Perf Group A history

[kenparry]

I have been trying to establish, out of idle curiosity, the history of the development of aircraft Performance Groups, particularly Perf “A”, with its rigorous treatment of engine-out operation. Web searches have produced little info.

I have seen a claim that Perf A originated with FAR25, which was first published in 1965, and that the first type certificated to FAR25 was the B737. Is that correct

This raises a few queries. What standard applied before that? To what standard were 1950s propliners, the Comet, B707, and DC-8 certified in respect of engine-out operation? I know that the FAA-approved B707 was rejected by the UK ARB, which required significant changes to the fin (vertical stabilizer if you’re from the US) and the rudder boost to improve engine-out go-around controllability.

Clearly, at that stage, there were major differences between national authorities. I suspect that alignment improved later – but the UK CAA still required changes to the Mmo of the B767 in the mid-80s.

When Perf A appeared, was it applied retrospectively to existing types? I recall that, maybe about late 60s, there were reductions imposed in the UK on the MGW of the DC-3, which were said at the time to make the type commercially unviable. Though, of course, it carried on flying for many years. I assume the change was to improve engine-out performance.

Further queries on performance criteria for some RAF types. I recall conversations with Shackleton pilots back in the 60s over a beer or three; they lived at times with a 20 knot gap on take-off between the stop speed and the go speed, thus giving in that speed band a horrid choice between a slow crash and a fast one. This was in the context of Gibraltar, with its short 6000ft runway, and Sharjah, rather longer but with temperatures frequently up to 40 deg C or so. Did that always apply? Or, at UK bases such as St Mawgan and Kinloss with longer runways, was there a higher level of safety?

Also, for the A-W Argosy, similar conversations in which I was told that at times they operated to “military performance standards”, whatever they were. The RAF Argosy first flew in 1961, so pre-dates (I think) Perf A. Did they operate to Perf A in normal times? What was the permitted degradation to “mil perf standards”, and in what circumstances could that be used?

I hope there may be some performance engineers out there with relevant knowledge.

[zzuf]

Ahh you need to correspond with John Tullamarine, I gave him my copy of the "Final Report of the ICAO Standing Committee on Performance" circa 1955. This will give a good background to the development of modern (civil) performance standards. Probably a good starting point.
D. P. Davies "Handling the Big Jets" has some detail on the development of jet transport performance standards, from a certification TP's view.
Old British military, probably AVP970, but don't really recall.
Cheers.

[kenparry]

A now deleted post included the statement "I've never heard of Perf A".

We have a terminology problem. UK national regs, specifically BCARs (British Civil Airworthiness Requirements) were the regs under which I first encountered public transport performance requirements, in the late 1970s. These were replaced by JAR 25 in the 1990s.

BCARs had several performance Groups, the highest of which, Group A, applied to aircraft with more than 19(?) seats or max gross weight of 5700kg. It required the aircraft to be able to tolerate the loss of the power of one engine at any time from brake release to the end of the landing roll. Other Groups applied to smaller aircraft and were less demanding.

Thus Group A was addressing broadly the same types as FAR 25, with broadly the same performance requirements.

The now deleted post included a link to a Flight Safety Foundation document that addressed some of the background I was looking for, such as the US codes under CAR4b that covered propliners and predated JAR25. No sign yet of the early US or UK jet certification standards.

I'm aware that AvP970 was the UK military design standard source. Its successor, DefStan970, is available online, but 1950ish AvP970 so far eludes me.

So, some progress; thank you.

[john_tullamarine]

Some thoughts .. noting that I don't claim to have much in the way of specific knowledge of what went on in the real olden days ..

I have been trying to establish, out of idle curiosity, the history of the development of aircraft Performance Groups, particularly Perf “A”

A specifically British thing.

If you run some searches you will find the odd document of interest to this question .. eg Chaplin's paper (https://www.aerosociety.com/media/4858/safety-regulation-the-first-100-years.pdf). I suspect that references to "Performance A" in the pilot fraternity refer more to the pilot licensing theory examinations rather than the certification airworthiness regulations.

I have seen a claim that Perf A originated with FAR25, which was first published in 1965

Can't speak authoritatively as to the commencing date, but I suspect somewhat before the mid-60s. I did my theory subjects around that time and, as I recall, references to Perf A in the British system were not new. As to the first Part 25 certification, that would take some research which I will leave to you .. you might start with the TCDS for a range of aircraft; the sheets will give you the certification basis for the Types involved.

What standard applied before that?

There was a range of standards developed prior to the jet age. A reference cited later in this post will give you a bit of a rundown on the specifics.

Clearly, at that stage, there were major differences between national authorities.

Indeed. In recent years the harmonisation of, especially, European and US standards has seen a more sensible life for the OEMs

When Perf A appeared, was it applied retrospectively to existing types?

Here you are thinking of BCAR Group A (if my memory hasn't atrophied too much), which was the heavy airworthiness regulations rather than Perf A, which related more to pilot theory exams. Perf A caused much head scratching and pencil and paper work in the graphs to get folks to the standard where they could pass the relevant exams.

As a general observation, rule changes are not retrospective as this, often, would kill the old Types which hadn't been designed with knowledge of the newer rule. So, for example, the 2 second ASD fudge factor in FAR 25 A/L 42 was not retrospective, so far as I am aware.

reductions imposed in the UK on the MGW of the DC-3

I have no knowledge of that exercise

Further queries on performance criteria for some RAF types.

I can't speak to the RAF but, based on comments from RAAF flyers of that era, I have the impression that the system was more on the basis of TLAR .. ie we got over the fence last week so it should still be OK this week.

for the A-W Argosy, similar conversations in which I was told that at times they operated to “military performance standards”

The RAF may well have operated to weights in excess of the civil requirements.

For this Type I had plenty of performance exposure, having looked after IPEC's performance for quite a few years. The Australian certification essentially rubber stamped the UK requirements. I would have to dig out some AFMs from the filing cabinets to check just what the UK basis was but the AFMs were bog standard Perf A data .. think FAR 25 in general. In fact, typical of many UK manuals, the performance charts were great to work with as they gave all the various limitations as separate information.

I gave him my copy of the "Final Report of the ICAO Standing Committee on Performance"

Indeed .. and, again, many thanks to my colleague zzuf for the donation. I had been hunting for a copy on and off for years. zzuf's copy comes from a line of well-regarded Australian airworthiness regulatory stalwarts, including his good self. Reading the names on the cover, I remain somewhat humbled in the presence of such august company.

Prior to the Standing Committee's report, things were a bit TLAR and simplistic in the performance arena.

The rapid developments post WW2 led ICAO to a desire to get ahead of the action and develop more appropriate Standards. The upshot was the development of the PAMC on Performance which applied a much more rigorous and statistically based rationale to aircraft performance.

If I recall correctly, the then-recently certificated F27 was the first Type to be (re-)certificated to the PAMC. Again some further information in later references.

A now deleted post included the statement "I've never heard of Perf A".

The post referred to was by IGh who chose to delete it for his own reasons. A pity, as the post was quite useful and, while I won't reinstate it out of deference to the poster, the main details were

(a) a link to a paper by Wagenmakers (https://flightsafety.org/fsd/fsd_feb00.pdf), a quite experienced and influential airline performance guru back in those days.

(b) reference to Wagenmaker's reference text (https://trove.nla.gov.au/work/18031507?selectedversion=NBD7102598)

(c) some references to the Boeing performance engineering manual Jet Transport Performance Methods. This document, which has been revised over the years, is a basic engineering reference text for the subject and is used by Boeing for training courses for airline ops engineering folks. For those who wish to read a copy, you shouldn't have too much difficulty finding a download on the net.

He made a reference to one edition (which I don't have) with credits to Walt ? Walt presumably is Walter Blake, a fine chap (now well-retired) but still active in the Industry. Walt, as I recall, was the Boeing performance boss for some years and is a very knowledgeable performance engineer quite apart from being a very nice bloke.

To give you some idea of how the very early regs were made, I recall some commentary by a chap at a training course (also attended by zzuf as a CASA - or whatever name it was that week - observer). This chap, whose name I can't recall, had been a junior engineer in the US regulator at the time. He gave two examples

(a) light aircraft maximum stall speed. Finger in the wind figure of 70 mph picked as being a reasonable balance for OEM design and crashworthiness survival.

(b) 50 ft screen height for light aircraft. This was based on a Curtiss demo at a military parade ground. The parade ground was surrounded by trees of about 50 ft height .. seemed as good an idea as any for starters.

While things had to start somewhere, one sees the eventual need for developing rigorous analytical approaches cf the PAMC on Performance.

[kenparry]

john t:

Many thanks, I had hoped you would be able to make a contribution. You may well be correct about Perf "A" being related to pilot licensing - that was the context in which I met it.

Yes, the post by IGh was helpful, and thanks to you for repeating the links that he had posted.

[Bergerie1]

John, I am requently amazed by your breadth of knowledge. Thank you. By the way I did my Performance A exam in 1962.

[Centaurus]

I recall from the mid 1950's at RAAF Base Townsville, a USAF RB66 "Destroyer" taking off on Runway 01 which was then about 8000 ft in length and at sea level. . In those days several RAAF airfields had Distance to Go marker boards on the flight strips. I believe that was because some military aircraft used Refusal Speeds as a rough check of take off performance. It not at a certain speed by a certain distance down the runway the pilot rejected the take off.

On this occasion the B66 was a long way down the runway in its take off roll when the pilot aborted using the braking parachute as well as brakes.
It taxied back and 20 minutes later tried another take off and this time it kept going.

I suspected the pilot wanted to get home to Guam for a date:E

Since then, I often wondered how many military aircraft of that era aborted their take off run unnecessarily; especially as marker boards v airspeed as means of judging acceleration, was only an approximation. Some years later the RAAF dumped the idea and marker boards disappeared from military airfields.

Again, in those days, there was no such thing as take off performance charts included in RAF/RAAF Pilot's notes for type. You simply opened the throttles and went. If the runway was short you pulled back the stick when there was no runway left and hoped for the best. Believe me, I am serious! The Lincoln Pilots Notes published a recommended lift off speed and a Take-Off Safety Speed (asymmetric).
For example the Avro Lincoln PN stated in part:
At 65,000 lbs ease (rotation hadn't been invented then) the aircraft off the ground at between 100-105 knots.
75,000 lbs ditto 105-110 knots
82,000 lbs ditto 110-115 knots.
There was nothing about how much runway you needed for take off and landing at various density altitudes.

For example. The length of Momote airstrip on Manus Island,north of New Guinea, was 5200 ft. There was no over-run except to ditch into the sea. It became a case of sticking up one's wetted finger to judge the w/v and go with full power +18 lbs/sq inch boost (manifold pressure) through the throttle gate. The plan was to lift off the ground before running into the shark infested water at the end of the runway. It was safer to keep on going with an engine failure or burst tyre because an abort would be fatal.

I well recall on one take off at Momote, nil wind, the rear gunner who (naturally) faced the rear seated between his two 0.5 calibre machine guns, could see the runway going away from him, remarking upon the four trails of propeller slipstream ruffling the surface of the sea as we built up airspeed at 100 feet towards asymmetric safety speed of 135 knots. Ops normal for that era, with take off performance considerations non-existent since there were no published charts. Things were so simple, then.

[kenparry]

Bergerie1:

By the way I did my Performance A exam in 1962.

Thanks for that; it takes Perf "A" back to before FAR25. Can you remember the sample aircraft type on which the paper was then based? Mine was somewhat later, 1978, and used the L1011.

Centaurus:

The UK's V-force used similar marker boards and go/stop speeds. I'm not aware of how the speed was calculated or how many variables were considered.

My military flying was all single-engine, so my initiation to the complications of multi-engine performance came only when I moved into the airline world. For the Hunter, in the late 60s, we had limited runway performance data, but it did include the ability to work out a ground roll distance. The engine failure case was simple, of course, in principle: stop or Martin Baker exit. However, there was no way of deriving a stop speed. The only info was a quoted "emergency braking speed", which IIRC was 100kt. Some problems with that: our only speed indication was IAS, and the single quoted speed took no account of what we then called take-off weight. I suppose the Perf A nearest equivalent is Brake Energy Limit, but that does account for at least some of the related variables.

[anchorhold]

On the subject of Performance A, do pilots qualifying through the MPA rating study Performance A or fo they just rely on the automation.

[PEI_3721]

Info U.K. CAA 2006 (JAR)
https://publicapps.caa.co.uk/docs/33/CAP698.pdf
CAA JAR-FCL Examinations Aeroplane Performance Manual

Engine out perf may have orrigionated from ICAO obstacle clearance studies?

P.S. The paper ‘Analysis if takeoff performance’ 1967 refers to BCAR. There may be other refs of interest.
http://naca.central.cranfield.ac.uk/reports/arc/cp/1034.pdf
References (#2) British Civil Aircraft Requirements 1966 section D

[Basil]

john_tullamarine, Re RAF Argosy Military Operating Standards, personally, was never required to use them and always operated to Perf A - Phew!
I can't recollect the approximate difference in what we could lift or TOD/TOR required but I know a man who probably can. Surprised he hasn't turned up to comment.

[john_tullamarine]

Biggest problem with the AW Queen of the Skies was its invariable, miserable, abysmal first segment performance. Far more than any other aircraft I have dealt with, first segment gradient was a regular constraint on TOW.

Then again, those main legs went up a long way to the cowls.

[tubby linton]

There is a book and a film by David Beatty called Cone of Silence about performance problems with a fictional airliner. The film uses an Avro Ashton.
Flight International reviewed the film in 1960 and made a reference to jet runway lengths ,ground stall effects and manual unstick speeds being again under close review.
https://www.flightglobal.com/pdfarchive/view/1960/1960%20-%200610.html

[eckhard]

Can you remember the sample aircraft type on which the paper was then based? Mine was somewhat later, 1978, and used the L1011.


I too took Perf A in 1978; the charts based on the L-1011.

I believe that the type used prior to that was the Britannia.

[JW411]

Basil:

I have indeed operated the Argosy to MOS including in the Middle East. Basically speaking, the MTOW went up from 97,000 lbs to 105,000 lbs. I have come out of Bait al Falaj at the latter weight so TORA wasn't a problem but an engine failure could not be contemplated.

[Basil]

Thank you,JW; now I remember those max numbers. I wonder what the approx difference would be on a limiting strip.

[aterpster]

Here's what I found on Wikipedia about the history of Part 25:

Part 25

This part contains airworthiness standards for airplanes in the transport category.

Transport category airplanes are either:
Jets with 10 or more seats or a maximum takeoff weight (MTOW) greater than 12,500 pounds (5,670 kg); or
Propeller-driven airplanes with greater than 19 seats or a MTOW greater than 19,000 pounds (8,618 kg).

A rather important section of this part, is the 121 - climbing guaranteed with one engine out for multi-engine aircraft.

The Boeing 737 and later types, and Airbus A300 series, are well-known airplane types that were certified according to standards set out in FAR Part 25.

Most of the Federal Aviation Regulations, including Part 25, commenced on February 1, 1965. Prior to that date, airworthiness standards for airplanes in the transport category were promulgated in Part 4b of the US Civil Air Regulations which was in effect by November 1945. Effective August 27, 1957, Special Civil Air Regulation (SR) 422 was the basis for certification of the first turbine-powered transport airplanes, such as the Boeing 707, the Lockheed Electra, and the Fairchild 27. SR 422A became effective July 2, 1958, and was superseded by SR 422B, effective August 29, 1959. Only a few airplanes were certified under SR 422A, such as the Gulfstream I and the CL-44. First generation turbine-powered transport category airplanes such as the DC-8, DC-9, and B-727, were originally certified under SR 422B. SR 422B was recodified with minor changes to 14 CFR part 25, which became effective February 1965.

[finncapt]

I did perf A in 1970 and can confirm it was based on the Brittania.

[kenparry]

aterpster: thanks, that pins down the code used for the early jets, and confirms the FAR25 date.

JW411 and Basil: Wow, I would have thought Bait to be exciting enough at Perf A without going heavier. I never went there, but heard you guys talking about it in the Muharraq bar. Who could authorise use of MOS? Was it Capt's discretion, OC Ardet, or higher up the food chain? And what was the level of need? Just a heavy load day, or something more urgent?

[Bergerie1]

tubby L

'Cone of Silence' was based on Harry Foote's accident at Ciampino:-
https://aviation-safety.net/database/record.php?id=19521026-0

I flew with him several years later on a Britannia.

Read my post about D.P. Davies and his comments in the second audio.

Nothing like blaming the pilot when a problem is not fully understood!

https://www.pprune.org/tech-log/602953-d-p-davies-interviews-certificating-aircraft.html

[Centaurus]

The early raising of the nose during the takeoff roll seemed to be a common technique in those days. At night, especially with the short nose on the Comet and also the DH Vampire series and the Canberra, meant it was a guess by the pilot how much nose up showed on the artificial horizon compared to the view outside seen from the cockpit when the nose-wheel was lifted off the runway at speeds well below the recommended main lift off speed.

For example, RAF Pilot's notes for the Vampire (another de Havilland type like the Comet) required: "As soon as the aircraft reaches a speed of 60-70 knots IAS, lift the nose-wheel just clear of the ground, then at 82-87 knots IAS ease the aircraft off the ground".

So a fair amount of back pressure would be needed to lift the nose-wheel at 60-70 knots and if that same back pressure was maintained during the remainder of the takeoff roll, it was inevitable with increasing airspeed the elevator was steadily becoming more effective. By the time the aircraft reached lift-off speed of 82-87 knots the nose would be considerably higher than "lift the nose-wheel just clear of the ground"

Is it any wonder the induced drag could have increased significantly while all this was going on, resulting in a longer take off run and a sense the aircraft was not accelerating fast enough.

There have been reports of this happening to civilian owned Vampires where the pilot held the nose-wheel too far off the ground instead of just clear while accelerating towards published lift off speed. During this writer's time on Vampires (both single and two seat versions) it was common pilot technique to "feel" the aircraft off the ground at speeds well in excess of the Pilots Notes recommended lift off speed because it was an uncomfortable squashing feeling to lift off at the PN speeds.

I have often wondered since those days, why it was seen necessary to lift the nose wheel relatively early in the take off run in the early jets, rather than at the rotation speed used in todays jet transports. Was it because a precaution against slush build up on the nose-wheel in winter weather in those climes?

[Vessbot]

I was wondering this myself 30 seconds ago while reading your post describing the history. The possibilities I came up with were lower speed limit on the nosewheel, or structural weakness wrt. bouncing, etc. And soft fields.

All those factors aside, my ideal conception of a tricycle takeoff is to rotate so as to reach the liftoff attitude at the same time as you reach the liftoff speed. Any sooner, and it's unnecessary induced drag. Any later, and you're using more runway than you need to.

[JW411]

kenparry:

Certainly not captain's discretion. I think we are talking AOC here. Load was usually local troops and equipment to Salalah. IFR safety height was impossible to achieve nor could we turn right and go down the coast via Masirah. Not enough fuel for that. So the answer was to turn left and follow the Bid Bid to Izki valley VFR out on to the plain. It would not have been good to do this every day of the week.

[kenparry]

I have often wondered since those days, why it was seen necessary to lift the nose wheel relatively early in the take off run in the early jets, rather than at the rotation speed used in todays jet transports. Was it because a precaution against slush build up on the nose-wheel in winter weather in those climes?

Me too. The JP, Vampire (T11 only for me) and Hunter all had that technique written in. Don't think slush came into it, as we did not fly from contaminated runways. Your Vampire speeds are lower than I recall; I do remember 100kt as the threshold speed for landing, and there's a vague memory of about 110kt for lift-off. This was the T11 with no external stores, and internal fuel of about 3300 lbs. (Was it? Long ago now. Might have been closer to the 2500lb quoted in the link below)

JW411: thank you for that. From what you say I conclude you were Aden based. My Gulf time was later, at Muharraq, where we had a standing detachment of Argosies from Benson.

[Goldenrivett]

why it was seen necessary to lift the nose wheel relatively early in the take off run in the early jets
Was it due to "nose wheel shimmy"? (like many supermarket trolley wheels)

The nose wheel does not seem to steerable from this description.
de Havilland Vampires - Aircraft Data (http://www.rafjever.org/vampires.htm)
"The undercarriage was hydraulically operated with a castoring nose-wheel. Brakes were the typical DH pneumatic style, fed by a pre-charged accumulator and using a single control column actuator and rudder application for asymmetric braking."

[kenparry]

Was it due to "nose wheel shimmy"?

Possibly. The types I listed in the previous post all had castoring nosewheels. Shimmy could be a problem on the JP at times. The JP (or Vampire - or both - it's long ago and memory is fading) had nosewheel tyres with a tread on each shoulder and a gap between on the wheel's central plane, described as "anti-shimmy" tyres.

[Sidestick_n_Rudder]

I remember David Beatty’s „The Naked Pilot” refering to the FCO Comet accident. He indeed quoted nosewheel shimmy as the reason for this technique.

[JW411]

kenparry:

I was on 105 in Aden 1966/67 and then did 5 x 3.5 month Ardet detachments in Muharraq finishing in October 1971.

Moving on; I have a friend who flew the Comet 4B with BEA. He talks about a technique called "skimming" the nosewheel at 80 knots on take-off. In other words, raising the nosewheel just off the runway. The object of the exercise was apparently to prevent water/snow/ice etc from going down the intakes.

[Centaurus]

The nose wheel off the ground technique is still taught by numerous Australian flying schools in several types of light aircraft including the Cessna 150/172 and their ilk. This is on runways. Reasons often include "take the weight off the nose-wheel because the nose wheel is the weakest part of the aircraft." Other reasons offered was minimising the chances of nose-wheel shimmy. Crazy ideas yet still taught by new instructors for some reason..

Several years ago I wrote to the Cessna company about this strange technique and they replied they had never heard of it before. Different if a soft field take off but unnecessary on hard surface. I once saw a Cherokee Six with a full load of passengers depart from a long runway at Essendon, Melbourne. As it taxied I noticed it was very tail down in attitude. The nose wheel was raised off the runway early in the take off run and the aircraft lifted off in a most frightening nose up attitude squashing while trying to gain speed. I was later told it was still nose up when it pranged into gently rising ground a mile from the end of the runway. No one hurt though.

[megan]

The procedure in the Comet airliner was to lift the nose wheel off at a speed of 80 knots, and hold it off while accelerating. V1 and V2 were used in the planning eg at the maximum permissible a.u.w. of 105,000 lb the critical speed at London Airport on the E.W. runway is 95 kt I.A.S. and the unstick speed 100 kt I.A.S.

The practise of lifting the nose wheel early lead to the first two crashes of Comets, both write offs. Inability to accurately assess attitude when holding the nose wheel off lead to the possibility of the aircraft either becoming airborne in a stalled state, or not being able to accelerate to sufficient speed within the confines of the runway, all due to excessive induced drag. I maybe wrong, but I seem to recall it was these two accidents that lead to the current procedures with introduction of the Vr call.

[Shrikered]

I have been trying to establish, out of idle curiosity, the history of the development of aircraft Performance Groups, particularly Perf “A”, with its rigorous treatment of engine-out operation. Web searches have produced little info.

I have seen a claim that Perf A originated with FAR25, which was first published in 1965, and that the first type certificated to FAR25 was the B737. Is that correct

I have read this thread with great interest and am chasing the references. I am trying to figure out how the V1, V2, MTOW and related calculations changed over time, and I'm mainly studying US practices including FAR 25. For example:

When was V1 first conceived? VR, V2?
When and where were they first made official?
What was done before then, for large prop aircraft?
Multiple accidents have shown that runway friction was not being measured consistently. According to AeroSafetyWorld, Nov. 2012, the FAA is investigating better ways to evaluate it. What other changes have been made over theyears in the details of calculations used for takeoffs and landings?
Do all airlines and pilots make the calculations consistently (presumably because of legal requirements)? Or, do some have what they consider improved methods?

Thanks to everyone who has contributed, and please PM or respond here if you might have additional information.

[aterpster]

I can only comment about V1 and jet transports. I did not fly piston transport airplanes. I went with TWA in January, 1964. V1 was already in use on the jets. Some years later the industry decided to reduce V1 down to Vmcg with long runways and light weights, on the premise the accident rate of high speed aborts had proven it was safer to continue the takeoff with an engine failure than attempt a high-speed abort.

[kenparry]

Do all airlines and pilots make the calculations consistently (presumably because of legal requirements)? Or, do some have what they consider improved methods?

I can only answer from my experience of one UK airline in the 80s/90s. We had a Performance Section within the Nav Dept, and they produced analyses for every runway on our route network to give V1/VR/V2 and RTOW for a range of temperatures (typically 0 to about +40 deg C) and wind components from 10kt tail to 20kt headwind. There were corrections available for pressure variations. Some runways had several pages for different flap settings. The resulting manual was about 3 ins thick on US letter-sized pages. It satisfied the regulators and gave the operating crews the data we needed.

Was it consistent across airlines? I don't know - we (UK) all worked to the same regs, but I suspect some Perf Sections had better analyses than others. These days, the same airline has a different system with all pilots having the manuals on iPads; tap in the numbers for your take-off, and the system produces V speeds, flap setting, trim setting etc in the blink of an eye.

After a couple of years on the line I had thought that I had a good grip of Perf "A", but at each Winter refresher course our man from the Perf Section would open my eyes to something new.

Regarding V speeds, look at CAR4b: here http://www.theiplgroup.com/CAR-PART4b.pdf

4b.114 lists takeoff speeds as V speeds in this 1953 document - and many of the other performance considerations of FAR25 are found elsewhere, though the requirements appear to be less demanding than those of FAR25. CAR4b is long - I have not yet read all of it.

[oxenos]

Can't give chapter and verse on the regs, but I can add a few memories

Mid sixties, the nose wheel was raised on the Mk 3 Shackleton at Rotate -10 knots. Certainly was not for nose wheel shimmy, we had a steering nosewheel. There was indeed a long gap between stop speed and rotate with no guarantees. On the Mk 2 Shackleton there was also a max brake speed. If an engine failed before stop speed but after max brake, it was close the throttles, wait for the speed to drop to max brake speed, then brake. Brake too early, and the pneumatic brake sacs would burst and leave you with no brakes at all. Happy Days.
1970, went on to the Nimrod. It was operated to Perf A, but we were told that in war we would be issued with Military Operating Standards.
The Nimrod performance was originally done entirely on D and X graphs, but later some simpler graphs were produced for our "regular" airfields.
I was on the same conversion course as one of the Flight Commanders, had no problems with perf, and so when we got back to the Squadron and he decided that there should be a performance "expert" on the squadron, guess who he chose? I went on the splendidly named "Long Performance Course" at Hullavington. Turned out I was the only person on the course who was not leaving the Air Force. All the others were, and were treating the course as a way to passing the performance exam for their licences.
The course and the exam were indeed based on the Britannia. It was necessary to extract figures from a set of graphs, and then use these figures to DRAW a graph from which you got your MTOW and V1/VR ratio.
6 years later I was the performance instructor on the Nimrod OCU, which in turn made me performance advisor to HQ Maritime. They asked what contingency plans I had for issuing the MOS to the Squadrons. None, says I, you've got them. Oh no we haven't says they. Turns out Hawker Siddeley had never been asked to produce them.

[JW411]

Centaurus:

Out of curiosity I've just got my Pilots Notes for the Vampire T11 out (I only ever flew the T11 and that was in 1961). It says:

Raise the nose wheel at 80-85 knots, maintain this attitude and fly the aircraft off at 110-115 knots at typical service load. At maximum all-up weight raise the nose wheel at 105-110 knots, the aircraft then becomes airborne at 120-125 knots.

It doesn't say why and I can't remember!

[kenparry]

JW: in that case, you are just the chap to revive my failed memory of the T11's fuel capacity - please?

[john_tullamarine]

Some thoughts ..

Some years later the industry decided to reduce V1 down to Vmcg with long runways and light weights

Fine, on first reflections. However, a poor overall risk balance due to the adverse effect on Vmcg of crosswind.

The problem can be mitigated, in the case of non-limiting runways, by increasing the V1 a bit to account for the crosswind. This, typically, varies from around 05kt/kt for twins to a bit in excess of 1kt/kt for quad jets.

By pushing up the V1, the handling should remain something similar to the nil wind (7kt for the old BCARs, as I recall) case.

they produced analyses for every runway on our route network

The bog standard approach for manual charts (whether produced as graphs - usually - or tabulations). Done by hand, a pain in the neck, so we generally crossplotted from the AFM to start with easier (read quicker) charts for analysis. When the use of electronic computers arrived, it was a tedious, but very useful, exercise to run a heap of regressions (or set up lookup tables) and then have the computer do in minutes what the human took many hours to produce.

but I suspect some Perf Sections had better analyses than others.

Presuming the basic ops engineering competence was similar, it really came down to how much money the bean counters would let be spent on the exercise. Very much a case of diminishing returns as one chased the last few kilos. If the exercise were done with a bit of care, the delta between the AFM and the RTOW chart data could be kept down to a very small figure .. but that took time and money so some skimped a bit here and there.

These days, the same airline has a different system with all pilots having the manuals on iPads

Essentially the same as earlier .. depending on source, the computer software will have been developed from the paper AFM (as we did in the old days) or (if from the OEM) will use software similar to that which would be used to produce the AFM data. Six of one, half dozen of the other.

There now arise two very significant problems -

(a) the crews tend to end up dumbed down (as in, they become adept at pressing buttons in the absence of understanding what the button might be doing) in the knowledge stakes and may have little basic knowledge of what it is the computer is doing for them. Particularly, for the older BCAR AFMs, the charts generally were a delight as each separate limitation was presented separately, making it a doddle to get a really good feel for what the aircraft could do.

(b) getting the obstacle data. The backroom folk spend (or should be) a lot of time and effort on this. Obviously, the line crew can't/don't so the question now becomes what does the tech services group provide for the exercise ?

but at each Winter refresher course our man from the Perf Section would open my eyes to something new.

A good airline that saw the value in investing a few extra dollars in crew training. We did the same with the IPEC (now a long-gone operator) boys.

The Nimrod performance was originally done entirely on D and X graphs

For those not familiar with these, the technique was a means of combining data sets rather than keeping things separate. Typically used for TOD/ASD analyses. Not as intuitive for the user's getting a feel for the aircraft but, in essence, just a different way of painting by numbers.

IGh's post is useful as it points to the good sources of superseded data on the web relating to the US rules, in particular. While it can get a bit tedious tracking things down, the results can be useful for those with the interest.

[aterpster]

Presuming the basic ops engineering competence was similar, it really came down to how much money the bean counters would let be spent on the exercise.

Have I previously told you about the fiasco at TWA that some obnoxious pilots got resolved in the 1980s?

[john_tullamarine]

.. c'mon, c'mon ... tell us all about it ....

[aterpster]

.. c'mon, c'mon ... tell us all about it ....

A bit incredible since that had a performance and engineering department that sold performance data to some other operators:

They taught: with an engine failure just above V1 you could fly the takeoff profile straight-ahead to 1,500 feet, afe, anywhere, anytime. That made no sense at Las Vegas to the west or San Francisco 19L/R. The union tech committee questioned it and were told "no problem, have a nice day."

We didn't buy that, so the head of the union went above the guy who ran the p and e department and gave a specific question: 727-231 taking off Ls Vegas Runway 25 at a temperature that just allowed MGTOW, engine failure just above V1, please provide altitude in one mile increments until reaching 1,500 afe. It took 31 miles. The airline would have crashed into the first ridge 8 miles west of the airport going straight ahead.

The result was pink takeoff performance sheets for airports where straight-ahead wouldn’t hack it. The pink performance sheet had an OEI flight track to avoid offending terrain.

This was in the early 1980s. TWA started operating jets some 20 years earlier. They were fortunate that an engine failure just above V1 didn’t bite them somewhere.

The solution was still the lousy +/- 300’ wide corridor. No one does that these days. Instead they use AC 120-91.

[kenparry]

Following the link given by IGh (post #37) to the US DOT archives, some reading during a snowy afternoon has led me to:

CAR 4 dated May 31, 1938: no mention of V speeds;

CAR 4 dated August 15, 1942: para 04.75320-T defines V1 and V2. This para did not exist in the 1938 document. The "-T" means the requirement applies to Transport Category airplanes. No mention of VR - interesting, as nosegear transports were around by then. (DC-4E 1938, C-54 Feb 1942, military though intended for airline use). I suspect that propliners did not need much of a "rotate" - just a gentle nose-up to leave the ground.

Still no mention of VR in CAR-PART4b of 1953. Perhaps VR came in with the jets?

That 1942 document also includes the first (that I have found) requirements for engine-out continued take-off from V1.

So, I think we can conclude that the beginnings of the modern performance certification date from the US DOT rules of 1942. Interesting - rather earlier than I had thought.

Also interesting is how long developments continued - e.g. aterpster's reference above to lateral clearance of obstacles. Is change still going on to these regs?

[aterpster]

Also interesting is how long developments continued - e.g. aterpster's reference above to lateral clearance of obstacles. Is change still going on to these regs?

In the U.S. the FAA encourages Part 121 operators to use AC 90-121. Last I heard all of them were.

If you are a 135 operator without a performance dept., and instead use APG, it will be a takeoff flight path predicated on AC 90-121. I suppose Jepp's data department does the same.

[Loose rivets]

I walked out of the performance exam 23 minutes after I walked in. It was somewhat to the surprise of the invigilators and the result of the most incredible bit of luck. I needed a pass in four days from scratch to get my first job and 3 quid for a fast track marking saw me on Channel Airways' DC3 training by Thursday. ARB's pre-passed on a mass of aircraft thanks to "Pop Speller's notes."

Despite seemingly having a natural ability with graphs, I had an instinctive belief that Performance A was a total nonsense. While on an extended vacation in Texas I stayed in aviation pioneer Col Carl Crane's* cabin in the Texas hills. He invited me to test fly his blind landing prototype on numerous occasions. While looking through his work I noticed a patent he held which advised pilots of the distance remaining/passed of the runway during T/O or Landing. I realised that his notion of selling receivers worldwide was unnecessary, as most aircraft had 75MHz receivers which fed audio to the pilots. I further realised that the tone could be replaced with not only the audio of distance readout, but a code which could be fed into a performance computer. I came back to the UK with an agreement that he would have the American market and I would have the rest of the world.

I sent extensive text and plans to the ?? government development board ?? (something like that.) They typically put in 50% on such projects. After some weeks I was knocked back to be told that they were already funding a project that was too similar to allow my application - which they were otherwise intrigued by. Quite a blow.

Co-location of DME's would have seen a quick end to multi lobe 75MHz transmitters. I smile now at my naïveté.

It seemed there was a BAC 1-11 charging up and down a runway somewhere using Doppler to measure the progress. It seems it worked . . . until it didn't, losing all reference to the progress. IIRC, their readout was a second needle on the ASI. Rather more basic than my making use of a multi-functional radar screen. I'd gained Marconi's interest in that at a science display in Alexandra Palace. Not many people were interested in TV screens in aircraft back then but there was a lot of redundancy in the radar unit's capabilities - not least of all a continuous wind velocity vector during approach.

The BAC 1-11 project devoured funds and was wound up.

Every so often over many years I would experience bewildering issues with performance. One vivid memory was being with my Fleet Manager, so it must have been a windless and fine day. We left Spit on what I deemed to be a marginal runway but the concrete slabs were so misaligned/heaved, that the crashing vertical input was thunderous. I pushed the power to the stops about half way along, and called V1 as we staggered over the rocks. The concept of energy-sucking vertical 'noise' being introduced was not in any graph I remember. Nothing was done. Nothing was ever said.

Forty years later I'm still reading of fence-scraping take-offs. It's not just my misguided gut feeling is it? So many V1 calculations have to be nothing short of ludicrous.


*There's a lot about Col Carl Crane on t'net. He fell out of the clouds in a biplane with a senator's son on board. He devoted much of his life after that teaching people they can't fly in cloud without instruments. He was a guest speaker at Randolph Advanced Instrument flying school in Texas where his talks left the young pilots spellbound.

[megan]

Still no mention of VR in CAR-PART4b of 1953. Perhaps VR came in with the jetsThe two Comet crashes I spoke of earlier occurred in 26 Oct 52 and 03 Mar 53, memory says these accidents brought about the introduction of the Vr concept, in keeping with your quoted 1953 date.

[JW411]

kenparry:

It says here: Fuselage tanks = 96 gals. Inner wing tanks 2x53 gals = 106 gals. Outer wing groups 2x64 gals = 128 gals. Total internal 330 gals.
Wing drop tanks 2x100 gals = 200 gals. (I never flew with underwing tanks).

[kenparry]

jw411: many thanks. At an sg of 0.8 for Avtur, that equates to 2640 lbs. So my memory had been subject to exaggeration. Not for the first time!

[Centaurus]

The course and the exam were indeed based on the Britannia. It was necessary to extract figures from a set of graphs, and then use these figures to DRAW a graph from which you got your MTOW and V1/VR ratio.

Drawing your own performance charts was part of the HS 748 type rating course at Woodford, Cheshire in 1966. The RAAF had bought two HS 748's VIP version and two crews were sent to UK for the course prior to ferrying both aircraft back to Canberra, Australia. The one month ground course was a bit "iffy" in our opinion but that is another story on its own.
The two day Performance part of the course was run by a white dust coated "gentleman" who was a bit younger than we thirty year olds. He looked down upon us Colonials with obvious disdain and we could pick that.

After going through the charts in the AFM he then led us through the construction of these charts. This was superfluous as far as we were concerned particularly as he treated us like children.
The second day was coming to an end and we were getting irritable with his attitude. He then handed out an examination paper requiring the drawing of performance charts. We all got varying answers and he got angry and sulked.
"You are not leaving this room until you get the answers right" he snarled.

"You can go and get F#*`cked" said our flight commander, a Squadron Leader who did not mince words. "Come on chaps, we are off to the pub" he said and we all trooped from the room muttering about two days of wasted time with this clown. The Chief Test pilot at Woodford, Jimmy Harrison, quietly agreed with us after he heard what had happened. We later heard via the grapevine that the white coated gentleman was pulled off lecturing Performance courses:ok:

[Centaurus]

Your Vampire speeds are lower than I recall; I do remember 100kt as the threshold speed for landing, and there's a vague memory of about 110kt for lift-off. This was the T11 with no external

The figures I quoted were from the Pilots Notes for the single seat RAF Vampire F1 with Goblin 1 or 2 engine. Although the RAAF built Mk 30 version had the RR Nene, we were issued with the original RAF PN until the RAAF publication was made available.

[kenparry]

Another snowy day, another opportunity to read into the background.

Little progress with the UK's BCARs; there seems to be nothing on line. The National Archives list thousands of documents, hardly any digitised, so the bulk only available by visiting Kew to search and read.

For the US records, though, very different. Following up the link to the DOT archives, there is a treasure trove of material, and a few hours of reading has found some gems.

SR-422A, issued July 2, 1958, and effective for Type Certificates dated after Sep 30, 1958, includes the following significant changes:
1. introduction of VR
2. introduction of all-engine TODR
3. take-off flight path extended to 1500 ft above the field.

The preamble to SR-422A includes lengthy discussion of its changes from the earlier SR-422. Found here DOT Online Database (http://dotlibrary.specialcollection.net/Document?db=DOT-CARS&query=(select+2131))

In 1959, SR-422B was introduced, effective Aug29, 1959. Strangely, this does not appear in the DOT archive. The only online record I have found is of para 4T.122, which refers only to landing distance. Was that all there was? I would have thought it to be a complete rewrite of SR-422A.

Some certification dates:
Comet, 22 Jan 1952. The Comet was not bought by any US operator, so was not certificated in the US.

B707-120, Sep 18, 1958 (USA). DC-8, Aug31, 1959 (USA)

Thus it seems that the B707-120 was certified to SR-422, the DC-8 to the later SR-422B.

[kenparry]

Centaurus: thanks for that. The F1 must have been much lighter than the T11. Incidentally, I spent my teenage years living about a mile from the end of the Woodford runway, watching Canberras, Shackletons, and Vulcans going past our house. That was what led me towards aviation - though I never flew any of those types.

[john_tullamarine]

Another snowy day, another opportunity to read into the background.

And, from the dates, one can see the tie up with the ICAO PAMC development mentioned earlier. As far as I am aware, that was the pivotal development in accelerating things in the performance arena.

[megan]

https://www.flightglobal.com/pdfarchive/view/1959/1959 - 2700.htmlNEW CODES FOR TAKE-OFF
Hammered out at the I.C.A.O. Airworthiness Committee meeting in Stockholm during July and August 1959 were new aircraft performance requirements that now only await the Organization's full approval before they become the basis for new national airworthiness codes. Among the most significant aspects of the Committee's work has been that on take-off for which a complete new range of terms—already airline pilots' jargon—has been conceived. The following explanation is condensed from the Monthly News Bulletin published by I.F.A.L.P.A.

The take-off case today is largely dominated by the requirement that the aeroplane must leave the ground at 1.15 (piston engine) or 1.2 (jets) times the power-off, free air stall. The essential difference in the case of the new code is that this condition is not required to be met until the aircraft is 35ft above the runway. Thus, between the surface of the runway and 35ft is a segment not protected by a 15-20 per cent margin above the stall. But other protection provided by the new specifications is claimed to be as good, and certainly to be adequate to prevent difficulty in getting airborne or sinking once in the air. But an exact comparison of the old and the new specifications is complex since the datum and the technique applied to take-off have been fundamentally changed.

The important speeds associated with the old take-off were (1) the stall (VSI); (2) the minimum control speed (VMC); (3) the decision or power-failure speed (V1); (4) the take-off safety speed (V2); (5) tlie initial climb speed (V2 or V2+,). There were various connecting links between these speeds but, as indicated above, the one usually dominant was that V2 had to have a 15 per cent (pistons) or a 20 per cent (jets) margin above the stall.

The important speeds associated with the new take-off are: (1) the stall (now Vs); (2) the minimum control speed (VMC); (3) the engine failure recognition speed (V1); (4) the minimum unstick speed (VMU); (5) the rotation speed (VR); (6) the unstick or lift-off speed (VLOF); (7) the minimum take-off safety speed (V2 min) and (8) the initial climb speed (V2). Space does not permit the full description of all these terms but the overriding points to be borne in mind are that lift-of (VLOF) is governed by 12 per cent above minimum unstick (VMU) and is—in the normal case—assisted by ground effect; and V2min (the old V2), with its 20 per cent (jets) margin above the free air, power-off stall, is now not specifically required till 35ft is reached.

The important newcomers are VMU, VR and VLOF and their connection is as follows. Unstick or lift-off and all speeds up to the 35ft height point (at which V2 min is reached) are mainly dominated by VMU. This is a speed used only in certification trials and is the lowest speed at which the aeroplane can be made to leave the ground and climb away safely. This should, of course, be substantially above the stall but it is difficult to relate to the stall because (1) most aircraft gain from ground effect (usually considered to apply up to half wing span height) and it may therefore be possible to unstick at or very near to the stall, which is a free-air power-off figure; and (2) the stall itself, as measured in U.S. certification trials, is not the same as the stall known to pilots and can, in fact, be much less (Comet 4, 5 kt; DC-7, 7 kt; KC-135, 9 kt). Anyhow, the fact remains that a test pilot must demonstrate that, for various weights, the aircraft can be made to unstick at VMU. This figure is then multiplied by 1.12 to give the lift-off speed. This is 'the speed at which the aircraft first becomes air-borne' and is the speed applied in the normal operational take-off.
The rotation speed (VR) is the lowest speed at which the nose gear should leave the ground. However, the rotation speeds currently being applied to jets are based on the American SR-422 code and thus associated with the old V2 (usually VR=V2 less 5-10 kt), while the new rotation speed (which will probably be applied to all jets in the very near future) is not directly related to V2 but to VMU, as described above. It is also related to the stall (1.1 VS) and to the minimum control speed (1.05VMC) but these will not usually be dominant.

Summarizing, once the nose gear has been raised at a speed between VR and VLOF the aircraft should climb away, reaching V2 min speed by 35ft or earlier. On leaving the ground the most likely dominant protection is the 12 per cent above the 'worst leaving condition' (VMU) and at 35ft the dominant protection is approximately the same as we have always had at the old V2, though with perhaps a small loss due to slipstream effect being greater than the 5 per cent allowed for as between the 1.15 VS1 (piston) and 1.2 Vs1 (jet) specification.

[Pugilistic Animus]

Before Vmu ground stalls were possible...that's the original takeoff accident

A bit off topic the original landing accidents in Jets was high sink rates on the Approach

[slast]

Just curious, aterpster - would the union guys have included e.g. Joe Meek, Tom Foxworth or Bill Melvin?
Steve

[aterpster]

Just curious, aterpster - would the union guys have included e.g. Joe Meek, Tom Foxworth or Bill Melvin?
Steve
Nope. Yours truly, John McCormick, and John Burns.

[Discorde]

A trick - sorry, I mean procedure - we used on the B732 was 'improved climb'. If excess runway length was available we would calculate a higher VR and V2 to get a better L/D to improve 2nd segment engine-out gradient. I can't remember how we calculated the V-speeds and extra RTOW from the T/O perf manual data. A graph maybe?

Another dodge - sorry, procedure - to achieve higher RTOW was to visually avoid obstacles penetrating the nominal engine-out climb-out path if the weather was suitable. IIRC Naples R24 was an example.

So, did IC arrive at the same time as Perf A?

[Tinstaafl]

'Improved climb' calculations were part of Perf A when I converted to a pommy ATPL in 1999.