Pilots, do you really not have acceleration data during takeoff?

[Gibon2]

Perhaps foolishly, I've been reading the thread in R&N about the EK tailstrike drama in Melbourne. I may have misunderstood, but several posts there imply that there is no acceleration display on the flightdeck, and thus no immediate way that the crew will know that the aircraft is (for whatever reason) not accelerating as it should during the takeoff roll. A couple of posters mentioned "rule of thumb" guidelines, e.g. 80kts by 1000' or some such, or even "feel".

I just wanted to ask: is this for real?? And if so, why on earth - after all the careful pre-takeoff performance calculations - is this simple, easily calculated bit of rather crucial information on actual performance not prominently displayed for easy reference?

Thanks for setting me straight if I've got it all wrong.

[Chippybus]

On Glass cockpits they have a trend vector arrow showing the acceleration on the speed tape. Larger arrow faster acceleration,smaller arrow less........:D

[Number34]

Looking outside seems to have worked for the best part of a 106 years

[wiggy]

Gibon2

Fair question.

chippybus

True about the trend arrow, but as I guess you know it's length varies with the derate, take-off weight and other factors. You'd expect a short trend arrow hauling out of, say, SIN bound for Europe and that's OK on a long runway...on the other hand is a long trend arrow long enough on a short runway? The bottom line is how do you know if the trend arrow you've got on the day is long enough for the runway length available??? ***** if I know.

[sharpclassic]

Surely this is what our performance figures are for? On my fleet, we get the take off weight, find the relevant box in the performance folder, find the V speeds and FLEX temp (engine de-rated setting), adjust for weather conditions on the day (wind, temperature and air pressure) and plug it into the aircraft.

Sadly, we have to trust the guys who worked out these numbers that they will work for the conditions of the day!

[Gibon2]

Thanks for the replies. Chippybus, yes the trend arrow is the sort of thing I had in mind - but from your description (and others in the EK thread) it seems that the arrow is just longer or shorter: it doesn't indicate the precise acceleration figure, and more importantly doesn't indicate if this figure is less than it should be. So you are still making a rule-of-thumb assessment - which strikes me as odd when the computer has all the data necessary to make an exact, objective assessment at every instant of the takeoff roll.

Number34:
Looking outside seems to have worked for the best part of a 106 years

Well, yes - except for all the times it hasn't. Looking outside and "feel" certainly have their place, but my question is why they aren't combined with objective real-time data that is already collected and available - but for some reason not presented to the crew. It's not as if any additional hardware would be required.

sharpclassic:

Surely this is what our performance figures are for?


If I've understood correctly, the performance figures determine how the aircraft should perform - what I'm curious about is why there is apparently so little real-time monitoring of how the aircraft actually is performing in comparison with those expected figures.

On the face of it, it would seem that if the crew had an acceleration indicator that flashed red (or whatever) if acceleration was significantly below that expected for that particular takeoff, they would be able to abort very early - and therefore safely. But perhaps there are other factors that mean this would not be a good idea?

Thanks again for your input.

[Scumbag O'Riley]

One only flies little planes, real airplanes one might say as they have a wheel at the back. Getting off the ground is just an issue with these aircraft as big ones and there is bugger all instrumentation in them like you get on these computerised marvels of technology.

So when when departing from runways at high altitudes/temperatures I keep an eye on the indicated airspeed while on the runway. If I haven't reached 75% of 'takeoff speed' (get that from a book) by half way along the runway (get the length from a book) I stop. Very simple and it works. No idea whether that applies in airline operations but no reason to believe it wouldn't, it's basic newton law of motion. I suspect this has all been thought of by airlines' flight departments and they don't do anything silly in their published operations. Jetplanes are also somewhat better designed and overspecified than real planes and have a lot of redundancy built in.

[Rainboe]

Take-off performance charts are very sophisticated taking into account all factors affecting take-off performance- aircraft weight, aircraft flap configuration, surface wind,air pressure, temperature, runway contamination, runway length and slope, local terrain, noise considerations. You can calculate minimal power to get airborne (for prolonging engine life) leading up to a 25% reduction of power. So, set the power, take-off at the right place, and it all falls into place! The calculations are good.

So these fancy 'acceleration detectors' are sounding perfect. They are useless. For a start, in varying wind on the take-off run, the airspeed leaps ahead or falls. If you are using an inertial system that only measures groundspeed, then again, it is affected by wind effects. A very large headwind component will lead to an early and very short take-off run with low acceleration.

I would suggest a crew that messes up a take-off calculation will be so hopeless that using an automatic acceleration 'meter' will be beyond their capability! The accepted wisdom these days is that rejects at high speed are more dangerous than getting airborne! The system works fine. In fact, BA uses communication with the main computer to calculate the take-off data for the pilot, so human error is removed from the system.

[Curious Pax]

I can see the argument as far as normal ops are concerned, however what if 'normal ops' are compromised by a technical failure - ie an engine is sluggish, or a tyre problem causes much slower acceleration than expected? Presumably there is middle ground between normal running and total engine/tyre/whatever failure?

[Rainboe]

Power checks are carried out on take-off. Binding wheels are not really a problem. How big is the problem? An airline as big as BA has no 'incidents' that I can think of in its history over the last 40 years, despite hundreds of aeroplanes flying an incredible number of take-offs a year. The answer lies in training- a fool crew can easily make mistakes. Well trained and experienced crews have no need of another system to make more complications and to go wrong. It's just another thing to monitor during take-off when there's already quite enough.

[Gibon2]

Rainboe, your forthright contribution is appreciated as always, but I fear you may be labouring rather beside the point. Yes, the takeoff performance charts are very sophisticated and take into account all sorts of factors - but the charts only tell you what should happen, not what actually is happening as you hurtle (or dawdle) down that runway.

The problem occurs when the aircraft does not perform according to the charts. This may be because the wrong weight was entered, or because there is something physically wrong, e.g. tyres, brakes, other systems (as Curious Pax pointed out).

To avoid the high-speed reject or the EK-style lift-off-in-the-weeds, it would presumably be good to know about this underperformance as early as possible. Scumbag gives us his rule-of-thumb technique:


If I haven't reached 75% of 'takeoff speed' (get that from a book) by half way along the runway (get the length from a book) I stop.


That's fine - but the computer can make this kind of analysis much faster and more accurately than you can. And it can do it from the first few seconds of the takeoff roll - no need to wait until you're halfway down the runway.

And we are not talking about an "acceleration detector". No gadgetry is required. The computer already has all the data: it has all the sophisticated performance chart information, it knows the actual speed at each instant (you read your V speeds off the display, right Rainboe?), it knows to the millisecond how long you've been rolling. I'm not sure about the interface to navigation data, but I guess it also knows pretty much exactly how many feet of tarmac you've rolled over so far. So it can tell you, in a flash of simple arithmetic, such things as: your instantaneous acceleration, your average acceleration, the time remaining (at current performance) until you reach V1 and Vr, the distance remaining until you reach V1 and Vr, etc. Furthermore, in the same flash it can compare these actual figures to the expected figures in the performance chart, and alert you if they differ by more than a certain tolerance.

The computer can do all this: it's only a matter of programming it to manipulate and display data it already has. So the question remains, why doesn't it? Cost? Rainboe Syndrome? Something else?

[Contacttower]

Gibon2 it can be done and judging from a post from the original thread they exist on some aircraft;

I used to fly Falcon 50's and 900's. We had an accelerometer on the PFD that read out in real time, as well as a time to 90 knots that we could figure out. Upon brake release you would look for a certain value. If you did not have at least that number something was wrong.(Not enough thrust, brake dragging, heavier than calculated, not enough tire pressure, etc).

We could literally abort takeoff at 5 knots based on the G number displayed. Or we could wait until we hit 90 knots. If it was at or less than computed time, go. If more, abort. Shouldn't be to hard to calculate and display on these airplanes as well.

It's quite often the case that business jets have a few extra bells and whistles that airliners don't, Learjets for example are one of the few aircraft to angle of attack meters.

[Rainboe]

Gibon, it sounds so good and sensible....except in real life it doesn't work so easily. Wind varies- I've seen speed falling for a short time on the take off roll as variable wind changes. sometimes it pauses for a few seconds, then increases rapidly. How many false warnings will result in unnecessary rejects? For all their bells and whistles, very few bizjets have even autothrottles. I was nearly the recipient of a wild Learjet in the ribs thanks to an out of control Learjet with no autothrottle control, so no, I am not impressed with the bizjet world! Any binding brake should be detected on taxi. You just turn onto the runway and set takeoff power and release the brakes. It's an amazing formula that's worked since aviation started- don't complicate it.

[Contacttower]

I'd have thought the key was in always having a point at which the aircraft should be at a certain speed. Some runways have maker boards in thousands of feet and most instrument runways have distance stripes at each end so it should be quite easy to check. I thought 0-80kts is normally timed anyway?

[BelArgUSA]

No acceleration table (that I know of) was ever used with airlines.
The only guys I know who did were KC-135 (B-707 USAF tankers).
They were converting fuel to noise and black smoke with J-57 engines.
Gross weight was 298,000 or 302,000 lbs at takeoff.
xxx
The civilian equivalent of these beasts were 707-120 with "wet" JT3C-6...
And these guys max gross T/O weight was... a mere 258,000 lbs.
xxx
I heard the only other guys playing these acceleration games are B-52 drivers.
With 8 engines, who cares if you blow one during takeoff...?
Runway length "remaining" markers... meant something for these guys.
xxx
I often did rotate in the "red lights", but I trusted my V1/VR speeds.
And still alive in my rocking chair, and reading Ppruner's "how can it be...!"
xxx
:8
Happy contrails

[galaxy flyer]

BelArgUSA and others

The USAF, in just about all operational planes use some acceleration check speeds, typically time vs. achieved speed. In the C-5, for standing take-offs, we had a time to make a specified speed, usually 100 or 120 knots. The tolerance on the low side was only 3 knots, but I never had a plane not make it. One memorable take-off during Desert Storm, I had one that on brake release didn't feel right and as the take-off progressed I knew it would be close, turned out to be 2 knots under check speed of 120 knots at about 40 seconds, as i remember.

GF

[Rainboe]

I'd have thought the key was in always having a point at which the aircraft should be at a certain speed.
Doesn't work! It all depends on the wind. In a 45 kt headwind, you will be airborne very quickly. In a 10 kt tailwind, you would go as far as the performance would allow before hitting that point. So which applies on the day? I don't want to be looking for marker boards or things painted on the runway, I want to be monitoring instruments, and I barely have time to glance outside.

The old way works: set (correct) power, release brakes and away you go. There is so much performance margin built in to the take-off calculation that it can't go wrong. However, you cannot protect from idiots. If they can mess up the above, they most certainly can just as easily mess up assessing speed at the right place. In addition, different types require totally different performance measuring points- a heavy 747 take-off doesn't compare to a light 737 or Q400 take-off, so what are we then into? Look for a purple/green/yellow board as appropriate? How do you account for aircraft weight?

Solutions for aviation problems from non-pilots are always insufficiently thought through.

[forget]

Solutions for aviation problems from non-pilots are always insufficiently thought through.

That leaves an awful lot of non flying engineers kicking their heels for the last 100 years --- or so. :bored:

[ProM]

I was think that forget.

In truth, most traditionally available solutions for "is the aircraft accelerating quickly enough to take-off" rely either on approxoimations or increased crew workload (at a time when they are too busy to have a relaxing cup of tea already)

One reason for this is the absence of digital information in the cockpit of runway remaining (except by making the crew look out of thw window).

The introduction of some of the next-gen ATC technologies could provide a means to fill that gap in data. However to absolutely and completely provide a total solution it would require either additional kit and workload at the airport end, or additional kit on airport and aircraft. So either way it would be expensive. So I don't see it happening for a while (although the introduction of next-gen at short & difficult runways could be combined with tower involvement to provide an additional warning for the most difficult cases)

[G SXTY]

a heavy 747 take-off doesn't compare to a light 737 or Q400 take-off

Rainboe is absolutely right. I like the logic of a device that tells me the acceleration is normal, but I just can't see it working in practice. There are just so many variables - particularly wind - that it would be fooled far more often than it got it right. As for something else to monitor during the take off roll - here's the problem. Our SOPs on the Q400 are for PF to advance the power levers slightly and call 'set power'. PNF advances them (slowly) into the detents, calls 'power set', 80kts both, V1 etc. At light weights it all happens so quickly that by the time power is set, you're practically at 80kts, and V1 comes up within a few seconds. There's barely time to monitor the engine instruments, never mind anything else. It goes like a bat out of hell, even with reduced take off power.

Compare the number of incidents caused by abnormally slow acceleration with the number of aircraft that successfully get airborne each day, and you have to say that the system works very well as it is.

[Kelly Hopper]

But the perf figures are already for a net aeroplane. That is to say the worst aircraft in the fleet on the worst day. They are factured down so that in reality you should always get better performance than the fugures suggest.
Sometimes I question them but the law makers say they are correct.

[Curious Pax]

However to absolutely and completely provide a total solution it would require either additional kit and workload at the airport end, or additional kit on airport and aircraft. So either way it would be expensive.

Not necessarily. Get a teenager just out of school, give him one of those signs that the ball-spotters at golf tournaments have to flag where a ball has landed, and stick him by the side of the runway. ATC then just have to give him a shout as the next departure lines up, so he can leg it to the correct point. All the crew then have to do is to make sure they have passed the spotty kid by 100 knots, and bingo!!

Might need some spotty youths in reserve for low viz procedures, as they would have to stand on the runway in that situation, and might get clipped by a passing wingtip.

[wclark1238]

I remember a very, very long takeoff roll when onboard a very well laden KC-135A. We were deploying to Darwen in Australia from Andersen AFB in Guam and had all kinds of maintenance/support personnel and equipment on board. We also will have been pretty well fueled as we were ferrying the B-52Gs to Darwen and all 3 birds were 'fed' en route.

The initial acceleration seemed (from my view through a side window in the back) slightly sluggish but this thing just kept on chugging along whilst it seemed reluctant to take flight. The wind on this day was from the opposite direction to normal so we were taking off in a (roughly) southerly direction rather than the more usual northerly. I'm watching landmarks passing by on the port side - oh look there's the phase hangar, there's the terminal building, there's the transient aircraft ramp, THERE'S THE OLD ALERT RAMP and we're still not up :eek: Soon after passing that and thinking that we'd be rolling onto the perimeter road she gently started to climb. I'm not sure of the length of the runways at Andersen - pretty damn long as it is an emergency landing strip for the shuttle - but on this day they were only just long enough for us or so it seemed from my vantage point.

Thank God for the re-engine programme and the emergence of the 'R' model. :D