AviatorAtHeart

What the hell were they thinking??

At 80m, the SIC said 'odchodzimy' which can be translated as 'break off approach' when used in that particular situation. So the question is, why did the plane continue to descend when at least one pilot wanted to gain altitude and not continue landing ?

jimjim1

Here is a plot of the cross section data from Google Earth that I posted the other day.

For details of method and data source see post 144.

http://www.pprune.org/rumours-news/4...ml#post5723588


I have plotted three scales, 1 to 1, x4 vertical and about x30 vertical. Scales are approximate, if you care use the raw data. I also put in a 3 degree "glideslope". Note that the reference point is an assumed touchdown point 400 meters from the runway end. I selected it partly because it is at a junction with a taxiway and is therefore easily referenced.

hasta.la.vista

This rate of descent was so high because the navigator was reading RA not baro altimeter! The terrain was rising while A/C was descending.

BMP

- Looking at the properly scaled terrain and
- Looking at the very good climb rate of this AC here:

YouTube - Tu 154M 101-go around


can you professionals confirm that this AC was not able to clear 30' trees several hundred meters after clipping a 12' tall sapling?

It would appear to me that:

1) when they did give the haircut to the sapling, they certainly knew they were too low (and this assumes 10 m of visibility)

2) This AC had ample power and responsiveness to climb the approx 30 meters in the 5 intervening seconds.

Ptkay

jimjim1:

Could you enhance your graphic analysis with the time scale
recalculated with the speed confirmed by CVR of 280 kph,
and the radio hight readings from the CVR above your terrain plot.

Also a glide path from the point of impact would be interesting.

Ptkay

According to Tu-154 FM you set DH on the RA at 60m
It gives a warning signal of 400 Hz.

It confirms, that the navigator was reading RA, because:
from 10:40:51 to 10:40:58 signal of 400Hz BPR sounds
and at 10:40:52 navigator reads 60m.

hasta.la.vista

At 10:10:07, as I understand it, they were setting RA with 100m setting - probably a trigger level for audible alarm, but this alarm actually started between 80 and 60 m read by navigator. What could be the reason?

Kulverstukas

AviatorAtHeart

Look at
10:32:55,8 Capt: Approaching for landing. If unsuccesfull, go around automatically
10:32:58,8 Eng: Automatically.
.
.
.
10:34:17,5 Capt: Unblock throttle
10:34:18,7 Eng: 7-0 set throttle unblocked
10:34:21,5 Capt: Automatic.
10:34:22,6 Eng: Automatic switched on

robdean

A translation to English of the last few pages of the CVR would be worth a big vote of thanks!

liider

I translated the last part of the transcript to English:

10:39:49,9 - 10:39:52,3 ATC: Approaching outer marker, on course, on glideslope.
10:39:50,2 - 10:39:58,0 Sound. F=845Hz. Outer Marker.
10:39:52,2 - 10:39:53,7 U: Outer.
10:39:54,1 - 10:39:55,0 Capt: (...)
10:39:57,1 - 10:39:59,3 U: 400 meters.
10:40:02,6 - 10:40:15,6 U: (...)
10:40:06,7 - 10:40:06,8 TAWS: TERRAIN AHEAD:
10:40:13,5 - 10:40:14,6 ATC: 4 on course, on glideslope
10:40:16,7 - 10:40:17,6 Capt: On course, on glideslope
10:40:18,6 - 10:40:20.1 U: (...)
10:40:19,6 - 10:40:21,1 Nav: 300.
10:40:22,8 - 10:40:25,6 U: (250 meters)
10:40:24,6 - 10:40:26,7 (Nav): (250).
10:40:26,6 - 10:40:27,8 ATC: 3 on course, on glideslope
10:40:29,6 - 10:40:30,3 U: (...)
10:40:31,2 - 10:40:32,4 ATC: Turn on the lights.
10:40:32,4 - 10:40:33,5 TAWS: TERRAIN AHEAD.
10:40:32,9 - 10:40:33,6 Nav: 200.
10:40:34,0 - 10:40:34,8 Capt: Turned on.
10:40:37,1 - 10:40:38,1 Nav: 150.
10:40:38,7 - 10:40:39,9 ATC: 2 on course, on glideslope
10:40:39,4 - 10:40:42,0 TAWS: TERRAIN AHEAD, TERRAIN AHEAD.
10:40:41,3 - 10:40:42,6 U: 100 meters.
10:40:42,6 - 10:40:42,7 Nav: 100.
10:40:42,6 - 10:40:44,1 TAWS: PULL UP; PULL UP.
10:40:44,5 - 10:40:46,1 TAWS: PULL UP; PULL UP.
10:40:46,6 - 10:40:49,2 TAWS: TERRAIN AHEAD, TERRAIN AHEAD.
10:40:48,7 - 10:40:49,4 Nav: 100.
10:40:49,2 - 10:40:49,6 (FO): Normal.
10:40:49,6 - 10:40:50,1 Nav: 90.
10:40:49,8 - 10:40:51,3 TAWS: PULL UP; PULL UP.
10:40:50,0 - 10:40:51,3 Nav: 80.
10:40:50,5 - 10:40:51,2 FO: Go around.
10:40:51,5 - 10:40:58,0 Sound. F= 400Hz. Decision Height.
10:40:51,7 - 10:40:53,4 TAWS: PULL UP; PULL UP.
10:40:51,8 - 10:40:52,4 Nav: 60.
10:40:52,3 - 10:40:53,1 Nav: 50.
10:40:52,5 - 10:40:53,4 ATC: Horizon 101.
10:40:53,0 - 10:40:53,6 Nav: 40.
10:40:53,7 - 10:40:55,5 TAWS: PULL UP; PULL UP.
10:40:54,5 - 10:40:55,2 Nav: 30
10:40:54,7 - 10:40:56,4 ATC: Height control, horizon.
10:40:55,2 - 10:40:56,0 Nav: 20.
10:40:56,0 - 10:40:58,2 Sound. F=400 Hz. Autopilot disconnect.
10:40:56,0 - 10:40:58,1 Sound. F=800 Hz. Inner Marker.
10:40:56,6 - 10:40:57,7 Sound. F=400 Hz. Autothrottle disconnect.
10:40:56,6 - 10:40:58,2 TAWS: PULL UP; PULL UP.
10:40:58,6 - 10:41:00,2 TAWS: PULL UP; PULL UP.
10:40:59,3 - 10:41:04,6 Sound of the collision with obstacles.
10:41:00,3 - 10:41:01,4 FO: (censored)
10:41:00,5 - 10:41:01,8 TAWS: PULL UP; PULL UP.
10:41:02,0 - 10:41:03,4 ATC: Go around!
10:41:02,7 - 10:41:04,6 U: Scream (censored)
10:41:05,4 End of the transcript.


U - unidentified voice

robdean

Big Vote Of Thanks for the translation!

criss

Just to clarify, final screams are not censored in the original version. As you can see, there is almost nothing coming from the Cpt in the final stages ; also no reaction from any other crewmembers to F/O call for a go around, he didn't take any action himself either.

SINGAPURCANAC

I read Russian version of CVRs and I have few questions relating more to atc .
-Smolensk ATCO is TWR/APP PROC or APS?
-When Smolensk ATCO cleared arriving a/c for APP? or
-Is it standard Russian ATC procedures to make approach to airport without approach clearance?
-where is TL given,together with airport QNH?
-is it standard procedure in Russian enviroment to allow aproaching a/c to descend to 100m/330ft and close as much as half mile to tochdown zone without asking for position and without landing clerance?
-If atco work as APS where is word: IDENTIFIED?

If this CVRs is real,than 50% of all mistakes belong to ATC,not to to ATCO itself,but ATC as system in Russia.
Because ATC is for safety purpose,and each phrase during the approach ,and especially when we are talking about low visibility approaches, each phrase is SAFETY NET,if you ommit them,chance for mistakes rises rapidly.

In normal proc enviroment there should be:
Polish 101,report POINT(initial aproach fix or something)
Smolensk ,Polish 101,passing POINT,
Polish 101, clear for TYPE of APP,report OM,
Somlensk .Polish 101, is cleared for TYPE of APP, wilco,
Smolensk ,Polish 101 OM,
Polish 101,Cleared to land ,runway 23,wind 230/02 knots
Polish 101, cleared to land.

Sounds stupid and boring but works in real life. each sentence is SAFETY NET,if used properly ,both by crew and ATC authority and operational staff.

Ptkay

I found this simulation on a Polish forum.

Done with the terrain elevations from GoogleEarth and RA (assumed) from CVR.

It makes sense to me.

AviatorAtHeart

As a side note, the CVR provided by Russian authorities is a copy of the original recordings. The Polish government has not received the black boxes and was not able to review the data contained in them independently. Keep this in mind when drawing conclusions based on the information provided.

ARRAKIS

Basically, nothing new. We already knew the key points.
They were using RA. I hope no more questions about that.
My estimation of the descent rate is around 6-7 m/s. They needed around 1 minute to go from 400 m to 0.

From the preliminary MAK report we know, when they reacted. 5 seconds before hitting the three, that cut a part of the left wing (10:40:59 here), which means 10:40:54. Reaction probably to visual contact with the ground (ahead of them?).

The second "100 m" was probably when they flew over the deepest point of the valley (10:40:48). After that, they had just a few seconds to react. 10:40:54 was too late.

If I'm not mistaken, when getting at 100 m, the PIC should, according to the Russian Tu-154M FM call "Landing" or "Go around". If not. The C/O should call a "Go around".

Arrakis

dukof

Hi,
First thanks to everyone contributing to this and previous thread. Although I'm not a pilot, but mechanical engineer, I'm very interested in all these details.

Pardon my lack of insight, but; I'm very surprised by the fact that there seems not to exist on this aircraft a "sink rate" alarm, working at lower altitude.. Isn't it normal for a fairly modern plane like this to have both a variometer as well as an automatic sink rate alarm?

criss

Singapurcanac, the "ATCO" was a Russian soldier without any ICAO endorsement, working according to military procedures, and not speaking English. You can't also call it "standard Russian ATC" if you mean civilian one - it's military. QNH is not given, they're working with QFE, which was given. He's not APS, so there's nothing about identification (btw, in the above post you have only the final part of the CVR).

Again, this is NOT a civilian ATC as per ICAO, so your questions and conclusions do not apply. PLF101 made an approach they had already made in the past and wasn't new for them (not approach to Smolensk, but this type of "military approach), and all would be fine had they not made 5 basic, crucial mistakes. You can't blame ATC for it.

Ptkay

Arrakis,

this is clearly visible from the simulation above

SINGAPURCANAC

thanks for further explanation, but you want to say that in Russia, SOLDIER could work in ATC,that there is no ratings, there is no separation between twr/app/proc/aps endorsement ,transition level is not effective ,there is no landing and approach clearance and so on?
even for Russian air force it is too much.

If atcos words correct that it looks like PAR approach,i have just said looks like, because he said,
2 on course,on glideslope,
3 on course,on glideslope,
heading seventy nine(yes we read correctly heading expressed with two numbers) and so on

I told you ,even for Russian military standards it is too much,if ,I said if ,tapes are "ORGINAL",