Thus, we were finally left with just the Britannia, which was an eminently sensible aeroplane and a task of great responsibility. We did over a year of flying on the Proteus 2s but in August 1953, just three years after the Proteus redesign was agreed, we got four flying in the second Britannia. At last we had an excellent engine, every part of which had been redesigned on my instructions apart from the reduction gear. This was an 11:1 ratio gear with straight-tooth spur gearwheels to reduce the 11,000 rpm of the power (LP) turbine to the 1,000 rpm of the big DH propeller. It was the only part of the original engine I liked, and it had never given a moment’s trouble. Plumb kept reminding me of our troubles at Derby with the Trent, Clyde and Dart turboprops, and how we had had to change to helical gears. I kept re-examining the design, and could see no reason to change, though I did authorise Plumb, as an insurance policy, to get a set of helical gears designed and tested.
One morning in February 1954 Doc Russell visited me with a delegation from KLM, potential customers for the Britannia which at this time appeared to have the world at its feet. I briefed them on the engine, and then they went off to fly in the No. 2 aircraft. Would I like to come? Not today, Russ, because there were things I had to do, and I had flown in the aircraft anyway. So off they went, and less than an hour later shiny G-ALRX was a wreck on the mudflats of the Severn estuary.
Cruising over Herefordshire one engine’s input pinion, in the reduction gear at the front end of the long propeller shaft, had stripped its teeth. Freed of all load, the big LP turbine had oversped, almost instantaneously breaking from its shaft and flying with the energy of a battleship’s shell through the cowlings and arriving in fragments on the ground. Mercifully it missed the fuselage and the wing tanks, but it did pass through the engine oil tank and the oil caught fire. The fire raged for 19 minutes as chief test pilot Bill Pegg headed back to Filton, but with only a few miles to go he judged that the wing spars might soon burn through and with great skill did a belly landing. The mud instantly covered the windows, and Russell thought they had actually gone deep under water. Fortunately nobody was hurt, but the aircraft was a write-off, largely because of misguided efforts to pull it off the mud using nothing but steel cables.
The telephone rang in my office a little after 10.30 that chilly morning and I was told of the crash, and of the fact that it was clearly caused by engine failure. My feelings can be imagined. A short time later the ‘phone rang again: ‘Lord Hives wishes to speak to you’. The familiar voice said ‘I hear that you are in trouble, Stanley. Do you want any help from us?’ My spirits soared as I stuttered, ‘Yes, please.’ ‘Right,’ said Hs, ‘I will send down the First Eleven, who do you want?’ I had the nerve to say Rubbra, Lovesey, Lombard, Howarth and gear-expert Davies. Next morning they were poring over the failure in my office. That was the last time I spoke to Hs, and I am very glad it happened. It was typical of the man that, though by this time he was calling me his ‘one great failure’, he never hesitated in his act of superb generosity to a competitor in trouble.
When we surveyed the failure it was clear that a resonant vibration had been set up which had swiftly eroded the gear teeth. Spur-gear teeth engage one by one with an impulsive force, and when the frequency of tooth engagement is equal to the natural frequency of vibration of any part of the gear train, trouble will result. Helical gears, with diagonal teeth, engage more smoothly with a sort of shearing action, and are free from this trouble. So Plumb’s wisdom was fully vindicated. Thanks to him we had several sets of helical-tooth gears on the shelf, and we were able immediately to build these into Proteus 3 engines and go on testing. Once these were fitted we had no more gear problems, though as a double safety insurance we added an automatic fuel cut-out which shut the engine down instantly if the torque in the LP drive-shaft fell to zero, as it would upon failure of the reduction gear. This device has never been triggered in many millions of Proteus 3 running hours to the present day.
We had no trouble Type Testing the first production model of Proteus 3, the Mk 705, for the Britannia 102 airliners for BOAC in 1954-55. But though the basic engine was fine, it was still able to bite us once more, in a totally unexpected manner. In 1956 a Britannia 102 was on final tropical clearance, flying over central Africa. Passing through clouds at about 20,000 ft the engines suddenly began to hiccough and stop, one by one. The Britannia drifted gently down, and the captain had no difficulty getting all engines restarted at a lower altitude; but what was the cause?
It took us a little while to identify what the problem was. With the aid of television cameras mounted inside the engine cowlings we were able to see that ice could adhere to the inner walls. When it reached sufficient thickness, it could break away and enter the engine compressor. The quantities were so large as to produce a compressor stall, or extinguish the flame in the combustion chambers, or cause both these effects simultaneously.
It proved difficult to locate the atmospheric conditions under which the ice could build up in this way. The conditions were not the same as those under which ice formed on the aircraft wings, so that the test aircraft was not always successful in locating an iceladen atmosphere. In the tropics where the sea level temperature is about 40°C, the ambient temperature reduces to zero at about 20,000 ft. At this altitude ice crystals can be found in the air because of the enormous amount of water evaporation at sea level. BOAC could easily have avoided these conditions by deliberately choosing to fly at a different altitude but again we were not able to persuade them to do this.
We quickly devised glow plugs which were fitted in a number of combustion chambers. These were heated by the burning gases and, when the flame was momentarily extinguished, the plugs continued to glow long enough to re-ignite the fuel/air mixture. Thus the engine suffered only a momentary bump, and full cruising power was restored at once.
These glow plugs were entirely effective and never failed to relight the engine. The aircraft was totally safe, and the only evidence of severe engine icing was a brief flicker of the cockpit instruments. However, we were unable to persuade BOAC to put the Britannias into passenger service with this palliative. The major flight test programme thus continued for two years to find a means of preventing the build up of ice inside the cowlings. Eventually a series of modifications to the cowling ducts overcame the problem, and the aircraft had a long and profitable service life.
Such was the apparent ignorance of our atmosphere that BOAC’s main argument was ‘If it ices up in the tropics, what will it do over the North Atlantic?’ We were forced to spend almost two years flying about looking for ice despite the fact that the problem had already been encountered in the only place where it existed. In cold regions, with sea level temperature of 0°C, there is extremely small evaporation, so there is little ice and that is near the ground and no problem to the Britannia. In temperate regions the icing is at around 7,500 ft. But in the tropics, with sea level temperature up to 40°C, there is massive surface evaporation and roughly eight times more atmospheric ice, and it is found at just the 20,000 ft level at which the Britannia cruised in such a climate. Piston airliners at much lower levels had never encountered it, and neither had the jets at around 40,000 ft. Even the Britannia could avoid it by flying slightly lower or slightly higher, with hardly measurable penalty in fuel burn, and in any case switching on the igniter plugs eliminated the problem even at the worst icing levels. BOAC seemed determined to turn what was really a molehill into a mountain, until the Britannia had become of such interest to Fleet Street that one observer commented ‘If it blows a fuse we hold the front page.’
The impasse between BOAC and Bristol delayed the entry into service of this splendid aircraft by more than two years, and gravely damaged its propects in the export market. In the meantime the stocks of aircraft and work in progress at Bristol caused a cash-flow crisis which came within an ace of bankrupting the whole company. Only the cool courage of Verdon and the financial skill of William Masterson kept the company afloat.
Suddenly
Israel’s airline, El Al, came to the rescue. They simply ordered three long-range Britannias and put them into widely publicized service between Tel Aviv and New York. On the first supposed direct flight out of New York we waited with bated breath to see if they would land at Rome to refuel. They didn’t. Next day the New York Times had a vast ad showing a big world and a much smaller world, and the headline ‘Yesterday our Earth became this much smaller’.
Subsequently the Britannia had the reputation of being the fastest, smoothest, quietest and safest propeller aircraft in the world. It gave many years of service to BOAC, the RAF and a small number of other operators, but the delay caused by BOAC was just sufficient to cause many other important airlines, such as TWA, to wait for the shorter-ranged but faster 707. Only about 80 Britannias were sold; just enough to save the Bristol Aeroplane Company. The programme was not helped by the decision of BEA to ask Vickers to build a duplicate of the Britannia, the Vanguard, which sold only half as many and proved a financial disaster to Vickers. This certainly converted me to the belief that there had to be mergers in the British aircraft industry.
As for the Proteus, the years of toil and tears receded as it established itself as a reliable and efficient engine. But I shall hate it to my dying day because of the way, whenever we thought we had the problems licked, a new one would suddenly emerge — at the worst possible moment. Yet this engine pioneered two non-aero uses for powerful gas turbines: propelling warships and generating electricity. It was Peter du Cane of the Vosper Company who came to see me about putting three Proteus into motor torpedo boats he was designing for the Royal Navy. He reckoned with 10,500 hp he could get a speed of 55 knots (well over 100 km/h). Three of these Brave-class boats were built, and with them we learned how to instal aero engines in ships, with sea-water separators in carefully designed inlets and jetpipes facing aft to add thrust. We changed many materials to avoid corrosion, and W. H. Allen of Bedford did a new reduction gear to drive the propeller shaft. The whole installation worked so well that 20 years later major navies changed over to gas turbines, with Rolls-Royce (having absorbed our Bristol team via Bristol Siddeley) right in the lead.
As for electricity generation, it was Bill Irens, by now Chairman of the SW Electricity Board, who saw the advantages of using on-the-spot gas turbines to make up local shortfalls in power supplied by the grid. The very first gas-turbine station was built at remote Princetown, near Dartmoor Prison, to feed 3.5 MW (megawatts) into the grid at times of peak demand. The station was unmanned, and was simply started and brought on line by a telephone call from Bristol. When it was getting short of fuel or oil it automatically rang up the controller! From this stemmed hundreds of bigger stations powered by single or grouped Avons, Olympus and many other gas turbines.
In turn, this led to worldwide use of high-power gas turbines for many industrial applications. Chief among these is pumping oil and gas along great pipelines.
I must, however, mention one good thing the Proteus did for me. We very nearly sold many Britannias to Trans-Canada and Canadian Pacific, and I spent much time in Canada on lecture tours. Each time I went to Montreal Dr Eric Warlow-Davies would throw a dinner party for me. Warlow came from Tasmania as a Rhodes Scholar, and was at the Oxford engineering school with me. Tall and spare, he was a determined disciplinarian who could inspire those around him. He joined the LMS railway at Derby, but at the beginning of the war I weaned him away to Rolls-Royce where he soon established himself as a brilliant mechanical engineer whose forté was investigating failures. No scratch or frettage was too microscopic to escape his attention, and he possessed the sixth sense of being able instantly to recognise whether such a mark would lead to a dangerous failure. In 1942 he was sent as Quality Engineer to the Merlin factory near Glasgow, and four years later went with Denning Pearson to Canada to oversee the Merlins in DC-4M airliners. In the early 1950s he set up the Rolls-Royce (Canada) factory outside Montreal to build Nenes for T-33AN Silver Star jet trainers. He did not like this, because running the whole plant cut him off from his first love, engineering.
At about midnight, after one of our (highly liquid) dinner parties, Warlow asked me if I could fit him into the engineering team at Bristol! I welcomed him with open arms. He joined us in early 1954, and our new managing director, Air Chief Marshal Sir Alec Coryton, instantly took to him. They shared a brusque manner and a passionate interest in veteran motor cars. It was agreed that Warlow would become Chief Engineer for current programmes, then the Proteus and Olympus, while I became Technical Director to work on future projects and such theoretical aspects as performance, research and planning — and notably also the icing problem on the Proteus. It was an harmonious arrangement, and Warlow lifted a giant load from my shoulders not only with the Proteus but also with getting the great Olympus into service with the RAF.
Chapter 8
The Olympus
When I arrived in Bristol in 1949 the Division had just completed the project design of the Olympus, an axial turbojet more powerful than any previously built in Britain. The project had already passed to the Main Design Office, supervised by two good designers, Alec Henstridge and Sam Blackman. I was relieved to find that I thoroughly approved of the basic design, though I knew from my Ilkley studies that it was rather large and heavy for the design thrust of 9,750 lb.
It was the first real two-spool engine in the world. The Rolls-Royce Clyde was the first two-shaft engine, having an LP axial compressor followed by an HP centrifugal, but the Olympus was the first to have two axial spools in series, each driven by its own turbine. As in the Proteus, the LP drive shaft passed down the centre of the HP spool. The advantage is that the overall pressure-ratio can be higher than anything possible with a single spool (in 1949), and I have already explained in the Axials story why it is not possible just to go on adding axial stages to the same spool. When a two-spool engine is started, the starter accelerates the HP spool only, and this runs up to speed and gets a good airflow going. The LP system accelerates much more slowly, and this eliminates the tendency for the early stages to stall.
To achieve a pressure-ratio of 12:1 one can take an LP spool working at 3:1 and add an HP spool with a ratio of 4:1. Compressors of such modest pressure-ratios were even then known to be stable and efficient over the whole running range from idling to take-off. But an additional problem is the presence of a sluggish boundary layer along the walls of the compressors, particularly on the inner surface at the roots of the blades. As the air passes from stage to stage through the compressor the layer starts at the inlet at zero thickness and gets thicker and thicker. The more stages of compression there are, the thicker the sluggish boundary layer, and the more does it restrict the effective outlet area. With a two-spool engine the boundary layer can be extracted between the spools, and the HP spool can start off with zero boundary-layer thickness, resulting in more efficient and more stable compression.
Bristol Engine Division was developing the Olympus at Ministry expense to power the great Vulcan bomber, the monster tailless delta whose project design owed much to Sir William Farren at A. V. Roe, who as W. S. Farren had taught me at Imperial College. The importance of the Olympus to the RAF and the UK deterrent force was obvious, but to me it was clear that this super turbojet would at one great leap put Bristol at least level with Rolls-Royce. I had plenty of time to study its design, and though the concept was excellent there were many details I did not like. I suggested to Frank Owner a number of possible detail changes, and ran into a brick wall. Frank would say ‘That is a matter for the Design Engineer’, and on each occasion I found Stanley Mansell unreceptive and obstinate. But at least the prototype engines were built quickly, and the first was ready for test in mid-1950.
When the first run was a few days off, an American, Roy T. Hurley, came to see me. He was then President of the mighty Curtiss-Wright Corporation, the major engine contractor in piston-engine days to the entire US Army and Air Force. Even in 1950 Wright was still pouring out vast numbers of powerf
ul Cyclone engines in various sizes, and was getting ready to mass-produce the pinnacle of aircraft piston engines, the Turbo-Compound. This was to enjoy a long run as top engine on the world’s long-haul air routes, quite apart from big military orders, but by 1950 Hurley and his board were in what they rather suddenly saw as a dangerous position. Unlike rival Pratt & Whitney they had not been dragooned into making turbojets, and their position was that they had millions of dollars in the bank but no capability in the coming field of gas turbines.
Hurley was ex-Ford Motor Company, and a very strong and energetic man. He had formed the view overnight that he must buy time and technology by getting a licence in Britain. Forthwith he had booked on the Queen Elizabeth and, Rolls-Royce being tied up with Pratt & Whitney, he had come direct to Bristol. We discussed the Olympus, and at the end of the day he asked if he could take out a licence. I took him along to Norman Rowbotham, not really believing he meant it. They got down to brass tacks, and Hurley then departed, saying he would return in two or three days to sign the documents.
The very next day I heard with horror through the industry ‘grapevine’ that Hurley had gone straight to our competitor Armstrong Siddeley and was negotiating a licence for their Sapphire turbojet, which was running neck-and-neck with the Avon. Rowbotham called Hurley in high dudgeon, stating that we did not do things like that in Britain, leading a company up the garden path and then nipping off to one’s competitor. Hurley replied ‘It’s OK, I’m going to buy both licences’. And he did just that.
Not Much of an Engineer Page 18
