From another forum:
There are quite a lot of discussion on the TU154 accident : people talking about instrument approaches procedures in Russia, heirloom from the soviet era :
There are three possibilities on the procedure they were shooting (but the clue is in the ATC observation that "they lost altitude very quickly" close in):
1/- A GCA approach, which would confirm the initial Russian controller who said that " they stopped acknowledging my messages and they lost altitude very quickly..." required vis is 1000 m.
2/- A very unusual procedure - apparently quite common in the ex-soviet world - in which, the pilots would use two NDBs for the alignment on centerline, while being talked down by a ground operator. I have never heard of that method until to-day, and it seems to me quite dangerous as the pilot has to deal with two mental pictures - one from his instruments on a visual cue, another from the required "correlation" of some height orders from the ground (aural / visual cues ). Required min vis is 1200 m.
3/- A dual NDB approach. The problem with this letdown is the absence of any glide slope information ; everything is done by stopwatch, down to an MDH of some 200m / 600 ft.
Contrary to what some may think, the lateral precision of that approach is very good (the deviation of some 45 m at 1500 m from the threshold amount to a precision of less than 2° (and Yes, on that sort of approach, that's "smack on the centerline" ! )
Another image I would have liked to see is a cockpit view of that approach, especially at the point where they started that strange "dive"... i.e. was there a visual clue that made them think they were closer to the runway ( and too high, of course ) ?
Dutch TV program Nova contains an interesting clip of the whole Tripoli crash site, filmed from another plane as it approached Tripoli's runway 09.
Fast Forward to 20 min 43 seconds, approximately. Disregard Dutch comment, it's not related to the clip itself.
Visible is what could resemble a first impact area plus signs of the plane sliding straight forward. The tail section plus smaller debris appears to be located some 200 meters after this first impact area (in front of the mosque described by others). It appears the plane was more or less intact before that point.
A large debris field can be seen after the tail-section with relatively small pieces and stretches for some 240 meters. One large section (wings?) appears approx 200 meters further on, with relatively little (if any) debris in between those 2 points (though it appears the ground is scratched in between). There's approx 730 meters between what appears to be first impact area and this large*object (wings?).
After carefully watching this clip a few times, I believe it's probable that the plane was intact and with its wings almost level when it hit the ground and slid for some 200 meters and eventually broke up (tail first) when it hit the road which runs North-West near the mosque. In the clip, this road appears to be slightly higher then surrounding terrain (i.e. a "berm").
*
For comparison, at the link below is a picture of the TU154 accident debris trail in which the Polish president died a short while ago.
Also apparently a CFIT during approach. TU154 crash site debris distribution (
http://bi.gazeta.pl/im/3/7765/m7765863.jpg )
You will appreciate that the length of the wreckage trail is dependent on the combined forward and downward vector. A high-speed accident with a large vertical component can leave a shorter trail than a low speed accident with a lesser vertical component. Other varying factors are the ground composition, obstacles and obviously the behaviour of the a/c structure during the impact sequence.
*
But it's likely that the TU-154 accident and the Tripoli accident were very similar in their origins. See explanation below.
*
Picked up this photo (below): of the TU-154 crash
.....showing that the big piece first torn from the airplane was indeed the tip of the left wing.
So thanks to this fact, we have confirmed that the reason the TU154 veered to the left and - apparently - ended up upside down, was a sharp low altitude left bank (and tree-top connection) which resulted in the airplane impacting the ground inverted.
Using that elevation/location data, here on Pprune some theorists have plotted the TU154's finals flightpath on a terrain profile and found out that, at the first tree impact, theTU154 was well below the airfield elevation and their subsequent trajectory was a 5% climb, insufficient to clear all obstacles.
That 5% value would be puzzling to anyone
unfamiliar with approach illusions, being way below the height and gradient from which we would expect a go-around/missed approach to commence; and with that puzzling final approach geometry, it would tend to make any stabilized final approach a lot more difficult than normal...even if they were well above stall speed... However this puzzling low altitude close-in on final approach could easily be explained away by a visual illusion caused by a runway upslope.... in limited visibility.
*
Yet it's hard to establish whether or not runway 26 at Smolensk has any
upslope - as the Russian military don't publish public charts/plates for their military airfields. I'll keep looking, but it's conceivably a common factor. The inordinately low altitude during the TU-154's approach would appear to support an
upslope visual illusion theory.
As with Aspen CO's 2001 Gulfstream III crash, superimpose a 2% upsloping runway upon a standard 3% approach profile and the pilot is inevitably being visually "conned" into descending into a subterranean approach - once he becomes visual with the first half of the runway. It's an illusion that's very difficult to perceive or even anticipate, so how can one avoid it - particularly in low visibility?
*
The TU154 Smolensk and Tripoli A330 accidents would appear to have much in common. That final "drive" and descent to the "now visible" runway, after an NPA approach, is totally dependent upon what the pilot perceives his approach angle to be....in limited visibility (i.e. with few cues). If he's unknowingly coping with (i.e. subject to) an upslope illusion, and consequently very low, any last minute line-up banking will put a wingtip perilously close to the ground..... but he won't realize that. It's a deadly combo.