by Damien Burke
Preoccupied with the difficult steering and with the speed building, Beamont inadvertently retarded the throttles past the point at which the fuel cocks were closed. The engines wound down and as a result the aircraft lost electrical power. The liquid-oxygen (lox) system, now without power, began spraying oxygen ‘snow’ out of the oxygen masks, and, when it was disconnected, lox leaked from the oxygen regulators. This was a serious hazard, as lox can ignite on contact with grease, and it could be pretty well guaranteed that various areas of the cockpit had been smeared with grease during the previous months. Luckily no fire broke out. The next day another morning engine run was followed by a further afternoon taxy run to check differential braking and run up to 60kt once more. This test was terminated owing to an hydraulic leak, and a second attempt later in the day had to be cancelled when the port undercarriage leg warning light lit up, indicating loss of brake pressure.
After a day in the hangar rectifying hydraulic snags and replacing some life-expired reheat fuel inlet pipes, the tests continued on 5 September, a Saturday. The intention was to carry out three faster runs, but in the event there was only time for two, one to 100kt (115mph; 185km/h) and one to 120kt (138mph; 222km/h), both with successful streaming of the braking parachute. Even on a Sunday the taxying tests continued, though a gusting crosswind meant that only one run was possible, in the middle of the day. The No.1 engine on this run exhibited ‘hunting’ behaviour, and the auxiliary intake doors were flicking in and out, so an afternoon engine run was carried out to try to diagnose the problem. The next day did not go well either, with a failed braking parachute stream at 140kt (160mph; 260km/h), resulting in heavy use of the wheelbrakes, which then became welded to the wheels and required wheel and brake replacements later. The engines were once again inadvertently stopcocked, and the aircraft had to be towed off the runway. Next morning the throttles were modified to stop a further repeat of the stop-cocking, but numerous attempts to start the engines for a run that day were frustrated by various problems culminating in the need to replace the constant-speed drive starter (CSDS).
Wednesday the 9th saw another brake parachute failure on a 140kt run. Heavy braking caused brake temperatures to rise to over 1,000°C, and the starboard forward main gear tyre deflated. The remainder of that day and the next was spent modifying the parachute and beak door, and fitting a crew-operated stopcock on the lox system to prevent further recurrence of the dangerous leak in the cockpit. The auxiliary inlet doors had continued to be problematic, so they were locked open at 40 degrees.
It fell off the back of a lorry, guv! – XR220 arrives
The 9th also saw XR220 delivered to the airfield, with wings on one trailer and the fuselage on another. The driver jack-knifed the latter trailer while turning right to enter B3 hangar shortly before five o’clock, and as the trailer toppled on to its side XR220’s 18 tons of fuselage partly parted company with the trailer, the rear end (at the forward end of the trailer) coming to rest on the ground, supported by the port tailplane spigot. This was fortunate, as had it rolled any further it would have smashed into BAC’s Lightning chase aircraft, parked just a few feet away. As it was late in the day and the resulting mess appeared stable, it was left overnight. Next morning an inspection was begun to see how much damage had been caused and determine how best to recover it.
Salvage procedures had been created for the type, but they did not cater for a situation in which a wingless fuselage would be lying on its side, partly attached to a trailer. They were designed to cope primarily with aircraft that had suffered landing accidents and were upright on their belly. Some specialized equipment was required, including a fuselage lifting beam, and these were fabricated at Weybridge in a matter of hours, being delivered on the 11th, when the salvage procedure began. After the covers had been removed, the fuselage was lifted with cranes, the trailer was taken away and air bags were placed underneath the aircraft. A ground rig was used to extend the starboard main undercarriage and the fuselage was then lowered until the wheels were on the ground. Then, with careful crane use, the fuselage was rotated until it was upright once more. The port undercarriage was then extended and the fuselage lowered further, a jack being placed under the nose before the nose leg was finally extended. Everything held together, and the fuselage was towed into the hangar six hours after the salvage operation had begun.
‘Recovery of a crashed aircraft’; unfortunately the recovery procedures did not deal with the problem of a wingless fuselage entangled with a trailer. BAE Systems via Brooklands Museum
The recovery of XR220 on 11 September 1964. BAC Weybridge came up with a successful recovery technique within hours of the accident, and very little additional damage was caused during the operation. BAE Systems via Brooklands Museum
Refuelling XR219 before an early engine run. At this point the navigator’s cockpit lacked a seat, and BSEL personnel manning the navigator’s position made do with a wooden crate. BAE Systems via Warton Heritage Group
Roland Beamont boards XR219 for the first flight. With the cockpits so high above the ground, access ladders were a must, though the specification did call for retractable on-board ladders to be used when necessary. BAE Systems via Warton Heritage Group
The Olympus 320 was not a clean-burning turbojet, and an engine run was announced not only by the deafening roar of the engines but also by the cloud of black smoke. BAE Systems via Brooklands Museum
While XR220 was being salvaged, XR219 continued to cause frustration. Fuel pump warning lights illuminated and the aircraft was shut down, whereupon fuel leaks from blown gaskets on the forward fueldraulic pumps were found. The next day it was found that the No.2 booster pump drive had sheared, and a decision was made to stop the taxying tests for several days and replace all the booster and fueldraulic pumps. Beamont had complained of poor visibility through the windscreen, so it was replaced.
Meanwhile, XR220 had been receiving a steady stream of visitors from various BAC departments to assess the damage. Initially this appeared to be nowhere near as bad as it could have been, being mostly limited to minor damage to the port tailplane spigot and main jacking points, and various dents in skin-panels, including the underside of the port intake and by the lower airbrake doors. However, beneath the cracked and dented panels lay a succession of damaged longerons and frames that would require extensive repair.
Late on the afternoon of Thursday 17th another attempt at taxying XR219 was made, but fuel vapour in the cockpit put paid to this. The next day things went from bad to worse. An engine run in the morning produced more fuel leaks, and melted droplets of metal were found on the ground below the aircraft. Replacing fuel pipes in the accessories bay and tracing the source of the metal droplets occupied the rest of that day and all of the next.
Taxying resumed on Sunday 20th with a 110kt (125mph; 203km/h) run on the shorter cross-runway at Boscombe, as the crosswind on the main runway was out of limits. The brake ’chute streamed successfully, but the parachute beak door hit the base of the fin as it opened. The aircraft was returned to the hangar again for another CSDS change which stretched through into Monday. Another run on Monday afternoon to 125kt (144mph; 232km/h) was again successful, as was a run to 140kt (160mph; 260km/h) on Tuesday, which surprised Beamont when he found he could rotate the nose and lift the nose gear from the runway at just 105kt (120mph; 195km/h) using 10 degrees of tailplane rather than the predicted 130kt (150mph; 240km/h) and 18 degrees. More metal had been found, though, traced to lagging from an oil pipe, and it was decided that filters needed to be installed in the lines between the engine and CSDS to prevent this metal harming the CSDS. Work then concentrated on getting the aircraft ready to fly.
By Friday the 25th further inspection of XR220, including fuel system pressure testing, had found several fairly minor fuel leaks in the rear fuselage and in the forward fuel collector box. On the plus side, a complete alignment check on the fuselage and the port tailplane spigot was passed with flying colours. More engine runs on XR219 had
resulted in more pieces of metal being found, both under the aircraft and also in the newly installed filters. This time the source was finally identified as the starter cart for the Palouste auxiliary power unit (APU), which had internal damage. This was repaired on the 26th, and an engine run in the afternoon cleared the installation, but now the all-important low-pressure (LP) drive shaft strain gauge on the port engine failed, so the first flight could not be made that day.
First flight
As the sun rose on the morning of the 27th, a Sunday, the weather did not look good. Mist blanketed the airfield and it looked as though XR219 would be staying put. An engine run first thing in the morning went well, however, and the intention was then to carry out a final fast taxy with brake-chute stream. However, the mist remained as thick as ever, and it was not until midday that the run could be tried. All went well, but the beak door once again lightly impacted the base of the fin on opening, though no damage was caused to either fin or door. It was decided that a temporary way of avoiding this problem would be to land, keep the nose up for aerodynamic braking and only release the brake ’chute once speed had decayed to below 150kt (172mph; 278km/h). Preparations for a flight later in the afternoon began in earnest. Crewingin at Boscombe Down at about 14.00hr, Roland Beamont and test navigator Don Bowen carried out their preflight checks and the assembled media on and off the airfield waited expectantly. As the aircraft began to taxy away from the ‘pear-drop’ area at Boscombe (a circular area at the end of runway 28, near the eastern perimeter), the weather had finally cleared completely and the sun was shining. Two chase aircraft, Canberra WD937 flown by John Carrodus and Lightning T.4 XM968 flown by Jimmy Dell, got airborne first, and then it was the TSR2’s turn to line-up on runway 24. A stiff breeze gave a crosswind of 11 to 15kt (12–17mph; 20–27km/h); beyond the 7kt (8mph; 12km/h) limit that had been agreed a month before.
Several taxy runs worked out various bugs in XR219’s systems before the first flight, the most serious problems being limited to braking-parachute failures. BAE Systems via Brooklands Museum
For this first flight XR219 was subject to a host of limitations. The AUW had to be kept below 74,000lb (33,500kg) so that an engine failure at lift-off speed (180kt (205mph; 333km/h)) would not make it impossible to climb away on the remaining engine (although 100 per cent power would be needed, and BSEL could only guarantee safety at up to 98 per cent), or impossible to land on one engine without reheat. This meant a restricted fuel load, and hence a short flight. In view of the short flight time, no attempt would be made to retract the undercarriage, because no time would be available to troubleshoot it if failed to retract or extend cleanly, which was a real possibility, as the required number of successful rig tests had not been completed. The auxiliary intake doors remained locked open, and the airbrakes were kept cracked slightly open, even in what should have been the fully closed position, as fully closing them could result in one or more of the airbrakes attempting to go past the fully closed point and damaging the mechanism. The short flight time would limit the flight’s usefulness to an assessment of the aircraft in flight in its take-off and landing configurations only.
Preparation for the first flight, with XR219 joined by Lightning XM968, the chase aircraft. BAE Systems via Brooklands Museum
Take-off for the first flight of the TSR2, Boscombe Down, 27 September 1964. With a change of government on the horizon, BAC gambled everything on a successful first flight of an aircraft that some considered to be far from airworthy. BAE Systems via Brooklands Museum
Beginning the first left-hand turn in the climb, shortly after take-off on the first flight. The Lightning chase aircraft stuck to XR219 like glue for this first flight, allowing the crew to compare instrument readings, and of course enabling BAC’s photographer to take a fine sequence of shots. BAE Systems via Warton Heritage Group
On the first downwind leg of the first flight, with Boscombe just visible in the haze below the aircraft’s undercarriage. No attempt was made to retract the undercarriage on the first flight, as not enough successful cycles had been carried out on the ground. Nor would the limited fuel load have permitted much time to troubleshoot had problems arisen in the air. BAE Systems via Warton Heritage Group
At 15.28hr Beamont released the brakes, and with twin spears of flame from the Olympus engines visible even in the bright sunshine the aircraft rolled forward. The wind was 11kt (12.5mph; 20km/h) at 240 degrees. Flaps were down 20 degrees and flap blowing was on. When the aircraft reached 120kt (140mph; 220km/h) Beamont began a steady rearward stick movement and the nosewheel left the ground three seconds later, by which time speed had risen to 150kt (170mph; 280km/h). Beamont had an impression of reduced acceleration, and ceased rotation for a few moments. However, his perception was in error and just over 23sec from beginning the take-off roll, at 189kt (217mph; 350km/h), XR219 became airborne, having used a whisker over 1,200yd (1,100m) of runway. By the time the aircraft had reached the magic 50ft (15m) of altitude, 1,688yd (1543m) of runway had gone by and the aircraft was flying at 206kt (237mph; 380km/h). The aircraft was flying well, and the expected standards of control and stability had been exceeded. Roll damping and fin effectiveness in particular were better than anticipated. Not everything was working well; the cabin air conditioning produced an icy blast of air when selected to ‘AUTO COOL’.
With Lightning and Canberra chase aircraft in attendance, XR219 leaves a trail of black smoke across the Wiltshire sky halfway through the first flight. BAE Systems via Warton Heritage Group
Short final to runway 28, with the BAC support crew and vehicles visible on the ‘peardrop’ area above the aircraft. The tent was the only shelter available to BAC personnel working on this area of the airfield. BAE Systems via Warton Heritage Group
Having climbed to 6,000ft (1,800m) with nothing falling off or blowing up, Beamont levelled off and continued with the flight as planned, entering a left-hand turn to proceed back towards the airfield via Salisbury, and flying past to the southeast before turning left abeam Stockbridge to make a pass over the airfield in the same direction as the initial take-off, having descended to 1,700ft (500m) to do so. From this pass he turned left once more to enter the circuit pattern and carry out his landing. No difficulties were experienced with lining-up on the runway and holding the approach path, though the aircraft was a little oversensitive in pitch. Approach speed was 200kt (230mph; 370km/h) with 35 degrees of flap, reducing to 190kt (220mph; 350km/h) at 250ft (75m), 182kt (209mph; 337km/h) at 50ft (15m) and flaring at around 30ft (9m) to give a touchdown speed of 174kt (200mph/322km/h). With just about everything going smoothly up to this point, it was a rude shock to Beamont when there was a sudden violent vibration from side to side just at the point of touchdown, sufficient to cause him to lose vision almost completely. Continuing to hold the nose-wheel off for aerodynamic braking, as discussed earlier, Beamont pulled the ’chute just a little too early, at 155kt (178mph; 287km/h), forgivably so given the mayhem he had just experienced, and the beak door once again impacted the base of the fin, denting the door. (A permanent modification using an hydraulic damper was later introduced to cure this problem. Meanwhile, a simple temporary fix was the work of true British genius; a metal strip clamped over the hinge slowed the opening of the door while the strip was being bent out of shape. It was easy to replace the deformed strip for the next use.) Wheel braking came in at 70kt (80mph; 130km/h) and the parachute jettisoned at 40kt (46mph; 74km/h), the aircraft rolling to a stop after using 2,050yd (1,875m) of runway. The aircraft had been airborne for just over 14min, but it was a successful first flight. An extract from Beamont’s report on the flight reads as follows:
Due to virtually complete serviceability this first sortie was carried out in full accordance with Flight Test Schedule No. 1. Stability and response to controls was found to be adequate and safe for flight under the test conditions, and to conform closely to predicted and simulated values. In this 1st flight configuration and under the conditions tested, this aircra
ft could be flown safely by any moderately experienced pilot qualified on Lightning or similar aircraft, and the flight development programme can therefore be said to be off to a good start.
Safely on the ground, much to the relief of everybody involved. The large braking parachute is seen here at fullest extent; on several occasions later in the flight test programme the parachute split open, and some thought was given to limiting it permanently to the smaller reefed deployment. BAE Systems via Warton Heritage Group
Roland Beamont and his navigator, Don Bowen, enjoy a well-earned cup of tea while the assembled press record their reactions to the successful first flight.
Such an extract was, of course, just the sort of thing BAC needed to send to the Ministry, but the long list of defects that needed to be dealt with before a second flight (most of which already existed before the maiden flight) was another matter entirely. The amount of instrumentation and data recording being carried out during the flight was unprecedented, and BAC also now faced the task of analyzing over 27 miles (43km) of tape recordings holding over 61 million data points.
Flying XR219 had been a huge gamble (and did not improve the ETPS’s opinion of BAC’s flight test organization), but it had paid off, relieving some of the pressure for visible progress. The company now had a lot of work ahead to get XR219 to a safer standard for continued flight testing (more than eighty modifications were planned), and to repair XR220. It had been planned to fly the latter aircraft in November, but repair was complicated because much of the damage was to the jacking points. Thus the aircraft could not be jacked up easily, and the repairers would initially have to work round this limitation. The slow delivery of required components to complete the build tasks outstanding on XR220 (which would have been necessary with or without the delivery accident) also frustrated progress, and on several occasions XR220 was even used as a ‘Christmas tree’, with parts being removed for use on XR219. The flight date for XR220 slipped back to December and then further, into early 1965.