Inexperienced Pilots of LX 3597

[TechFly]

NigelOnDraft I understand your point about RA setting.

Nevertheless when you join an airline with your past experience and pilot techniques you must follow SOPs.

It is difficult, as other collegue was pointing out, but it is the reality.

I was in such a situation different times: normally I try to use the good sense.

Coming back to the discussion on the RA setting my believe is:

1. RA, theoretically, is to be used with approaches demanding its setting (see previous well explained posts);

2. My company XXX is requiring the RA setting by the pubblished SOPs (DH or 150ft on Prec. App. and 300ft for NPA/visual);

Now the question is: who is right and why?

If you follow #1 you don't follow the SOPs.

If you follow #2 you follow the SOPs but maybe you are apporting items disturbing the crew and the situational awareness.

I think that both have some good and bad aspects: it depends on the approach profile (see terrain orography vs. path), meteo, crew spectations, ...

I will look forward for your comments.

It is important to clarify that this discussion is respecting the crew involved and the only purpose is to provide a constructive discussion.

My condolences and sympathies to the people involved.

[DownIn3Green]

I've had IT with this RA discussion!!!

On the generic 727 checklist I've used on many contracts, "Radio Altimiter" is there on the "Approach" check. Even though we are not and never have been (on these contracts) CAT II or III qualified, the unintiated still insist "I'll just set it as a reminder" or some other insane response.

It is not required or AUTHORIZED even as a reminder, and to use it can cause a pilot to bust minimums on a Cat I ILS or Non-Presision Approach.

Period.

Not to accuse the crew of this Crossair Flt if that was their procedure, but aren't we meant to learn from out errors?

[411A]

Yes, but in some cases DownIn3Green, it is the situation of....don't confuse us with facts, our mind is made up.

[Ignition Override]

Downin3Green, and others who can't stand the idea of a radar altimeter being set at all (even if not prohibited by my company): this is not to point the finger of blame for any accidents at ATC, but---

Years ago during the approach to NAS Fallon in Nevada, a Navy Reserve (D)C-9 crewmember had his RA set to some setting. Suddenly, while being vectored either at night or in IMC, the light came on with the needle quickly turning. The crew shoved the throttles forward and climbed, without asking questions. It turned out that the Fallon Approach controller had given a certain vector to the wrong aircraft. This was in the Navy's "Approach Magazine". This is also how Frank Sinatra lost his mother along with the rest of the crew in a Learjet in southern CA many years ago.


I.E., if ATC vectors you away from a SID or STAR, how can you always know whether mountains/towers are at or above your altitude and on your present heading? What is wrong with an RA set to 1,500' or 2,000' if your company does not prohibit this? We have no information to tell us the minimum vectoring altitudes while blindly following ATC clearances. We have only a general idea where the highest mountains are at night/IMC. If a pilot or ATC mistake is made, or Salt Lake, Denver or Alb... Center's (or XXX Approach!) radar goes down, who cares what the cockpit RA needle is set to? Does anyone set the RA as procedure into Eagle, Steamboat, Hayden CO etc? A Convair 580 was vectored into a tall tower years ago somewhere around North Dakota, and luckily the wing held together. I rest my case, at least for now.

Going to annual training tomorrow, will ask our IP what he/she thinks. Will check y'alls' responses when back en mi casa.

Hasta la vista.

[ 03 December 2001: Message edited by: Ignition Override ]

[411A]

Yeah, but the Navy doesn't have GPWS so they made the most of what they had.
The upcomming requirement for EGPWS "should" make these CFIT incidents/accidents mostly a thing of the past, I hope.

[NigelOnDraft]

Re RA setting comments...

Maybe I could clarify my view! I am not saying that the RA should ONLY be set on an approach with a RA minima. However, I am saying that IF you set it in other circs, it must be very clear exactly what you are expecting to do if it goes off...

The Navy crew, as mentioned, set it, and acted on it. Sort of DIY GPWS as has been suggested.

We in BA have the RA go off at 2500' automatically. This is early enough for most terrain as a decent alert that the numbers (Baro Alt, terrain, position) tie in. 300' is a bit low and late for this sort of check!

What "concerns" me is the attitude of setting it, and not knowing why! It could lead to a mind set of going through minima (after all, someone said here "NPA minima is usually ~300' AAL") and awaiting the RA warner.

Bearing in mind the Baro MDA here was nearly 1000' AAL (2390' v 1416'), I stand by the question of what was the purpose of setting the RA to 300'? And if an SOP, what is the corresponding SOP if it goes off?

NoD

[propjock]

Where I earn my daily bread ( and wine) we have introduced a procedure to set the RA at 200' for all CAt1 and NPA approaches.
This because "...studies conducted by the Flight Safety foundation clearly indicate that the radio altimeter may largely contribute in avoiding CFIT ..." ( O.M. pt A)

I have my doubts about this because as said before, the "minima" call might be confusing and I suspect some people do not get the point in the difference between a setting of 200' as a procedure and a RA set as minima...
but, as it is a SOP...

On the other hand we have the concept of "monitored approaches", where below certain WX values, the F/O flies the aircraft while the captain monitors ,takes over and lands if he is satisfied and has visual clues.

Also we are not allowed to fly level at the MDA, we fly a 3° slope and at MDA we go around if we do not have enough visual clues to perform a safe landing...( wedo consider the MAP for turns after G/A ect, of course)

Just to tell you how we do it here in Lux...

[NigelOnDraft]

I'm going to take a look at the plates and minima later, if I can. However, as has been alluded to earlier:

1. Reqd Visual Refs for a NPA are minimal - for us "1 element of the approach lighting system".
2. Here the MDA was ~1000'AAL = > 3NM from Touchdown.
3. From MDA / MAP there is no "glideslope" guidance, not even a table of Alt v DME.
4. If you are visual (night - with just 1 element of the approach light system) at 3NM, and set up even a slightly high RoD (with no real benchmark), the first you know of the impeding doom is the rising ground obscuring your view of that 1 element of lights you had. Suspect its all over by this stage... Even with a 300' RA warner (very late with RoD on AND rising ground? )

I would be very interested to see the exact layout of the terrain in the last 5NM or so for ZRH 28...

In addition, what are the Vis minima for this approach? With an MDA of ~1000'AAL, surely in the order of 3NM? And what were the minima for ZRH "enforcing" the compulsory 28 approach after 2100Z? I seem to recall the vis in the METAR as 4500m?

Possible airmanship grounds for one briefing in these circs that to go below MDA, one wishes to continuously see the PAPIs (does 28 have them?), and keep a satisfactory number of whites!

NoD

[ 03 December 2001: Message edited by: NigelOnDraft ]

[Few Cloudy]

Nigel, one more time - the 300ft Radio Altimeter setting is NOT for terrain warning or minimum purposes, but for the Power and Rate check which some Swiss airlines do.

It seems that Crossair have their Radio Altimeter programmed to give a "Minimum" call at bug, which I think is a BAD thing when it is not being used for the designed purpose.

PS Newspaper info has the flight south of track on final (steady standoff) and descending early in mid approach. We will see.

RW 28 has a steeper than usual PAPI.

[DownIn3Green]

IO,

I agree with you in your view, the RA is a valuable tool for that situation.

I will use it when cleared for a visual approach sometimes. For instance setting 1,000' or 1,500' depending on the situation. However, this is not a "minimums" reminder, which if using the RA as such when not authorized can lead to "confusion".

[rolling20]

Firstly my sympathies to the crew and passengers in this tragic event.
id just like to add to the balanced views on this incident,that this seems to have the "hallmarks" (crew wise) to the Little Rock crash in June 99!
There i believe the Captain was a veteran of (i believe) some 30 years experience,and the F/O was a newly qualified member of AA flight crew.. i am unaware of his exact length of service!
AA was pilloried at the time in the press for allowing such a situation on the flight deck to exist..ie it was termed as a Major General being crewed with a 2nd Lieutenant and all the problems inherent with such a crew pairing!
I understand AA changed their crewing procedures after this tragic event.(altough im unaware of any NTSB recommendations at this time)
However there must be literally scores of similar crewing scenarios all over the world.
If it is seen as a major problem under any workload situation,then airlines unilaterally will not allieviate the potential problem that may exist!
Perhaps we may see from the eventual report on this accident(or Little Rock) the recommendation that airlines in general adopt a more "balanced crew" when it comes to CRM!?

[nimbuscumulus]

Ratherbeflying,


I like your approach of this unfortunate event. Just a couple of things:

Yes, there was a visual reference, on the plate it refers to a papi to be followed when visual.

Because of that, we must assume that the crew was never visual with the runway of it's lights. The crew did however report (actually the captain did) visual with the ground below. The metar shows few clouds at 600' feet. Also the visabillity was reported to be 3500 meters. this is about 2.2 miles...
Another aircraft reported visual before at 2.4 dme.

One thing I don't understand though:

If the Captain was flying the approach and the f/o was monitoring, why was the cpt looking outside? I thought it is normal for pf to stay on the instruments 'till the pnf calls visual. Maybe this was a contributing factor in the accident as the Captain might have tried to keep visual reference with the ground.

Ofcourse no judgement here, I'm just trying to understand and learn like everyone else.

[Dagger Dirk]

While icing cannot be ruled out (and is never likely to be proven/disproven as a factor), an unanticipated stall is less likely to have been the main cause. It is far more likely that the captain "chucked away" the instrument approach based upon fleeting ground contact in the turn onto the finals course - and decided to press on visually. You could uncharitably say that they perhaps threw CRM, airmanship and SOP's out the same window through which they caught an enticing glimpse of ground - during the finals turn.... but I won't.

At night "going visual" early is always fraught because there is a dearth of info to stop you going low. That's why it is invariably safer to fly that approach to the point where the runway environment is in crystal-clear view. But the fact that they were probably loaded with ice could have given them a much higher rate of descent for the SOP attitude and power set, particularly with the heavy bleed air load sucking thrust - and it's obvious that they were not playing it safe by maintaining the minimum descent altitude of 2390ft anyway. The combination of icing and the lack of true visual cues was likely to have been the lethal cocktail. To put it another way, when your aerodynamics get mussed up by rain-ice and you're carrying a much greater load (of ice) but are blissfully unaware of it, your flight-path is likely to be a whole lot more downish-vectored than would otherwise be the case. Airplanes don't always go where they're pointed!!! And if you're not flying the clocks, you won't necessarily notice that high sink-rate. Why should you? It can be very subtle. I think that's possibly what caught them out.

The captain should not have been looking out, it is always the PNF's duty to declare visual, runway in sight - and that shouldn't happen turning a seven mile final in those conditions. The hazard is that any such declaration is usually the straight precursor to continuing visually (and implies that the PF has all the visual cues that he needs and that a missed approach is not now on the cards). Maybe they don't load up the simulator's performance with rain-ice and demonstrate what a lead-sled an aircraft can suddenly become. Think of it as the FW equivalent of settling with power/vortex ring. There you are pointed at the runway, which may well have been "visual", but you are wholly unaware (until the RADALT cooks off with a sudden unexpected warning) that somebody pressed the down button on your elevator (i.e. the power and attitude you'd habitually set was actually appropriate for a much lighter and cleaner airplane than your flying ice-cube).

http://www.bfu.admin.ch/de/html/Presse_LX3597_1e.htm
At 21:04 hours UTC an aircraft immediately preceding CRX 3597 passed weather information to Aerodrome Control. The pilot stated that the runway became visual at about 2.2 NM DME distance.

At 21:05 hours UTC, CRX 3597 reported to ATC that they were established on the VOR/DME runway 28. Shortly after, the crew completed their final checks in preparation for landing.

As the aircraft approached the minimum descent altitude, the captain commented on this and stated that he had some ground contact. A short time later, the 500 foot mechanical voice announcement activated. At 21:06 hours UTC, the radar altimeter "minimum" call activated. Just after, ATC cleared flight CRX 3597 to land. At the end of the transmission, the CMR called for a go-around and a cavalry charge indicated the autopilot disconnect. Then the F/O also declared go around. One second later the CVR began to record a sound of impact. A short time later the recording of the CVR ends.

When you re-read this excerpt above, in the context of my lead-sled/ice-block scenario, two things come to mind:

a. It fits the bill (both in sequence/ timing and in them being apparently caught totally unawares by that unexpectedly high sink-rate). The DFDR's rate of descent analysed against that to be normally expected from their configuration, power and attitude should be quite definitive in this regard - and permit conclusions to be drawn.

b. If the aircraft ahead of them genuinely became visual at only 2.2DME (1.3 nm from the threshold) it should legally have been still at 2390ft on QNH [i.e. 974ft above field elevation (HAA)]. They would then have required about 1700ft/min rate of descent to make a 50ft TCH (threshold crossing height). That's a real plummet.

In the prevailing weather, with two octa's at around 2000ft QNH (and the main base about 900ft higher), they would have been in and out of cloud at MDA (and would still have been "in and out" at up to 400ft below MDA).
The reason why they suddenly decided upon a go-round could have solely been that 300ft preset RADALT warning - but it could also have been that they suddenly dropped below the base of the scud and saw what a flat trajectory they had to the runway lights and PAPI. They hit the trees at 1784ft on QNH and that was 606ft below their MDA (yet still 368ft above airfield elevation. It would have looked (quite suddenly, as they dropped below that scattered/broken cloud-base) to be an alarmingly shallow approach. And that's the trap for those who think that they can sight-see their way in at night through the murk.

[scrubba]

let me get this right: most of the poms and half of the yanks think that the radar altimeter is your enemy and should only be permitted to operate on some(but not all) ILS approaches. OK, you're entitled to your thoughts but please stay over there and try not to over-influence all of the visiting ozmates.

flight safety foundation (in concert with the FAA and other reputable bodies) has done a lot of work on CFIT. they ran an ALAR (approach and landing accident reduction) task force which recommended (#6) the use of rad alt on all approaches as a protection device.

clearly you experts have forgotten what it is like to operate dumb aeroplanes (mind you the rj is semi-imtellegent) and what it takes to get things organised in npas. i don't know why swissair/crossair requires 150/300ft on the radalt, but i know why it is often done here. basically you fly the airplane to a minima which you brief and remember/set dumb bug, having beeen warned by your partner when you are 100ft above it. at that minima, if not visual, you either decide to go round or in the case of an npa you may decide to continue level to the map looking at your circling options. important to note that, in those circumstances, the radalt "minima" call should never activate. our sop sez: "if the radalt minima call or light activates, you must immediately commence a go round unless you are visual."

why 150/300? for us, the ils minima are not less than 200ft (a 50ft buffer) and the minimum terrain clearance for npas and cicrcling is 400ft (a 100ft buffer).

for the ******s who reckon the noise/light is a distraction, step out side your tiny squares and think of it as a startling warning that is your only gpws mode left when in the landing configuration. and for the others who calculate np approach profiles based on low steps, we calculate a 300ft (or more if required) per mile profile to arrive over the threshold at 50ft, which means that the "distraction" should only activate at 1.2 miles from touchdown, at which point you wanna be stabilised on an acceptable flightpath or you are outta there RFN.

now, just in case some of you missed working out how disagreeable i am, let me be blunt: the radalt is a highly desirable part of every instrument approach. seems a lot of you need to spend a bit less time on pprune and a bit more time catching up with professional aviation things like CFIT/ALAR programs.



[ 09 December 2001: Message edited by: scrubba ]

[Few Cloudy]

Night time noise restriction over Black Forest area requires 28 to be used. The ATIS can be misleading as it only changes 20 minutes after the hour and the RW changes on the hour.

[EDDNHopper]

soddit,

there is an earlier thread where this has been discussed. the use of rwy28 for noise abatement has been introduced after a very long quarrel between Swiss and German authorities/public, and refers to German territory.

[Jumb0lin0]

Scrubba,

You are spot on. You set the dumb bug on the alt, and at that point you go round, BUT IF THE RAD ALT TALKS TO YOU SOONER YOU GET OUT OF THERE. For some to suggest that you might be inclined, if you set the radalt, to wait for the radalt to signal and ignore the altimeter, is plain stupid.

The radalt is a very very handy, last ditch, survival tool.

[vheijens]

It is not the experiece that matter. That only thing that matters is: "Are you qualified to do the job?".

I'll prefer to fly with somebody who has "only" 200h flying at an European Flight School. Thene some American who has 1500h "instructor" time, where he's European student does all the flying for him.

[Hooking Fell]

It appears Capt. Lutz, who was in control of LX3597 during the CFIT has been seriously disoriented on at least one previous occasion.

On 21 March 1999 he aborted an approach into LIMW upon discovering that this was not LSGS where he was supposed to land. Weather was clear on that day.

More details at http://www.tagesanzeiger.ch/ta/taOnl...l?ArtId=147998

[ 11 December 2001: Message edited by: Hooking Fell ]

[Ignition Override]

Interesting points. An unfortunate MD-80 crew was attempting a VOR approach a few years ago in night/IMC into Windsor Locks, CT (Bradley, near Hartford), and they had an old altimeter setting (the ATC tower had been evacuated due to very strong winds). They dragged at least one engine through tree branches on a small hill and barely made it to the runway. The FAA wisely decided to redesign the minima for that VOR 15 approach, due to the one hill on the final approach segment.

During a LOC or BC LOC approach, we push the GPWS light when it alerts us with "glideslope". In this situation we have no back-up for the barometric altimeter, unless someone sets the radar altimeter as a "technique".

If we land with flaps 25 instead of flaps 40 (i.e. tail de-ice inop, single engine etc), we override the GPWS. Potential trouble if no glideslope is available, or tower forgot or was unable to switch it over for your final approach.

Last night we landed in Billings, Montana, which has some low hills. I used the RA anyway as technique (set at 2,000') and told the FO during the descent checklist. Today the only problem was using manual pressurization (not electric) due to the auto having failed (no luck with cb's). Luckily, the FO had done it before (was recently 747 SO) and volunteered to operate it, thus I decided not to return.

[ 11 December 2001: Message edited by: Ignition Override ]