Polish President and wife killed in Tu-154 crash

[Northwester]

Originally Posted by Peter_K

You seem to be very selective in choosing your facts, rejecting those that do not support your theories young man.
First Polish experts arrived on the site on the evening of April 10th. The team of Polish investigators remained there until April 28th. By that time all the recording devices were recovered and the examination of the wreckage was no longer necessary.

So why would one of the main objections from Polish side be the refusal of Russian side to provide the documentation of the crash site from the day of the crash?

[Northwester]

Originally Posted by Peter_K

You seem to be very selective in choosing your facts, rejecting those that do not support your theories young man.
First Polish experts arrived on the site on the evening of April 10th. The team of Polish investigators remained there until April 28th. By that time all the recording devices were recovered and the examination of the wreckage was no longer necessary.

I don't know if you have noticed it, young man, but my theories change as new facts are revealed. I try to look at things that still require careful examination.

[Northwester]

Originally Posted by Evan

At this point the autothrottle is still engaged. The PF adds pitch, and therefore drag, so the autothrottle adds power (N1). I don't see a plot for AoA. The Indicated Air Speed drops momentarily, then levels off as thrust takes effect and remains steady until the autothrottle is disengaged by the actual go-around.

If you look closely at N1 plots, you will notice that during autothrottle engagement, the power fluctuates quite a bit more than with manual flight, as the computer is more active and precise in adjusting to power demands.

In this case, the PF added marginal pitch, then held it. The autothrottle added a marginal degree of power. It may have been enough to slow the rate of descent somewhat, but it was nowhere near the go-around requirement.

When the actual go-around happens (at the moment of the first tree strike), the autothrottle and AP both disengage as designed.

So to make sure I understood it correctly, you are saying that the increase of thrust was balanced by the increase of pitch, since both the speed (283 to 280) and altitude decreased in that time interval?

[Northwester]

It has been just announced that Polish investigators are going to meet with some guys from FBI on Wednesday. The purpose of the meeting has not been revealed.

[Northwester]

I just noticed that the speed of the plane at the first ground contact was 269km/h. Considering that the plane ploughed the ground with its tail and wing for about 22m before hitting the ground, the speed at first impact was about 260km/h. If the first part of the plane came to a complete stop in 0.2 seconds, the max decceleration forces acting on passengers would be 37g.

[Geebee]

Originally Posted by Evan

A bit ghostly.

I'm certainly glad these aren't my tax dollars at work.

Unfortunately my tax dollars are!

There is 1 other important element to add here that hasn't been shared yet. 2011 is a year of parlementary elections. If the governing majority would just sit back and accept the report as it is, definitely it would be used against them in the fall elections. So they have to do what they have to do.

[Evan]

Originally Posted by Northwester

So to make sure I understood it correctly, you are saying that the increase of thrust was balanced by the increase of pitch, since both the speed (283 to 280) and altitude decreased in that time interval?

Based on the FDR plots and the report analysis, it looks to me that the autothrottle was reacting to a moderate pitch input by the F/O at around 80m RA, and added power (N1). The control column movement there is obviously human interaction. Notice how, previous to this, the column position mirrors the N1. Now look at the IAS. There is a slight lag while thrust takes effect, as the added drag reduced the airspeed, but then the AP overrides the input and the airspeed plot becomes a flat line until the autothrottle disconnects during the actual go-around.

Also, compare the thrust lever plot to the N1/N2 plot. You see no pilot initiated thrust input until the go-around attempt. Before that, all thrust commands are coming from the autothrottle.

[Gabriel]

Originally Posted by Northwester

I just noticed that the speed of the plane at the first ground contact was 269km/h. Considering that the plane ploughed the ground with its tail and wing for about 22m before hitting the ground, the speed at first impact was about 260km/h. If the first part of the plane came to a complete stop in 0.2 seconds, the max decceleration forces acting on passengers would be 37g.

Except that:

- You are considering a constant acceleration. Probably the deceleration of the fuselage was small while the the wingtip and tail were dragging the ground (if only because the wing and tail are not designed to withstand 2 Gs in the horizontal direction, let alone 37), and then become immense as it hit the ground itself.

- The seat will brake loose at much less than 37g, and then the passenger with seat and all will be launched against whatever where it will finally decelerate in just a few cm.

- The plane was inverted, so probably the passengers (before the seats broke loose) helped decelerate the fuselage, and not the other way around, by dragging their heads against the ground through the destroyed top of the fuselage.

- 37 Gs is most likely not survivable with a lap belt. Impacts with a brief peak of 40 Gs are considered survivable with no permanent injury if secured with a properly designed and fitted 5-points harness.

[Northwester]

Originally Posted by Evan

Based on the FDR plots and the report analysis, it looks to me that the autothrottle was reacting to a moderate pitch input by the F/O at around 80m RA, and added power (N1). The control column movement there is obviously human interaction. Notice how, previous to this, the column position mirrors the N1. Now look at the IAS. There is a slight lag while thrust takes effect, as the added drag reduced the airspeed, but then the AP overrides the input and the airspeed plot becomes a flat line until the autothrottle disconnects during the actual go-around.

Also, compare the thrust lever plot to the N1/N2 plot. You see no pilot initiated thrust input until the go-around attempt. Before that, all thrust commands are coming from the autothrottle.

That seems to be fine, except I see the first control column movement and pitch adjustment at 10:40:45, and thrust increase at 10:40:47. At that time the plane is still at 100m RA, not 80.

[Northwester]

Originally Posted by Gabriel

Except that:

- You are considering a constant acceleration. Probably the deceleration of the fuselage was small while the the wingtip and tail were dragging the ground (if only because the wing and tail are not designed to withstand 2 Gs in the horizontal direction, let alone 37), and then become immense as it hit the ground itself.

- The seat will brake loose at much less than 37g, and then the passenger with seat and all will be launched against whatever where it will finally decelerate in just a few cm.

- The plane was inverted, so probably the passengers (before the seats broke loose) helped decelerate the fuselage, and not the other way around, by dragging their heads against the ground through the destroyed top of the fuselage.

- 37 Gs is most likely not survivable with a lap belt. Impacts with a brief peak of 40 Gs are considered survivable with no permanent injury if secured with a properly designed and fitted 5-points harness.

So, in your opinion, if the plane hit the ground on its belly with the same speed, generating the same g forces, all seats would brake loose and jetisson passengers forward? What would be a statistical survival rate in such a case?

[Gabriel]

Originally Posted by Northwester

So, in your opinion, if the plane hit the ground on its belly with the same speed, generating the same g forces, all seats would brake loose and jetisson passengers forward?

Yes. Absolutely.

What would be a statistical survival rate in such a case?

Close to zero, except by those exceptional cases where extreme good luck (miracles for the believers) takes place. Being inverted just substantially reduced the chance of this good luck to materialize.

[kris]

Originally Posted by Northwester

I just noticed that the speed of the plane at the first ground contact was 269km/h. Considering that the plane ploughed the ground with its tail and wing for about 22m before hitting the ground, the speed at first impact was about 260km/h. If the first part of the plane came to a complete stop in 0.2 seconds, the max decceleration forces acting on passengers would be 37g.

Nope. That would be average value. The truth is the time (and distance) needed to decelerate may be much longer, but the bodies would travel in air with mostly zero deceleration, only to hit something solid and stop rapidly.
This is comparable (not equal, but comparable) to a high-speed car crash. While someone survives from time to time, such crashes are generally non-survivable.

Originally Posted by Northwester

Right after, at about 10:40:46 I see N1 values start increasing. The plane is at 100m altitude still descending. The speed is decreasing. If N1 values are increasing then we also have an increase of thrust. No change of angle of attack. But the altitude and speed are decreasing. How can that be explained?

I can't see anything weird there (note: the pressure altitude plot doesn't say much), but pitch seems to increase starting at ca. 10:40:45, with the AP still controlling the aircraft. Also see vertical accel, control column and trimmer position. It looks to me they were in fact pulling up, and without AP in the way (neutralizing them with the trimmer) they would go around successfully. Did they really forget about AP? No, that doesn't seem probable.
Another possibility is they wanted to arrest the sink rate (in preparation for landing), and I think they accomplished it, even with the AP engaged. Makes sense, the throttles weren't advanced, they were still on A/T, and they weren't pulling hard enough. I think it could eventually work, they only didn't know about the hill. Would mean they tried to go around only at the last moment.

[Evan]

Originally Posted by Northwester

That seems to be fine, except I see the first control column movement and pitch adjustment at 10:40:45, and thrust increase at 10:40:47. At that time the plane is still at 100m RA, not 80.

No, it happens at the 60m Nav call, which is 60m RA or 10-15m QFE.

The pitch before then are very minor adjustments probably from the autopilot pitch channel.

It could be the beginnings of a go-around, but it is not followed by any thrust command for 4 seconds. Wrong instincts.

See the attached crop from the plot on page 156 of the report.

Study the plots and ye shall find the truth.

[Northwester]

Originally Posted by kris

Nope. That would be average value. The truth is the time (and distance) needed to decelerate may be much longer, but the bodies would travel in air with mostly zero deceleration, only to hit something solid and stop rapidly.
This is comparable (not equal, but comparable) to a high-speed car crash. While someone survives from time to time, such crashes are generally non-survivable.

You misunderstood me. I assumed no gradual deceleration. I assumed that the first part of the plane would hit the ground with full 260km/h speed and stop almost immediately, in 0.2 seconds. 0.2 seconds because the travel direction is mainly horizontal and the ground is soft. If you have a stop watch, see how long 0.2 sec is - just an eye blink. So 37g is the max. value for the first part of the plane hitting the ground.

I can't see anything weird there (note: the pressure altitude plot doesn't say much), but pitch seems to increase starting at ca. 10:40:45, with the AP still controlling the aircraft. Also see vertical accel, control column and trimmer position. It looks to me they were in fact pulling up, and without AP in the way (neutralizing them with the trimmer) they would go around successfully. Did they really forget about AP? No, that doesn't seem probable.
Another possibility is they wanted to arrest the sink rate (in preparation for landing), and I think they accomplished it, even with the AP engaged. Makes sense, the throttles weren't advanced, they were still on A/T, and they weren't pulling hard enough. I think it could eventually work, they only didn't know about the hill. Would mean they tried to go around only at the last moment.

I agree, both options are possible, but "go-around" more likely. After someone says "anything visible?", PIC says "go-around", then co-pilot repeats it. They either screwed up the execution of go-around, or had some other problem.

[Northwester]

Originally Posted by Evan

No, it happens at the 60m Nav call, which is 60m RA or 10-15m QFE.

The pitch before then are very minor adjustments probably from the autopilot pitch channel.

It could be the beginnings of a go-around, but it is not followed by any thrust command for 4 seconds. Wrong instincts.

See the attached crop from the plot on page 156 of the report.

Study the plots and ye shall find the truth.

We might be talking about 2 different things. I am talking about when the change is initiated. See attached.

[Northwester]

Originally Posted by Gabriel

Yes. Absolutely.


Close to zero, except by those exceptional cases where extreme good luck (miracles for the believers) takes place. Being inverted just substantially reduced the chance of this good luck to materialize.

Well, I don't know specifics of these accidents but according to NTSB (2006 data) out of 69 IN-FLIGHT COLLISIONS WITH TERRAIN/WATER 28 were fatal. That's 60% non-fatal.

[Evan]

Originally Posted by Northwester

We might be talking about 2 different things. I am talking about when the change is initiated. See attached.

You are looking at autopilot inputs. There is no (0) force applied to the control column until 10:40:51, just after the F/O calls out to go-around, between 80 and 60m RA NAV callouts. Everything before that is automation (AP & AT).

Originally Posted by Northwester

Well, I don't know specifics of these accidents

Yes, and you would need those specifics to draw any comparisons to this accident. Most survivable CFIT crashes are wings-level-shiny-side-up-low-approach-angle affairs. This most definitely wasn't. I don't understand how you can expect a human being to survive this kind of trauma. I also don't see where you are going with this.

[Evan]

Originally Posted by Geebee

Unfortunately my tax dollars are!

There is 1 other important element to add here that hasn't been shared yet. 2011 is a year of parlementary elections. If the governing majority would just sit back and accept the report as it is, definitely it would be used against them in the fall elections. So they have to do what they have to do.

I know, I know. They are putting on an expensive show for political survival. Bush did this for years in Iraq. But it's hard to watch. The thing is, if they accept the legitimacy of the FDR data, the cause is quite obviously pilot error. By conducting these test flights, they are effectively showing that they suspect the FDR data is corrupt, and making the accusation through actions instead of words. The reason they won't come out and make this accusation in words is because they would then have to back up those words with credible evidence instead of mere conjecture. And when I say credible, I mean an overwhelming discovery that could overrule the flight recorders, which are, in lieu of such discovery, considered the most credible evidence we have. Something like a mind-control death ray or a airplane-sucking robotic tree.

What other overwhelming evidence could they expect to discover in conducting these tests? It's all political pantomime designed to defer a domestic condemnation of the party, at the expense of international relations.

[Northwester]

As Gabriel said, in order to prevent future accidents, the investigation has to look at all FACTORS and CAUSES of the accident. We might be getting very close to the cause, but many factors are still left out. It is not OK that the Russians can publish full autopsy report of one selected passenger in order to justify their opinion, but refuse to include ATC recordings in their report.

[Northwester]

Originally Posted by Evan

You are looking at autopilot inputs. There is no (0) force applied to the control column until 10:40:51, just after the F/O calls out to go-around, between 80 and 60m RA NAV callouts. Everything before that is automation (AP & AT).

Which graph shows what force is applied to CC?