Simulator assessment - a matter of personal opinion

[Centaurus]

Circling approach in simulator. Due to simulator visual display cut-off limitations, the turn to base and final is done using MAP mode until half way around final when the runway comes into view. In other words, the turn is on instrument indications and not physically looking outside as you would be doing in a real thing.

Within about 20 degrees from the extended centre-line the PAPI now appears through the visual display and indicates aircraft is high (all whites). The runway is sea level and length 3500 metres; more than enough for jet transports. The pilot has to make up his mind. He can choose to go-around because without becoming unstable he cannot increase rate of descent to get back on the PAPI visual 3 degree slope.

Or, he can judge the aircraft can land safely further in to the runway within normal stable approach parameters. He chooses the latter and touches down 1000 metres in from the landing threshold with literally a mile of runway remaining when he stops with normal braking.

The simulator instructor gives him a FAIL because in the instructors opinion the pilot should have gone around again because he was too high to be stable by 500 ft. Yet the pilot did conduct a stable approach to touch-down and well within the landing distance available.

The basic cause of the high approach was the lack of simulator visual fidelity in the first place placing the pilot in an invidious position where to play the game the base and final turn was made using instrument flying and navigation displays instead of visual flying.
Comments appreciated.

[Bealzebub]

He chooses the latter and touches down 1000 metres in from the landing threshold

Got away with it today, but that isn't the planned touchdown point.

The simulator instructor gives him a FAIL because in the instructors opinion the pilot should have gone around again because he was too high to be stable by 500 ft.
You can understand why.

The basic cause of the high approach was the lack of simulator visual fidelity in the first place placing the pilot in an invidious position where to play the game the base and final turn was made using instrument flying and navigation displays instead of visual flying.


Simulators are less than perfect in this regard, most people would acknowledge that. However they let you walk away unscathed whatever decision you make. The real world is often much less forgiving. If you find yourself high or fast or both in either domain and for whatever reason, it is the choices that you make that matter. This illustrated choice when it was felt the pilot was in an "invidious position" was to press on with an obviously high approach that gave him a touchdown point three times further from the threshold than it should have been.

What the pilot did (according to you) was to conduct a "stable" approach on an incorrect profile to a grossly incorrect touchdown point and judged it to be Ok because on this occaission the runway was long enough to accomodate the poorly executed approach.

If you had been the instructor/examiner what would you have done?

[BOAC]

Far better, is it not, to make sure you get the 'g/a' call, freeze the sim, discuss and THEN continue if time will not allow a g/a. Surely also it needs to be asked - why was it high at the 30 to go point? Was it flown correctly?

[Mad (Flt) Scientist]

The simulator instructor gives him a FAIL because in the instructors opinion the pilot should have gone around again because he was too high to be stable by 500 ft.

Had the FAIL been because of "failure to conduct a good circling approach" then I think the pilot could reasonably feel hard done by, since the simulator itself prevented a good circling approach by virtue of its limitation.

But if the fail is because of something else, then it doesn't really matter WHY the pilot was high - he's being assessed on his reaction to being high. That could have been induced by any number of means, some of which could be representative of real world conditions. You can end up high on a real approach - what decision would the pilot make?

Non-pilot opinion, but it really comes down to what the assessment is against. If I, while learning to drive a car, am doing the parallel parking exercise and some swine nips in and steals the space while I'm setting up, I doubt I'd be failed if I drove off and said to the instructor "cant park there". If I persevered nonetheless, and either rammed the other guy or got into a fight in the street over 'my space' I think I'd deserve to fail the test. Even though the 'space stealer' isn't part of the test setup, my reaction to it is an indication of my behaviour that CAN be judged.

[welliewanger]

I agree that "practically" the pilot did the right thing. However "legally" you have to retain visual contact with the runway during the circle. Since (due to simulator inadequacy) this is impossible - and - in a sim check you should do everything "by the book" everyone should go around when performing this procedure in the sim.

IMHO the pilot demonstrated good airmanship and judgement. The trouble is that, for some people, boxes have to be ticked and hoops jumped through.

[Otto Throttle]

The original poster does not state their airline, but from my point of view (and that of my employer), a stable approach requires you to be following the correct approach profile, vertically and laterally, amongst other things. Four whites on the VASI/PAPI is not on profile.

A landing 1000m beyond the touchdown point would also be considered a flight safety violation.

Did the pilot in question have a 'should' gate in mind for the mid-point of the base leg in order to assess his vertical position regardless of simulator visuals? If not, hopefully a lesson learned. If he did, was he high at this point and why did he choose to continue?

[Northbeach]

With all the instructing you have done over the decades, you would probably agree that if one "wants" to one can FAIL just about any pilot during any simulator session.

Slow to rotate past Vr, FAIL. 11 knots fast/slow after multiple altitude, airspeed and commanded configuration changes-"sorry Charlie"-FAIL. During a Non-precision approach, I say you started your descent prior to being established on the final approach-yes you did I am the instructor and I say so-you FAIL. It goes on and on.

In a simulator without approved visuals capable of representing circle to land to FAIL a pilot for the performance described is weak in my opinion. Especially
If the rest of the ride was good.

This situation is another reason your instructors need to have their own ego in check.

[Sky Wave]

In a simulator without approved visuals capable of representing circle to land to FAIL a pilot for the performance described is weak in my opinion. Especially
If the rest of the ride was good.

But the fail was not for the visual manoeuvre, the fail was pressing on to a long landing when a go-around should have been the correct course of action.

Over the years, many aircraft have crashed through pilots pressing on regardless. There have also been many runway excursions caused by long landings.

Yes, this runway was 3500m but what if the decision was taken on a shorter runway, what if there was a tailwind, wet runway, downslope, late reversers selection etc.

Our company is extremely worried about long landings because people try and hold off for a nice smooth touchdown. One day someone is going to do it with the wrong conditions and they'll end up in the dirt.

Sorry for the person concerned, but unless there's an emergency or a very serious reason not to go around (lack of fuel) there's no excuse for landing 1000m into a runway.

[lederhosen]

It does rather depend on what you are testing or required to test. We use simulators with excellent visuals and some with the limitations you describe. We also train the circling exercise mainly as an instrument procedure using map mode and lnav on the 737. The transition to visual flight particularly somewhere tricky like Salzburg can be interesting but usually works fine.

If this is the guy's only misjudgement it is tough to fail him on this alone, particularly if you dont let him try it again. If he is generally borderline then that is another matter. I agree about the importance of landing in the touchdown zone, but then again in the real world people do not land like they do in the simulator. However we have failed people for not being able to get into a position to land after several attempts usually due to finger trouble with the fmc. So the answer is it depends.

[compressor stall]

But to what azimuth is a PAPI accurate?

I believe that they are designed to 10 degrees, but in Australia only flight tested to 5 degrees. Thus anything outside that should be ignored?

[cavortingcheetah]

This reads like the scenario you can get away with in a turbo prop but not in a jet. One which you might think about when flying the former but never entertain when commanding the latter. But this time it was a simulator check and maybe what the instructor wanted to see was a judgment call? What he got was a landing but all pilots can do those.

[lederhosen]

As a TRI on the aircraft type I believe Centaurus is refering to I cannot see much benefit in deliberately setting a scenario where you cannot reasonably expect the trainee to get down due to the limitations of the sim. If you want to see a go around then by all means obscure things on final or instruct him to go-around at some point earlier.

More likely are either finger trouble from the instructor in setting things up like the winds and a sensible airport or the trainee not following the procedure. If you are at normal circling height, fly the correct time from abeam and are fully configured the autopilot should fly you round into a position you can intercept the glide. If not the examiner is playing games.

I am sure 411A would have loved to tell us how real men did it in times past on the mighty threeholers. These days it is a pretty straightforward procedure, but one people seem unnecessarily nervous about, probably because they do not do them often enough.

[Capt Chambo]

The simulator instructor gives him a FAIL because in the instructors opinion the pilot should have gone around again because he was too high to be stable by 500 ft.

The right decision IMHO.

The approach was not stable at 500' HAT (VMC)
The landing was not made within the touchdown zone

No comfort should be taken from the fact that landing worked out "OK" in the simulator. Landing over runs have been a major cause of aeroplane accidents/incidents in recent years. QF 01 Bangkok, AF 358 Toronto, AIX 812 Mangalore, AA 331 Jamaica Bay, all spring to mind.

Speculating now but, you don't mention whether this was a multi-crew operation and whether the PM had anything to say. I mention this because "if" it was a multi-crew flight, and "if" the PM was calling "high" or "all whites" or "Go Around" and the candidate continued then this would raise further question about the candidates judgement.

[welliewanger]

To all the nay-sayers here. I have one question:
How long would the runway have to have been to allow this to be a safe landing? The original poster said that the aircraft stopped with "literally a mile of runway remaining" If a mile of spare runway is inadequate, would two miles be sufficient?

It sounds to me like the approach was stable, but the pilot had to adjust his aim point further down the runway.

[Bealzebub]

How long would the runway have to have been to allow this to be a safe landing?

Of course that is the problem. Turn the question around for a moment and ask how short did the runway need to be to ensure a safe landing? The performance figures for landing make certain assumptions, and one of those assumptions is that you will touch down at the correct touchdown point. It becomes very difficult to establish performance criteria if you allow a situation to develop where you are not complying with those criteria. If a 10,000ft runway has a displaced threshold of 2000ft which would broadly equate to the situation illustrated here, then there is effectively only 7000ft of runway from the touchdown point remaining. In a situation where this is known it may well involve different planned application of braking or other retardation. Weather conditions may well make the use of the shorter runway length unacceptable. All of that would be planned for in advance.

Simply flying a high approach and creating this reduced effective runway length hasn't been planned for, and therin lies the potential problem. Ask how short the runway would need to be, in order to be acceptable in this scenario?

What is wrong with simply throwing it away and having a second attempt to get the fundamentals right. It might be much harder to do, but is likely to show far better overall judgment when it comes to commercial air transport operations.

As has already been pointed out, there have been too many instances of aircraft landing deep and then running out of tarmac to justify what is really just a normalization of deviance in these circumstances.

[fireflybob]

Treat the simulator exactly like the aeroplane! If you're not stable etc Go Around!

There isn't a sim visual in the world that is precisely the same as the real world but you have to play with the cards you are dealt, so just play the game!

Be careful about "rationalising" why you took a particular course of action because "it's the simulator"

[cavortingcheetah]

It hasn't been mentioned and perhaps the cockpit set up or pre flight brief precluded such a thing but perhaps the co-pilot/captain might have been in a position to prompt a go around?

[lederhosen]

Most of us seem to agree that unstabilised approaches and long landings are a no no. If the exercise cannot be reasonably carried out due to the limitations of the sim then the checker is on shaky ground. Put it this way if your upgrade was delayed by at least a year because of this you might feel aggrieved. After all not being able to get down off a circle to land might also be construed as a fail. We have certainly asked people to come back for not being able to complete one. It is an interesting one. What is your opinion Centaurus?

[The African Dude]

he cannot increase rate of descent to get back on the PAPI visual 3 degree slope.
In most cases this would represent a very significant increase in the ROD (consider 3-400fpm while configuring on base versus 1000fpm stable criteria, for example) which in turn suggests a serious profile problem.
Circling is a visual maneuver so what prevented the PF from increasing their track mileage by flying through the extended centreline, at ATC's approval, keeping the runway in sight? N.B. This ignores the question of why the aircraft ended up so high in the first place, and ignores the fact that this is a simulator because that should not affect the decision to continue or otherwise.

That said:
A commercial operation provides safety margins which the passengers pay for. These margins are achieved through, in part, adherence to company SOPs. If they state you have to be fully stable at, say, 400ft for a circling approach and you instead choose to ignore the SOPs and point your passenger jet at an arbitrary point on the ground then you have made the decision to ignore the in-built safety margins based on a hunch that may or may not work out.

Inextinguishable fire: may be (but not necessarily) a different situation.
Normal ops: doesn't matter if you think it was "safe". You don't own the aircraft and must play by the rules which are in place for good reason.

[de facto]

Was the captain given the chance to conduct a proper briefing for the approach?or was it hop we go 5 miles to mda...fly and think as you go...
Non precision app key of success or at least safe outcome is a proper briefing so both pilots are on the same page on how it will be flown and when automatics will be disconnected,and how will the missed ap/go around be flown if above or below mda....

Poor briefing leads to poor decision!

Now if the captain had the chance to do a non rushed briefing and it hapened as mentionned:
A FAIL for the circling exercise for sure(attempt1)but repeating the item after a quick chat on why:
He thinks a 'stable' ROD leading him 3000 feet from threshold is acceptable.
If he knows his expected landing distance.(based on weight/ab setting/wind...average skills).
Ask if he believes his action of continuing was due to his experience or sim rush to complete the exercise.
If the pilots believes all is OK , a FAIL and a serious debriefing is required.

Too many overruns by these 'no prob i can do it' egoes.... Time to weed them out if they dont get the potentential danger of their actions...

A -800 at 65t will need more or less 2000m with ab3(including average skills/not over flaring),a Vref +5 at threshold height(as it is in most cases).
So touchdown at 3000ft +6600ft =9600ft....so a mere 1000ft remaining.
A bit of tailwind and manual braking needs to save the day...:rolleyes:

[Doors to Automatic]

One dot high (or PAPI equivalent)at decision height = go-around. No ifs or buts.

Follow SOPs to the letter was a lesson I learned at interview stage -even though In my case my flying career never went any further!! :p

[Slasher]

I've NEVER accepted a visual-based approach in any sim with
limited screen visuals where my licence (and income) are on
the line Centaur, much less accept the following of any damn
box-generated magenta or green line within the circuit when
my head should be outside watching the runway/terrain like a
bloody hawk. That's not me trying to be an inflexible hardass,
but trying to keep an uninterrupted paycheck and unblemished
resume.

If the visuals were up to par then likely this event wouldnt've
happened. However in this situation Id've gone around - sim
checks are not only a test of simulator-handling skills and sim
technique but of SOP knowledge irrespective of practicality or
any successful exercise outcome.