The V8: Birth and Beginnings
By Jack Phillips
How many design engineers are fortunate enough, I wonder, to find themselves in the position similar to mine in 1954? I had a first class set of tools in the shape of a Rolls-Royce trained team of designers who were both loyal and enthusiastic, plus Harry Grylls (Gry) as a boss who gave broad brush instructions of requirements and left the design to the designers.
Although Gry would have frequent ideas for detail design changes, and many times during my career made suggestions which resulted in great and fundamental improvements, he left me alone for long periods and steered the trend of engineering, rather than, as so often happens, dictated his wishes
He thereby established a repartee between us which encouraged not only respect for his ability but, in addition, a feeling amounting almost to affection for a man who during my most vehement moments would say: "Don't rock the boat!", or to a radical idea: "Where's your rat hole if it all goes wrong?"
I had known Gry since the early 1930s, and as I had donned the mantle of Charles Jenner (Jnr), the previous motor car engine designer, Gry questioned many of my ideas, but responsibility for the design of the successful 'B' range of military and commercial engines which had sold in their thousands to the British Army, Germany and NATO, I suppose gave him some faith in my judgment. In fact, without Gry's encouragement I would not have left the Aero Division in Derby for Belper immediately after the war.
The starting-gun of the' L' engine (code name of the V8) was fired one Monday morning early in 1954 - can it be more than 45 years ago? I remember quite clearly the usual request when my office phone rang. Gry, on the other end: "Can you come along a minute?" Almost always the same words summoned me to his office along the corridor next door to 'Doc' Smith (the late Dr Llewellyn Smith (LS) then Joint Managing Director).
Gry had a spare kneehole desk with a tooled green leather top which he told me had belonged to his father; his own workaday desk was larger and more prosaic, and I used to half sit on the corner of his father's furniture when discussing progress.
"Our car engine is running out of steam," he said, "after all it has been in production almost twenty years. We need an engine which will fit within the bonnet of the present (Silver Cloud) car. Go away and think around a 50% increase in power and torque."
The decision was made quite logically. We had to increase the swept volume over that of the 4.887 litre in-line six-cylinder by a considerable amount, even if we increased the specific output per litre. That ruled out any in-line cylinder arrangement on the score of insufficient under-bonnet length. A double banked arrangement was the simplest solution. A V6 was known to be more difficult to make as unobtrusive as the six in-line, and would need excessively large diameter cylinders; a flat eight would have excellent balance but would limit the steering lock, and hence give an unacceptably large turning circle, but a 90° V8 would provide perfect balance for primary and secondary out-of-balance forces and couples, and would be short enough to allow all auxiliaries to be mounted in front of the cylinder blocks.
When I suggested this solution to Gry I found, to my chagrin, that he had already asked the Advanced Project Design Office in Derby to have a look at a possible V8, and had a rough preliminary sketch of a bottom-end design, ie a crankshaft with balance weights.
At least both Gry and I had come to the same solution, and he could not be blamed for wanting to have an independent assessment. He proceeded to lay down a broad specification for the future power unit. He said it must weigh no more than the then present production in-line six-cylinder; it must use the same radiator frontal area, and be no more costly.
All these stipulations with a 50% increase in power and torque, together with the usual reliability and unobtrusiveness expected of a Rolls-Royce power unit, was a formidable challenge to the engine design team. We had plenty of ideas, clean sheets of draughting film, and an endless supply of rubbers to enable us to change our minds as much as necessary to evolve a satisfactory design. Someone once said, very truthfully, that engineering design is 10% inspiration plus 90% perspiration; design offices are always full of rejected schemes, almost always by the designer himself!
Rolls-Royce car engines have a much longer production life expectancy than other makes, which in turn means one must have an unique design philosophy to cope with the needs of future car development.
Firstly, one has to provide sufficiently large bearing areas to allow an increase of 15% to 20% in power and torque during its life expectancy.
Secondly, whenever any re-arrangement of auxiliaries occurred I have always tried to reduce the overall height of the engine by at least 1 ½ inches. This latter prerequisite was thoroughly justified later when it allowed the engine to fit under John Blatchleys beautifully styled Silver Shadow bonnet without difficulty.
To meet the weight and power requirements the engine had to be "all aluminium", and have a potential increase in swept volume of 20% over the six-cylinder. For installation purposes the engine speed was kept similar to that of the previous unit, though the potential power increase would inevitably increase the top speed of the car and hence the speed range of the engine. Few, if any, advantages were to be gained from using an overhead camshaft arrangement; moreover, Rolls-Royce engines must have a camshaft drive that is both silent and must last the life of the engine. A chain drive does not meet these requirements, though I must concede that modern technology could probably provide an acceptable solution.
The compact combustion chamber compared with the side valve exhaust arrangement of the six-cylinder engine enabled an increase in compression ratio and thermal efficiency to give a theoretical increase in horsepower of between 12 and 15 per cent, which was in fact achieved in practice. The consequent reduction in specific fuel consumption was an added bonus. (Figure 1 is a good illustration of the complete head and coolant circuit - the spark plugs were centrally disposed.)
The decision to adopt top-seating wet liners for cylinder bores was taken for a number of reasons. I wished to reduce oil consumption to a minimum, and to achieve good figures cylinder bores must suffer as little hot distortion as possible - bottom seating liners do suffer distortion by the clamping load applied by the cylinder head. The coolant capacity must be small to speed warm-up and aero engine experience had taught us how to seal the liner skirts.
We had, by now, crystallised our ideas around a light alloy wet-linered OHV V8 engine with push rod operated valves and so we commenced project design. Project design is a preliminary coalescence of ideas on paper to see if the whole gells together, and it was my practice to start from the heart of the engine and work outwards, hence the approximate shape of the power unit evolves automatically.
A short stroke engine was necessary to reduce the overall width of the V engine; to increase crankshaft stiffness and to provide ample cylinder diameters to give room for good big valves and hence good breathing. The bore and stroke chosen was 3.8 x 3.5 inches, giving a swept volume of 5.2 litres.
The crankshaft bearing loading was limited to a moderate maximum pressure of 2,500 lb/sq inch at 4,500 rpm, for both main bearings and the side-by-side connecting rod big ends (see Fig 2a).
Almost all high performance V8s are designed with a single plane crankshaft, ie the same arrangement as a four-cylinder in-line engine. The advantage of this lies in the fact that it gives even firing per bank of cylinders and hence is easier to carburet nicely, but the resulting out-of-balance is inadmissible in a luxury car.
The two-plane shaft (Fig 2b) which the Rolls-Royce V8 employs has an arrangement of crankpins rather akin to two pairs of bicycle pedals, the centre pair lying at right angles to the outer pair. This gives very good balance but necessitates a somewhat complex induction system to produce even pulse intervals at each of the twin SU carburettors.
A good deep skirt to the crankcase and heavy ribbing between the two banks of cylinders at the upper end ensured the very stiff assembly necessary to give smooth running, and a large base to which the gearbox bell housing could be fastened. Incidentally, there are quite a few modern engines where the joints between the engine and transmission are insufficiently rigid to prevent excessive flexure and hence rough running.
Figure 3 illustrates the way we tie the upper ends of the V together, and Figure 4 shows the heavy ribbing in the crankchamber.
Loads between the cylinder head and main bearing caps were accurately calculated at this stage, and final designs of piston, connecting rod and crankshaft completed. It was then that we looked into the load-carrying path and decided to use cylinder head studs which would penetrate as far as possible into the crankcase structure and almost meet the main bearing studs, hence the amount of light alloy subjected to tensile stress fluctuation was minimised and the structure lightened.
As a matter of interest, the way in which we calculated these loads was to draw a diagram of cylinder pressures throughout the compression and expansion strokes in order to obtain piston loadings at 50 intervals of crankshaft rotation. Thirty-five years ago we had only mechanical calculators; no electronic computers - we had therefore to calculate the vectored loads laboriously and plot the derived information shown in Figure 2a (called a polar bearing-load diagram).
The cylinder heads, which were identical, were based on a modified hemispherical combustion chamber and, as mentioned earlier, had a central plug and so a very short flame travel. Project work on maximising the stiffness of the large exhaust rocker arm for the minimum inertia occupied some time, as did the evolution of the precise combustion chamber shape to give smooth combustion. The very short exhaust ports in both this and the later head designs were a major contributory factor in reducing the heat to coolant from 76% of the power output in the case of the in-line six, to only 54% in the V8. This helped me to meet the need to use the same radiator frontal area for both engines, even with a vastly increased power output.
By now we had completed all project work and detail design was under way. Although it had been proved theoretically that differential expansion (between the cylinder block plus the head and the push rods) would compromise the camshaft design, we went ahead to commence cutting metal.
The first two or three crankshafts were machined from solid billets of steel by the Experimental Department under the wizardry of Stan Smith, and we ordered one hundred connecting rod forgings from the drop forge, with extra balance weights to assess the variation in weight likely to be experienced.
Another interjection here - Harry Grylls for years had maintained the opinion that if a nut face and its bolting face were truly square with the thread no locking was necessary on a big end nut. To prove this all Experimental car engines (six cylinders) for several years ran with unpinned big end nuts with no sign of trouble.
It will not have escaped the reader's notice that the V8 has no locking on the big end nuts! Nor for that matter have I employed any bolt head traps; a knurl on the bolt stem prevents rotation and is so positioned as to lie in an area of minimum stress.
By the time the first engine had been produced and assembled we were well into 1956 and encountering the usual development troubles. We needed more power, which was not available without sacrificing the good (but not perfect) idle quality. I remember Stan Dean, an Experimental Test Driver, christening it 'the gutless wonder', so we had a V8 versus a six in-line race up Pyms Lane, Crewe, from a standing start, which dismissed the 'gutless' part of his remark! The fact was that the V8 was much smoother than the six in-line and its gentility was mistaken for lack of power.
Other shortcomings were evident. It was heavier than we had hoped and was a shoehorn fit with the bonnet width; the steel camshaft and chill cast tappets which had been completely satisfactory in the six-cylinder were a complete disaster in the V8; the tappets were too noisy, and so on . . . Figure 6 shows the engine at this stage of development.
I had been to Detroit to see General Motors about hydraulic tappets, and had bought several engine sets of Buick tappets which were the right diameter. Coupled with this, the crankcases were modified to accept cast iron tappet blocks (as now) and steel push rods were substituted for the light alloy ones.
These measures largely silenced the valve gear, and with a change to the chill-cast camshaft and the hardenable iron hydraulic tappets we were in business.
From time to time, amid the rush and hurly burly of engine design, one has a chance to sit, sometimes in the office, sometimes in the wee small hours, and question why things are the way they are.
All Rolls-Royce aero engines, and the PIII car engines, were made of copper-aluminium alloys - a lovely alloy to machine, but not the best for corrosion resistance. The V8 was designed around this material, but why not use a silicon/ aluminium alloy, with its much greater anti-corrosion properties?
I suggested to Gry that we should change to LM8, a silicon/ aluminium, and he agreed but wondered if we could accept the worse machinability.
To get the best out of men one needs to be a bit of a psychologist, and Stan Smith, the Experimental Shop Superintendent, was a person who had to be handled with tact. I told him about the wish to change the material and said I doubted his ability to machine it. He immediately rose to the bait, saying: "You give me the ******* stuff, I'll cut it." And he did, though admitting to some difficulties. So we changed to the new material for all subsequent prototype engines and, of course, all production to date.
Ron West and his valiant development engineers worked hard to develop the induction system to give good distribution throughout the running range, but three problems remained in addition to size and weight. The oil pump, which ran at engine speed, was audible; the idle still 'fluffed' now and again; and the ignition advance at maximum speed was excessive, which denoted slow burning.
The weight and width problems were faults of design, as was the slow burning (technically the rate of flame propagation). At this point the 10% inspiration came to my aid and a new design of cylinder head was started to halve the weight of the previous head and increase the combustion squish area to 33%. This project was a design experiment, but on one of his infrequent visits to the Drawing Office LS became interested, and we were told to proceed to fruition. This new head reduced the overall width of the engine by seven inches, reduced the weight by thirty pounds and halved the ignition advance requirement. Figure 7 shows the new head and cast iron tappet blocks.
A solution to the oil pump moan proved extremely hard to find. We tried many different gear forms for the pump gears and all to no avail. Progressive reduction of pump speed effectively cured the noise but left insufficient oil pressure at idle speed to feed the hydraulic tappets.
The poor performance at idle was the result of differential thermal expansion of the light alloy pump casing relative to the steel gears, and the only answer was to make the casing of cast iron. This, with a reduction in speed to about 40% of engine speed, did the trick and we had a silent lubrication system.
The demand for an increase in torque called for a larger swept volume, and the bore was increased from 3.8in to 4.1in, the stroke from 3.5in to 3.6in, giving a swept volume of 6.23 litres. No alteration to the crankcase was needed; we had allowed for these demands in the original project design!
By this time we had begun to concern ourselves with exhaust gas emissions and the need for extended plug life. The plug bosses in the new head were positioned immediately over the jet from the main coolant galleries, and although accessibility left a lot to be desired, we were rewarded with plug lives in excess of
24,000 miles - twice the ultimate American demand.
Pressed-in valve seats were included for the first time in a light alloy Rolls-Royce engine; previously they had been screwed in à la Merlin. The result was a notable improvement in idle quality. Charles Caisley, our carburetter and induction development engineer, had previously discovered that the small annular gap at the base of the screwed-in inlet seat was responsible for the occasional 'fluff' or hesitation in idle speed.
Valves were in-line and operated by tubular push rods having hardened steel ends swaged in - a very light assembly. It is not generally realised that the main criterion in push rod design is by how much the rod reduces in length when the end load is applied, and since, for a given load, mild steel deflects no more than the best alloy steel there is no point in employing expensive rod material - hence the bundy tubing rods (Figure 8).
As this engine specification (Figure 9) went into production we had a major panic - one of our forged light alloy bearing caps broke. Investigation showed that the cap had been machined wrongly, thereby putting the radius under the main bearing bolt boss in tension rather than in compression as intended. In spite of my protestations, Gry would not believe this could account for the vast difference in fatigue life and ordered that all bearing caps should be solid blocks until a correctly machined cap had been endurance tested. Of course he was right not to take a risk. I asked what would convince him. His reply - 500 hours full throttle! This was more than five times the life we had managed to obtain from contemporary American V8s. We started endurance at 4,200 rpm full throttle. Five hundred hours is a long time, with the exhaust system red hot from the exhaust flange to the tail pipe, but we did it without any failure and Gry was convinced we were right.
Although Figure 10 does not depict the actual test engine, it does show the test-bed configuration used during endurance testing, and in colour would give a graphic impression of the high temperatures produced during high power development, in fact the exhaust valves run continuously at bright cherry-red temperatures.
With the engine now in production and powering the last of the Silver Cloud series, we in the Design Office were heavily involved with the preparations to meet the proposed Californian, and later American Federal, exhaust emission requirements.
The spark plug requirement from the US proved to be only 12,000 miles, which allowed us to shift the plug to a more accessible position above the exhaust manifold at the expense of a small reduction in life and so allay the justifiable criticism of inaccessibility (Figure 11).
A succession of small, but very important, modifications continued through the Design Office. Cam forms, induction tracts, liner fits, as well as the proposed brake pump for the power braking for SY (code name for the Silver Shadow). Improvement and refinement never stops. We were never satisfied, even though the engine was by now quiet, refined, powerful and unique in its ability to stand continuous full throttle running without distress.
At that time it had the highest power to weight ratio of any car engine, weighing only 2.7lb/ hp complete with all auxiliaries, exhaust manifolds and bell housing. I have no doubt our successors at Crewe have improved greatly on those figures.
In spite of being accused from time to time of over-engineering the V8, it has stood the test of time. After all, more of these power units have been made than all the previous Rolls-Royce car engines in total.
Nor should it be forgotten that, at the same time as this programme was being enacted, we designed and developed an opposed piston two-stroke multi-fuel engine, the 4-litre F engine and an experimental overhead camshaft engine giving almost 70hp/ litre. A busy period, and one which perhaps could be the subject of a future article.