Handley Page’s Victor was a stablemate of the Avro Vulcan in the RAF’s strategic jet bomber order of battle. Built around a crescent wing, the Victor was the most sophisticated of the three British V-bombers. Although the Victor B.1 had greater range and ceiling than the firstgeneration Vulcan, it was not long before the Victor B.2 was created to counter improving Soviet air defences. The new variant came about from the classic ‘stretching’ process, in that it featured bigger engines, more wing area and higher all-up weight. It was capable of safe operation up to 60,000ft (18,288m) and blessed with far more reliable internal systems. The first contract for Victor B.2s excluded provision for a prototype, so a B.1, XH668, was brought forward on the Handley Page (HP) production line to assume the role. It took the B.2 designation after being fitted with new Conway engines rather than Sapphires and larger mainplanes. Painted in overall anti-flash white to reflect the light from nuclear blasts, XH668 made its first flight in the hands of the firm’s deputy chief test pilot, John Allam, on February 20, 1959.
TESTING TIME
Victor XH668 was transferred to the Aeroplane and Armament Experimental Establishment (A&AEE), at Boscombe Down, Wiltshire, on August 17, 1959 for preview handling trials. Personnel selected to undertake these were Sqn Ldrs R J Morgan and G B Stockman as pilot and co-pilot respectively, Flt Lt L N Williams as navigator and Flt Lt R J Hannaford as air electronics officer. They were accompanied by Bob Williams, HP’s chief flight test observer, who was well versed in the management of Victor B.2 systems. Both pilots were given a familiarisation flight in XH668 by John Allam, on the evening of August 17.
The schedule for Wednesday August 20 called for a climb to 52,000ft, an hour of tests including high-speed turns to reach the fringe of wing buffeting (and to pass just beyond it), a rapid descent down to 35,000ft using airbrakes, further manoeuvres at that altitude and then a drop to 10,000ft. There followed more high-speed turns before returning to Boscombe after a trip lasting almost three hours.
Airframe XH668 got airborne from runway 06 at 10:35hrs. Having cleared with local control, the crew changed its communications to Boscombe approach. A warning of an aircraft passing ahead was acknowledged by XH668 at 10:38hrs, but Boscombe had no further radio contact with the Victor crew. The approach controller watched the prototype climb away on radar, heading approximately 020°.
“A WARNING OF AN AIRCRAFT PASSING AHEAD WAS ACKNOWLEDGED BY XH668 AT 10:38HRS, BUT BOSCOMBE HAD NO FURTHER RADIO CONTACT WITH THE VICTOR CREW”
Although Boscombe Flying Orders stated that prototype aircraft should maintain communications with the ground, XH668 was regarded as a new mark of an existing type, doing nothing that had not already been done by HP, and so it was not technically considered to be a prototype. Furthermore, due to weather vagaries and the large distances covered at high speed, preflight submission of a precise flight plan was not expected. Therefore, Boscombe had only a rough idea of when and where the crew intended to operate. It was only because Sqn Ldr Morgan had arranged with Boscombe’s Performance Division to photograph the Victor’s landing run that anyone noticed anything amiss as early as they did. At around 13:10hrs a telephone call was made from the Performance Division to Air Traffic, to ask for XH668’s estimated time of arrival. After an element of confusion because no one on the ground was exactly sure how long Sqn Ldr Morgan intended to stay airborne, ‘overdue’ action was not taken until 15:03hrs. By then, XH668 had long since disappeared from the sky… and in a catastrophic manner.
INVESTIGATIONS
A board of inquiry was convened at Boscombe on August 21. Having no information, it sought help from the public at large. The disappearance of Britain’s latest bomber conjured all manner of theories, including hijacking on behalf of an unfriendly agency along the line of James Bond’s Thunderball. Many people wrote to say they had seen an unfamiliar shape “flying at a great height and very fast”, but most sightings were either in the afternoon or in unlikely places such as “low flying over Kensington”.
The Royal Navy diving support and submarine rescue vessel HMS ‘Reclaim’ was involved in the hunt for XH668.
A page from the Royal Aircraft Establishment (RAE) accident report on XH668, showing the recovered port pressurehead unit. Note the pitot tube is bent more than 90°
The inquiry had more luck when Capt Yendall, Master of the small coaster Aquiety bound from the Mersey to the Thames, heard a BBC radio broadcast that an aircraft was overdue from Boscombe Down. Earlier that day, when the ship was off St David’s Head, Pembrokeshire, the captain and two of his crew had been on the bridge because there were extensive fog patches. At around 11:40hrs the vessel ran out of the fog and about five miles away from The Smalls (a group of islands with a lighthouse, west of Marloes Peninsula), the captain, mate and helmsman observed a large column of water and spray about 50ft high, some five miles away. This sighting was followed almost immediately by two sharp reports, similar to rifle fire. Capt Yendall passed his sighting to the Ilfracombe port radio station, but there was no way of knowing if it involved the loss of XH668.
The Air Ministry asked its UK radar stations to check their photographic records for the period XH668 was airborne. On August 22, RAF Wartling in East Sussex found a radar track that ended abruptly at approximately the same position and time as that reported by the Aquiety’s captain.
Two days later both radar trace and Capt Yendall’s statement were in the hands of the inquiry. The track drawn from the trace was then found to be that of a single aircraft that began at a height, position, direction and time corresponding precisely with that of the Victor when last seen on radar by Boscombe. Because the track was continuous, and the responses were good, even at the extreme range of 240 miles (386km) from Wartling, it was possible to be sure the aircraft was above 35,000ft. The area covered by the track corresponded with where Sqn Ldr Morgan normally carried out this type of testing, and the track ended 17 miles in distance, and three minutes in time, from the column of water reported by Capt Yendall. The final radar response from the Victor showed it to be in a turn to port; the corrected position of this last response and that of the splash were just ten miles apart. This was sufficient to convince the inquiry that at least a substantial portion of XH668 entered the water at the reported location. Their suspicions proved to be founded on August 25, when the first white fragment of Victor rear radome was found by a schoolboy on Whitesands beach, St David’s, Pembrokeshire.
WORKING THE CLUES
The inquiry concluded there was an adequate supply of oxygen on board and, as individual crew members were wearing air-ventilated suits, G trousers, partial pressure jerkins and helmets, they should have been protected from the physiological effects of high altitude. No single fault could cause total loss of electric power on the B.2, and nothing pointed towards structural failure because the entire flight was planned to remain within design limitations. There should have been no fuel tank or battery explosion because the Victor was flying with unpressurised fuel tanks, and a low-voltage battery that was not crucial to the maintenance of all electrical power.
Sabotage did not seem credible, but the plotted track of the missing Victor passed close to the missile range at Aberporth which, around that period, was testing the Bloodhound surfaceto- air missile designed specifically to cope with high-flying jet bombers. It must have been a relief there were no launches from Aberporth on August 20, nor was there any simultaneous activity in adjacent danger areas.
By September 1 an extensive air and sea search had been narrowed to an area 16 miles long by 12 miles wide in a south-southwest direction, centred on the estimated position of the splash noted by Aquiety’s captain, and now marked with a datum buoy. This 192 sq miles (487km2) approximated to Greater London, from Tottenham to Streatham and from Woolwich to Ealing. St George’s Channel, where the splash had been seen, was nearly 400ft deep. Undaunted, the Royal Navy carried out a co-ordinated sweep of the sea bed with ASDIC, the submarine detection device. When suitably high intensity echoes were found, wreckage was retrieved by fishing trawlers specially hired for the purpose. The number of trawlers would rise from four to 16, supplemented by a special salvage ship, Twyford, which carried a diving chamber, underwater television and huge grabs capable of lifting 20 tons.
NET RESULT
On January 5, 1960, the Picton Sealion, a trawler fishing out of Milford Haven and independent of the search itself, let down its nets as she approached the designated search area. When the catch was finally examined, it contained a crumpled piece of shiny corrugated metal bearing just a small trace of marine growth. At last the sea had begun to show co-operation but where, along the vessel’s 12-mile trawl, had it scooped up what was later identified as part of the starboard wing trailing edge? Twyford was moved out to sweep along either side of the line of trawl, but by early March nothing further had been found.
“ON AUGUST 22, RAF WARTLING IN EAST SUSSEX FOUND A RADAR TRACK THAT ENDED ABRUPTLY AT APPROXIMATELY THE SAME POSITION AND TIME AS THAT REPORTED BY THE AQUIETY’S CAPTAIN”
In March the trawler Clyne Castle, searching north of the Smalls, brought up four pieces of bent and twisted metal, seven fragments of fibreglass and three lengths of plastic-covered wire. By noon the next day, all 14 finds had been identified as wreckage from XH668. Ten days later the trawler Forards was fishing further west than usual, when it recovered a piece of engine and some outer skin from a bomb bay door. This was the first indication the Victor had begun to break up before hitting the sea.
Dr Percy Walker, head of the Structures Department at Farnborough and the man tasked with overseeing the investigation into the accident, authorised two trawlers to search the new western area. However, while they concentrated on looking for pieces that must have broken away at altitude before falling into the sea, the real search effort swung towards where it looked increasingly likely the mass of the wreckage had hit the water – where Clyne Castle had made its initial find. By the end of March, 100 pieces (just over 1% of the Victor by weight) had been recovered.
At Farnborough, a full-size skeleton of the B.2 prototype had been erected and was waiting: such was the importance afforded to the whole exercise that the second Victor B.1 prototype had been dismantled at HP’s Radlett, Hertfordshire facility and its major components taken to Farnborough for comparison. The plan was that XH668 would be, as much as possible, rebuilt from its broken remains. In fitting the jigsaw back together it was hoped to discover why it broke up in the first place. It was soon possible to pinpoint by Decca navigation equipment the area in which XH668 was most likely to be found. As the total tally reached 275 pieces on April 7, a marker buoy was dropped on the new search datum. It was more than five miles north of the original datum and Aquiety’s estimated splash position.
Victor XH668 weighed around 63 tons when it crashed, of which 20 tons was fuel. The maximum recoverable weight was optimistically put at 40 tons but, having retrieved half-a-ton per day in June, a definite pattern was emerging. Nearly all the wreckage was concentrated in a relatively small circular area not more than 200 yards in diameter, but Farnborough was becoming just as interested in wreckage being found in the western outer area. It was there that pieces of both port and starboard wingtips – the former with the pitot tube still attached – were recovered showing remarkably little damage. This pointed almost certainly to portions of aircraft having broken away before the main mass hit the water.
One of the smallest items collected was the co-pilot’s wrist watch. Given the shattered condition of his ejection seat, his timepiece must have been in the aircraft when it struck the water. The hands on the face had stopped at 11hr 30min and 46sec, and experts concluded that at this time the watch had been stopped by a single violent blow.
PITOT PROBLEMS?
By September, Dr Walker’s team had a reasonable understanding of what had gone wrong with XH668. Knowing the pilots were fully conscious and not injured in any way, and having recovered motors, generators and hydraulic pumps that showed signs of running just before final impact, the inquirers considered loss of control brought about by other factors. The Victor and its crew had been sent to operate up to Mach 0.97. As the aircraft approached the speed of sound, aileron control would have virtually disappeared because the power flying control jacks would have been unable to supply the heavy forces required to operate the down-going aileron. This would not have been critical when flying straight and level, but the crew was briefed to carry out steep turns, and buffeting would have restricted the amount of extra lift the elevators could generate. The combined effects of all these factors meant a spiral dive could have developed, from which recovery became less and less likely as speed increased.
The issue rested on what caused the Victor to exceed Mach 0.97 in the first place, and given that a skilled test crew was involved, the inquirers had to find a cause that distracted the pilots in a major way at the critical time. Among the recovered wreckage were the wingtips that came adrift during the descent. Victor B.2 wingtips were, more or less, self-contained and detachable units, and each carried a long, slender tube projecting some 6ft ahead of the wing leading edge. They comprised both static and pitot pressure systems. When the wreckage of the port wingtip was re-assembled it was noted that the pitot tube was bent through more than 90° at the end of the supporting sleeve that constituted part of the mounting. This damage was probably caused by flutter late in the descent, but it was significant that, despite the severe strains imposed, the pitot tube base remained firmly held in its mount.
The re-assembled starboard wingtip told a different story because, although structural damage echoed that on the port side, the pitot tube was missing in its entirety. A Victor pitot tube was held in position by a form of chuck, in which a conical sleeve was tightened against two tapered collets. There was nothing wrong with the design, but the starboard pitot tube had come cleanly and neatly out of its mounting without any signs of damage. Furthermore, when the tapered sleeve for the starboard mounting was recovered from the sea with the collets still in position, the inside of the sleeve was found to be covered in protective paint, which had clearly been applied too freely. It was surmised the sleeve had originally been tightened adequately but only against localised areas of paint. As the paint wore away or became distorted, progressive loosening of the collets’ grip upon the pitot tube could reasonably be expected. Nothing similar was found on the port installation.
MECHANICAL HELP
The pitot systems were in many ways identical, but the one to starboard was unique in serving the stall detector and Mach trimmer. Wind tunnel tests a decade earlier had predicted the crescent wing stall would be fairly vicious, so HP developed large accumulators of stored energy, which thumped down leading-edge flaps within a second on a signal from a pressure ratio switch. The latter device calculated lift coefficient from tappings below the wings. While the flaps travelled, warning lights illuminated in front of both pilots. Ultimately, the stall was not as bad as feared, so the nose flaps would eventually be locked up on the Victor B.2, but in 1959 XH668 still had them. In addition to pressure ratio switch signals, the stall detector received pitot and static pressures from the starboard pitot tube.
A Mach trimmer was incorporated to counteract the nose-down trim change caused by air compressibility at high Mach numbers. As the aircraft accelerated from Mach 0.85 to 0.95, the Mach trimmer raised the elevators through an angle of approximately 5° without altering the stick position (and irrespective of the pilot’s control inputs) in response to pressures conveyed from the starboard pitot. Dr Walker’s team started by assuming XH668 was flying buffeting turns around Mach 0.94 at 52,000ft. Suppose then, the starboard pitot pressure tube started to come adrift slowly under the effect of the buffeting. A leak would gradually develop, causing a fall in indicated airspeed on the co-pilot’s side. Assuming this was noticed, the captain’s instruments would still be functioning properly and maybe the assumption was made that water in the starboard system had frozen.
So far so good, but then the pitot tube dropped clean away. Zero speed would definitely have been recorded, but as the tube detached, the stall detector would have lowered the nose flaps. The first the pilots might have known about this was when the warning lights came on, but what would they have thought? At Mach 0.94 and 52,000ft, a Victor B.1 would have sailed along very comfortably like a knife through butter, but in similar circumstances, a B.2 would have been buffeted and shuddered around most uncomfortably because of its larger wings and bigger engine air intakes. In such circumstances the warning lights and lowering nose flaps would only have confused matters, because they would have been activated for no reason as far as the crew was concerned.
“FACED WITH STALL WARNING INDICATIONS, THE NATURAL REACTION WOULD HAVE BEEN TO LOWER THE AIRCRAFT’S NOSE”
CONFUSED READINGS
Faced with stall warning indications, the natural reaction would have been to lower the aircraft’s nose. Unfortunately, distractions would have masked the most sinister complication. At Mach 0.95 the Mach trimmer would have been fully out, but on receipt of a spurious lowspeed signal it would have steadily lowered the elevators and pushed down the aircraft’s nose, which was very efficient aerodynamically. Moreover, the elevators would be lowering through almost 5° at a time when the pilot might have been pushing the stick forward himself.
The combined nose-down movement – which might have been increased if the co-pilot had instinctively reacted to an apparent fall in his airspeed indicator before realising what was happening – would have pushed the Victor quickly beyond human recovery. It is most likely the throttles stayed open and the airbrakes closed, because both pilots were concentrating all their hands and efforts on trying to overcome the forces opposing the aileron jacks to level the wings, as an essential preliminary to dive recovery. It would have been futile, and the crew was doomed as the spiral tightened.
The Victor’s flaps were found to be down and its Mach trimmer actuator virtually fully retracted when it hit the sea. To prevent recurrence, the mountings of all Victor pitot tubes were modified so the collets were locked permanently, and therefore could not vibrate loose under any circumstances.
Right at the bottom of St George’s Channel, a Victor B.2 pitot head lies buried in the sand, as it was not recovered. But after spending £2 million at 1960 prices (the cost of a new Victor B.2), Britain not only restored the bomber’s credibility with the crash investigation, but also gained the record for the greatest overt aviationrelated salvage operation thus far. Sadly, though, no trace of the crew was ever found.
In the absence of an flight data recorder, the only record of what went on inside XH668 that fateful day was a fragment of the crew’s last conversation, heard purely by chance at Boscombe. It was a very weak and mangled transmission, and by a freak of radio it told the inquiry the crew was listening to the BBC radio drama Mrs Dale’s Diary as the aircraft went down. That might sound mundane, but if the loss of XH668 teaches anything, quality control and effective supervision are not just high-tech preserves: they apply right across to the chap with the paintbrush.