PART TWO: POWER AND ENGINEERING 322 1920–1945: The high speed compression ignition engine Compression ignition engines for application to transport require a compact, lightweight, high speed (piston speeds exceeding 7.6m/s; 1500ft/min), high efficiency internal combustion engine, with solid injection of the fuel. Such an engine was developed between 1920 and 1930. At the time it was a widely held belief that high speed in the compression ignition engine was impossible, because the combustion processes were too slow. In addition, Diesel had found, and experience in the 1920s confirmed, that solid injection methods required careful study, otherwise the engine ran roughly, with poor fuel consumption. The requirements for a successful high speed, solid injection engine were elucidated as a result of careful studies of the details of the combustion process. One of the most important contributions to this work was H.R.Ricardo. He found among other things that careful control of the air motion in the cylinder was important, consequently the introduction of solid injection coincided with development of many different types of combustion chamber. 1945–1986: Power, efficiency and cleanliness In the years following the Second World War the history of the compression ignition engine has been dominated by the development of engines of ever- increasing specific power output (kW/kg or hp/lb mass), which has been due to the rising mean effective pressure as a result of the universal adoption of turbo- charging (Figure 5.16). Consequently, attention has to be paid in design to handling high mechanical and thermal loads. The design of the compression ignition engine has settled into a pattern where most low speed (less than 250rpm), large stationary or marine engines operate on the two-stroke cycle. High speed engines, with speeds in excess of 1200rpm are four-stroke, and medium speed engines (250–1200rpm) usually operate on the four-stroke cycle, but occasional examples of engines using the two-stroke cycle can be found (e.g. General Motors locomotive engines). Because the fuel consumption of the compression ignition engine is not as sensitive to pollution control measures as the spark ignition engine, a number of passenger automobiles were offered in the mid-1970s with the former. This interest seems likely to revive as oil supplies dwindle toward the close of the twentieth century. The need to develop a compression ignition engine with a high thermal efficiency has resulted, among other things, in a re-appraisal of the combustion chamber form. In particular, the direct injection combustion chambers, with their lower heat losses compared to the indirect injection chambers (e.g. Ricardo Comet), have been receiving considerable attention. However, prob STEAM AND INTERNAL COMBUSTION ENGINES 323 ably of greater importance is the clear indication, as has already occurred in the spark ignition engine, of a trend towards microprocessor control of the engine. Marine compression ignition engine Approximately half the world’s tonnage of shipping is now propelled by compression ignition engines, and the percentage is increasing. This form of power was first applied in 1912 to ocean-going shipping in the Danish Selandia. This was a 7500 tonne freighter fitted with two Burmeister & Wain four-stroke, single acting, eight-cylinder engines of 1340SkW (1800shp) each. There are a number of advantages in marine applications of the compression ignition engine compared to the reciprocating steam engine, (a) The oil fuel is easier to handle and store than coal: it can be carried in the Figure 5.16: Historical trend of the performance parameters of two- and four- stroke cycle compression ignition engines 1900–85. The curves are based on data in J.W.Anderson, ‘Diesel—Fifty Years of Progress’, Diesel Progress, vol. 14, May (1948), pp. 5–331, and Anonymous, ‘Fifty Years of Diesel Progress, 1935–1985’, Diesel Progress, vol. 37, 50th Anniversary Supplement, July (1985), pp. 12–182. PART TWO: POWER AND ENGINEERING 324 bilges, whereas coal bunkers use cargo space, (b) The oil fuel has a larger heating value than coal (about 25 per cent), and weighs less, (c) The compression ignition engine is about half the weight of a boiler and reciprocating steam engine of the same power. The 1920s was the heyday of the four-stroke single-acting crosshead engine with air injection, and probably the most important manufacturers of this type of engine at that time were Burmeister & Wain of Copenhagen and their licensees Harland & Wolff, the Belfast shipbuilding and engineering company. However, by the mid-1920s these two companies had turned their attention to the four-stroke cycle double-acting engine, and they jointly developed an experimental single cylinder engine that produced about 745SkW (1000shp), about 150kW (200hp) more than the then probable maximum value for a single-acting four-stroke engine. However, in spite of considerable engineering effort, this type of engine was never successful. The main difficulty was to design the complicated lower cylinder head to withstand the thermal stresses induced by its form. By the mid-1920s the two-stroke single-acting engine started to compete seriously with the four-stroke engine. The first engine of this type in an ocean- going vessel was fitted in the Monte Penedo, which was built for Brazilian owners, and had two 630SkW (850shp) Sulzer engines. By 1935 the two- stroke engine had essentially eliminated the four-stroke engine from consideration for marine use when engines of more than 1500kW (2000hp) were required. (By 1930 Burmeister & Wain and Harland & Wolff, established producers of four-stroke engines, were offering the two-stroke type.) Inevitably, demands for even greater power output directed the attention of owners and builders to the double-acting two-stroke engine. An engine of this type was first built by Krupp in 1909; this firm and Maschinenfabrik Augsburg-Nürnberg were each commissioned by the German navy in 1910 to produce a double-acting lightweight engine for use in warships. The MAN engine was a 8950SkW (12,000shp) machine with six cylinders. Work ceased on the project in 1914 and it was broken up in 1918 by order of the Allied Control Commission. However, this type of engine was revived in the ‘Deutschland’ class of pocket battleship (the Graf Spee was a member of the class), which were fitted with eight double-acting two-stroke solid injection nine-cylinder engines each rated for 4550BkW (6100bhp) at 450rpm. The double-acting two-stroke engine suffered somewhat from the faults of the four-stroke double-acting engine, except that the use of port scavenging reduced the number of valve openings required in the lower cylinder head. Although this type of engine was installed in ships up to the mid-1950s, it has now, because of the lower head problems, been displaced by single-acting and opposed piston types of engine. The opposed piston engine was introduced in 1898 as a two-stroke gas engine by W. von Oechelhäuser and was adapted to burn oil by Hugo Junkers. It had STEAM AND INTERNAL COMBUSTION ENGINES 325 been applied to ship propulsion in Germany without success, but in spite of this the British firm of William Doxford & Sons, marine steam engine builders, decided in about 1912 to investigate the possibility of using it for marine service. The first production engine (four cylinders, 2240IkW/3000ihp, 75rpm) went to sea in 1921 in the 9140 tonne Yngaren (Swedish Transatlantic Lines). The Doxford engine (Figure 5.17 (a)) has probably been one of the most successful marine internal combustion engines ever produced, but in spite of this, production ceased in 1980 (undoubtedly as a consequence of the shift in the centre of shipbuilding activity to the Far East). Opposed piston engines were also built by Burmeister & Wain- Harland & Wolff, and Cammell-Laird, among others. The development of the marine compression ignition engine since 1945 has concentrated on increasing the power output per cylinder, the thermal efficiency, and the ability to burn low-quality fuels. Two-stroke single-acting engines of this type (Figure 5.17 (b)) produce up to 3505bkW per cylinder (4700 bhp per cylinder) with a specific fuel consumption of about 157 grams/ BkWhr (0.258lb mass/bhphr) which for a fuel with a heating value of 44,184J/ gram (19,000btu/ lb mass) corresponds to a thermal efficiency of 52 per cent. This performance makes engines of this type the most efficient of all thermal prime movers. Wankel engine Rotary internal combustion engines in which all the processes of the cycle occur in one structural containment (or body) are attractive because they have the inherent dynamic balance characteristics of the turbine, but avoid the need to provide separate components for each of the cycle processes. The first attempt to produce an engine of this sort is probably due to an English engineer Umbleby who proposed the adaptation of a steam rotary engine that had been invented by J.F.Cooley of Alston, Massachusetts, in 1903. This particular engine was essentially the last in a long line of steam rotary engines that appears to have its origins in a proposal due to Watt in 1782, which was the last of various types of rotary engines considered by Watt and his partner Boulton between 1766 and 1782 in a frustrating, and eventually unsuccessful, attempt to obtain rotary motion without using the piston and cylinder. Rotary steam engines never successfully competed with the reciprocating steam engine or, for that matter, the steam turbine. In general the necessary technological sophistication, materials and financial resources were not available. This is confirmed by the development history of the most successful of the rotary internal combustion engines produced so far, the Wankel engine. Substantial resources have been devoted to the development of this engine since its original conception (1954), but it still (1988) cannot be said to be the equal of or superior to the reciprocat ing engine, except in regard to the previously noted characteristics of good balance and high power-to-weight ratio. PART TWO: POWER AND ENGINEERING 326 Figure 5.17 (a) Doxford opposed piston two-stroke compression ignition engine, c.1950. The figure shows two cylinders with a piston-type scavenge pump between the power cylinders. The arrows trace the path of the scavenge air through the pump, to the inlet manifold, through the intake ports, traversing the space between the pistons in a spiral uniflow pattern, and leaving through the exhaust ports into the exhaust manifold. In the right hand cylinder the pistons are at inner dead centre and their faces are so shaped that an almost spherical combustion space is formed. Reproduced with permission from W.H.Purdie, ‘Thirty Years’ Development of Opposed-piston Propelling Machinery’, Proceedings of the Institution of Mechanical Engineers, vol. 162 (1950), pp. 446–64). STEAM AND INTERNAL COMBUSTION ENGINES 327 (b) End section of a two-stroke, single acting, crosshead type compression ignition engine, 760 mm (30 in.) bore by 1550 mm (61 in.) stroke, with power output per cylinder of 1340kW (1800hp). This is actually a single cylinder experimental engine (type 1RSA76) built by Sulzer Brothers in the late 1950s. The engine is fitted with a turbo-charger (top left-hand) supplying air (through an after-cooler, shown dashed) to the inlet plenum located on the right-hand side of the engine. The lower side of the piston acts as a scavenge pump. Loop scavenging is used with incoming air introduced through the right-hand ports and the exhaust leaving through the left-hand ports. Super-charging is by the pulse (Büchi) method. Direct fuel injection is by the injector mounted in the cylinder head. The piston is water cooled with coolant supply and removal through the hollow piston rod and a swinging link connected to the crosshead. Reproduced with permission from A.Steiger, ‘The Significance of Thermal Loading on Turbocharged Two-Stroke Diesel Engines’, Sulzer Technical Review, vol. 51, no. 3 (1969), pp. 141–159). The rotary engine is an elegant concept that undoubtedly appeals to the aesthetic sense of the engineer. It has attracted a number of inventors, but the difficulties of translating the idea into a working engine have been such that only the engine conceived by Felix Wankel in 1954 has been developed to the point where it has been extensively used in automobiles (most notably by the Mazda Motor Corporation, formerly Toyo Kogyo Co. Ltd). In the rotary engine the piston and cylinder of the reciprocating engine are replaced by working chambers and rotors enclosed by a stationary housing. The rotor and the working chamber are specially shaped to allow the working substance that is trapped between these two components to undergo the processes of some internal combustion cycle. The exact shapes of the rotor and the working chamber depend on the kinematics of the linkage that connects the rotor to the drive shaft. In the case of the Wankel engine the working chamber is shaped like a figure of eight with a fat waist (actually an epitrochoid), and this contains a three-cornered rotor that orbits the drive-shaft so that all three corners are permanently in contact with the walls of the working chamber. Ports allow for the admission of the mixture (carburetted engine), or air (engine with in-cylinder injection), and the exhaust of the products of combustion. For a spark ignition engine the necessary spark plug is located at an appropriate point in the wall of the working chamber. When working on a four-stroke Otto cycle with spark ignition, the space between the rotor and the inner surface (the working surface) of the working chamber continually changes in size, shape and position. These volumetric changes provide for charge induction, compression, expansion and exhaust. The cyclic sequence of these processes is followed by all three faces of the rotor, so a single rotor engine is acting like a three-cylinder reciprocating engine. As in the two-stroke reciprocating engine, no valves are required on the inlet and exhaust ports, since the gas flow is controlled by the rotor motion (cf. PART TWO: POWER AND ENGINEERING 328 piston motion in the two-stroke reciprocating engine). This feature of the engine, among others, makes it mechanically quite simple. The Wankel engine had its origins in a rotary compressor that was invented by Felix Wankel between 1951 and 1954. He operated a research and development company that carried out contract work for various German engine manufacturing companies. He had little formal engineering training, but had served German industry since 1922. One of his most successful inventions prior to the Wankel engine was a rotating disc valve, invented in 1919, that replaced the conventional poppet valves of the reciprocating engine and required the solution of particularly difficult sealing problems; this background was probably of great assistance in the development of the Wankel engine, where the sealing at the rotor apex, its point of contact with the working surface, is especially critical. The German NSU company obtained the rights to the Wankel disc valve for use in racing motor-cycles, and at about the same time (1954) they also considered the use of a Wankel compressor as a supercharger. It was only a short and obvious step to investigate the possibility of adapting the epitrochoidal chamber and triangular rotor of the compressor for use as an internal combustion engine. NSU and Wankel then spent the period between 1954 and 1958 in an intensive development effort on the Wankel engine. Ultimately a car, the NSU Wankel Spider (1964–7), was produced. The Curtiss-Wright company in the United States acquired the American rights to this engine, as have a number of other companies in various parts of the world. The most successful of these licensees has been the Mazda Motor Corporation in Japan. This company has manufactured since 1967 a line of very successful cars using their version of the Wankel engine. Experience with the Wankel engine has shown that its fuel economy and emissions performance are not as good as those of the reciprocating engine. The former is strongly affected by heat losses (the rotary engine has a larger combustion chamber surface-to-volume ratio than the reciprocating engine), by friction at the rotor seals, and by difficulties in optimizing the inlet port configuration for all engine speeds. The Wankel engine has a lower oxides of nitrogen content in the exhaust gases, but more hydrocarbons than the reciprocating engine. Initially the high hydrocarbon level was reduced by oxidizing this component in a thermal reactor (‘afterburner’) located at the exhaust port, but more recently a two-bed catalyst in the exhaust system has been found to reduce the hydrocarbon content of the engine exhaust gases. The Mazda company appears to be still interested in the Wankel engine, so it appears likely that this type of rotary engine will continue to be built for the foreseeable future. Figures 5.14 and 5.16 summarize the progress in internal combustion engine performance from 1900 to 1986. The graphs refer to spark ignition and STEAM AND INTERNAL COMBUSTION ENGINES 329 compression engines respectively, with performance data for both two-stroke and four-stroke cycle engines included. The important effect of the fuel injection method and the use of supercharging on the performance of the compression ignition engine should be noted. GAS TURBINES The gas turbine appears to be the ideal prime mover, both mechanically and thermally, since it involves only rotary motion, with its consequent advantages, together with internal combustion, which avoids the drawbacks associated with steam boiler plant. Because of this superiority to other forms of prime mover, the gas turbine has been of interest to engineers from an early date, with the first patent being issued to one John Barber in England in 1791, although it is very unlikely that a practical device based on his proposed machine was ever built. The gas turbine can, in principle, be based on either of two air standard cycles, see Figure 5.18: the constant volume heat addition (Lenoir) cycle and the constant pressure heat addition (Brayton or Joule) cycle. The latter is used exclusively as the basis for modern gas turbines, although considerable effort was expended in the early years of this century on developing machines based on the Lenoir cycle. Both of the air standard gas turbine cycles incorporate a heater, an expansion turbine and a cooler (see Figure 5.18). The constant pressure cycle (Figure 5.18 (a) ), requires, in addition, a compressor. The lack of this latter component in the constant volume cycle (Figure 5.18 (b)) gives it two advantages over the constant pressure cycle: the air standard cycle efficiency is higher for comparable conditions; and it avoids the need for a high efficiency rotary compressor (very difficult to build with satisfactory efficiency), which are used in gas turbines working on the constant pressure cycle. The disadvantage of the constant volume cycle is the unsteady nature of the heating process, which leads to unattractive mechanical complications. Working gas turbines have been built in which heat exchangers are employed as the heaters and coolers, respectively, and the working fluid (usually air or helium) is continually circulated in a closed system, i.e., a closed cycle. A commercial closed cycle machine (12.5MW), using air as the working fluid, which is no longer operational, was placed in service at St Denis, France, in 1952. More recently a 50MW helium machine, using coal combustion as a heat source, was built at Oberhausen, West Germany. The motivation for using helium is the possibility of using the gas turbine to extract energy from a power-producing nuclear reactor, because, unlike air, helium is not activated in its passage through the reactor core. PART TWO: POWER AND ENGINEERING 330 The majority of gas turbines are based on the so-called open cycle (see Figure 5.18), where air is drawn into the machine from, and discharged to, the ambient. In the practical open cycle gas turbine, the heater is replaced by a combustion chamber (or combustor) in which a fuel (liquid, gaseous or solid) is burned in the air supplied by the compressor. This means that the expansion turbine handles the products of combustion, which, because of their temperature and tendency to react chemically with the materials of this component, has resulted in a major effort since 1940 to develop suitable turbine blade and disc materials. For non-aircraft gas turbine applications the products of combustion are discharged at low velocity through an exhaust stack to the surroundings. In the aircraft turbojet the gases leaving the expansion turbine are accelerated by discharging them to the ambient through a nozzle, and this provides the desired thrust. Figure 5.18: Schematic arrangement of the components of the air standard gas turbine cycles. In the open cycle the heat removal process is omitted. An actual gas turbine working on the open cycle would replace the heat addition process by a combustion chamber, (a) Constant volume heat addition (Lenoir) cycle, (b) Constant pressure heat addition (Brayton or Joule) cycle. STEAM AND INTERNAL COMBUSTION ENGINES 331 Early attempts (1900–10) to build gas turbines based on the constant pressure heating cycle were unsuccessful because it was impossible, at that time, to build a compressor of sufficiently high efficiency. The strong influence of the compression process on the cycle efficiency is a result of the large volume of air that must be handled by the compressor. This, in turn, results from the need, because of material limitations at that time, to limit the cycle maximum temperature (T M ) to 538°C (1000°F). As an alternative to employing large amounts of air, the combustor and expansion turbine can be cooled. This was the approach used by Holzwarth, who constructed a number of gas turbines based on the constant volume heating cycle between 1908 and 1933 (see below). He was motivated to use this method because of the inadequacies, already referred to, of contemporary compressors. The necessary compressor efficiency requirements that had to be attained by the early gas turbines can be demonstrated as follows: suppose an air standard cycle thermal efficiency of 20 per cent is desired, which is comparable to that achieved in actual steam power plants in 1910 (see Figure 5.10). Further suppose the maximum cycle temperature is to be 530°C (1000°F), which is attainable if the materials used in the expansion turbine are chosen with care, and if the turbine blades are cooled (although it is doubtful that appropriate methods were available at the time). Then the efficiency of the compressor (assumed, for simplicity, equal to that of the expansion turbine) would have to be 88 per cent. This is high for a centrifugal compressor now, and was certainly not attainable in the early years of this century when typical efficiencies were near 60 per cent. The most successful of the early gas turbines were based on the constant volume heating cycle in which high pressures at the expansion turbine inlet are obtained by constant volume heating instead of using a compressor. In theory, no compression work is required, which avoids the limitations, mentioned above, of the constant pressure cycle, but in practice it is desirable to employ a compressor, albeit of very modest performance, in order to introduce a fresh charge into the combustor and to scavenge the ‘end gases’. The apparent advantages, in the cycle thermal efficiency and in the compression process, that made the constant volume cycle attractive to the early gas turbine engineers, unfortunately brought with it some very undesirable complications. Constant volume combustion must take place in a closed chamber with valves for admitting the air and discharging the products of combustion to the expansion turbine. Furthermore, the unsteady combustion process is not well matched to the steady flow expansion process. In spite of this a number of moderately successful turbines based on the constant volume cycle were built. . development of the Wankel engine, where the sealing at the rotor apex, its point of contact with the working surface, is especially critical. The German NSU company obtained the rights to the Wankel. chamber and triangular rotor of the compressor for use as an internal combustion engine. NSU and Wankel then spent the period between 1954 and 1958 in an intensive development effort on the Wankel. with performance data for both two-stroke and four-stroke cycle engines included. The important effect of the fuel injection method and the use of supercharging on the performance of the compression