It had nothing to do with the British ARB
It had everything to do with the ARB, in particular their chief test pilot D. P. Davies. To put that myth to bed,
"Flight" 25/12/59
Some time ago it became clear that the A.R.B. was not prepared to give approval to the Boeing 707 for B.O.A.C. purely on the basis of compliance with American performance requirements. The history of this situation is a long one, and it is closely bound up with the gradual reconciliation of British and American performance requirements. Although these are now similar they are not identical, so it does not follow that because the aero-dynamically similar Boeing -320 has been certificated to S.R.422B requirements that these can automatically be accepted by the A.R.B. In fact, there are some 17 special conditions which the A.R.B. have applied to American jet aircraft; and there are also five special conditions which apply to all American aircraft, piston or jet, swept or straight wing. It is one of these five—that concerned with performance—which is presenting something of a stumbling block.
A significant section of the requirements is that which concerns unstick, rotation and lift-off speeds in relation to the stall. The two authorities, working from different assumptions, calculate the respective speeds in different fashions; and, although the differences are well understood, jet take-off is critical enough (there is now more than a measure of doubt that the airfield performance demanded of the 707-120 by S.R.422 is sufficiently stringent) for the A.R.B. to require compliance by demonstration to British civil airworthiness requirements.
One difference between the S.R.422 and B.C.A.R. requirements is that the latter require demonstration from the ground of an aircraft's ability to unstick from a nose-high attitude. It was thus important for Boeing to show that the 707 was limited by its geometry from entering a ground stall condition during take-off, and this was the reason for the protracted tests at Edwards Air Force Base with a boiler plate and wooden-runner tail bumper. These tests, of course, take time. They are important because on them may rest the amount of payload that the -420 will be able to carry for a given length of concrete; the A.R.B. require double the margin between the minimum unstick speed and the start of rotation if a ground stall can occur. To an aeroplane moving at the take-off speed of the Boeing 707 this increased margin represents many feet of concrete.
The situation is further confused by changed thoughts within the F.A.A. Over the period of Boeing 707 certification, requirements have been revised from S.R.422 to S.R.422A and then to S.R.422B. At the time of writing the situation remains unresolved and delivery date of the -420 to B.O.A.C. Is still uncertain. The conflict between safety and economics is considerable because every percent increase in take-off (i.e., rotation or lift-off) speeds imposes on the operator the need to reduce payload or use more runway. The A.R.B.'s requirements are laid down and are firm, but the application of performance requirements is nearly always "supertuned" to strike the fine balance between practical operation and safety. Boeing must adjust the 707-420 flight manual to meet these requirements while maintaining pressure all the while—by demonstration and by argument—to avoid unnecessary performance margins which could constitute a commercial penalty for the operator.
"Flight" 5/2/60
While Boeing still claim on-the-nail or early delivery for every 707 so far, by the time BOAC's first -420 reaches London this claim will no longer be valid. It came as something of a shock, although not as a complete surprise, that following flight tests by D. P. Davies, the Air Registration Board's chief test pilot, British approval of the 707 was still withheld. The issues are quite separate from those affecting take-off performance.
Unease at the aircraft's behaviour is apparently concerned with the three training incidents which occurred last year, two of which resulted in fatal accidents. In each case, it was lateral behaviour of the aircraft at slow speeds that was called to question, although it is believed that the initiating actions were quite different in each case. The two incidents in which engines were shed—over France in February and Seattle in September—were apparently the result of demonstrations, possibly inadvertent, of conditions outside the normal flight envelope. On the other hand the American Airlines' accident over Long Island in August during a two-engine approach in a gusty crosswind could not be explained in this way, although again it was suggested that had a different technique been adopted the accident might have been averted.
From the violent rolling manoeuvre that developed on each occasion it may be inferred that roll and yaw control demands higher-than-average pilot skill under asymmetric thrust, at low speeds, and in adverse conditions—a conclusion that is borne out by Boeing's intention to increase the -420's fin height by a drastic 35in, to add a small ventral fin and to provide full rudder boost on the Boeing 707-120. (The -320 and -420 rudder is already power boosted beyond 15° travel to counter additional engine thrust, and the Qantas -138s were modified similarly before delivery.) Another 707 modification apparently found desirable is duplication of the yaw damper, and this again points to Boeing's persuasion that low-speed flight is a regime where control could be improved. It should be made clear that although the -420s are to be modified at ARB's behest, Boeing have conducted on their own account a similar programme with other aircraft in the 707 series. The first modified 707-420 will be ready for trials in the early part of this month and it will again be flown by Mr Davies before it is finally accepted and handed over to BOAC.
But while it now seems probable that all future 707s may be modified as part of normal design improvement, it is not clear if retrospective action is contemplated, nor who would pay for it if it was. All deliveries made so far have been under FAA certification or validation, and unless this certification standard is changed (and there is ample precedent for this in the past history of 707 approval) it would at first sight seem to be incumbent upon airline customers to request and pay for modification action themselves.
What is involved is a matter of opinion about pilot skill, and perhaps this makes the modifications under discussion different from mere design improvements. As a matter of policy Boeing may thus take responsibility for the modification upon themselves. After the accident at Long Island last year it was questioned whether, with the lessened likelihood of a pilot experiencing full asymmetric turbojet failures, a training demonstration of this kind should now be needed. And the FAA demanded that this type of exercise, tacitly admitted to be a difficult condition with a swept-wing aeroplane, should be attempted only at altitude.
The ARB's insistence on modification suggests a philosophy that if a redeemable failure case exists, it must be demonstrated, and pilots must be trained to handle it. Indeed, looking back on the FAA's "high altitude only" rule, it is easy to argue now that this could only have been thought of as an interim measure and that the FAA too must be concerned that low-speed asymmetric control must be improved until it is no longer a hazard.
Looking at the particular modifications a little more closely, the fin extension (and rudder boost in the case of the -120) are obvious ways of improving asymmetric-thrust control at low speeds. This is adversely affected by sideslip or yaw of the swept wings, because rudder tab effectiveness is reduced. Another gain will accrue from the ventral fin. It has incidentally been suggested that this fin could act as a tail bumper and so reduce any penalties that may result from an excessively nose-high ground angle. The full take-off story has yet to be told, but it now seems to have been successfully demonstrated that the geometry of the 707-420 is such that ground stall does not occur and that only revision of the flight manual requires to be completed.