Tagron

BOAC, that would be my understanding too, though I have to say I am not fully familiar with current systems. "Terrain Ahead" is predictive and therefore may be database dependent though not necessarily airport specific.

The point is, "Pull Up" is a demand for instant action . Do it, then ask the questions later.

RetiredF4

As far as i understand the system it works with two data-bases, one terrain and obstacle database and one airport database.
When the aircraft is landing at an airport, the system uses the airport database, otherwise it would not recognize the aircrafts landing intention and give out continous alarms.
If there is no database for the airport, exactly that happens, alarms where there is no real trouble.
Now how do you decide which alarm is real trouble and which one just due to missing airport database? As it looks like, the crew of Polish 01 didn´t pay any attention to the system.

But i´m sure, some expert can explain it a lot better than myself.

franzl

BOAC

I agree - I was being 'economical' with my words and following the query about the airport database - I would assume a terrain model of the area existed.

FullWings

Hmmm. "Distracted" by hard EGPWS warnings while below MSA in IMC... Not a lot you can say about that, really.

At least we didn't hear the equivalent of "shut up, gringo" in Polish but the overall effect was the same. What's the point of a terrain alerting system if you ignore it?

IMHO there's not a lot to learn from this accident apart from that if you bust minima in poor weather and don't react to time-critical warnings, you're probably going to crash. They got that bit right...

SadPole

Some recent "news" on the subject allegedly from the "new crew" that says it cannot possibly be the pilots' fault. Well, they say that the transcript cannot possibly be real and must be cut and pasted together from other flights because:

• the CVR holds 30 minutes and the transcript presented shows 38 minutes and there is no service record showing that the tape was substituted for a thinner one (this part is unimportant)
• they say that the way they land this plane you cannot go below DH because they land it on autopilot and there is no way to go below DH entered into the autopilot (????)

While I have a very low opinion of the "new crew", this second part got me thinking and I went looking for Tu-154M flight manual to figure it out how they set the ABSU autopilot to bring them down to the DH. All I could find was the manual translated by the simulator guys, but, there is some interesting info there.

• The autopilot has three vertical mode buttons (M- mach hold mode, V- IAS velocity hold mode, H-altitude hold mode. So far so good.
• But, where do they set this DH? Well, according to the instruction I found, they set it on the radar altimeter: "RV5 Radar altimeter. The clickspot on the left is used to set the green index during approach to circling altitude and later to the decision height. It triggers the ‘H’ (altitude hold) signal light (on autopilot???) with corresponding sound when the radar altitude passes this altitude."
• And then: "During the flight on the glidepath, ABSU will change the settings of controls on autopilot signals radar altimeter RV-5" and "During the approach, the longitudinal channel may be located in manual mode or in the stabilization of a given pitch angle, or in the "H" (altitude hold)."

Please note that from early on all the old Tu-154 pilots would instantly recite (even before the transcript was released) that you never, ever fly on autopilot below 100 meters as some sort of holy dogma that they would never, ever dare to offend. What do they know that the "new crew" did not? Is it perhaps the slight little problem that autopilot being directly driven by RA is not going to be too hot while flying close to the ground over say a deep ravine?

I know it sounds NUTS, but if this is how the friends of the dead guys say they fly that thing and this is how this ABSU thing works, the following scenario is possible:

• They know they are a little too high so they increase the angle of descend.
• They set the RA Altitude Hold to 100 meters decision height and enter the altitude hold mode thinking the autopilot will level them off at DH.
• The autopilot gets them there, with slight little problem that it locks onto the falling edge of the ravine because they either forgot about the ravine or thought they should be further ahead, or thought the radio altimeter will filter out the ravine. And for a while, they are actually happy while they fly parallel to the falling edge of the ravine thinking they leveled off at DH, and even ignoring the TAWS pull-up warnings generated as a result of data from forward looking radar. .
• Only when the RA starts reading the rising edge of the ravine, the bliss disappears, because, suddenly they seem to be descending at a very high rate.

Note that if only they were several hundred meters further and RA never got to read mostly falling edge of the ravine, the scheme would work (even as stupid as it was).

Does this make any sense whatsoever?????

ARRAKIS

There are a few holes in your theory.
RA was set to 60 m. You can find it on the transcript.
The autopilot thing is a little bit tricky as it wasn't a standard Tu-154M. We don't know how it was coupled to the FMS and other new installed equipment . What amendments were added to the original, Russian FM and how the crew was trained to use the new equipment.

Arrakis

Alice025

To me, of the whole congregation (crew and ATC) - during the flight btw the two Markers - TAWS and 2nd Pilot look the most brainy.

I don't know what was fed into TAWS in the absence of formal maps, but somehow TAWS was of the same opinion that the 2nd pilot - "go around, up, up". I read somewhere it's not only maps that the TAWS feeds on, but is straightforward connected to the altimeters on board. Is it connected with the vertical speed?

May be TAWS can scream warning of low height + high vertical speed combination. ?

There were counted in 101 photos seven altimeters of all types in the flight deck.

Alice025

Now, the requested translation. (brace yourself for the Russian English)

"Here is rolled out a T-shape anthennae. Seventy-meter three-ray half-wave vibrator. As the Beacon is Near, the height of anthennae lifting is 7 meters. Stationary PAR-10C is located inside the building. Anti-weighters?
Contra weight balancers? are located undeground.
Nearby the building we've got the anthennae of the marking radio ?"light-house!"? E-615.

On the building is a lantern KNC-4Y (code-neon light ? "light-house")
It issues by Morse alphabet by method of ? sparks? the letters, given to the Far Beacon of this direction of landing.

(mind it which letters and given by whom I don't understand even in Russian - Alice)

Overall, there also exist a thing called "umbrella anthennae" (but I may be wrong ab it) for the Beacons. I've seen this umbreall thing only once.
It looks like a mast 10 m (Near Beacon) or a 22 metre mast (Far Beacon) and the rays are stretched out in an umbrella kind of shape go, from upwards to downwards. This thing was delivered together with a device type PAR-9 (on the base of YAZ-452) but this very PRC did not became a mass serie production becase it proved too capricious in operation.

It can be located along the direction of the landing and can be located across the direction of the landing.
Overall, I've got several photos of the Far Beacon, takes ab 10mb, if you wish give me you mailbox I'll send them to you.

______________
There is a checkered black and white building that has the "lantern" on top, this is the same lantern on top of the vehicle at the Smolensk radio beacon and it looks as if the same T-mast antennas that are mentioned at the building, are the same at the radio beacon at Smolensk."
_______________

"In military aerodromes, as a rule, all equipment is coloured khaki.
While this code neon light light-house burns red and is thus seen from all sides.
It's seen at night in ? PMY "P - meteo conditions" approx. at the distance of 50-60km. It is switched on normally at night or can be in day time in case of bad visibility.
It looks like a tube because there inside it are neon lamps - type of usual day-time light neon lamp bulbs."


"About the Marker (from SU-24) (from the plane? Su? Alice)

: Sound signalling MRP-56P can be heard in the back-ground of signals R-862, R-864, ARK-15M. Far Beacon emitts signal in the zone at Height=200 - 300 meters - 800 meters by course and 1,200m in area perpendicular to the course;
Near Beacon - at height = 50-90 meters - 600 meters by course and 800 meters in area perpendicular to course.
Height when it works when flying above the Beacon is not more than 2,000 meters."
_________
Picture of the light house thing
_________
Post #29 shows it clearly and prior post #21 talks about the PRMG 76U and the PRMG-5
_________

Karel_x

RetiredF4 -The translation of CVR is not fully accurate. I will try to explain and comment it:

10:09:55,2 N /(What kind of landing? ) procedure.
10:09:57,4 KBC It is not known yet
10:09:59,3 N / INT Landing data.
10:10:00,6 2P / 2P Partially recorded
10:10:06,2 N /We have a course of 72 working hours GIC working in induction gyroscope mode
10:10:10,3 N / INT PB-adjuster. PB. (Altimeter adjustment?) PB (in latin RV)=radar altimeter
10:10:11,9 KBC 100 meters.
10:10:14,1 KBC The course of the band through a minute. ??? (I tell you) Heading of RWY in a moment
10:10:15,0 N / INT Fuel.
10:10:16,8 2P / 2P About 11 tons on land.
10:10:17,9 IP / B / And I admit it. I admit it. (???) Cofirm
10:10:20,6 2P / 2P Okay, I do not set yet.
10:10:23,0 KBC We will establish a 2-5-9, with the other side. from that direction (landing from opposite direction?)
10:10:48,2 KBC The course of the band 259 is installed. heading of RWY is set

Fox3WheresMyBanana

Karel x, many thanks for the more easily understood translation.
It is possible it was the autopilot following the RA and therefore maintaining a false 'level' in the critical moments. When I said it was under the guidance of the PF, this could have been how he set the autopilot.
Karel x. Is it possible you can pick out from the transcript for us:
1) where the actual approach decided upon was, and what it was?
2) whether there's any briefing or mention of the ravine by the crew?

Lastly, if ATC had glideslope information, why are they calling "on glideslope" when it isn't? Are there wave propagation errors due to the fog/whatever? IIRC, PARs can send you way off both track and glideslope in certain conditions. I remember several occasions when mates on PARs have punched out of cloud to be staring at buildings, and it wasn't always Air Tragic's fault.

vakakaaa

10:39:37,3 KRL / Д Полоса свободна. (Runway is free)

Remainds me of our many AFIS airports. One AFIS+Tu-154 incident from Finland: Onnettomuustutkintakeskus - B 1/2005 L

AFIS - Airport Flight Information Service, providing information to, but not control of, aircraft using that Airport.

Tagron

Sadpole (#570) I will try to answer the questions you raised about the use of radio altitude. The usual caveat applies about my lack of Tu154 specific knowledge, but in this case these comments should be common to all types.

There is no direct interconnection between the autopilot and the radio altimeter in the way you describe. The altitude hold facility is a function of the barometric altimeter, and pitch modes or indicators are a function of barometric vertical speed.

All instrument approaches are published using barometric heights and descent paths. One reason for this is that it would not be possible to fly a stabilised approach, or fly level, over undulating terrain using RA. Radio altimeters usually have a maximum usable height of 2500 ft.

DH for non-precision approaches is always barometric. RA DH is used for ILS Cat2/3 and may be published for ILS Cat1, though in that case there may be a requirement to fly a coupled (automatic) approach. Autoland uses the RA height in various ways, such as rejecting the glidepath signal at 50 ft RA in order to perform the flare.

Other than the above specific cases, crew use of RA is for the purpose of height awareness, not control. Low level RA heights are called by a crew member as part of normal procedure, or automatically, and in large aircraft are used to help judge the flare.

So the DH alert should have had no function on the Smolensk approach, but they had to set the alert somewhere, and (my guess) they chose 60 RA only for that reason. Its function on that approach should have been only advisory.

Hope this helps.

BOAC

Fox3 - your mates must have been very unlucky! In 17 years of PARs, NEVER have I been anywhere other than spot on, and on many, talkdown to touchdown - on c/l and numbers. Ah well, that's the RAF for you

RegDep

Could you please elaborate a bit. What implications would you see, then?
Best
RD

vakakaaa

Ciao RD old chap. Dont have anything wise enough to add at the moment. One important QUESTION however:
-- Does anyone definitely know the official status of Smolensk Korsazh ATC ???

Alice025

Important addition to "does anybody know....?" SINCE LAST SUMMER.

(when the military have moved out; the aerodrome ceased to be military, and is known now - at least if to judge by what was heard around post accident as "in joint military -civillian use")

Karel_x

Fox3WheresMyBanana:

1) I did not find any explicit notification about landing procedure in CVR transcript, neither in the cockpit discussions nor in communication with ATCO. Only a few remarks:

10:24:51 ATCO: Temperature +2, QFE 745, there are no condition for landing
10:25:01 PIC: Thank you but if possible, we do the approach and if not wetter (conditions) we make GA
10:25:12 ATCO: 101, after checking/trial approach, have you enough fuel for diverting?
10:25:19 PIC: We have
10:25:19 ATCO: Roger
10:25:22 PIC: Request for descent, please
10:25:25 ATCO: 101, heading 40, descent to 1500
...
10:30:01 PIC: No ILS, heading 259, ARK ready, 310/640 was set, 5-6-autothrotle [two NDB]
...
10:34:56 ATCO: 500m…did you land at military airport (already)? [= Are you familiar with military landing procedure?]
10:35:03 PIC: Yes, sure [and nothing more...]
...
10:35:22 ATCO: Polish 101, at 100m be ready for GA
10:35:29 PIC: Exactly
...
10:39:38 ATCO: RWY free
10:39:40 ATCO: Landing (clearance) later, 120-3 m [wind]

2) It is hard to say, it is not clear what “hole” /dziura/ means. It may be the valley beneath glide slope:
10:30:45 PIC: The worst thing is that there is a hole, there are clouds and a fog fell

3) I can’t explain what ATCO’s calling “on GS” exactly means. I suppose that he has information from landing radar witch accuracy on height was approximately +/- 50m, so he need PIC’s height backreports. On distances 2km and more, ATCO can be sure that 101 is safe on/above GS. There is the problem to harmonize OM&MM signal with distances reported by ATCO (for example - signal OM/6,1km/ came 4sec after ATCO’s report 6km). It looks like the difference between ATCO’s reports an actual location of 101 was constant (320m).

vakakaaa

Short history of XUBS in Russian and in English, respectively: ---- for you, Alice, a big bush of roses !

Ú-Ãàçåòà - Àýðîäðîì Ñìîëåíñê-Ñåâåðíûé ------- Smolensk North Airport - Wikipedia, the free encyclopedia

SadPole

Targon wrote:
I understand this is how it is supposed to work, but apparently NOT on Tu-154. I have checked everything I could find in regards to altimeters on Tu-154. In another part of the documentation it says that Tu-154 autopilot has no connection to the barometric altimeters at all, so when cruising at high altitude, in the H-altitude hold mode the autopilot only keeps the plane leveled. It is possible to set min/max on the digital barometric altimeter, but only for the purpose of sounding alarm when the range is breached.


I know it sounds REALLY REALLY GOOFY, but I tried to figure out where this "never ever land on autopilot" Tu-154 dogma came from and this is where it led me.

The reason guys like me didn't like Soviet Union was not because we could not live without sex-shops or McD's but because we did not like politics driving everything, including engineering (See Chernobyl or K-19 for premiere examples of what it leads to). The very reason this whole affair gets to me so much is because it proves that we still have not gotten rid of that idiocy neither in Poland nor in Russia.

Now, I am not trying to blame Tu-154 design for this mess. If they flew it, they should have known everything there was to know about it, especially all possible quirks. And I do hope someone gets to Tu-154 documentation and disproves it, because, as I said, it sounds completely NUTS.

SadPole

That's the part of the CVR that got me thinking these guys were setting "their" Decision Height on Radar Altimeter. See my long previous posts in this thread. Yes, sounds really, really STUPID but that's where reading the transcripts, interviews with the pilots, and the Tu-154 documentation led me to.