Habsheim

[Linktrained]

O. W.


Thank you for the tree height. Perhaps this was not clear enough to whoever planned this flight.


( Someone born in the late 1840s advised her R.F.C. pupil pilot son "not to fly too high or too fast"... Probably she was thinking of 20 ft. and 20 mph. would be enough. Advice which he did not follow !)

[HazelNuts39]

Hi gums,

the magnitude of the phugoid oscillations is a function of the thrust change from that required for level flight. If thrust had been set for a gradient of 4 degrees the amplitude of the speed oscillation would have been halved and 100 ft achieved after 10 seconds.

A gradient of 8.5 degrees with landing flap and gear down near stall speed is quite healthy for a transport airplane.

[gums]

Good points, Okie.

I also remind folks of Sioux City and the Carmel accidents where the crews used only power once they established a "trimmed" AoA/airspeed that the plane "liked".

I gotta tellya, if we gave out a "Medal of Honor" for airline pilots, the Constellation pilot would have been a sure fire recipient. He was the last troop to try leave the plane once he did his duty to the pax. Sully stayed behind and made it, this guy didn't. Amazing the survival rate.

1965 Carmel mid-air collision - Wikipedia, the free encyclopedia

[Saint-Ex]

Strange how we ignore OO-DLL at Baghdad. Perhaps because "it was just a cargo flight." The crew had to cope with an A300 severely damaged on the left wing by a missile strike so causing extra drag from the damage and still landed at BIAP at less than 2g. A DH structural engineer estimated the wing would have departed after a further 20 minutes.

[DozyWannabe]

@OG - I stand corrected, my apologies.

[CONF iture]

Originally Posted by Owain Glyndwr

Consequently I see no reason why the point I was emphasising :
"the EFCS behaves as a damper of the oscillations, commanding appropriate variations of angle of attack in a way that, when the aircraft is slowing down, makes it pitch downward and vice versa"
should not be a valid explanation of the reason why alphamax was not developed at Habsheim.

Simply because you omit the necessary initial conditions :
"Additionally, in these high angle of attack situations, with a tendency towards phugoid movement, studied by longitudinal stability, in which the aeroplane oscillates between two kinetic and potential energy levels"
Such oscillations were taking place in Bilbao due to atmospheric disturbance.
It was not the case in Habsheim.

Air Safety Week July 18 2001
To change the outcome in dynamic wind conditions near the ground, Airbus plans to modify the software to eliminate pitch rate as a controlling factor in alpha protection.

Again, you omit to take in consideration the initial conditions : "in dynamic wind conditions"
Was it the case in Habsheim ?

Originally Posted by gums

I bet that the pilot would have donated vital parts of his anatomy to gain 100 feet in about 1300 feet of horizontal travel, ya think? 7 seconds at 115 knots +/-

He didn't need 100 feet but just a tenth of it to survive the situation he put himself in and come back for a landing. It was all in the aerodynamics with still 2.5 deg to grab to get to alpha max and engines already at 83% N1.

[awblain]

The tweak to the flight control laws that damps the phugoid oscillation seems wholly irrelevant to the Habsheim situation. It's also seems irrelevant to Bilbao. A control input to ensure the smooth gain of height on a 20s timescale isn't going to put someone into a hole they're flying towards.

"due to atmospheric disturbance"… then including the tweak to the flight control or otherwise won't stop it. It's not going to make the oscillations worse.

At Bilbao, if the aircraft is being pitched around so that it enters a protection mode, then it's possible that at least briefly it needed protection - not smoothing the flight path, but avoiding stall. This might have caused it to impact the ground, but what if it had been allowed to stall and crash instead?

Is it wise to fly either into a forest, or into horrible record-setting windshear? A non FBW aircraft wouldn't help you in the circumstances either, unless you recognize the impending doom early enough, and maybe the A320 improved the situation in terms of reducing the number of pieces of wreckage to sift.

Would it be possible to argue that without a yaw damper someone might have squeaked themselves out of a particular sticky fix they'd embedded themselves into? Overall, is it a bad thing to have this mechanical protection?

[Owain Glyndwr]

Originally posted by Con_fiture

Simply because you omit the necessary initial conditions :
"Additionally, in these high angle of attack situations, with a tendency towards phugoid movement, studied by longitudinal stability, in which the aeroplane oscillates between two kinetic and potential energy levels"

Initial conditions are irrelevant - the system must be stable for any initial condition and in any atmospheric conditions.

If you want confirmation that phugoid oscillations between two kinetic and energy levels can exist without there being any turbulence just look at the graph that HN39 posted recently. That had at best neutral stability and had no additional phugoid damping - there was no turbulence present. It shows why the phugoid damping terms are necessary.

The point about the software changes made after Bilbao, which you seem to have misunderstood, was that they were made to prevent unnecessary engagement of Alpha Protection in turbulent conditions. The laws that govern the motion once Alpha Protection is engaged have been unchanged from the word go. Other than the logic governing the point of entry into alphaprot the same laws were in place at Habsheim as at Bilbao.

You are simply adding confusion by trying to conflate the two events.

You seem to be using the fact that the system description I quoted came out of the Bilbao report to suggest that it ( the system description) is somehow not applicable for Habsheim. Can we go back to discussing that accident and maybe you could tell us where/why you think the system as described in the Bilbao report, especially that part that describes the effect of speed feedback does not apply for Habsheim? For example, do you have any positive evidence that the basic laws were changed after Habsheim and before Bilbao?

[roulishollandais]

Dear Owain,

You are simply adding confusion by trying to conflate the two events

It is easy to be confused when theory of systems are not very familiar. Many pilots and engineers would be happy, feel much more comfortable, and work safer by increasing mutual transdiscipliniraty despite it needs, once again, a sharper selection, and still more time for formation.


Theory of systems and pluridisciplinarity help in all domains of modern life. Isn't ?

[HazelNuts39]

Phugoid

Originally Posted by CONF iture

Such oscillations were taking place in Bilbao due to atmospheric disturbance.

The phugoid oscillation as such has no relation with atmospheric disturbances. The initiating disturbance can be anything, a gust, longitudinal control input, thrust increase ...

QUOTE

Phugoid

The long-perioid of phugoid mode involves a trade between kinetic and potential energy. In this mode, the aircraft, at nearly constant angle of attack, climbs and slows, then dives, losing altitude while picking up speed. The motion is usually of such a long period (about 93 seconds for a 747) that it need not be highly damped for piloted aircraft. This mode was studied (and named) by Lanchester in 1908. He showed that if one assumed constant angle of attack and thrust=drag, the period of the phugoid could be written as: T = p V2 U/g = 0.138 U. That is, the period is independent of the airplane characteristics and altitude, and depends only on the trimmed airspeed.

P.S.
The formula for the period should read: T = π*√2*U/g, where U is the trimmed airspeed (TAS) and g the acceleration of gravity. With U in ft/s the period in seconds is T=0.138*U .

[Owain Glyndwr]

Dear Roulis

I would be very happy to work towards a better mutual understanding between engineers and pilots - I have been trying to do that all my working life.
But I have found that one essential element for success is that we listen to one another - and yes, I know that is a two edged sword.

[gums]

I am with OG on discussion, protocol and such.

@Nuts

How can you "trim" to TAS?
Secondly, what is happens to phugoid phenomena when thrust is greater than drag? I would imagine it would be there, but the time constant/perios/amplitude would be interesting.

I did not get a major in aero, only had the so-called "general" engineering degree, but had maybe 20+ semester hours of thermo and aero and related. Then many hours of physics, chemistry, astro, mechanics, EE, computer science and even civil engineering.

I flew 6 different jets and saw many things that verified about all of the theories I had been taught. But I never had a course about human performance until I actually flew the jets. And that course had very harsh results if you got an "F". Also flew the first "pure" FBW operational jet in the world, while recognizing that a large portion of the Concorde flight control system was FBW. In my little jet, if the electrons went away it was a nylon let down, or worse.

Since I jumped in here with AF447, I have learned an awful lot about both technical and human factors for the commercial pilots. I appreciate all of your acceptance of my "contributions.

[HazelNuts39]

gums,

How can you "trim" to TAS?

You trim to CAS or Mach, then calculate TAS to find the phugoid period.

Secondly, what is happens to phugoid phenomena when thrust is greater than drag?

It's shown in the graph I posted recently. AFAIK the period does not change, but the natural damping could be different if thrust changes with airspeed.

[gums]

@ Nuts

I can see "trimming" to a CAS/AoA, but not a mach or TAS. For a small altitude delta, the TAS or mach would not vary much. I only flew 2 jets that had a "mach hold" using otto ( F-102A and F-101B) , otherwise we humans had to maintain the mach for our climbs or level flight. In my 1800 hours +/- in straight-wing jets with no otto, I never noticed the phugoid once I had the sucker trimmed. Well, maybe +/- 50 feet and 1 or 2 knots until I had the thing nailed.

I saw your neat graphic for the phugoid, but seems the condition for the model was thrust equals drag. So my question is, "what happens when thrust is greater than drag", especially for a go-around that should have been executed in the incident? I would have preferred a graph of unrestricted FPA using the configuration and commanded gee/AoA.

As with others here, I think the phugoid considerations for the 'bus FCS had negligible effects upon the incident. However, I was surprised that the jet even had some phugoid elements to the laws. Always learning, ain't we?

[HazelNuts39]

Originally Posted by gums

I saw your neat graphic for the phugoid, but seems the condition for the model was thrust equals drag.

I described the initial condition as level flight, constant speed, i.e. thrust equal to drag. What matters for the phugoid is the thrust set at t=0, which exceeds drag by sin 8.5 degrees times weight.

I would have preferred a graph of unrestricted FPA using the configuration and commanded gee/AoA.

Sorry, but I must be misunderstanding what you are saying here. The FPA is unrestricted and the graph is independent of configuration. The only 'aerodynamics' that goes into the calculation is the fact that at constant AoA lift is proportional to airspeed-squared. "gee" is not commanded but is proportional to airspeed-squared. AoA is 'commanded' to remain constant throughout.

As with others here, I think the phugoid considerations for the 'bus FCS had negligible effects upon the incident.

What matters is not the phugoid but the control law that reduces the commanded AoA when speed decreases and/or pitch increases.

[gums]

@ Nuts

Sorry, but I must be misunderstanding what you are saying here. The FPA is unrestricted and the graph is independent of configuration. The only 'aerodynamics' that goes into the calculation is the fact that at constant AoA lift is proportional to airspeed-squared. "gee" is not commanded but is proportional to airspeed-squared. AoA is 'commanded' to remain constant throughout.

Yeah, my understanding of the plots was an 8 degree FPA or so. No biggie. I fully understand the relationships between AoA and lift and velocity.

My concerns are that many folks don't understand the actual relationhips of the gee command versus the AoA once at the "limits". You can command all the gee you wish, but the jet will only give you so much due to the FBW control laws. Been there, done that.

Thanks for the civil discourse. And out.

[Owain Glyndwr]

Originally posted by awblain

The tweak to the flight control laws that damps the phugoid oscillation seems wholly irrelevant to the Habsheim situation. It's also seems irrelevant to Bilbao. A control input to ensure the smooth gain of height on a 20s timescale isn't going to put someone into a hole they're flying towards.

It's not really irrelevant to Habsheim because although the longer term effects of the 'tweak' could not affect the motion very much in the short time available the speed feedback that provides some of the phugoid damping became operative as soon as Alpha Protect mode became active. Seeing the steady loss of airspeed during the last few seconds this term led to a nosedown signal which offset the pilot's command for more AOA and restricted the aircraft to 15 deg AOA rather than the 17.5 deg of alphamax. As I understand it, it is this restriction that Confiture claims to be a factor in the aircraft striking the trees. The significance of holding AOA below alphamax is best established by a performance discussion rather than one on the system design. As gums has pointed out, when flying close to the aircraft lift limits the available gee for a pull up is very limited.

[awblain]

OG,

OK, but in both cases, if the alternative was a stall, avoided by the system, rather than a flying impact with the ground/trees, I'm not sure that would have been better. Almost everyone did walk away.

Could either crew have honestly said "if only we'd had another 1.5 degrees, our finely planned and executed manouevre would have been a spectacular success."

It's also not clear that the phugoid is avoided by changes to the early control actions, or as it builds a few seconds later. Making a legal excuse based on the flight control systems action just suggests you'd be vulnerable to the question "So, Capt. Asseline, did you know this about the flight control software you were using before you crashed at Habsheim?"

[rudderrudderrat]

Could either crew have honestly said "if only we'd had another 1.5 degrees, our finely planned and executed manouevre would have been a spectacular success."

It's a bit like the Costa Concordia (Costa Concordia disaster - Wikipedia, the free encyclopedia) crew saying, "If only we had turned 50 feet earlier".

One has to ask, "why push so close to the limits?"

[Owain Glyndwr]

awblain

Don't get me wrong, we are on the same wavelength. I am only pointing out a mechanism whereby the observed AOA restriction can be explained. Whether it made any difference is another matter. When you write:

Could either crew have honestly said "if only we'd had another 1.5 degrees, our finely planned and executed manouevre would have been a spectacular success."

my answer is no. It would have been possible to gain a little more height at the expense of speed if the AOA had followed the pilot's command without any restriction but keeping sidestick movement and timing and throttle movement and timing unchanged, the height gain at the back end would have been no more than half of that required for clearance. Let us remember that the RA height over the last few seconds was only 30 ft and the average tree height 39ft.