As the rotor went round at about 28,000 rpm, cutting down the width of the vanes to about two-thirds would make a substantial reduction in the stress. Hs went on, ‘As for the intake, I like the look of that too, so we are going to make it all for an engine test’.
Things then began to happen. Rubbra and Lovesey came to see me to get the dimensions of the rotor and diffuser, and the shape of the intake. The Main Engine Design Office went into gear to produce the definitive drawings for the revised engine. Thus were born the Merlin 45 engine used in Spitfires, and the Merlin XX used in Hurricanes and many other aircraft, and for me the impossible had come to pass. I had changed a Rolls-Royce engine designed by the great Henry Royce himself.
About this time, I was visited by the two senior testers, Orme and Arnold, who always seemed to operate as a pair. Both were horny-handed sons of toil, but they came to tell me that they had working for them a young man, G. L. Wilde — inevitably known as Oscar — who was a mathematical wizard, and a good tester to boot. Would I have a talk with him? Although they would be loath to lose him, they felt that better use could be made of him.
I saw him and was much impressed, and longed to have him on my staff. But how did one do this? I consulted Frank Allen who said there was no problem — if Orme and Arnold were willing to part, and I was willing to have him, then it could be fixed immediately. And so it was that Oscar joined me and became a valued and loyal colleague. He was a man of great enthusiasm and imagination, who was eventually to rise to the very top in the hierarchy of Rolls-Royce engineers.
About this time, I began to realise that it was not sufficient to consider the supercharger by itself. The important thing was the engine-supercharger combination, and it was essential to match the output of the supercharger to the demands of the engine. But what were the demands of the engine? Surprisingly, I could find little information about this, and so decided that we must measure the air consumption of the Merlin on test. Oscar said there wasn’t a hope of doing this on the experimental beds because the schedule of testing was very strictly laid down; but why didn’t we talk to the chaps on production test?
At that time, in early 1939, the production of the Merlin was just getting into top gear, and the rule was that each engine had to do a 2-hour endurance test before despatch. There were plenty of engines doing this routine test, and when we went along to see the chief tester he welcomed us with open arms. We could take what measurements we liked, provided it did not interfere with the passing-off tests. My own men went and made the measurements, and in a few days we had complete sets of air consumption curves for a considerable number of engines, running over the full gamut of varying power outputs.
In the case of the Merlin, the air was first sucked through the carburettor, where the petrol was sprayed in to give the correct mixture of fuel and air. It went on to the supercharger, where the mixure was compressed and, of course, heated by the compression, and delivered into the induction pipe which lay between the two banks of six cylinders with symmetrical off-takes on each side to each of the 12 cylinders. The pressure in the induction pipe was known as the boost pressure and was measured in pounds per square inch. The higher the boost pressure the higher the engine power, because more of the mixture of air and fuel would be forced into each cylinder.
The principle on which all four-stroke engines work is very straightforward. On the first downward stroke, the appropriate mixture of air and fuel — about 15 to 1 by mass, 9,000 to 1 by volume — is drawn into the cylinder through the open inlet valve(s). These valves then shut, and in the following upward stroke the piston compresses the charge. At the top of this stroke, the sparking plug ignites the compressed mixture, and the rapid rise in pressure which results forces the piston down again to give the power stroke. Before the bottom of this stroke the exhaust valves open, and on the next upward stroke the piston forces the burnt gases out through the exhaust pipe. The cycle then repeats itself. Thus there is one power stroke for every two revolutions of the crankshaft; in the case of the 12-cylinder Merlin there would be six power strokes to each revolution of the crankshaft.
In the theoretical examination I made of the relationship between the power output on the Merlin and the quantity of air/fuel mixture which it consumed, I started from the premise that the internal power generated in the cylinder must be proportional to the quantity of air/fuel mixture consumed. But not all the power generated in the cylinder appears as useful power at the propeller shaft. Some of it is used to drive the supercharger at the back of the engine, and some is absorbed by the friction of driving the pistons up and down, the crankshaft round and round, the valve gear and other parts.
The final answer, which was obtained by a combination of theory and experiment, showed that the Merlin generated an internal power in the cylinder of 10.5 horsepower for every pound of fuel/air mixture it inhaled per minute. This figure was independent of the rpm, or of the boost pressure from the supercharger.
On the theoretical side, I deduced a formula which accurately gave the quantity of charge which the engine would inhale under any condition, either on the ground or in the air. Now at last it was that, armed with the various formulae, we could calculate the power that the Merlin would give in any condition of flight.
Prior to this, we had calculated the horsepower from some very suspect and empirical formulae which had emanated from the Royal Aircraft Establishment. These formulae overestimated the power, and caused a constant argument between Mitchell, who designed the Spitfire, and Camm who designed the Hurricane. Both designers claimed that, from the measured speeds of their aircraft, the Rolls engines could not possibly be giving the power that was claimed for it. In fact, both aircraft were 20 - 30 mph slower than expected at about 20,000 feet, a fact that was kept very secret at the time since the war was imminent. Fortunately, the Germans were equally bad at estimating the power of their engines, and their aircraft were also slower than both the calculated and published figures.
The Americans took their usual line, and decided to build a special test bed at Wright Field in Ohio where their engines could be tested on the ground in conditions which simulated those at altitude, and in this way they were able actually to measure the power output of their engines. By the standards of the day this was a huge installation, with refrigerating plant to reproduce the cold air temperatures at altitude, and large exhausters to reduce the air pressure and take away the exhaust gases.
In comparison with this massive attack by the Americans, we had only a few simple formulae which enabled us to calculate the power of the Merlin while sitting at our desks. When, late in the War, a Merlin was actually tested in the Wright Field installation, the measured powers agreed exactly with those we had calculated several years before. From then on we adopted as our motto “The Pen is mightier than the Spanner”.
What was far more important, we were now able to reconcile the performance of the Spitfire and Hurricane with the power output of the engine, and were then able to predict accurately what would happen if we changed the power of the engine by changes to the supercharger.
It is seldom economical or desirable to use a supercharged piston engine at full throttle at sea level, because the resulting power would overstress the engine and its cooling system. In the case of the Merlin, the power at sea level was initially limited to about 1,000 hp, and to obtain this power the throttle would only need to be partially opened. As an aircraft climbs to higher and higher altitudes, the atmospheric pressure decreases, and the density of the air inhaled by the engine diminishes. If nothing is done, the power of the engine would fall off proportionately. But by starting at sea level with the throttle partially closed, as the aircraft climbs the throttle can be gradually opened so as to maintain the power output of the engine. Eventually, an altitude is reached at which the throttle is fully open, and this is known as the “full-throttle height” of the engine.
The function of the supercharger is to force a heavier charge of air/fuel mixture into the engine’
s cylinders. The extent to which it does this is measured by the boost pressure. To give its 1,000 hp, the original Merlin required a boost pressure of between 6 and 9 lb/sq in. The engine was fitted with an automatic boost control which kept the boost pressure constant as the aircraft climbed. It did this by a servo system which automatically opened the throttle as the altitude increased. Thus, the engine automatically maintained its 1,000 hp up to the full-throttle height, which in 1939 was about 16,000 ft (4877 m).
There was another vital reason for keeping the boost pressure constant at a predetermined value, and this was the need to prevent the engine detonating. The ideal situation is that the charge of fuel and air should burn smoothly in the cylinders. If too much charge is forced in by too high a boost pressure from the supercharger, then detonation can begin and, instead of burning smoothly, the charge literally explodes, and causes shockwaves, like those at the nose of bullets, to bounce around inside the cylinders. These waves are of such intensity that serious mechanical damage can be caused to the cylinder head and pistons, which for lightness are made of aluminium, and thus can be relatively easily damaged.
The onset of detonation can be controlled by the octane value of the fuel, which in 1939 was limited to 87. Just before the Battle of Britain, small amounts of 100-octane fuel became available from the USA and this enabled us to open the throttle further on the Merlin and, in fact, to obtain nearly 2,000 hp without detonation. Thus, the 100-octane fuel made a crucial contribution to the performance of the Spitfire and Hurricane in that battle, as did the work of Lovesey, Rubbra, and their teams, which enabled the Merlin to withstand double its design power for short periods without mechanical failure.
To obtain the increased power, the pilot had to override the boost control which was normally limiting him to 1,000 hp. To do this, he had to pull a knob in the cockpit, and break the seal on it. So we always knew when he had done it! But in the Battle of Britain, 1,000 ft of extra altitude or 5 mph in speed could mean the difference between shooting down the enemy or being shot down by him, such was the equality between the performances of the Bf 109 and our fighters.
Thus, with the advent of 100-octane fuel, we were for the time being released from the nightmares of detonation. We could concentrate on improving the mechanical integrity of the Merlin to withstand higher power, which was Lovesey’s job, and improving the performance of the supercharger so that the power could be increased and also maintained to higher and higher altitudes, which was my job.
Unfortunately, the results of the work of my team, which I have so far recounted, did not come into fruition in the RAF squadrons in time for the Battle of Britain. In that epic encounter all the Hurricanes and Spitfires were fitted with the original Merlin III, as designed by Royce. Moreover, all the engines that fought in the battle were made at Derby, because the great factories at Crewe, Glasgow and Manchester, which were subsequently to produce more than 100,000 Merlins, were not even built.
We did not discuss it, but I must say that I expected Derby to be flattened by the Luftwaffe in the first weeks of the war. Thank God, Hitler made one of his several crucial mistakes, and Derby was spared to equip the fighters of the RAF.
I still remember the relentless pressure exerted by Hs in those critical months, and the response he got from the workers in the factory, who willingly worked 18 hours a day, seven days a week to produce the engines. Hs would regularly tour all parts of the factory before some of us were even out of bed, and since everybody knew him, and he knew everybody, the effect on morale was great.
It was Hs’ custom to hold a technical review with the senior engineers every Monday afternoon starting at 2.00 pm. On his right hand would sit Elliott as Chief Engineer, and on his left Swift as Chief Production Engineer. Directly opposite him would sit Lovesey, Rubbra and myself, and gathered around would be liaison engineers with the RAF to report troubles in service, material experts from the laboratories, HPS speaking for the Experimental Shop, and Dorey for the Installation and Flight Testing establishment at Hucknall airfield near Nottingham.
Hs was a great chairman. He had the happy knack of being able to carve away the undergrowth and immediately get to the nub of the question. He never allowed time to be wasted by individuals arguing together, because he knew his man, and would turn to the appropriate expert.
‘Well, what have we got to do, so-and-so?’
He would accept ‘I don’t know yet’, but woe betide anyone who tried to dissemble or cloak his manifold sins and wickednesses.
Once he had an answer, HPS would hold up his hands ‘Give me the drawings, give me the drawings’, and Swift would be put on notice that a modification was coming on production.
Of course, on occasions we were all nonplussed. Hs would look around. ‘I don’t know how you can sleep in your beds’, he would say, ‘Get your jackets off, and bring me the answer’.
The Merlin was a water-cooled engine, and we were always troubled by minor leaks at the various joints between the engine and the radiator situated under the wing of the Spitfire. This was Dorey’s province, and these leaks used to exasperate Hs, who would taunt Dorey by saying he had only to make the plumbing as good as his lavatory!
On one occasion in the early days of the war, someone in the RAF had filled a Merlin cooling system with water which had been fetched in carboys originally containing nitric acid, and which still had traces thereof. Even though the acid concentration must have been very small, the mixture played havoc with the aluminium cylinder blocks. The Aeronautical Inspection Directorate of the Air Ministry decided that, to avoid any possible contamination, there must in future be an official specification for the water used. This specification was the first item on Hs’ Monday meeting, and the AID were there in force. Hs sat in his usual place with his head in his hands listening to the prolonged discussion on how to write the specification. Suddenly he looked up exasperated and said,
“I’ll write your bloody specification for you. You have got to be able to drink it’.
These simple anecdotes, out of context as they are, now sound trite and elementary. But in the grim atmosphere of the war, they were just sufficiently humorous to give all enough lift to go away and get on with our jobs with zest and energy.
The initial work that I had done on the supercharger and its air intake appeared on production in 1940 in the form of the Merlin 45 for the Spitfire and the Merlin XX for the Hurricane, Mosquito and Lancaster. The effect was to increase the full-throttle altitude of the engine from 16,000 ft with the Merlin III to over 19,000 ft with the Merlin 45. I knew that this particular blower had now reached the limit of its development. Any further improvement would be much more difficult, and could give only a small gain, unlikely to justify a change to production. Where did we go from here?
At that very time, Lovesey and I were called to a meeting at the Air Ministry in London. We were told that Rex Pierson, Chief Designer at Vickers, had designed a capsule which fitted in the nose of the Wellington bomber and in which the pilot and bombaimer could sit. The idea was to pressurize this capsule, so that the aircraft could fly at very high altitudes in excess of 30,000 ft (9144 m), with the crew comfortably seated with an air pressure corresponding to 10,000 ft in the capsule. It was one of the first pressure cabins.
The standard engine for the Wellington was the aircooled Bristol Hercules sleeve-valve radial. To boost the power of the Hercules to get the Wellington above 30,000 ft, Bristol had decided to fit an exhaust-driven turbosupercharger, similar to those fitted by the Americans to their aircooled engines.
The high-altitude Wellington project was considered of sufficient importance to justify asking Rolls-Royce to provide an insurance policy by turbocharging a Merlin. This proposal was not as straightforward as it sounded. At Hucknall, under Ray Dorey and Harry Pearson, a great deal of work had been done on the Spitfire and Hurricane by taking the exhaust from the Merlin and ejecting it rearwards through very short exhaust pipes, where it acted as a means of jet propulsion equiva
lent to about 150 extra horsepower. With an exhaust-driven turbocharger we would lose that effect, and were loath to do so. I argued that, to obtain the necessary power, all we had to do was to raise the full-throttle height of the Merlin from 16,000 to 30,000 ft, and that to do this we needed two superchargers in series at the back of the engine, driven by the same gears that existed on the standard Merlin.
There was one obvious snag. Due to the high compression of the charge, its temperature would become very high, and the old bogey of detonation would rear its ugly head again. Also, the engine power formula indicated that the high charge temperature would actually reduce the power of the engine. The solution was obvious. We had a water-cooled engine, so we would add an extra water-cooled ‘intercooler’ after the two superchargers which would cool the charge to 100°C before it entered the cylinders. Calculations showed that in these ways we could double the power of the Merlin at 30,000 ft from 500 hp to 1,000 hp. Now the task was to determine the dimensions of the two superchargers.
At this point, a happy thought occurred to me. The Rolls-Royce Vulture engine, which had 24 cylinders and was much larger and heavier than the Merlin, gave, by virtue of its size and capacity, 1,000 hp at 30,000 ft. Since the power of an engine depends approximately on the amount of air and fuel it consumes, obviously the Vulture supercharger had the right capacity to supply the necessary air as the first stage of the proposed two-stage blower for the Merlin. No design effort on this component was necessary.
It was now that the advantage of the two independent ends of the supercharger test-rig paid off. On the one end we fitted the Vulture supercharger, and on the other the Merlin blower. A long sheet-metal pipe connected the outlet of the Vulture with the inlet of the Merlin, and thus we were able to run the two blowers in series, and measure their combined performance. The result was so good that no further calculation or testing was necessary, and we were able to go to Rubbra and start the Main Engine Design Office on the task of combining the two superchargers together as a single compact unit suitable for fitment on the rear of the Merlin.
Not Much of an Engineer Page 7