Polish President and wife killed in Tu-154 crash

[Northwester]

One of the "faults" of the Tu-154 crew was entering the glideslope too late, too high, that resulted in excesive vertical and airspeed. Our friendly pilot has some problem with that. I already showed his suspicions about the Vy in my previous post. He started looking at how to determine the glideslope start from the published data and found some issues. There is a reason why the terrain contour does not extend beyond 8000m on pic 1. He extended the line and superimposed contour from Google Earth on pic 2 and discovered a discrepancy. Then when he calculated the altitude in relation to the RWY based on the real contour, he got something amazing (heavy blue line on pic. 3). He also realized that entering the glideslope was shifted on the graph by about 500m (pic 4).

That has other consequences. The beginning of glideslope is 9200m from the RWY. It took 28 seconds from there till OM. That is 9200 - 28x80 = about 7000m. The OM is supposed to be 6260m from RWY. He maintains that the plane was on a perfect 3deg 10' glideslope that was shifted to point not to the threshold of the RWY but at some field near the MM.

[Northwester]

He also examined the airspeed. He compared the graphs with what F/O was reading from his instruments. On the attached pic when F/O is saying "3-3-0 reducing" the ghraph shows 366 km/h. At 10:38:37 F/O says "and reducing 300". The graph shows about 340. At 10:38:48 F/O says "and flaps 36, we have 280". The graph shows airspeed above 320. F/O repeats at 10:38:56 "Flaps, 280". The graph airspeed at that moment is 318, then drops to 302. All of that happened before entering the glideslope. For the 3 deg 10' glideslope and 280 km/h airspeed we have 4.3 m/s Vy. Looks like a perfect approach.

[Northwester]

The crew was also accused of not reacting to the TWAS alerts and warnings. It is strange that on one hand the crew was accused of not pressing "TERR INHIBIT" button to silence the alerts when approaching an airdrome that was not in the TAWS database, and on the other hand they were accused of not reacting to them. They chose to leave the alerts as advisory which is perfectly correct during a landing procedure. Tu-154 was executing a full auto precision landing using a TLS system.
Besides, when you look at the Alert points and corresponding Terrain Conflict points, 3 out of 4 alerts were not generated according to a proper algorithm. Especially puzzling is Alert Record 37 and Terr Conflict 37. This itself warrants a separate investigation.

[Northwester]

One clear proof of an automatic approach on an instrumental glideslope is shown on this graph. In the time interval 10:40:53.5 - 10:40:55 PIC's hands are on the control column so he could not touch the "DESCENT-CLIMB" wheel. Look at the trimmed position of the column. It is crystal clear.

[Northwester]

This is what most likely happened in the last few seconds of the flight. At 10:40:51 (point A) the plane is at 60m alt. The PIC starts gradually pulling the column in preparation to switch to manual. The plane is on a precision approach on intrumental glideslope proceeding to land. They see the fake search lights so everything is cool. The PIC pulls the column to 1/2 of the value necessary to switch to manual (points B and C). It takes about 3.5 seconds. At altitude of about 35m the PIC pulls the column just beyond the value necessary to overpower A/P (50 mm is necessary, he takes it to 56.6 - point D). The NAV continues to read the altitude at every 10m as required in SOP. Then the PIC RETURNS the column to the previous position (point E). He is not trying to execute GA. He is just switching to manual before landing. At that time the fake search lights are turned off, he can see the ground but no RWY. He slams on the throttle and about 1/3 of a second later aggresively pulls the column to over 100 mm (point F).
Whether the plane hit any trees or not is not important here. It certainly did not loose any part of the wing in a colision with the birch. At TAWS 38 the plane is still ascending, with ZERO roll.

[Gabriel]

This is what most likely happened in the last few seconds of the flight. At 10:40:51 (point A) the plane is at 60m alt.

So what on Earth were they doing 40m belos the minimums?

The PIC starts gradually pulling the column in preparation to switch to manual.

40m below the minimums and they are just gradually pulling the column (still with the AP engaged?) in preparation for a go around? Brillunt!

The plane is on a precision approach on intrumental glideslope

Show me the precision approach chart or they weren't.

They see the fake search lights so everything is cool.

Yes, I think that "cool" is the right word. Like jumping from a helicopter with a flight suit and landing on a pile of cardboard boxes is cool. As for airplane operations, you mus go around at the minimums unless you have the runway environment in sight. I wouldn't call the gleam of a "gate" of search lights in the midle of solid fog "runway in sight".

The PIC pulls the column to 1/2 of the value necessary to switch to manual (points B and C). It takes about 3.5 seconds. At altitude of about 35m the PIC pulls the column just beyond the value necessary to overpower A/P (50 mm is necessary, he takes it to 56.6 - point D).

To be honest, I don't know the procedures in this plane. In the rest of them, for a manual go-around it is to disconnect the AP with a button especially provided for that and manually fly away. To force the AP to disconnect by smoothly pulling up until several seconds later you overcome the disconnect force is a very bad idea, because in the meantime the AP would have been applying trim to fight you, leaving the plane well out of trim. And, buy the way, since you were already at 35m, it sounds like a good idea to go around right away instead of playing with the yoke and AP. And why would the FO do that if they had briefed that the go-around would be made on AP?

Figure the scenario. They are flying an unpublished precision approach, have briefed an unaproved go-around in AP, they are 35m above the ground, 65m below the minimums, with no runway in sight, still descending fast, and they have not yet initiated the go around.

The rest of what you say is irrelevant until you can produce the most minimal cue that there were fake search lights and that they were turned off.

[Northwester]

To be honest, I don't know the procedures in this plane. In the rest of them, for a manual go-around it is to disconnect the AP with a button especially provided for that and manually fly away. To force the AP to disconnect by smoothly pulling up until several seconds later you overcome the disconnect force is a very bad idea, because in the meantime the AP would have been applying trim to fight you, leaving the plane well out of trim. And, buy the way, since you were already at 35m, it sounds like a good idea to go around right away instead of playing with the yoke and AP. And why would the FO do that if they had briefed that the go-around would be made on AP?

The fact that the PIC returned the control column to the previous position after disconnecting AP shows that he was not even thinking about GA. And he could only not be thinking about GA at that altitude if he was going to land. And he could only be executing landing and be at that altitude without ground visibility if he was on precision approach and instrumental glideslope. Period.

[Jonathan_Creek]

[..] At TAWS 38 the plane is still ascending, with ZERO roll.

Nothwestern, this is just my subjective opinion but I feel like you are getting overexcited and are posting a lot of material that is of average use. I'm afraid that unintended consequence of your deep involvement may be that you will discourage others from this conversation by flooding them with data. I recognize your right to post as you please, but I kindly urge you to consider being more specific or asking precise questions.

Now to your zero roll "discovery" you kindly highlighted in red box.
In my previous post I pointed to Appendix #4 to the official Polish report. Full document is available here

http://mswia.datacenter-poland.pl/protokol/ZalacznikNr4-TechnikaLotniczaIJejEksploatacja.pdf

"TAWS DATA EXTRACTION FOR NTSB IDENTIFICATION: ENG10SA025" starts on page 414.

• Appendix C (Pages 452 to 457) indicates that Roll Angle was 0.000000 deg for each TAWS entry.
• Appendix A (Page 443) defines Roll Angle:
  Filtered roll angle, in degrees.
  TAWS filters the roll angle input for use in the bank angle limit alert. Bank angle limit is a configurable alert, which was not configured in this aircraft. For this configuration, the source of the input to the filter in order of priority is:
  1. PRIMARY FMS (L325)
  2. SECONDARY FMS (L325)
• APPENDIX E - EXTRACTED CONFIGURATION LOG (Page 464) indicates that Bank Angle Limit was in fact disabled.
  Bank Angle Limit
  Value.......................................... 35
  Status.........................................Disabled

Mystery solved.

[Northwester]

Nothwestern, this is just my subjective opinion but I feel like you are getting overexcited and are posting a lot of material that is of average use. I'm afraid that unintended consequence of your deep involvement may be that you will discourage others from this conversation by flooding them with data. I recognize your right to post as you please, but I kindly urge you to consider being more specific or asking precise questions.

Now to your zero roll "discovery" you kindly highlighted in red box.
In my previous post I pointed to Appendix #4 to the official Polish report. Full document is available here

http://mswia.datacenter-poland.pl/protokol/ZalacznikNr4-TechnikaLotniczaIJejEksploatacja.pdf

"TAWS DATA EXTRACTION FOR NTSB IDENTIFICATION: ENG10SA025" starts on page 414.

• Appendix C (Pages 452 to 457) indicates that Roll Angle was 0.000000 deg for each TAWS entry.
• Appendix A (Page 443) defines Roll Angle:
  Filtered roll angle, in degrees.
  TAWS filters the roll angle input for use in the bank angle limit alert. Bank angle limit is a configurable alert, which was not configured in this aircraft. For this configuration, the source of the input to the filter in order of priority is:
  1. PRIMARY FMS (L325)
  2. SECONDARY FMS (L325)
  
• APPENDIX E - EXTRACTED CONFIGURATION LOG (Page 464) indicates that Bank Angle Limit was in fact disabled.
  Bank Angle Limit
  Value.......................................... 35
  Status.........................................Disabled
  

Mystery solved.

Jonathan, I am not going to comment on your subjective opinion about my posts.

I am well aware that the Bank Angle Limit was disabled, but not the Roll Angle Input.

[brianw999]

Whatever the figures are/were little old simplistic me who openly admits to knowing nothing useful about airliner dynamics in the landing phase interprets this....

Figure the scenario. They are flying an unpublished precision approach, have briefed an unaproved go-around in AP, they are 35m above the ground, 65m below the minimums, with no runway in sight, still descending fast, and they have not yet initiated the go around.

As.........dead.........they just don't know it yet !

[Gabriel]

The fact that the PIC returned the control column to the previous position after disconnecting AP shows that he was not even thinking about GA. And he could only not be thinking about GA at that altitude if he was going to land. And he could only be executing landing and be at that altitude without ground visibility if he was on precision approach and instrumental glideslope. Period.

Look, these days I can take a Cessna 172 with a WAAS capable Garmin G1000 and make a coupled approach (localizer and glide slope) on autopilot to an airfield where the only navaid is a windsock. Still, it's NOT A PRECISION approach unless I have a precision approach chart to follow. It's not even an instrument approach (and hence I have to comply with VMC minimums) if I don't have an instrument approach chart.

If I have a non-precision chart (like a VOR, NDB or GPS chart), then I can follow the GPS simulated glide slope, but I have to respect the non-precision minimums.

Further, I don't know if you are aware, but at least in the western world there are other conditions that must be met to enable a precision approach beyond having a glideslope, including having a visual glideslope indication (VASI or PAPI) to make the transition, having a certain runway lighting, runway marking, and approach lighting. But nothing of this is really important. The important thing is if they had a precision instrument approach plate with a minimums lower than 100m. They didn't because it didn't exist.

Then, unless they had a precision instrument approach chart with a minimum lower than 100m, they had absolutely nothing to do 60m or 35m above the ground without the runway in sight.

So it doesn't matter what fancy instrument they had, they were NOT in a precision approach and they were NOT allowed to go one inch below 100m (unless they had the runway in sight) until you can produce the instrument approach chart that they were using and that says that they were allowed.

By the way, could you please post a bit more of that parameters chart? I want to see barometric altitude, vertical speed, and airspeed.

[Northwester]

Look, these days I can take a Cessna 172 with a WAAS capable Garmin G1000 and make a coupled approach (localizer and glide slope) on autopilot to an airfield where the only navaid is a windsock. Still, it's NOT A PRECISION approach unless I have a precision approach chart to follow. It's not even an instrument approach (and hence I have to comply with VMC minimums) if I don't have an instrument approach chart.

If I have a non-precision chart (like a VOR, NDB or GPS chart), then I can follow the GPS simulated glide slope, but I have to respect the non-precision minimums.

Further, I don't know if you are aware, but at least in the western world there are other conditions that must be met to enable a precision approach beyond having a glideslope, including having a visual glideslope indication (VASI or PAPI) to make the transition, having a certain runway lighting, runway marking, and approach lighting. But nothing of this is really important. The important thing is if they had a precision instrument approach plate with a minimums lower than 100m. They didn't because it didn't exist.

Then, unless they had a precision instrument approach chart with a minimum lower than 100m, they had absolutely nothing to do 60m or 35m above the ground without the runway in sight.

So it doesn't matter what fancy instrument they had, they were NOT in a precision approach and they were NOT allowed to go one inch below 100m (unless they had the runway in sight) until you can produce the instrument approach chart that they were using and that says that they were allowed.

By the way, could you please post a bit more of that parameters chart? I want to see barometric altitude, vertical speed, and airspeed.

I see your point. But keep in mind that this is Russia. I am presenting info from an experienced Russian pilot, and they were landing at a military airdrome. All of their actions in the last seconds indicate that they were executing a fully automatic landing. For such procedure Russian standards are: Decision Altitude - 30m, Horizontal Visibility - 350m (see pic).

One more important issue is the radalt altitude alert setting. There is a record in the CVR when they set the alert to 100m. Exchange of commands, and a confirmation. That would indicate that they were going to descend to 100m and then decide to land or GA. But the alert went off at 60m. Resetting of the radalt alert means that they changed something about their approach. It is reasonable to assume that they received some new info and figured out they can descend lower. Likely the info was about a different mode of approach. Unfortunately there is no record in the CVR about any conversation relating to resetting the radalt alert, but we know it had to happen and it was not done by some crew member without communicating it to the PIC.

What is also important is the fact that 3 days earlier (April 7th, 2010), both, Polish Prime Minister's and Putin's planes landed in Smolensk using some kind of precision guiding equipment installed temporarily. Such equipment could have been still in place on April 10th. Our friendly Russian pilot speculates that it was TLS. From the crew point of view the TLS approach does not differ in any way from an ILS approach.

[Gabriel]

I see your point.

But decide to ignore it.

They were in a non-precision approach because they were following a non-precision approach procedure. I care very little what equipment was in the ground or airborne (as long as it was the minimum required for than non-precision approach). I don't understand Russian, but I guess that what you posted are the minimums for established procedures, or the criteria to design such procedures. Not an actual minimum that says "never mind if the runway is not appropriate, the ground equipment not calibrated, there obstacle clearance not enough, and there is no published procedure: as long as you receive the glide-slope signal go down to 30m".

Again:
Anything that they went beyond the published procedure is illegal.
Even further, the ATC cleared them to descend only to 100m and repeatedly advised them that they had to go around by 100m. So other than busting the minimums, they also busted the clearance.

Your latest descriptions of the events are reckless and careless flying, busting minimums, busting ATC clearance, very stupid, and suicidal (as proven).

But I really don't know that that was the case. I still think that, somehow, they had the intention to go-around at 100m but somehow failed at that.

[Gabriel]

By the way, would you please honor my request and post the part of the parameters chart where the airspeed, baro altitude and baro vertical speed are displayed?

If you don't, I will have to think that you decide to post only the part of the info that sustains your hypothesis and hide what could contradict it. I don't know if "confirmation bias" was mentioned in this thread before.

[Northwester]

By the way, would you please honor my request and post the part of the parameters chart where the airspeed, baro altitude and baro vertical speed are displayed?

If you don't, I will have to think that you decide to post only the part of the info that sustains your hypothesis and hide what could contradict it. I don't know if "confirmation bias" was mentioned in this thread before.

No need to get upset. Didn't get to it yet.