Bill Waterton, chief test pilot on the Javelin, notes in his book "The Quick and the Dead", that the tailplane was found to be trimmed in the opposite direction to be expected to enable recovery. One previous stall had been carried out, and the fatal stall was with flaps down. He notes that elevator control was "sluggish" at low speed. At the time of writing the book (1955) results of the accident investigation had not been released, even though he the chief test pilot had asked for a copy from the Air Ministry.
Found the following on
Aircraft Profile #179. Gloster Javelin
High altitude flying on the first two prototypes had shown that the high speed handling characteristics could be improved by modifying the outer wing. From just outboard of the guns, the leading edge sweep back was reduced thus increasing the tip chord and reducing the thickness/chord ratio. The decrease in taper ratio reduced the span wise flow and consequently improved the tip stalling characteristics and the lift coefficient at high subsonic Mach numbers. A modified set of wings was fitted to WD808 in time for a first flight on 28th May 1953, But only a fortnight later disaster struck again, this time with grim results. Peter Lawrence was flying WD808 on Nth June, doing stalling tests with the C.G, further aft than normal. During a test at a high angle of attack, the "super stall" condition developed, where the nose of the aircraft pitches up and the forward speed decreases rapidly. Below about 80 knots the condition is virtually stable and, despite all Lawrence's efforts, he could not regain control. Unfortunately he must have misjudged his height and, when he finally ejected, he was too low for his parachute to deploy completely. Thanks to the information he had given over the radio and the auto observer records, a complete dossier was made out.
Again testing was held up while an answer was sought. Wind tunnel tests, at high angles of attack confirmed the "super stall conditions. The main cause of the trouble was tip stalling which rapidly spread to the rest of the w ing at high angles of attack. Having the flaps down aggravated the condition so the front edge of the flaps were cut back to give a semi slotted effect. This improved things but did not cure the tip stalling. Model tests with leading edge droop showed no improvement but wind tunnel tests on slotted wings gave promise of a cure. Consequently a set of wings with slats of constant percentage chord was test flown and the stall became innocuous. But the weight penalty (130-180kg) and the cost were not acceptable to cure a conditions which need never occur.
Consequently a stall warning device was developed which proved effective. A small metal flag with a rotating mast was fitted at about mid span of the outer wing. At rest, a spring held the flag at right angles to the line of flight and a set of contacts were closed. When the aircraft took off, the airflow moved the flag round into line of flight position and opened the contacts. When the undercarriage retracted, a circuit to a sound box, operating on the intercom, was made except for the flag contacts. Should the airflow over the top surface of the wing start to break down, the disturbed flow would no longer hold the flag against its spring, the contacts would close and the pilot would get a warning note in his headphones.
Apart from this audible warning there is little else to warn the pilot of an approaching stall. If the nose is held up below 185km/h the speed will drop off rapidly and finally the aircraft will yaw uncontrollably to one side and roll in the opposite direction before falling into a spin.
Interesting to note the CofG further aft than normal - too far aft perhaps? An SR-71 was lost during test because the CofG was too far aft and control was lost.