Not Much of an Engineer

by Stanley Hooker

  With the new breed of HBPR engines which became visible in the first half of the 1960s the first thing to do was to take an existing type of core and add an extra turbine downstream, driving via a long shaft to an enormous fan at the front. This fan, in essence a multibladed propeller, runs at a tip Mach number of about 1.5, so most of the area of each blade is supersonic. It has a design pressure-ratio of about 1.5, so it supercharges the entire airflow entering the core engine by this amount. Thus, if the core had a pressure-ratio of 16:1, a figure within previous experience, the overall engine pressure-ratio became 24:1 when supercharged by the fan, giving a ‘boiler efficiency’ higher than ever before achieved.

  So far we have increased the pressure-ratio and reduced the jet velocity. The third objective was a further major improvement in ‘boiler efficiency’ gained by running much hotter than ever before. A combustion temperature of about 1,250°C was sought, roughly 150° above the current state of the art.

  These were the basic objectives of all the new HBPR fan engines, which were pioneered by the TF39 built by General Electric for the Lockheed C-5A Galaxy for the USAF, and by the Pratt & Whitney JT9D produced mainly at company expense for the Boeing 747. It was clear that such engines would have an enormous impact, especially on commercial aviation. Their great size and power made possible so-called ‘wide-body’ airliners carrying more than twice as many passengers as their predecessors, for unprecedented long ranges, and with an enormous and welcome reduction in both specific fuel consumption and noise.

  By 1967 Boeing’s chief rivals, Lockheed and McDonnell Douglas, had offered the airlines large trijets to meet the specifications of American Airlines and other US trunk operators. These needed engines in the 33,000 lb class, compared with 40,000 lb for the C-5A and 747, but the aircraft grew in weight and more thrust was needed.

  Sir Denning Pearson (Psn), Managing Director and Chairman of Rolls-Royce, and commercial director David Huddie decided to get into the US wide-body market. The RB178 had been developed for the European Airbus A300 into a very large and powerful engine, the RB207; but this was never built because Psn ordered the entire resources of Derby to be devoted to a smaller engine, the RB211, sized to the requirements of the Lockheed L-1011 TriStar. The final contractual commitment was that RB211 engines would be delivered at a fixed price from September 1971, fully certificated at 42,000 lb thrust. In March 1968, on this basis, the L-1011 went ahead with large launch orders from Eastern Airlines and TWA, plus a specially formed British financial group called Air Holdings whose purpose was to underwrite 50 of the US jets for sale to non-US carriers.

  General Electric, in fact, not only secured the propulsion contract for the McDonnell Douglas DC-10, almost a duplicate of the TriStar, but also launched the development of a much more powerful version of their basic CF6 engine, the CF6-50. This fitted the Airbus A300B and also enabled McDonnell Douglas to offer a much heavier DC-10 for long-range routes. Rolls-Royce could see that this would lead to DC-10 sales far outstripping those of the TriStar, but had their hands so full with the original RB211 that they had no hope of funding a more powerful version.

  Thus, fairly soon, Rolls-Royce was in an inferior marketing position. They were striving to build an engine of 42,000 lb thrust in competition with a much larger US rival who was offering engines at thrusts up to 49,500 lb. Moreover, whereas both the American rivals had extensive running experience, with the JT9D on the verge of airline service in the 747, the RB211 existed only on paper. Rolls-Royce had additionally been compelled by market pressure to abandon the policy laid down by Hs that no engine should be sold for airline service without extensive military experience. It is ironic that when the RAF got its first RB211 engines in 1983 they were secondhand ex-British Airways!

  Obviously, the RB211 was a much greater challenge than Rolls-Royce had ever previously faced. The operating pressures and temperatures were outside the company’s previous experience, and the size of the fan blades, casings and rings was more than double anything seen at Derby previously. There were two key features of the RB211 that could be, and were, used by Rolls-Royce to sell the engine to Lockheed instead of a US rival, and used by Lockheed to justify their choice to the airlines. One was the use of three shafts, and the other was the Hyfil fan blades.

  The reason for the three-shaft, or three-spool, layout appears later in a Technical Appraisal I wrote in the crucial month of February 1971. By that time the Hyfil fan blade, on which so much hope had centred, was a thing of the past.

  This was a bitter blow not just to Rolls-Royce but to Britain. In the early 1960s British workers at the RAE at Farnborough had pioneered the conversion of polyacrilonitrile yarn into fibres of pure carbon so strong and stiff, yet extremely light in weight, as to open up a whole new world of structural possibilities. Carbon fibre became all the rage, one of the main planks in what Prime Minister Wilson called ‘Britain’s white-hot technological revolution’. But actual applications were hard to find until Rolls-Royce decided that carbon fibre material was ideal for the RB211 fan. The fan was huge by previous standards, over 7ft in diameter. Each pound saved in the mass of its blades could be multiplied many times in reducing the weight of the hub, bearings and main engine frame. Aerodynamic calculations showed that with a sharp-edge ‘lenticular’ profile the blades could reach surprisingly high efficiencies, at pressure ratios of 1.5 to 1.7 and tip speeds over 1,500 ft/sec (Mach 1.5).

  To make each blade vast numbers of strong carbon fibres, thinner than human hairs, were bonded by resin adhesive into thin flat plies which after being cut to shape were stacked in a pile and bonded into one solid mass. The blades were smooth, black and amazingly strong and light. They enabled the RB211 to give the TriStar considerably greater payload than the rival engines with metal fans.

  On test the Hyfil blades worked well, and enormous money and effort went into their development and tooling up for mass production. But when prototype fans and complete RB211 engines were subjected to the mandatory birdstrike test, in which 4 lb chickens from the supermarket are fired at high speed into the engine running at take-off speed, the blades proved to have an Achilles heel. The weak resin plastic could not take the sudden impact on the thin leading edge.

  As a fallback programme an all-titanium fan had been designed, and with great reluctance this had to replace the new-technology Hyfil construction. At a stroke this wiped out the engine’s weight advantage over its rivals.

  As for the vital HP turbine, this again went well beyond the previous state of the art. Derby had been one of the pioneers of aircooled turbine blades in the Conway bypass jet designed in the mid-1950s. All Derby blades were forged in a Nimonic high-nickel alloy, able to retain adequate strength at about 900°C. In the Conway the peak gas temperature was about 1,100°C, so the cooling air holes and passages inside the blade had to keep the blade at least 200°C cooler than the surrounding gas. In the RB211 the gas was going to be at 1,250°C. This may not sound a very big increase over 1,100, but the strength of the blade is halved for every extra 25°C!

  To make a blade at Derby the rough metal bar from which the blade would be fashioned was first pierced by a large number of long but small-diameter drilled holes. Just where these had to be was found by a laborious process of trial and error, so that, after the rough blank had been forged, the holes were all in the correct positions. To forge the blade it was squeezed while almost white-hot between two dies having the correct shape to form the blade. Finally the whole blade was machined all over to precisely the correct size and profile. The price of each blade was inevitably very high, and the whole process was an engineer’s nightmare because microscopic variations between the cooling air holes in successive blades meant that the cooling would be uneven.

  At Bristol we had abandoned forged blades soon after 1960. Instead we cast our blades using the famous lost-wax process used 500 years earlier — by da Vinci, for example — to produce precisely repeated sculptures. Casting enabled us to use materials basically stronger than Nimo
nic alloys, too hard to be forged or machined. We cast our blades to the finish dimensions, with a mirror finish, without any machining apart from precision grinding of the root. Cast blades proved to be much stronger in tension than forged and machined blades, and despite their brittleness and susceptibility to thermal stress we proved on the Proteus that they are superior. On the Olympus 593 and Pegasus we produced large numbers of eminently reliable blades with radial air cooling passages operating at some 1,250°C, much hotter than any production engine at Derby at that time.

  Derby was aware of the Bristol experience, and planned to switch to such blades eventually, but designed the RB211 to use forged and machined Nimonic. Predictably, the HP turbine blades were prone to failure. Yet a further problem was that, though potentially an advantage, the three-shaft layout posed severe problems in oil leakage and bearing failure.

  This all added up to plenty of problems, but far more serious than these was the fact that, when the first RB211 engines ran, their performance was way down and totally unacceptable.

  As the months went by in 1970 it became obvious to me that Rolls-Royce was in trouble over its contract with Lockheed, and yet this was the contract on which the company’s future was to be based.

  On the financial front things went from bad to worse. On a really successful programme the money paid out on development and getting into production is so enormous that it takes something like 15 years for the programme to move from the red into the black. At the fourth or fifth years many millions of pounds have to be spent on production tooling, raw materials and finished parts which begin to fill the stores with ‘work in progress’. But if there is any minor hiccough in development, production is thrown into utter confusion making engines and parts that cannot be delivered to the customer, while the stores fill up with costly parts of which some or all will have to be scrapped.

  By mid-1970 it was clear that Derby was deep into this situation and would be unable to fulfil its contractual obligations to Lockheed. The company’s enormous assets were no help, because what it was fast running out of was cash and credit to pay the wages and the bills of the thousands of subcontractors and suppliers. The government saw all this and in 1970 at last had to act by removing Psn and Huddie and appointing Ian Morrow to take charge of finance. Hugh Conway was appointed managing director at Derby, and Lord Cole, ex-chairman of Unilever, was appointed caretaker chairman. I pressed Hugh to reorganise the engineering set-up, because obviously it was the engineering problems that had to be put right first. I recalled Hs’s axiom ‘If the engineers are wrong, then we are all wrong.’

  Finally at the end of 1970 Hugh agreed that I should go to Derby to survey the situation. To give him his due, he also insisted I should be accorded the status of Technical Director at Derby. I flew up to East Midlands airport and was met by car and taken to the colossal ultra-modern main office block at Moor Lane, which I had never seen. I was filled with trepidation. Would I be able to do anything useful? Would I be accepted back into the fold?

  When I arrived at Moor Lane I was greeted with delight by my old friend Chief Commissionaire Phillipson, who looked more imposing and immaculate than ever. As he ushered me into the main foyer I could hear the voice of Hs saying ‘Go to it Stanley, tell us what we have to do and we will do it as easy, as easy, as easy.’ When I later went in to lunch, everyone crowded around to welcome me, and Freddie Morley said, ‘Welcome back, you old bugger. It has cost us 63 million to get you, but it’s worth every penny’. He was, of course, referring to the price paid for Bristol Siddeley.

  I had been allocated a small empty office in the main engineering block. The first thing I did was to call in the performance engineers to give me a run-down on the reasons for the shortfall in engine performance. I was disturbed at the lack of data, and the scrappy nature of the analyses. They claimed that the efficiency of the HP turbine was 65 per cent. To that I said ‘Rubbish, turbines can’t be made that bad. It takes a genius to get above 85 per cent but it also take a lunatic to make one worse than 75 per cent.’

  I then enquired about the speeds (rpm) of the three shafts. Again I was bemused to find that these were way off the design values. It is always possible to adjust shaft speeds by altering the areas of the nozzle guide vanes between the various turbine stages which swirl the gas into the turbines. We went to work and calculated the changes in NGV area necessary to achieve the correct shaft speeds. These modifications were instructed to the shops. We got the test results of the first modified engine at the most opportune moment, in mid-February 1971. What follows is the first Technical Appraisal which I delivered to Hugh Conway in February 1971, very slightly abridged.

  Technical Appraisal

  1. The RB211 is a high bypass ratio, high-compression ratio, high-temperature fan engine competitive with the Pratt & Whitney JT9D and the General Electric CF6 engines made in the USA.

  2. The take-off thrusts at which the three engines are on offer today are: JT9D, 43,500-45,000 lb; CF6, 40,600 lb for the DC-10 and 49,500 for the A300B Airbus; RB211, 42,000 lb for the Lockheed 1011 TriStar. The specific fuel consumptions and weights of the three engines are competitive. The comparative order of basic size or air consumption is: CF6, 1.0; RB211, 1.10; JT9D, 1.20.

  3. The RB211 is unique in that its fan, intermediate (IP) compressor and HP compressor are each driven independently by its own turbine. This arrangement allows the three components — fan, IP compressor and HP compressor — to be driven each at its own optimum speed, and thereby improves the overall aerodynamic efficiency and flexibility. In particular, the independence of the fan speed, and the ability to raise this speed, the fan being the major thrust-producing component in the engine, allows further thrust growth as the turbine entry temperature is raised by future development.

  4. As a typical example, the desirable speed relation between the three spools of the RB211 is, in round numbers; Fan, 3,600 rpm; IP, 6,800 rpm; HP, 10,000 rpm.

  5. The situation in regard to the thrust of the RB211 as it existed in January 1971 vis-a-vis Lockheed was:

  (a) A contractual obligation to produce 40,600 lb take-off thrust up to a day temperature of 84°F.

  (b) Because of weight growth during the manufacture of the prototype L-1011, and because the RB211 was itself overweight (38,441 lb for a ship set as against the original guarantee of 34,566 lb), Rolls-Royce undertook to produce a thrust of 42,000 lb up to 84°F. This proposal was discussed with Lockheed.

  (c) Notwithstanding the weight growth of the aircraft and the engines (much larger, in fact, on the aircraft) it was the opinion of Rolls-Royce that an acceptable thrust for entry into service of the L-1011 would be 38,500 lb up to 84°F. This proposal was not discussed with Lockheed. Summarising: 40,500 lb was the contractual commitment, 42,000 lb was offered, and 38,500 lb was considered acceptable by Rolls-Royce for entry into service.

  6. For the start of the first prototype L-1011 flight programme, Rolls-Royce delivered to Lockheed five Batch 1 engines rated at 34,000 lb at a Turbine Entry Temperature (TET) of 1,167°C. For the second prototype, Rolls-Royce delivered six Batch 3 engines rated at 34,200 lb at a TET of 1,202°C, with a contingency or emergency rating of 36,200 lb at a TET of 1,232°C. These engines flew in the second prototype on 15 February 1971.

  7. At this stage, it was clear that the engine performance was sub-standard, in that the TET was too high for a given thrust. Modifications were, therefore, made to engine No. 100011, and on test at Derby in early February 1971 the following figures were demonstrated: 37,000 lb at 1,167°C TET; 39,430 lb at 1,227°C.

  8. On another engine, a new design of HP turbine blade discharging its cooling air from the trailing edge was tested, and this improved design reduced the TET at a given thrust by 28°C, or increased thrust at a given TET by 2,000 lb. It is intended to cast this type of blade by the lost-wax process, and to conduct further testing on the bench in October 1971, with a view to incorporating into production engines in mid-1972.

  9. Following the successfu
l improvements demonstrated on Engine No. 100011, a further change was been made to the LP nozzle guide vane areas, and this is predicted to give 41,500 lb at 1,227°C. This engine will be on test on 17 February 1971.

  10. Summarising the performance results (all quoted for the test bench on a Standard Day):

  Batch 1 Engine 34,000 lb at 1,167°C.

  Batch 3 Engine 34,200 lb at 1,202°C.

  Engine 100011 37,000 lb at 1,167°C.

  Engine 100011 (February 1971) 39,340 lb at 1,227°C.

  Engine 100011 with modified LP NGV (predicted) 41,500 lb at 1,227°C.

  Engine 100011 with cast blades (predicted) 43,500 lb at 1,227°C.

  11. Actual thrust developed depends directly upon TET. It is anticipated that a TET of 1,227 ± 20°C can be adequately cleared by bench running.

  12. In the future, the new cast turbine blade will not only be better aerodynamically but will be cast in a material (MARM 202) which has better high-temperature properties and will allow at least 50°C increase in TET. This will increase the thrust at 1,277°C to 47,000 lb on a standard day, and this figure must be regarded as the potential growth in thrust of the RB211 without major redesign.