Aviation of World War II

Aviation of World War II

Home  Custom Search Russian
Flight 12 1940 The Aircraft Engineer

The Aircraft Engineer

The Two-stroke Diesel Engine

On March 14th, 1881, Sir Dugald Clerk filed his patent specification, "Improvements in Motors worked by Combustible Gas or Vapour." This was the birth of the two-stroke or two-cycle engine. Despite certain obvious advantages, it is only in recent years that the two-stroke has been closely studied. In 1932 Sir Dugald Clerk wrote: "The two-cycle engine is worthy of great attention, and the best type will be in the form of a heavy-oil engine of the compression-ignition type." Sir Dugald Clerk's opinions carry great weight, and there is a rapidly growing opinion that the two-stroke has been neglected tar too long.

Mr. W.S. Burn, in his paper on "The Application of the Two-stroke Heavy-oil Engine to Aircraft Propulsion," read before the North-East Coast Institution of Engineers and Shipbuilders in Newcastle-on-Tyne earlier in the year, definitely comes down on the side of the two-stroke. At the beginning of his paper he outlines the potential advantages of the two-stroke heavy-oil engine over the four-stroke petrol engine. In his summary the author gives his opinion that reduced engine weights are fundamentally possible, and that among the advantages of the two-stroke are: greater reliability ; less fire risk ; greater radius of action ; cheaper running costs; elimination of electrical ignition and consequent elimination of radio interference ; no carburation troubles; greater simplicity of control; easier overhauls, easier cooling; greater power for a given weight; better performance at altitude and more reliable starting.

A formidable list of advantages which was by no means accepted by those whom Mr. Burn describes as suffering "from a particularly strong petrol complex."

Past Experience

High spots in Diesel aero-engine development are represented by the following makes: Junkers Jumo; Packard; Deschamps; Guiberson ; Clerget; Coatalen , Salmson ; Talhot; Zod ; Beardmore ; Bristol Phoenix ; Rolls-Royce Condor; and various British experimental sleeve-valve types. The chief developments have been made abroad, and the greatest progress has been made by the two-piston Junkers engine, the only commercial type.

Practically all of the various aero forms of engines have been experimented upon, chiefly the four-stroke type, from 250 to 1,000 h.p. The weight so far has been usually unattractive at about 2(1/4) to 3 lb. /b.h.p., and the fuel consumption at about 0.38 lb./b.h.p./hr., although the Coatalen engine claims to be as low as 0.3 lb./b.h.p./hr.

In describing briefly the Junkers engine the author pointed out an interesting development, namely, the reduction of the total stroke-bore ratio which is 5 :1 with the Jumo 206 (the latest type), 6:1 in the Jumo 205, and 7 : 1 in the Jumo 204. The scavenging excess-air ratio has, it is stated, been reduced from 1.6 to 1.3 ; the b.m.e.p. increased to as high as 135 lb./sq. in. As the type is fundamentally heavy, due to the necessity for two crankshafts and crankcases, the progress made is very creditable.

"An investigation of diesel aero designs to date seems to indicate," said Mr. Burn, "that the chief inadequacy has been in the exact fuel injection, and perhaps in the lack of air-controlled movement or turbulence—to give a regulated and complete combustion with high m.e.p., low maximum pressures, and good fuel consumptions. Nearly unlimited air supply will only enable high mean pressures with inferior fuel consumptions."

Marine Development

"It is curious," said the author, "that while the four-stroke marine diesel engine has been almost completely replaced by two-stroke single - and double-acting engines for some years past, for reasons of reduced weight, cost and fuel consumption, the aero-engine has gone steadily along on its uneconomical four-stroke petrol-consuming course, with stcp-by-step improvements until consumptions of 0.44 lb./b.h.p./hr. are possible.... Commercial aero-engines have increased no less than three times in power output from given cylinder dimensions in fifteen years, a figure almost identical with that of certain well-known large slow-speed marine diesel engines; but while in the former case the improvement is due to increases in b.m.e.p. and r.p.m., in tie case of the marine engine it is chiefly due to change of type, the mean pressure and revolutions remaining almost the same due to heat stress and propeller limitations respectively.

"The overwhelming need for a suitable heavy-oil engine is demonstrated only too clearly by figures for a large Transatlantic flying boat given in a lecture by Mr. Gouge," said Mr. Bum. "With six 1,300 h.p. engines and an economic cruising speed of 237 m.p.h., the total weight of 163,000 lb. is made up as follows: —

Power units 23,320 lb. Equipment 8,700 lb.
Tankage 2,670 lb. Fuel 62,350 lb.
Structure 54,110 lb. Pay load and crew 11,850 lb.

"Discussing these figures," the author continued, "the prime importance of improvement in fuel consumption is shown by one glance at the fuel weight. Engine weight would appear to be almost a secondary matter. Another striking feature is the large power required to propel such a small useful load. With regard to engine weight, there is a tendency to require much greater reliability and reduced maintenance, and to halve the increase in specific output. Weight reduction will therefore become increasingly difficult to obtain. There will still be considerable scope for improvement in aerodynamic efficiency, but the greatest scope will be in radical reduction in engine drag, and improvement in fuel consumption by the adoption of C.I., the two-stroke cycle, exhaust efflux and even cooling propulsion. A cleaned-up leading edge, and the use of integral fuel tanks, will tend to more economic wing structure." Mr. Burn summarised the above improvements which can be effected in the propulsion of aircraft in eight sections, the first being the fitting of the engine completely within the wing to obviate interference drag and to facilitate the use of pusher screws.

Junkers Jumo engine

Fig. 1. The Junkers Jumo opposed-piston, two-crankshaft heavy-oil engine.

Mr. Burn rushes in where aero-engine angels have, as yet, trod with the greatest discretion, and not only discusses what should be the correct basic aero-engine type, but proceeds to design such an engine. Considering the many engine types in service, the variation in existing design indicates the wide possibilities and suggests the mapping out of a new course guided by fundamentals rather than practice.

"The suggestion that some form of flat engine will be the ultimate solution becomes more and more insistent.... The present tendency of increasing the number of cylinders is evidence that present engine development is going to seed.

"A review of the whole gamut of petrol engine types reveals that even from weight considerations alone it is necessary to leave the petrol engine system and adopt some other combustion system which is free of cylinder size limitations for reasons of specific output and fuel economy.... The type trend will, therefore, be to compression ignition, the two-stroke cycle, and the smallest number of horizontally opposed cylinders.... Any in-line type can be adapted to be a "flat" opposed-cylinder type, which would lend itself to meticulous fuel-injection control as arranged in, say, the Coatalen engine."

Various Valve Types

Various Valve Types

Fig. 2 shows the poppet-valve type, with a single exhaust valve combining a good gas flow with a dual-turbulence insulated combustion chamber in the cylinder cover, and yet possesses a reasonable gas flow for exhausting. The short piston and short length of cylinder will make for low weight. The figure also shows the single-piston double-flow scavenge type, so popular in marine engines, which should not be ruled out. It is significant to note that a rather similar design was patented by Junkers. For comparison the differential-stroke opposed-piston construction, showing the very straightforward cylinder construction possible, is given.

Said Mr. Burn: "I will propose what I believe is an entirely new engine type to the aeronautical world, even if it has as a basis a type well known to the shipping industry, namely, the slide-rod opposed piston engine. I will also be a renegade to the aero-engine fashion and endeavour to give the aeroplane designer just what he wants—that it, an accommodating sort of engine, and one that will not interfere with his desire for perfect aerodynamic forms."

The author believes the pusher type of aeroplane will be the future type, and from that argues that an engine must not only be completely located in the wing without nacelles and with the neatest of propeller bossings, but must be in the trailing edge section as near to the trailing edge as possible. Hence the engines must be not only of horizontal form, but of as small a depth as possible.

Argument used to justify the use of engine nacelles is that in the case of the high-wing monoplane they afford a location for the retractable undercarriage, as in the D.H. Flamingo, but doubtless other means will be found to do this. One suggestion to enable the undercarriage to be carried efficiently is to accommodate it in a small cantilever variable-incidence low-wing, as shown in Fig. 4, when some useful and controllable lift is obtained with relatively small drag, at the same time eliminating all interference drag on the main wings.

D.H. Flamingo

Engine Accessibility

The thickest section of the wing should be reserved, in the larger aeroplane, as a passage way to gain accessibility to the engines. As a tractor screw at the leading edge must be about 25 per cent, of its diameter forward of the leading edge, placing the engine towards the trailing edge and using pusher screws will eventually entail the shortest shafting and thereby reduce weight. Once the engine is established in the wing, with pusher airscrews, the tendency will be to encourage the high-wing monoplane.

As a major advantage of the oil engine is that it imposes no limitation on cylinder size, and as there is ample wing span, an oil engine can be of almost unlimited length in a horizontal plane. "I am firmly of the opinion," said the author, "that once the decision is made to depart from the present conception of an aero engine being either a compact radial or in-line type of minimum frontal area, and to start making the fullest use of the one dimension which gives unlimited scope, i.e., horizontal length, the diesel engine for aircraft will rapidly go from strength to strength.

"The engine I am going to propose for 2,100 b.h.p. will have only six opposed-piston cylinders, three on each side of the crankshaft, but there will, of course, be twelve pistons in all, six controlling the air inlet and six the exhaust. The diameter of the proposed cylinder is 7 in., and the combined stroke 12 in. at the above power. The b.m.e.p. is 150 lb. per sq in., and the r.p.m. 2,000."

Bristol Phoenix diesel engine

Fig. 3. The Bristol Phoenix diesel engine established a world's record by reaching a height of 27,450ft.


In the course ol a detailed description ot the author's proposed design, he stated: "The most important problem of the two-stroke cycle is that of the complete evacuation of exhaust products and their replenishment with pure air. The propulsion of the air into the cylinder must be done with a minimum of power; therefore the air route must be as short and direct as possible. These and other requirements are best accomplished by using ports in the cylinder bore, arranged as far as possible completely around its full circumference. The proposed horizontal arrangement of cylinders offers the greatest advantage for the two-stroke type of cylinder, inasmuch as there is complete freedom to utilise two large-area, highly efficient exhaust pipes per cylinder, with an excellent gas flow. As a correctly-shaped combustion space is a fundamental necessity of any type of heavy oil engine, this should be arranged chiefly from the point of view of obtaining the best fuel and air distribution. The problem is to obtain the ideal combustion space and scavenging system without exposing either the pistons or cylinder bores to excessive heating effects. This inevitably necessitates some form of double-piston construction.

"Figs 5 and 6 show the horizontally opposed in-line arrangement proposed. The six- or eight-cylinder engines tor the "wo-stroke opposed type have definite balance of advantage (over the twelve-cylinder vee and nine-cylinder radial).

"It is obvious that such an engine is going to require an entirely different design technique. For example, the side tension rods and bearings; the sleeve side-rod guides; the ported, oil-cooled cylinder liners; the composite and oil-cooled pistons; hot combustion-space parts; the three-throw crankshaft with quite different lightening arrangements; the special fuel injection equipment; scavenge air pumps; and the altered exhaust technique. The mere fact of having an accessible engine whose parts can easily be renewed and replaced piecemeal, instead of removing the entire engine, will tend to a different conception of designing.

"The cylinder jackets are made separate for each cylinder, aluminium alloy castings, in two parts, bolted together in way of the combustion chamber. The liners are in two pieces, bolted together about a central belt with the protecting combustion ring held between them.

"Oil engine pistons have a number of design considerations requiring quite different treatment compared with normal petrol engine practice. For example, the piston crowns should be at as high a temperature as possible to reduce heat losses and maintain the highest ignition temperature. An oil-tight gudgeon is essential. The proposed piston is one of composite construction, the crown and skirt being in one part from an alloy steel pressing, surrounding a forged, light alloy central portion in which are formed the gudgeon inside bearings. A heat-resisting steel cap prevents heat travelling down via the outside in which the piston rings are held. All three pistons are held together by a single central bolt for reasons of possible differential expansion. This is shown in Figs. 5 and 6, while Fig. 7 gives a comparison with the Junkers design.

"The design of an efficient scavenge pump is the most difficult feature of any two-stroke engine; therefore, to start with, the scavenge pressure and excess volume required should be as low as possible."

Mr. Burn indicated that there is some doubt whether the air displacement and pressure characteristics of the centrifugal blower (which is usually employed) with variations of speed, conforms with the desired requirements. An advantage of the centrifugal blower is that it has been continuously developed for a number of years as a supercharger and can deliver up to 15 or 20 lb./sq. in. pressure. It must be kept in mind, however, that the fuel consumption and output of the Junkers Jumo were, until recently, seriously handicapped by the relatively inefficient centrifugal blower, and in the present state of the development of both radial and axial flow blowers there is a definite incentive to develop the simplest form of piston displacer type of scavenge pump. Even the most efficient type of Rolls-Royce blower, said to be 73 per cent., is not as good as can be readily obtained with simple parts from a displacement-type pump which has an efficiency of 85 to 90 per cent. Noise and the complete dependence on the engine are other defects. Constant air pressure rotary blowers of the Rootes type, even after years of development, consume the excessive figure of 15 per cent, of the total power developed by the engine. A normal figure for a large marine oil engine does not exceed 4 to 5 per cent., using plain reciprocating blowers with automatic or mechanically-operated valves. The pump design proposed will provide: —

(1) A blower of maximum efficiency by giving as far as possible a uniflow movement of the air and at the same time, making full use of the kinetic energy imparted to the column of air by the flap.

(2) The simplest possible system of moving parts. The oscillating flap or displacer type of piston being chosen as enabling the smallest and lightest operating rods to be used with a minimum of friction or lubricating surfaces.

(3) A timed air supply.

The layout of the engine in the wings is shown in Fig. 8. Mr. Burn summarises the expected advantages of the proposed engine compared with accepted petrol engines as follows: —

1. At least 25 per cent, improved fuel economy on a test-bench basis, and still greater economy in service.

2. Geometric form of engine suited to efficient aerodynamic design.

3. No greater weight at greatly reduced revolutions and mean pressures without greater maximum pressures, and hence, ultimately, much greater reliability, reduced service charges and longer life.

4. Much higher degree of exhaust efflux propulsion than any existing engine type.

5. Reduced cooling losses, and the chance of a slightly positive propulsion from the oil cooler.

6. More completely balanced engine dynamically.

7. Almost unloaded main bearings.

8. All connecting-rod bearings are unidirectionally loaded to give permanent silence and freedom from vibration, even if bearing wear takes place.

9. Outstanding accessibility without equal in present aeroplane engines.

10. Reduced weight of wing structure and fuel tankage.


The discussion was opened by MR. K. O. KELLER. He could not help feeling that the author's technical knowledge outran the practical difficulties. It must be understood that low fuel consumptions were more easily obtained in a slow-running engine than in a high-speed engine, and for this reason he had to dampen the author's enthusiasm to reach a consumption of 0.32 lb. per b.h.p. In the engine put forward by the author he proposed that scavenging air should be 10 per cent, in excess of stroke-bore volume. He ventured to say that this would account for a few per cent, increase in consumption. The author claimed to obtain maximum combustion efficiency with a low compression pressure. Practical experience would tell him that the reverse was the case.

DR. T. W. F. BROWN did not agree with the author's thesis that an oil engine for aircraft can be developed from considerations of large marine diesels. The 10 per cent, excess air for a b.m.e.p. of 150 lb. per sq. in. was far too low. Very careful experiment would be necessary to ensure that the shape of the compression space would enable reasonable ignition lags to be obtained; otherwise there was no hope of a fuel consumption being obtained that would be better than at present for the petrol engine. The author stated that the fewer and larger the number of cylinders per unit cylinder volume, the lower the weight. This, taken to its logical conclusion, would mean a single cylinder unit, which was absurd.

MR. JOHN NEILL remarked that the author suggested that the Junkers Jumo oil aero engine was in advance of anything in this country. It was interesting to note that the first German bomber brought down near Edinburgh was fitted with a Junkers Jumo using petrol.

MR. A. ORTON was sure the author was on the right lines, and drew attention to the Kadenacy two-stroke system. Engines of this type with a b.m.e.p. of 142 lb. per sq. in. and 1,500 r.p.m. were working regularly without the use of any air pump. With air under a moderate pressure, 200 lb. per sq. in. had been obtained. He was doubtful of the proposed increase in cylinder size and reduced r.p.m. compared with present aero-engine practice. He thought it would lead to a decrease in b.h.p. per litre and b.h.p. per unit weight.

MR. S. CAMM fully supported the case made out for the need for an intensive programme of research and development to produce a successful C.I. two-stroke engine. He agreed with the size of the proposed engine, thought its proposed form excellent, though it would be a great advantage if it could be made even shallower. He fully sup ported indirect cooling and felt that this would be the ultimate method. He thought the proposed oil cooler was inadequate, and suggested that a ducted cooling system as used on present aircraft was probably best.

MR. A. GOUGE also thought the author had put forward an extremely good case for the C.I. engine, and that research on it should be actively pursued. The consumption of the modern petrol engine was in the neighbourhood of 0.42 lb./b.h.p./hr. while that of the contemporary C.I. engine* was now about 0.38 lb./b.h.p./hr. Allowing for increased weight of the latter, the gain in range appeared small.

With high wing loadings, take-off was extremely important. Did the author see the possibility of increase in take-off power from the C.I. engine, such as was possible with the petrol engine?

He thought the proposal to instal the engine completely within the wing would prove extremely difficult. Pusher airscrews had their disadvantage, perhaps the most serious of which was a Joss in the case of four-engined aircraft of approximately 15-30 per cent, of lift at take-off. DR. F. W. LANCHESTER completely disagreed with the author s dictum that the fewer and larger the cylinders (within definite limits) per unit total cylinder volume, the lower the weight per horse power. But Dr. Lanchester did believe that the diesel would win part of the field, at any rate, in the long run, and that the two-stroke was inevitable.

Conclusions to be drawn from the full and lively discussion were that on the whole there was a general agreement that the two-stroke diesel deserved more intensive research, but the author had underestimated the mechanical and other difficulties to be overcome before a successful result could be obtained.

As the author is building the engine he advocates, it is to be hoped that the medicine he will have to take on the way will not be so nasty as many of the speakers in the debate suggested it would be.

Bristol Beauforts

HANDELS SCHIFF HUNTERS: Bristol Beauforts as torpedo carriers are taking a heavy toll of German merchantmen.

admin. * - diesel engine (also known as a compression-ignition or C.I. engine)