Andrasz (#532) makes an interesting observation about the mismatch between the changing Russian rules and those likely to have applied to the Polish government flight. It scarcely bears repeating that if either the old Russian rules or the current western civilian concept of approach ban had applied the accident could not have happened because the approach could not have been flown. Provided of course the rules were adhered to.
But this does not really explain why the accident happened. The crew should have been able to conduct a PAR to limits and either land if suitable visual reference were obtained, or if not then go around. Conducted properly it would remain a safe operation, rules notwithstanding.
And if they were absolutely determined to land regardless by busting limits, surely the best chance of success would be obtained by scrupulously following the PAR down to DH/MDA then maintaining heading and descent rate towards the threshold.
It is clear this did not happen. The location of the wreckage and controller’s statements indicated the flight went well below the PAR glideslope and did not react to warnings and corrections. I find it hard to believe the crew would see any advantage from deliberately going so low so far from the runway and below the PAR DH, while hugely increasing the risk profile. What would be a likely DH - 300ft QFE perhaps? (We used to operate to 200ft in the UK many years ago. I imagine no-one on this forum can give us likely Russian data?)
I am not convinced that altimetry is the real issue. The PAR controller passes glideslope deviation advice and instruction derived from his own radar returns, rather than raw altitude. This is unlike SRA where the information passed is distance to touch down and an advisory altitude check for that point.
Ptkay’s point #539 about the terrain at Smolensk is worth noting and especially the Russian description of the “death trap”. But pilots are surely aware of the possibility of radio altimeter fluctuations on final approach at many airfields due to undulating terrain. It should be irrelevant to the conduct of the PAR, just as with ILS .
I have made the assumption that the approach was PAR because that is what the Russian statements appear to be saying and it would have been the most logical procedure in the circumstances . We need to know why the approach went wrong. Could they have been flying some alternative non-standard procedure (GPS ?) using PAR only for monitoring ?
I'm sure there will be a lot of soul searching on all sides once the investigation is concluded. By now we seem to know reasonably well what happened, but as Ptkey put it some posts earlier, when it comes to the why, the swiss cheese seems to have gone rotten to the point of hardly having any cheese left...
There appears to be a massive (and possibly deliberate) confusion as to what rules exactly applied to the flight, what authority did ground control have to issue binding instuctions to the pilot, all shaded with the very evident pressure on both the crew and the ground controllers to complete this flight. When it comes to establishing direct cause, the question of why did the aircraft get 60m below the glideslope pales in comparison to why was this approach commenced in the first place.
Perhaps, this was a poorly flown PAR approach. I would feel safe in saying that the PAR approach is not flown/practiced as often as the ILS. The PAR approach is more demanding and requires precision (thus the name). If the pilot had become used to having the AP capturing the localizer and GS on the majority of ILS approaches, it may be safe to say that his hand-flying precision skills may have suffered.
The controller also plays a vital role in the PAR. One small slip-up from the controller can be quite dangerous, where as if you were on the localizer you would be able to see that the controller was flying you through the course.
Finally, if the pilot was doing a PAR, did he know the PAR minimums? Sounds straight-forward....but it is shocking how many pilots don't know what or how to find the DA on a particular PAR. You can get spoiled flying from ILS to ILS around the world and when something new is added everything can go to hell in a handbasket real quick.
I am no expert, so please be respectful with opinions/comments. I just feel it may be useful to look at the obvious.
Ptkay’s point #539 about the terrain at Smolensk is worth noting and especially the Russian description of the “death trap”. ...
Could they have been flying some alternative non-standard procedure (GPS ?) using PAR only for monitoring ?
I still think, this is worth something, because it certainly was an
"alternative non-standard procedure".
The contact with the ATC was broken, if the terrain profile shown in some animations is correct, then, at the NDB antenna point, where they first hit of the small birch tree, they were in the shade of the road on the top of the hill and invisible to the PAR radar.
Maybe this "alternative non-standard procedure" was just scud running according to GPS and radio altimeter, with wrong terrain data in the GPS, wrong, or non existing maps and rising terrain: it is a "death trap".
Been there done it before, but with a small single engine prop, and the approach lights were full Calvert. To follow the "undulating terrain" and finding the runway was not an issue by that speed.
The opinion of the Russian pilot, was by a fighter pilot, who described such "alternative non-standard procedure" going well in an agile a/c, but certainly not in Tu-154.
I was just quoting opinions from some Russian speaking forums. I take no responsibility for correctness.
<<What would be a likely DH - 300ft QFE perhaps? (We used to operate to 200ft in the UK many years ago. >>
In the UK PAR talkdowns usually continue to 1/2 nm from touchdown - just over 150 ft height - but can go to touchdown when required. I suspect that if most pilots saw nothing at 1/2 nm they would pull-up pretty smartly.
The question of descending out of cover behind high ground is interesting. It happened long time ago with an aircraft inbound to Northolt when it descended below the PAR glidepath. The resulting "emergency" climb prevented a major accident. Radio communication should be OK until some while after the a/c dropped out of radar cover at that short range.
You can get spoiled flying from ILS to ILS around the world and when something new is added everything can go to hell in a handbasket real quick.
This was basically the conclusion of the CASA accident report, regarding pilots.
Also the controller was blamed in that report for poor performance.
In case of this accident, the cooperation between controller and pilots must have been severely disturbed by lack of common training, understanding of procedures, and probably cultural barrier...
Tagron, there seems to be very little data published for this military airfield, but to illustrate, here's an excerpt from the Russian AIP about PAR minima for Vnukovo (UUWW, one of the proposed alternates for the flight):
Quote:
The control shall be commenced from the moment of detection of the aircraft position blip on PAR display in the vicinity of final turn and terminated at 500 m before RWY beginning.
NB: UUWW uses 3° glideslope, while 2°40' is more traditional for Russian airfields.
Regarding little or no fire, the engines, as the heaviest, must have been thrown far forward by the cartwheeling fuselage, away from wings full of fuel. Also most of the fuel might have been already spilled after the wings hit the trees 1 km before final impact.
Just my guess.
You don't need running engines to ignite a fire in a crash impact. There are lots of sparks from just metal friction.
Key element is fuel atomization.
Wet ground reduces the chance of a fire but does not eliminate it.
I'm not aviation expert, but it looks that some 'well known facts' do not fitt to strict logical path.
1. Pictures in Picasa album have coordinates ???-??????? Picasa - ?????? ?????? - ?????????????...# 2. First birch tree top was cut approx at 2.5 km distance from runway. 3. Tower operator states, that he noticed deviation from glide slope when a/c was at 1.5 km distance from runway. 4. Is that logical ? At 2.5 km. plane had to be about 130 metres high on 3 deg. glide slope. What is expected descent rate - 5 m/s ? It translates to 26 sec. to descend and forces me to expect that warning must have been issued at least 10-15 sec. earlier. Also to cover distance 2.5-1.5 km= 1km with 300 km/h approuch speed there is another 11 sec required.
So, it it was PAR approuch, tower response came 20 sec to late, if my math is more less right. How to explain that ? What, if the pilots calculated the risk but their risk equation included timely warning ? Not PAR approuch ? But tower issued warning to abort landing - they had some data available. Very strange.
Last edited by Ground Brick; 14th Apr 2010 at 19:04.
I'm an outsider, for a short time here only, sorry. Simply interested whaat foreign pilots think of it, as compared with our pilot's forum. I'm Russian, and no relation to aviation, but naturally interested, why the Poles broke on our land. As we were a host country. This is the link to the local blog with the discussion. Many maps of? something. that you find here important. I'll translate you the conclusion on wghich the discussion stopped at a Russian forum, if you wish. That is, until new data, of boxes, becomes available.
Here is the crew total flight time in hours by 36 splt www:
- (captain) kpt. pil. Arkadiusz PROTASIUK - total time: 3528h (on Tu-154M - 2937h)
- (first officer) mjr pil. Robert GRZYWNA- total time: 1939h (on Tu-154M - 506h)
- (navigator) por. pil. Artur ZIĘTEK - total time: 1069h (on Tu-154M - 59h, as navigator)
- (flight engineer) chor. Andrzej MICHALAK - total time 330h
Sounds similar [thinking wise] to that Avianca 707 Crash on Long Island in the early 90's "Get-Home-Itis"
