Our feel was strickly biased according to the commanded Nz we had set using a roller wheel or the hat switch on the stick. let go of the stick and the jet assumed the "trimmed" gee, and rather aggressively, heh heh.
@ rouli
Probably the best flight instrument ever invented for us was that flight path marker in the HUD. Ooooops, the 'bus didn't have one. We had it back as far as 1968 - 1969 in the A-7D. No air data required. The sucker used inertial data only, and showed where the jet was gonna impact the ground or climb above the ridge or.... Invaluable for an instrument approach. And with AF447, it would have shown that the aircraft velocity vector was approaching zero pitch well before they flew into the stall. My lef failure video shows the ILS "cross bars" which I used to help in the dubious WX I had to deal with. You can also see where I mistakenly had my vector slightly short of the runway, as I expectd it to move down the runway when I got close and "flared". Hell, was 30 knots above approach speed, but happy I still had roll control.
http://www.sluf.org/warbirds/lef-landing.m4v
For now, I can still see a procedure change and maybe a tweak to the 'bus flight control laws and reversion sequence.
A year or more ago I raised the issue of "standby gains" to use a given value for dynamic pressure should the normal sensors go off to wonderland or the system deems the values FUBAR. Unfortunately, the 'bus laws are not designed to provide the AoA/speed "feeling"most pilots are used to. Only time we had that in the Viper was gear down, and even that mechanization was a kludge.
The 'bus laws are primarily a Nz command ( just as ours were, and best I can tell) and looks to me that AoA and other inputs only come into play when approaching a stall AoA. And then the system appears to ignore AoA if the dynamic/static pressure inputs are deemed to be unreliable ( and the infamous 60 knot deal), or we are in sub-mode "c", of mode "b", and the beat goes on.
As 'doze has pointed out over and over, and on end, our FLCS was designed for a different mission, but I can tell you that we would like to have had an airspeed/dynamic pressure feedback when gear up to "feel" more like a normal jet. Our feel was strickly biased according to the commanded Nz we had set using a roller wheel or the hat switch on the stick. let go of the stick and the jet assumed the "trimmed" gee, and rather aggressively, heh heh.
@ rouli
Probably the best flight instrument ever invented for us was that flight path marker in the HUD. Ooooops, the 'bus didn't have one. We had it back as far as 1968 - 1969 in the A-7D. No air data required. The sucker used inertial data only, and showed where the jet was gonna impact the ground or climb above the ridge or.... Invaluable for an instrument approach. And with AF447, it would have shown that the aircraft velocity vector was approaching zero pitch well before they flew into the stall. My lef failure video shows the ILS "cross bars" which I used to help in the dubious WX I had to deal with. You can also see where I mistakenly had my vector slightly short of the runway, as I expectd it to move down the runway when I got close and "flared". Hell, was 30 knots above approach speed, but happy I still had roll control.
http://www.sluf.org/warbirds/lef-landing.m4v
For now, I can still see a procedure change and maybe a tweak to the 'bus flight control laws and reversion sequence.