The Perfect Engine

The Perfect Engine

It does not exist. But luckily we all have different ideas of perfection.

“If you chase perfection, you often catch excellence.” William Fowble

Perfection is as hard to attain as infinity is to reach; a million miles is as nothing to infinity and perfection can only be reached by an eternity of trying.

I will make the hypothesis that just as beauty is in the eye of the beholder so can be perfection. It depends on what one thinks is important and what one can ignore. To begin, can we ignore, or at least accept, that as converters of energy stored in liquid fuel all kinds of engines are far from 100% efficient and, therefore, intrinsically far from perfect? If we can, what are the other parameters by which to gauge perfection? Some of the candidates as parameters might be the level of efficiency, specific power output, power/weight ratio, noise, exhaust emissions, size or shape, reliability/longevity, ergonomics (its feel). Does it do what was specified for it to do?

It depends upon what you mean by perfection

Do you remember the stories of the racing engine builders who could get the crankshaft and crankcase bearings so perfectly in line that it would spin under its own volition – like a perpetual motion machine? Well, that would be perfection! And I used to love it after a race to come to a stop by the van in the paddock and shut the throttle for the G50 or 7R racing engine to tick over at 200rpm. The engine was so hot there was, perhaps, a trace o automatic ignition and virtually no internal friction due to the hot, thin oil

Those engines were a sort of perfection – to me – at the time.

Jim Boughen, who had worked with my father developing the 7R AJS and G50 Matchless, built superb engines (albeit perhaps not quite perpetual motion) for the bikes that I rode for Tom Arter and when the engine did that I thought, “Perfect”. It depended, too, on getting the air jet and float chamber perfectly adjusted. The twist grip had not the slightest bit of slack which also contributed to a perfect bump start. In those days there were no clutch starts with running engines. I would push with all my might on the handlebars with the front brake on; crack the throttle open the precise amount and as the flag dropped, brake off, three steps, drop clutch, , engine fires, leap onto the saddle and first away!

Similarly, to make a perfect engine Jim had to be vigilant that all the parts and his careful assembly were perfect. In those days wider machined tolerances than today’s digitally controlled and checked machine tools can achieve, his ‘care’ was to choose parts that went together with the right clearances or interference fits. I remember the first 7R AJS engine of his I used. I called it a ‘little 500’ because it was so quick. When he finished building an engine he would turn it over using his timing disc on the crankshaft. This engine would not go over TDC and he thought that he must

have made a mistake with the piston squish clearance. Upon removing the cylinder head he found all was well and after refitting the head he simply gave the crank a little more effort than usual to turn it over. This particular engine had a particularly good compression. The roundness of the cylinder as assembled, the fit of the piston and rings in the cylinder bore were perfect. That was the critical difference.

Similarly, the poppet valves and valve seats had a radius instead of a conical shape with a virtual line contact, gas tight seal of the combustion chamber and it is only with perfect alignment of the valve seat cutter with the valve guide and precise concentricity of the valve seat and stem that this can be achieved. The clearance of the valve stem in its guide must be close to sticking to give the valve head a chance of a perfect landing after letting gas in or out of the combustion chamber. A racing engine running at 10,000rpm may have its valves open for 5 milli-seconds– so there isn’t much time to waste for the valve to shuffle about on its seat!

Digressing for a moment – thinking about that time of 5 milli-seconds. Even in these days of the atomic clock, nano-engineeing and quantum mechanics (whose recent theory is that the electrons of the atoms in our bodies are simultaneously in us and at the other side of the universe) the workings of something as relatively mundane and old-fashioned as the internal combustion is incredible and impossible to comprehend. As an example of limit of human observation and, therefore, understanding, I designed a research engine with a glass cylinder barrel about twenty years ago so that you could see the piston and valves going up and down; except that you couldn’t because they are moving too quickly for the human eye to follow, even at tick-over speed, and even at 200rpm. It is why it is o fascinating to watch the slow action replays of activity in nature, science, engineering or sport taken by the camera whose eye can see – record things that happen at high speed. Actually, that is what the scientist or engineer does at a leisurely human pace when calculating how metals and fluids behave at high speeds in very short time frames.

One of the most important features for the performance of an engine is the design of the inlet port but unless you have gas flow equipment and many, many hours of time available to you, what you have is what you use. Jim Boughen once added 15mph to the top speed of an early Norton Commando because he knew how to perfect the shape from standard inlet ports.

Perfection is, indeed, hard to define and the closest we can come is that the best in a particular period is the perfection at that time. Perhaps we would do better to ask, “What is the optimum engine”. We would still be in trouble to define it but to reach that goal, whatever it is, engineering has developed better machine tools and techniques for making things. Camshaft design and manufacture is better now so that valves ‘floating’ out of control is almost unheard of even at higher engine speed than of yore; cylinder bores are more perfectly round; pistons and rings seal the combustion chamber more nearly absolutely; oil resists the engine’s rotation less.

I think it is a credit to the designers of the engines of the ‘Classic’ era that the engines are still so conducive to perhaps an eternity of development to approach perfection.

The John Player Norton Monocoque

Like so many ‘original’ ideas, the 1973 John Player Norton was not entirely original and, of course, it was not all my own work. It was created by a team led by John McLaren and managed by the late Frank Perris who, to his credit, accepted and approved my basic concept. And the concept was to give the motorcycle the smallest frontal area as possible by utilising one part to perform as many duties as possible – for one part to take the place of several parts.

I knew that the reason riders on racing bikes put their chins and chests on the tank was to reduce frontal area but I learned quite early in my racing career that raising my chin off the tank by 5 cms would take 100rpm or 4 kph from my top speed. In my first GP at ~Hockenheim a high, big capacity fuel tank added two seconds a lap compared to using a sprint tank. And I bought a book on land vehicles aerodynamics which described a motorcycle and rider as a ‘bluff’ object – a fairly large frontal area but short length and it said that reduction of frontal area was, therefore, more effective than a ‘streamline’ shape.

The 1973 John Player Norton was a development from the 1972 model and both owed hugely to my wonderful Arter Special Matchless G50 on which I first tried and proved most of the 1973 features such as the cast magnesium wheels, disc brake designs, engine cooling air ducting as well as the pannier fuel tank the 1972 bike had which improved the ‘feel’ of all three bikes so much. I had always wanted to get rid of the frame and to make a ‘stressed’ fuel tank to carry the loads instead; for 1973 I had my chance. Without the pesky frame tubes cool air could be taken to the carburettors over the top of the engine and the position of the centre of gravity could be controlled more easily and, most importantly of all, the frontal area could be minimised. We used stainless steel because it does not lose its strength as badly as aluminium does when welded. We visited the MIRA wind tunnel three times with our ideas for the fairing and, with the screen at a precise height, the best drag-area measured was 0.222 square metres.

As I write this, it is interesting to note that the Arter Special Matchless G50 was the step towards the 1973 John Player Norton F750 ‘Monocoque’ and that, in its turn, is the platform upon which Peter Williams Motorcycles is based. The target for PWM is to make modern bikes and, in particular, to make what we call the ‘Real Monocoque’. I mention above that I had always wanted to use the fuel tank a part of the chassis but the top aim is to use the bodywork as the chassis. The idea is from college fifty years ago! And at last I met some like-minded people who wanted to make motorcycles – better motorcycles. Greg Taylor was one of them and he and I are now partners in PWM. We realised that my ambition could only be realised with cash – by capitalising on the success and reputation of the motorcycles I used to race. There were only four John Player Norton ‘Monocoques’ made and as they were one of the first to have cast magnesium wheels and the first to demonstrate the advantages of the twin-spar frame configuration they are unique – and valuable. We figured that we could sell replicas for a fraction of the value of the originals and we could make them at a profit.

It has been a huge task to ‘reverse engineer’ two of the original JPN ‘Monocoque’ bikes but the use of modern measuring and manufacturing techniques has resulted in an extraordinarily high quality product. Every part was remodelled and redrawn on CAD by Greg and me. As an example of the quality overall, the high specification stainless steel sheet of the fuel tank/chassis was laser cut to an accuracy of +/-0.2 mm whereas the originals were cut with shears. Who knows what accuracy John McLaren would have achieved? The cut panels were bent with modern accuracy and fitted to the jig by Gavin Tappenden and welded with modern equipment and extraordinary skill and dedication by

The Crankcase Breather

I think we all know the expression, “It’s run out of breath”, when an engine is asked to do more than its performance will allow. An internal combustion piston engine does actually breathe. The action of the piston going up and down in the cylinder is like our chests expanding and contracting to draw air in and exhale carbon dioxide. It is on the firing stroke the engine does all it’s work to drive us forward on our bikes. It does so because the burning of the fuel and air in combustion chamber creates very high pressures to push the piston down and turn the crankshaft around. A little of the hot gas above the piston leaks past the piston and piston rings into the crankcase because of the high pressure. With the engine running at 7000 RPM you can imagine that it would not take long for a very high pressure to build up in the crankcase. Unless it is given a route out of the crankcase it stays swirling about in the cases. It can be argued that this potentially high pressure may not detract from the engine performance because the work done against the crankcase gas to compress it is returned when the gas expands again as the piston goes on its way back up its stroke. But there are several reasons for having as little gas as possible in the crankcase.

1/. If the pressure gets very high gaskets and seals will rupture. Even with internal pressure only a little about the outside ambiant pressure oil leaks will be encouraged through the tiniest casting porosity.

2/. The oil scavenge pump will prefer to pump gas to the detriment of oil and the crankcase will start to fill with oil as well as with gas.

3/. We also want as little oil as possible swelling around with the gas because it causes drag on the rotating parts. For example, the speed of the periphery of the crankshaft and the big-end of a 500cc Manx Norton rotating at 9000 RPM is around 80 mph and the drag can be considerable so both volumes above and below the piston have to be allowed to breathe as effectively as possible. I mention the Manx Norton because in my experience it was the worst engine for breathing arrangements. In my day the Manx had a crankcase breather pipe at the front, at the drive side and from the timing chest. These various pipes provided nothing more than ventilation with no mechanism to actually expel gas. I blocked all the pipes off and drilled the drive side main shaft and fitted the G50 flapper valve. This is almost the ideal place for the crankcase breather to be because it is at the centre of the whirl-wind in the crank case so almost all the oil is centrifuged away from the breather. While I still had the old exposed hairpin valve springs I needed hardly any oil absorbent felt draped around the engine as many the Manx Nortons of my contemporaries had. Nor did my rear tyre glisten with oil as theirs’ did!

As I said above, a high pressure in the crankcase will damaged gaskets and oil seals so, conversely a depression should help and indeed should actually discourage oil leaks.

The limitation of the G50 flat flapper valve is that it is no more than a non-return valve and has very small hole for the gas to go through. Of course, it is not the only mechanism that can get rid of blow-by gas and creative depression in the crankcase. The normal oil scavenge pump has a capacity of somewhere between one and a half and two times that of the feed pump. That is why you see the oil return to your tank in globs and why the oil tank itself has to breath. The Cosworth V8 DFV racing car engines used the roots type scavenge pumps that were more gas pumps than they were oil pumps and had such pumping capacity that they created a huge gale through the engine which entrained the oil and took it to the oil tank but had to have a swirl pot to separate the oil from the air in the tank.

There is actually a fourth reason for letting the crankcase breath well. The leaked blow–by gas is nasty noxious corrosive stuff and it is best to have as little as possible in the crank case just as it’s bad for us to have smoke in our lungs.

Nowadays it is unacceptable to breathe the crankcase and ventilate your oil tank to the atmosphere. Car and bike breather systems now redirect crankcase gases into the inlet tract to ensure that there are fewer and fewer unburned hydrocarbon emissions as possible.

Polar Torque

There are so many things that influence the behaviour of a motorcycle. The tyres, friction, profile, suspension system, springs, damping and matching the front to the rear.

Steering geometry, weight distribution, length and height, all these elements have significant and interactive effects on the performance and enjoyment of riding a motorcycle on the race track and on the road. Today engineers have all sorts of instrumentation to assist the set-up of the bike but in the end it is the riders’ experience that counts.

I gave an example recently of a steering geometry set-up that I developed for a Norton Production Racer that was a really significant benefit to a Commando in racing trim but was a disaster for road riders. It is an example of how I believe, for me and my style of riding at least, a nearly dynamically unstable set-up is best for a racing bike.

Another example was with weight distribution on a succession of Norton Twin racing bikes.

The very first one was a friend’s Norton Dominator 500cc Special which had the engine well forward. The bike was proof of the theoretical fact that the more weight there is on the front wheel the more stable the bike will be. (Modern Moto GP riders all sit very much more forward than we used to for this reason. The next was the Dunstall Dominator – less weight on the front but still very stable – but a heavy steering without being weavy. Then there was my series of 1971 to 1973 Norton’s each of which had the weight further and further rearward. The 1973 JPN was perfect because it was in the right place.

Another example: Have you ever ridden your classic bike with the headlamp on the forks as standard and then ridden it without the headlamp? You can feel a real difference. It feels better because you don’t have to do so much work to steer were you want the bike to go; it does it or you!.

You would think the weight of a headlamp was negligible compared to that of the forks and the wheel itself so it is very surprising that the change is at all significant. But in a similar way I could feel the difference when I positioned the brake calliper behind the fork leg rather than in front. I really noticed the difference when I moved the forks closer to the steering head centreline by putting the wheel spindle in front of the fork leg and using a smaller offset of the yokes on production racer test. I certainly noticed the difference in the feel of the steering when I reduced the mass of the front wheel that swings around the steering head axis when I went tubeless on wagon wheels.

Motorcycles and Mechanisms

Motorcycles and Mechanisms Sammy Miller‟s Museum

There is nothing new under the sun – but the future is based on the ideas of the past for mechanisms, engines and motorcycles as for everything else. See the future in the past at Sammy Miller‟s Museum.

British people who ride foreign motorcycles should visit our motorcycle museums to see the first versions of the bikes, engines, and suspension systems that they use today. They are often British.

Walking around a museum is good for the soul. Taking notice of the exhibits of what has gone before makes me realise how little I know and makes me hungry to learn more. Whether it is natural history or science museum I cannot but marvel at the miracles of nature or the seemingly endless diversity of human thought and ideas. Sammy‟s museum at New Milton is exceptionally good in this way for me, of course, because of my fascination with mechanisms and motorcycles. Each one of his exhibits has an unique story; this is not just a jumble of motorbikes. It is a catalogue of the thought and engineering skill that has exercised thousands of people over more than one hundred years in the ever continuing search for the perfect motorcycle.

I see representations made of metal of both genius and foolishness although – to give some exercise to the bee in my bonnet – most of the latter was down to management. Many of the exhibits are of engineering ingenuity that people did not want to buy or were just glorious failures. People buy things because they best serve a purpose not just because they are different. They buy things which have an aggregate of cost, function, appearance, convenience and availability that is better than another product‟s. All these elements can be identified here. Overall you see the anthropological evidence everywhere of the willingness of the engineer – especially in the earlier part of the 20th century – to try new things not just because they were convinced that they would work better but because they wondered were where it might lead. They hoped a new horizon might open – that the new mechanism might provide a new vantage point to see the next way forward.

The first things I wanted to see on my visit were the rotary valve designs by Aspin and Cross. Before going the whole hog and to forget the reciprocating piston engines in preference for gas turbine engines these engineers thought they could at least try to do away with the reciprocating poppet inlet and exhaust valves that required so many parts and then seems to be a barrier to gas flow even when open. It seemed wrong that four-

Leading Link Front Fork

In my last newsletter I mentioned that I am working on a leading link front fork design. It is something that I regret very much that I failed to get round to using leading link fork  on my Arter Matchless Specials nor the John Player Norton.

In fact a long time ago I did make a leading link fork and it was a long time before I started racing. It was for my bicycle when I was fourteen in the last of the years when a motorcycle with a single cylinder four-stroke engine could win a World Championship. The motorcycle was the Moto Guzzi which to my eyes was the most functional, purposeful, beautiful thing in the world (apart from Dierdre Scott at school that is). You could see that the Moto Guzzi was very light as well as having the best aerodynamic bodywork on the Grand Prix starting grid. Everything on the bike was simple including the leading link front fork.

Because the Moto Guzzi front fork looked so simple and I wanted desperately to take up the impossible task of making my bicycle be like a racing motorcycle I decided to make a leading link front suspension for my bike. The idea was attractive because the first step was after removing the front wheel to reverse the fork so that the bend of the fork’s tapered tubes which normally sweep forward sweep rearward instead. The distance between the now misplaced wheel spindle lugs and the required wheel position is just right for a pair of short swing arms.

I learned a lot from that project including how to lace the spokes of a wheel (because I had to have a drum brake) and have always been convinced the leading link is the best type of front suspension. At about the same time I went to Silverstone and watched Geoff Duke leave everyone standing in the 350cc race on his lightweight Ken Sprayson frame Norton and Reynold front fork. Soon after I started racing I was fortunate enough to ride the ‘works’ Greeves which, of course, had their leadin0g link fork with rubber in shear suspension. I had just taught myself to use the front brake going into corners and right up to the apex gradually releasing it as I banked over. I confess that I was not at that stage of my racing career as analytical as I was later (e.g. my mechanic often told me off for being unable to tell him what the rev-counter was reading) but I am sure I found it easier than with telescopic forks.

When we watch MotoGP it is clear that they really do need leading link front fork suspension. When a bike is being leaned over to fifty-five degrees the sliding action of the telescopic fork would not be able to compete with the swing arm movement of the leading link. We never hear riders complaining about ‘chatter’ with the rear wheel although even the rear suspension has a difficult job to work with its pivot at such an extreme angle from the horizontal.

But there is another feature of the leading link type of fork that I strongly believe is in its favour. In fact, it is more than a belief; I know it. The leading link fork is also fundamentally lighter than its telescopic counterpart. I know that the lighter the front wheel/brake / fork assembly can be there is a benefit to the bike’s steering. More precisely, if the moment of inertia about the steering head (or head stock) axis can be reduced the rider will feel the benefit.

Here are some examples starting with fun but the ridiculous. Again, long before I took a motorcycle around a race track I spent almost all my time on my bicycle and sometimes a friend needed a lift. Passengers had the choice to ride on the saddle, the cross-bar or the handlebars (not on my special ones above). The steering and the technique in each mode was different, of course, but the most awkward was with my friend on the handlebars. (Three friends on handlebars, crossbar and seat with me on peddles is not worth reporting!)

When later, I went out on my old Velocette to see if it would run after some work on it I before I fitted the headlamp to the front fork. The ‘feel’ of the motorcycle was different, and better, more ‘directional’.

I took remembered these experiences when I with Norton and had a chance to mae some experiments which led to the John Player Norton telescopic fork with the front wheel spindle on the front of the sliders. This brought the significant mass of the front for 1.25 inches back towards the steering axis. To maintain the required steering geometry the position of the front wheel and brakes assembly which is the major mass is unchanged, of course, but the steering was smoother. Why is this?

I think we can why if we go back to the TV and watch MotoGP. For some reason we have occasionally to see part of the face of one or other of the riders through the visor of his helmet.

Featherbed Frame

Frames and Featherbeds

I have strong views about the motorcycle frame.

It has been at the centre of discussion about motorcycles ever since I can remember. The first one to enter my consciousness was the Featherbed for the Manx Norton. I was in the Isle of Man to watch my first TT with the rest of the Williams family. Even then I could feel the excitement that is always in the air at TT time. There is an anticipation that great deeds will be done during those first two weeks of June. In those years soon after The War it was by far the greatest motorcycle event in the world.

Without a doubt part of the excitement that year was about the Featherbed. I was just a kid and I could not see what a feather bed had to do with a motorcycle. Even now I am not certain about it! Is it the Norton bike or the frame that is the Featherbed? Or is it the Manx Norton –the most successful motorcycle the world has seen?

Nor can I see that it could be just the Featherbed frame that was entirely responsible for the lovely precise Norton steering. The suspension, steering geometry and weight distribution contributed a lot. When I used to cycle to Brands Hatch I would see some frightening antics down Death Hill by blokes on bikes that seemed to steer dreadfully. But none of them were Nortons. Some might say that they were not quick enough to exhibit poor handling! But later, before I started racing, I rode some of the makes of bikes that performed those awful contortions. Sure enough they were discouraging to ride at any great speed on the open roads!

I have to say that I did not ride that many bikes on the road. I had an ES2 for a time that steered beautifully (and it got me out of big trouble once when I had to bank over to an unbelievable angle!) Certainly the Featherbed frame set a seal of excellence on Nortons for twenty years until – for good or bad depending on who you are – it was usurped by the Commando.

Damping

Damping

So crucial to speed that I broke lap record thanks to it, and my dad risked his job over it. But it can only keep getting better.

Where there is a mass and a spring there must be a damper. Over the 30 years before the period of my racing career Isle of Man average speeds went up about 16 miles per hour. Since then they have gone up by about 22 miles per hour. This contradicts the law of diminishing returns and it exemplifies the impressive advances modern technology has made in suspension and damping. There does not appear to be an absolute limit to lap speed in the Isle of Man.

Several developments have enabled this disproportionate speed increase; power has tripled and the development of tyres has had to keep pace. But so has the application of vehicle dynamics without which motorcycles would be lethal.

The springs and dampers are just part of the study of vehicle dynamics. If the two wheels of motorcycles were connected to the chassis as on conventional bicycles the dynamics would be relatively easy but they are not so easy to understand when for reasons of comfort and control you support the chassis on springs and it heaves and pitches as the wheels roll on bumpy surfaces.

These actions have to be steadied by dampers.

Poor dampers turn some classic bikes into camels and lurchers – we have all had at least one. Dampers control the speed of the oscillations of the unsprung masses of the wheels relative to the sprung mass of the chassis and rider forced by the bumps in the road.

My KSS Velocette was a lurcher. It had standard girder forks and naturally it had an adjustable friction damper. I used to play with it tightened up and loosened off and got myself in some uncomfortable situations in the process. The trouble with friction dampers is that the greatest resistance to movement is actually when the suspension is not working, i.e. when it is  not moving. This is not a good thing. You don’t want the damping to be working against you, tending to prevent the suspension from working and nor do you want the damping to remain constant as the road gets more bumpy. The moving parts of friction damping wear and heat up; the clamp load varies with ambient temperature the friction coefficient varies with humidity and so on and so on. The other problem is that you can’t vary low and high speed friction damping independently as you can with oil dampers. On the other hand hydraulic dampers only start working when they are moving and the faster they move the greater the resistance. This is a good thing. I cannot imagine how they used to get those girder fork bikes around the TT so fast. They were hard men.

I managed to get Koni to supply me with the adjustable rear dampers and they made some adjustable dampers for the front forks on my Wagonwheels G50 Matchless. Strangely, but not unusually with racing suspension in those days the Konis were initially disappointing but the stopwatch said I was going faster. (I used to test the bike round Thruxton) so I relied on the watch alone. As we played with the settings I went faster and the road seemed to get smoother until lap times went below the 750 lap record. I like the Kois by then.

The Yamaha monoshock was just coming into vogue towards the end of my career. I really admired the design as well as the reasons for it. One of them is the Square Law. The volume of the fluid inside the damper increases as the square of the diameter so when you double the diameter of the damper the volume of the fluid is quadrupled. One monoshock contained considerably more fluid than one pair of our old Konis. This is an even better thing. Springs control the displacement of the wheels in response to bumps and dampers control its resulting velocity relative to the sprung mass, ie the rest of the bike and the rider. The speed control of the masses can be more precise if there is more fluid available to be pumped through the internal jets.

My dad was very aware of how his riders had to do battle with the dreadful ‘jam pot’ rear suspension units on AJS and Matchless bikes. He made a little test jig like a large camshaft to find how fast it could push the units up and down before they went out of control – like valve bounce in the engine. He compared Girling, Armstrong and Woodhead Monroe units. At the end of the swinging arm was a small 10 inch diameter wheel with a rubber tyre which

Commando Long Studs

Re-designing the Commando barrel

How something as simple as a stud can make an engine more powerful

In 1969 I started working at the Norton Villiers Performance Shop at Thruxton and racing the Production Racer Norton. We were housed in an old WW2 RAF workshop cum offices next to an aircraft hanger. I tried to tell them that production bike racing is not racing at all – not in my book anyway. Eventually, in 1970, it was agreed that I could design a race bike to compete in the new F750 Championship and I could have the excellent Peter Pykett to build and maintain it. I wish I could work now at the pace I worked then. I learned from my father how to design cams and we started the famous ‘S’ series of Norton cams and I designed a “through-bolt” barrel.

One of the weak points of the Norton 750 twin was – and still is –  the base flange at the rear of the cylinder barrel. The problem was compounded because the studs that clamped the flange down to the crankcase, and the crankcases themselves were also weak. When we increased the power of the engine, one or other would invariably fail.

This also ruled out the possibility of using a short stroke crankshaft to get more power by increasing engine revolutions as we’d have needed a bigger bore to maintain cylinder capacity, removing even more metal from an already weak component. There was no option to move the position of the studs further away from the cylinders to maintain a decent thickness of metal.

The solution seemed to me was to effectively move the flange from the bottom to the top of the barrel. That’s to say to use longer bolts, or studs, and effectively clamp from the top of the barrel, what I’d call a “through-bolt” barrel.

I have often found that when you have three problems in engineering design, when one is solved, strangely, the others often go away too. This was not entirely the case this time but they were helped by the consequence of the high flange because long studs are much better than short ones. And we could also use small diameter socket hexagon sleeve nuts at the top of the stud, allowing us to keep a decent wall thickness between the nut and the cylinder. So there were several advantages that reduced the risk of metal failure.

You can tighten long screws and studs tighter and more reliably than short ones. This is because they are like tension springs of very high rate. One way to “torque” a stud to give a required load or tightness is to measure the increase in its length as you turn the nut. The more the stud stretches the more tightly it will clamp together the faces of two or more parts. Like a long spring, a long – stretchy – stud gives only a small extra load when it is stretched a lot. Problems due to thermal expansion will be reduced by using a long stud, and they can also be more accurately tightened – the difference between two different torque settings will require you to turn the fastener on a long stud through a greater arc than a short one.

We had one barrel cast in aluminium using the standard sand casting tooling. To create the new upper flanges we scraped sand away in the mould. The primary reason for the aluminium barrel was to save weight but the long term idea was to save weight AND to increase the bore diameter and coat the bore with Nikasil. Norman White tells me this barrel still exists – complete with foundry mark.

I must say I always had reservations about the prospect of success from the aluminium “Big Bore”. It was highly likely that the leverage of force on the upper flanges – or bosses – hanging off the side of the barrel – might distort it. It might put the bore slightly out of round. This does not happen with the bottom flange set-up but it’s always a game of balancing evils and finding the “least worst”.

However, when Norton needed to increase the capacity of the production engine to 850 (actually 828cc) the through bolt design was adopted for the iron barrel. This barrel was also used on the 1975/’75 short stroke Thruxton Clubman 750cc production racers. These had very good power but unfortunately I never got the chance to use it.

Overmatter

The limit is reached when you keep tightening a screw and it does not seem to get any tighter.  Unless you have stripped the thread, what has happened is that the yield point and the elastic limit have been exceeded. When you then inspect the screw the shank has “necked”.

Some modern car cylinder head screws are tightened to just beyond yield point, i.e. when you take them out they are slightly longer than when they were fitted. They can be used three times before risk of “necking”.

The forces due to the compression and firing strokes of each cylinder are reacted by the fasteners and the metal structures of the barrel and the crankcase.

One of the ways to get more power is to increase engine revolutions by shortening the stroke and increasing the bore

Unfortunately there was no option to move the position of the studs further away from the cylinders to maintain a decent thickness of metal, so we couldn’t increase the size of the bore.