Frank Metcalf Aspin 1913 - 1976
Patents 1936 - 1974
This is all about Frank Aspin and his ideas for a rotary valve engine
Somewhere in Warwickshire
Somewhere in Cumbria
Some art work survives
Some test data survives
My own attempt
Recovered thought lost
In my view Frank Aspin did quite a lot for the Internal Combustion Engine though, rather like the Romans, you cannot see much now without actively looking. It is difficult today to see the diversity of technical innovation, and the uncertainty of direction that was obvious and evident in the 1920's and 1930's. It was far from certain that the poppet valve that is used today would actually be the mechanism of choice. Indeed, the great Harry Ricardo in his book "The High Speed Internal Combustion Engine" thought that the sleeve valve engine would probably be the victor. It ran higher compression ratios on the same fuel, gave better torque, and was less prone to detonation. So it was in this environment that Aspin (and Cross and Minerva and many others) thought that the rotary valve engine should have their chance, others thought that the sleeve valve engines had already won. During WW2 more horse-power was generated by sleeve valve engines than any other type, examples are the Bristol Hercules engine, and the Napier Sabre 24 cylinder 3,500 hp engine.
This article is about Frank Aspin the development engineer, not about him as a person. One Company investor in the Aspin Company was a friend of my mother, and I "inherited" some papers from him. I have tried to get help to build an Aspin engine based on a 600 Yamaha engine, but solving the many problems that plagued him with new materials technology and plastics, building it better by design. I asked for help (just information, not money) from several large concerns involved in the motor trade, they just did not want to know me. I tried several Polytechnics to have a student project started as a joint venture, again no money involved, but they did not want to know either. Why am I surprised when it has happened to so many previous British endeavours before. I tried to get a Regional Development grant and was told that I was wasting my time, all dirty engineering was going to India and there would be no money for me.
Our story starts with Frank Aspin acquiring a Rudge Ulster single cylinder motorcycle engine and converting it to a simple 250cc rotary valve head. This was in 1933, and legend has it that it cost 10/- (that is ten shillings in old UK money, or half of one British pound Sterling). He claimed 18 bhp at 7500 rpm, 31 bhp at 10,000 rpm, with 14:1 compression ratio, and maximum rpm at 14,000. He established "F M Aspin & Co" in 1936. In May 1937 the Belfast Telegraph reported that Aspin, "a young engineer claims to have invented an engine that will enable cars to travel four times the distance on the same amount of fuel". Also in 1937 it was reported in Motorcycle that the Aspin engine was running at 12,000 rpm with a compression ratio of 17:1, but both Aspin and Cross reported that their engines ran well at compression ratios over 20:1. By 1960 Aspin claimed that with further work to be able to give 50 bhp. So, double the revs and four times the fuel economy, such things dreams are made of, and inevitably there were disbelievers.
But before we look at the detail of Aspin's work, I wish to explore the circumstances surrounding his work, and in particular what patents existed in the twenty years prior to his first engine. The ideas laid down in the period 1910 to 1930 are surprisingly sound, and showed a great deal of foresight by some of those engineers. Foresight and understanding that is not always apparent in the patents submitted in the 1980 to 2000 period.
The story of the true rotary valve internal combustion engine starts in 1902 so this is the centenary year for the Rotary Valve, but this section is about the preceding four hundred years leading up to that moment. If you stretch a point then Leonardo Da Vinci was about the first to consider the internal combustion engine when he penned a sketch of a gunpowder powered atmospheric engine. I think that we are grateful that the engine was never built at that time. The brave soul who did was named Christian Huygens and he did it in 1673. In an atmospheric engine the combustion resets the mechanism and the atmospheric pressure does the work. The atmospheric engine became famous years later to pump water out from the Cornish tin mines. The operator of the Huygens motor had to open up the cylinder after the power stroke had finished, and add a new charge of gunpowder to the cylinder and reseal the cylinder. Firing the gunpowder (how I do not know) re-cocked the mechanism.
By 1794 the engine is no longer atmospheric but Robert Street needs to use a lighted taper to fire his turpentine powered engine at every revolution. In 1799 a French patent is granted for an internal combustion two-stroke engine to run on coal gas. So, let me get this straight, after 300 years of development we manually fire each revolution with a lighted taper. Hummm.....
Things are about to change!
We jump to 1860 to see the first useful (IE: non-prototype) engine. Manufactured by Lenoir in France, England, USA and Germany, and over 400 were sold. It was double acting and thus fired on every cycle, so maybe it should be a "one stroke". Developments come thick and fast now, Otto is producing atmospheric engines in 1864, not exactly smooth running, but good enough to beat off the commercial competition and win a Gold Medal at the 1867 Paris Exhibition. However, the engines were EXTREMELY LOUD. In 1876 we see the birth of the modern engine, Otto makes a breakthrough when he designs and tests an engine that compresses the mixture before ignition. Yes, you heard that correctly. Until 1876 no engine compressed its fuel charge! He builds a flat-4 test engine that is so unexpectedly powerful that it blows to bits on the test rig. (Aspin himself does this trick too in years to come).
Two years later in 1878, Dugdale Clerk uses an auxiliary compression pump to produce a 3hp motor, and Karl Benz produces a 1hp motor. Dugdale Clerk said of his own engine "It had nothing in it but slide valves; that is, the whole function of the engine was performed by slide valves, so that I had some experience of sliding surfaces and from my experience I did not like them at all". The birth of the "modern" two-stroke motor occurs in 1892 when Joseph Day patents a two-stroke engine with crankcase compression. The three port two-stroke motor is born, yippee.
At this time, most of the engines were using technology derived from the gas and steam engines, and this was gradually replaced by rotating rotors driven at uniform speed. While many thought that the Rotary Valve and the Sleeve Valve were technologically superior, the reality is that by 1900 the humble poppet valve have proven itself to be reliable in poor operating environments, and was universally adopted. The Sleeve Valve in particular was expected to be the dominant design, even in the 1920's the legendary Harry Ricardo predicted exactly this in his book "The High Speed Internal Combustion Engine". Sleeve valve had been adopted by the Daimler Company of Coventry for their luxury cars, and sleeve valve engines produced more horsepower in WW2 than any other design, and the mighty Napier Saber engine produced 3600 horsepower at combat power from 36 litres to power many a British WW2 bomber. Sleeve valve engines were reliable, mechanically quiet, and expensive to build.
Rotary valves are, by their very nature, usually subject to the heat and pressure of combustion, high friction is certain parts of the cycle, and difficult to lubricate without polluting the exhaust. All of the useful patents will be trying to solve one or more of these issues, and I will be commenting on the patents with this criteria in mind. The position of the valve(s) can create a dead space which will not be scavenged, thus diluting the incoming charge and reducing power, which later shows up and increased fuel consumption.
In order to understand the era better I propose to look at a selection of patents that just preceded the first 250 development engine that Aspin made in 1933. This should provide us with a list of ideas that were under discussion at the time. the key to all this is how to deal with combustion heat and pressure whilst still keeping a gas tight seal and not incurring crippling high friction losses.
|The first significant patent that I can find is that of Vallillee in 1911. His is a flat disc type rotary valve derived from the gas engines. The valve is driven not by chain or gears as in so many applications, but by two rocker arms actuated by the crankshaft, which ratchet the rotor through levers so that the inlet rotates half a turn after induction, and the exhaust rotates a quarter of a turn after expansion, but there are two exhaust ports. An important detail is that pressure is bleed from the high pressure combustion chamber to the back of the valve so that there is no residual pressure on the valve increasing friction. Friction is the Achilles Heel of the rotary valve engine. By moving the rotors in occasional movements and by bleeding the pressure, the author does not turn the rotor during maximum pressure and minimises the friction by bleeding pressure. Good plan. The periodic movement rather than continual movement means that the ports open and close quickly and allow their maximum orifice for the longest possible time. The question of lubrication is ...err... not discussed, but the point of a patent is to discuss one particular aspect of a mechanism not necessarily to address every problem.
|Another very good patent for such an early time. The valves are flat discs, as they were for Vallillee, and they move periodically. This time they are driven by "Maltese Cross" or Geneva Stop mechanisms. No valve movement occurs during the compression/expansion phase. The valves move 45 degrees each time the Geneva-Stop engages, and the actuating pegs are positioned to give correct valve timing.
|I like this one as well, what a fine period this must have been to be experimenting in. The rotor
is now conical rather than flat, but the rotors (yes, two per cylinder) are moved by levers (as Vallillee)
and do not rotate but oscillate in opposite directions. The rotors are conical and concentric, one fitting
inside the other, and the outer moves faster than the inner by a ratio 2.5:1. There is no valve movement
|This patent is about lubrication of the rotor. The design of the rotor is immaterial, but he recognises that lubrication is an issue. He does not recognise that over-oiling will be an issue. The design proposes what will become a standard way to distribute a thin film of oil in a rotating rotor, and for all I know it was already a standard way. The oil supply feeds a series of short slots in the housing, these slightly overlap with slots in the rotor. The overlap of slots allows the rotor to pick up some oil which then moves down the length of the slot in the rotor and lubricates the rotor/housing interface. There is no mechanism for clearing excess oil nor draining it away to a catch tank, so it has to be burned. I guess that did not matter much in 1921.
|Another interesting patent, mostly about rotor oiling using the technique described for Francis above,
and he introduces the idea for epicyclical gearing for altering the rate of rotation of the rotors, so
that the rotors open and close quickly and extend their full open period. His rotor is positioned so
that there is no residual force from the compression/combustion phase, keeping frictional losses to
a minimum. In doing so, he creates a combustion space and flame path that is truly horrible.
Never mind, the epicyclical gearing makes up for that.
Dennison tries to "provide a rotary valve which will efficiently control the admission and exchange of fuel" and "provide a rotary valve which may be used for the control and distribution of matter under pressure. In his discussion he acknowledges the lubrication and friction difficulties of existing designs, in particular their propensity to seize after running under load for extended periods. He deliberately circumvents this by having the combustion forces act perpendicular to the valve axis of rotation, and act of both sides of the rotor. The rotor is lubricated by oil reservoir and slots, light sealing pressure is created by a spring in the drive mechanism. The long transfer passages to the valve, which form parts of the combustion space, give a diabolical combustion shape, the space cannot be scavenged, and the residual exhaust gases manifestly dilute the incoming new charge. A minor disadvantage with rotary valves is that they are partially open for most of their duty cycle, and Dennison suggests epicyclical gearing as a way of improving breathing and keeping the valves open longer.
|Just about frictional loading caused by combustion. He suggests bleeding pressure behind the valve,
alternatively he has a piston exposed to the pressure transmitting force to compress an oil reservoir
which presses on the back of the rotor. This is quite neat I think since the geometry of the pressure
piston and the surface area of the rotating valve can be adjusted to give the characteristics
Echard clearly understood the dangers of friction as this was the focus of his patent. "The object of the patent is to provide a means which are applicable to flat or conical rotary valves in order to obviate all undue friction of the valve member upon its seat while under pressure arising from explosion or expansion." Well said, Sir. He advances three ideas for how the crippling friction due to the conic section being driven up into its tapered housing should be nullified. The combustion pressure is bled to the chamber above the rotor so that combustion pressure presses downwards on the roof of the rotor. The angle contact and friction of the rotor are offset by the area of the roof to give neutral effect. The sealing pressure of the rotor is applied by a preloaded spring in the drive mechanism. Suggested alternatives are to isolate the rotor roof from the raw combustion gas, and to apply that pressure without associated heat by using a combustion facing piston to transmit the pressure through an oil reservoir. This should allow the head to run cooler at the expense of an oil reservoir and some mechanism to keep it topped up. Echard does not discuss how (or if) the rotor is to be lubricated.
|Focus is on frictional loading (as above) but solved this time by arranging for the combustion pressure to force the rotor down a helix and off its seat. Reasonable idea but I do not see what stops the rotor lifting too far and the oil film breaking down. Lift the rotor too far and the seal of the rotor across the inlet and exhaust ports is broken.
|Definitely looks like a forerunner to the first Aspin engine, this patent describes how to achieve good gas sealing using a rotor with 120 or 160 degree included angle conical rotors. Oiling is by the overlapping slot type of design, and the load is taken on the rotor face. I think that friction during the compression/combustion phase would have been significant. The Reverend W.H.Peacey was the rector of Mordiford near Hereford, and also an engineer. He rebuilt a 1929 Sunbeam Model90 single cylinder 500cc engine but with rotary valves, maybe inline with his patent but that is a guess.
|Definitely looks like a forerunner to the first Cross engine. We have a rotating cylinder valve opening the combustion chamber to inlet and exhaust ports. The design allows for the cylinder to be the length of a multi-cylinder engine and to operate all the cylinders in the same fashion as a modern overhead camshaft. Moreover, the cylinder is water cooled to prevent distortion, and the hot water helps vaporise the incoming charge. The cylinder is slightly tapered and rotates inside an outer tapered sleeve so that a small axial movement of the stationary outer sleeve will adjust the running clearances. In a modern version, this axial movement would be controlled by oil pressure or engine management system to give correct running at all times.
|Very similar to 1928_Luyckx above but concentrating on valve sealing. That fact that there is a patent focused on valve sealing and not on the basic design, leads me to think that the design is already in use though not necessarily patented in mid 1920s.
|Even by this early juncture, the primary players had a grasp of the problems even if they did not offer a complete solution. By grasp, I mean that they understood that these problems existed, that they were serious and needed to be solved, and they had a measure of their nature. All the significant ideas used in the next 70 years are already in place and discussed by 1930, including the flat discs of the DMW, the vertical cylinders and cones of Aspin, and the horizontal ported cylinders of Cross. Rotor friction is the sin, and many designs succeed to not have any rotor movement at all during the high pressure phases. Those that do have rotor movement, have schemes to lift the rotor by hydraulic or mechanical means are suggested, bleeding pressure so that the high pressure is on both sides of the rotor is common practice. Keep in mind that the goal is to deal with combustion pressure without incurring substantial oil burning or excessive friction. With hindsight, we can say that a compete solution is still being sought by a handful of souls in the twenty-first century, but most of the large scale hope for rotary valve engines has evaporated.
Frank Aspin trained as a draftsman at the Lancashire Steel Company, so was knowledgeable of materials. He would certainly have had access to the patents listed above and drawn some conclusions from them. In 1933 he started experimenting with a 250cc head and barrel fitted with a rotary valve, built onto the Rudge Ulster crank and crankcase modified with strengthened main bearings. The basic design may have drawn on the earlier work by the Reverend Peacey from Cheltenham who had similar ideas, but the Aspin design had a steeper cone angle. The Rudge engine was a popular choice in the day, as Cross also selected this engine to build his prototype. In the prototype engine, Aspin made a cylinder head (presumably from solid) with a cylindrical rotor having a port cut in its side driven from the top using helical gears, the motor dimensions were 67mm bore and 70.5mm stroke. The diagrams with this section show the three stages of rotor development during the formative years, starting with a cylinder then proceeding with a hybrid conical section driven at its mid-point to a full conical shape driven from the top. In the first design, the cylinder rotated at half engine speed and the sparking plug was in the top. Sealing at the bottom is by a small beveled edge, and there is a piston ring around the top. Gas tightness around the port depends on quality of fit, so machining and distortion were important. His first patent was applied for in 1935 and granted in 1937, and claimed a "high power engine capable of high power without preignition and large ports for improved gas flow. A revised patent (applied for in 1938 and granted in 1939) was submitted to combat problems with wear and friction with respect to gas sealing. It advocated the use of tapered roller bearings " .... aforesaid patent in which unusually high compression and explosive pressures are obtained, it becomes difficult if not impossible an anti-friction bearing of such dimensions to avoid overloading." This, in a nutshell, is the problem that Aspin never did solve. Sealing and friction are the two life-long bug-bears of the rotary valve engine. During the next six years the patents come thick and fast, sometimes with such minor changes that you wonder why they were submitted at all, and why they were granted. The thought "why was that patent granted" occurs to me quite often when examining the rotary valve patents granted, particularly concerning those in the period 1980-2000.
The first design from Aspin was of a straight forward cylindrical rotor with conical top spinning in a regular fashion co-axially with the bore, without any compensating mechanism for combustion pressure, and without significant oiling arrangements. Compared with 1921 the goalposts had moved considerably. Granted that the petrol was not much better, and preignition was still a major problem, but the rev limit had climbed from around 2000 rpm to 7000 rpm, and metallurgy had improved. The higher compression ratios and combustion pressures have caused significant sealing problems. With a running clearance of 0.0015" the rotor would certainly have been pushed against its conical housing with some force and friction would indeed have been a problem. To improve the port sealing, Aspin had the inlet and exhaust tracts in the cylinder head as sprung loaded tube so that the tube was forced against the rotor. In addition, slots in the sealing face of the rotor housed sprung loaded sealing bars to further prevent pressure loss around the ports. Lubrication was by pressurised oil feed to the thrust races (which supported the vertical load) and onward through a bleed hole to an annulus in the rotor head leading to vertical slots that hold the sprung loaded port sealing bars. The oil is smeared around the rotor and is eventually burned. The problem that was well known, and described by Dennison some twenty years earlier, was that lubrication was a serious issue.Enough lubrication for low revs and low power is totally inadequate for high revs and high power, and seizures are inevitable unless the supply can be precisely controlled and the excess is reclaimed not burned. To be effective the rotary valve has to be able to perform along side the best-of-breed poppet valve engine, and beat it on performance or cost. Performance is the headline seeker, attracting news coverage and investors, but the performance was elusive it seems.
The first design was quickly superceded by a design of part cylinder and part conic, again with the spark plug in the centre. The combustion space in Mk2 is eccentric permitting an increase in the area of the port. This rotor was made of phosphor bronze running in a cast steel insert, which itself was a press fit in the alloy cylinder head. The performance was good and with a running clearance of 0.006" the oil consumption high, and the spark plugs had a bad time of it. In the patent relating to engine#2 he contends that improved bearing stiffness is needed to maintain an oil film around the rotor and avoid overloading the thrust bearing, and maintain a good gas seal. This engine had a 3/32" steel shell over an aluminium shell forming a 60 degree included angle cone whose lower diameter was 3 5/8"
Another patent (submitted in 1939) is devoted to rotor distortion by liquid cooling the rotor. It did not dramatically reduce the problems but it did prevent the central location of the sparking plug, Aspin was forced to move it to the side and position it so that the port would uncover it at the appropriate time. This also means that the sparking plug was shrouded from the main combustion space for much of the expansion stroke, and a substantially "hotter" plug needed to be used. one that was markedly less prone to oil fouling. Engine#3 retains the 60 degree cone. The spark plug being at the side is thought to fire a relatively rich mixture, the larger fuel droplets flung to the outside though overall the fuel:air ratio was quite lean. The rotor is now case-hardened nickel steel running in an aluminium bronze housing. Performance was again very good but the motor needed careful warming through since aggressive use when cold would be certain to cause a seizure of the rotor.
Just a few months later than the patent above, the next patent suggests that a rotor with a wavy groove is needed. Really? A wavy oil groove is the subject of a whole patent? Low pressure oil is fed through the conical housing to a wavy groove in the rotor, thence to vertical grooves in the conical housing to lubricate the whole rotor. Additional measures are required to prevent loss of lubricant at the port boundaries. Sharp faced grooves apparently scrape oil away from the ports. I say apparently as a personal note as I do not belive that after 10,000 miles of operation that such edges would still exist. I generously conclude that the surplus oil moves up the cone to lubricate the top of the rotor and up to an overflow to lubricate the drive gears and bearings.
1940 and another patent, back to a solid rotor and multiple thrust bearings in a water cooled head. OK, reasonable idea, but water cooled heads and blocks are not new and are well established in production. I cannot see what is being patented here, and yes, I have read it twice. 1940 is the end of our Early Years section, and with the war effort well under way, it seems a good time to review what Aspin was doing outside his patents. I will also review what the press said during this time.
During the Early Years, and indeed for many years after, Aspin was able to build an engine with good reliability but at the expense of burning oil, if the design or clearances of the rotor were such to stop any oil being burned and clean up the exhaust then the reliability dropped out of sight. By detail design Aspin tried to find a middle way including special metal coatings on the rotor, nothing was very successful, the whole situation is too marginal. It needs to be simply better by design, it needs to be more obviously correct and less marginal and better able to take abuse. Fabricated rotors, water or oil cooling, dual input ports with the first port to open being open to atmosphere (remember Helmut Fath's URS sidecar?), all were tried with benefits and drawbacks.
|The Early Years period ends with the Aspin modified Velocette 350CC being
tested without success by the Velocette factory in Hall Green Birmingham in 1936. Frank Aspin was a
superb salesman, and his staff worked many additional hours to make the deadlines, and Frank had
sold Velocette on developing a rotary valve engine for which they could retain exclusive rights
for twelve months. Velocette needed some results just then as they had recently had some
disappointing TT results, and Harold Willis heard some magic words.
It turned out many years later that Frank had sold the same exclusive deal to BSA, but at the time nobody found out. The Velocette worked on a 350cc single, and the BSA worked on a 250cc twin. At the BSA they were more focussed on their new Gold Star engine project, and as these projects were "secret" they did not appear in the press, so few details emerged. The photo to the left is an Aspin BSA head, but whether it is *the* head from BSA is unknown.
The only report that I have came from a sceptic so the view is slanted, but it is clear that despite several suggestions from Frank Aspin, the factory was not able to have the engine make better power and reliability then the current best of breed. The engine was given to Aspin to make of it what he could, at his own expense. Encouraged by press reports, and persuaded by the claims of 35 bhp expected by Aspin, the factory has spent considerable time and money on this experiment. The standard best of breed 350cc Velocette produced 22 bhp at that time. The engine was to have been tested to destruction on a test rig but the Aspin works never really got around to it. It was lying in a corner some years later when spotted by a Jack Ashworth, who was loaned the engine for a 20,000 mile extended road test provided that the engine was not opened up or stripped, and that the engine was returned to the works at the end of the test. Together with the journalist "Wharfedale" the engine was mounted in a BSA M20 chassis and featured in several articles. When the test was completed the bike was ridden back to the Aspin works, but the staff were too busy to be bothered with it, and it was left to Wharfedale and Jack to dismantle it. At least with their journalistic background they thought to take a photo! The engine was in excellent order (no mechanical problems since this engine burned oil!) and in the rebuild it was decided to fit a higher compression piston at 10:1. That came as a surprise to me as the original development engine ran at over 12:1, so why build the Velocette engine at only 8:1 ? No wonder the engine was "soft", but it could pull high gears up hill from practically zero revs. To my knowledge the rebuilt engine never features in any further report of Motorcycle. Regrettably, Jack Ashworth died a few years ago and I cannot therefore ask what happened to that bike.
Laurie Bond seduced Norton with a similar dream nearly twenty years later, and again the final result was a rotary valve motorcycle engine that was less reliable and less powerful than the best of breed existing engines.
While on the topic of the Velocette factory in Hall Green, we should discuss the most vocal of the "Aspin unbelievers", a young engineer named Phil Irving. He did not believe a word of it. The idea that a 350 Rudge could rev to 14,000 rpm was not credible. His views and my comments are below.
It may all have been a mistake. It was certainly not a scam, as Frank was fully committed to the project, and he was funding it with his own money rather than primarily asking for other people's money. I summarise here what is the sceptic view of his work. Look again at the original claim for a modified Rudge Ulster, over 30 bhp at 14,000 rpm. this is far and away above any power or revs from a Rudge of the time or indeed any other motorcycle engine of the time. they would literally have blown themselves to pieces. Nor would the main bearings and big-end's have taken kindly to 14:1 compression ratios, but these can be (and were) strengthened in the Aspin development engine. The real difficulty lies in the fact that there was no independent testing of Frank's work, and that all the articles and news reports are adapting press releases made by Frank personally. Any engineer or scientist who does not expose his work for peer testing risks being labelled a sham.
Here we have an engine running at 14,000 rpm where all contemporary engines, including the Rudge 350 on which it was based, could only make half that. The sceptic view is that Aspin had the wrong rev counter drive fitted. Not such a silly suggestion, actually, though you would think that you could easily tell the difference between 7000 and 14,000 rpm by ear without a rev counter. The mathematics of bhp means that if you get the revs wrong then the bhp is wrong for any given torque measured on a dynamometer. If the rev counter drive was 2:1 not 1:1 then the real revs are 7000 not 14,000 and the real power is 19 not 38. Some of the real world testing showed the engines to be lack-lustre and prone to seizing when worked hard.
The first major article in the press was Motorcycle in 1937 and it aroused much interest. This and all subsequent articles were based on material provided by Frank Aspin, and to my knowledge no independent tests were ever made on his engines. Many of his engines were very satisfactory and quite mundane, not producing much more power than one of a conventional design. The engine that caused the excitement was Aspin's first 250cc development engine. The article describes it as having "phenomenal performance" and "comparable with a 500cc Super Sport engine of normal design". A later article in Motorcycle by Warfedale followed up, but no real details of this remarkable engine were ever released. Some details of a later aero-engine were released. It was a 1731cc flat-4 giving 80 bhp at 5000 rpm. This power is 46 bhp/litre and is well down on the stunning 120+ bhp/L given for the Rudge test engine. The article explains for the aero-engine "For road use, much higher [rpm] speeds could be used, of course, and the power would be almost doubled." So we have improved a measly 46 bhp/L to an estimated 84 bhp/L by higher crankshaft speeds and appropriate port timing. Actually 84 bhp/L is more than excellent for the period, casting some doubt whether it really could have been achieved. Even the mighty 900cc Kawasaki Z1B of 1975 only managed 95 bhp/L and that was a landmark engine and more than 30 years in the future.
I cannot understand why no independent tests were run on the 250cc development engine, why magazines and manufacturers were not clamouring to confirm the tests. Maybe they were, if so, no comments of complaints were printed. Companies like Ford or Vauxhall could easily have built test engines of their own. The reporter Warfedale remained active and interested in the Aspin engines, writing articles including a 20,000 mile road test. The journalist LJK Setright was a long time supporter as well, and wrote appreciatively of his work, and even accompanied him of a trip to Italy to meet with Fiat. He wrote "I have no hesitation in saying that the Aspin arrangement could be the answer" and added "he is one of those perfectionists who always want to defer finalisation of design because he has another idea that should make it work even better. Sometimes I think that he is his own worst enemy". The entrepreneur and industrialist Mr J C Bamford (of JCB fame) was a fan. I had several letters and phone calls from him and he was most helpful. He did not however still have the Wolesley 6 car that had been Aspin converted. Even as late as 1945 the journalist Costello said "the design is undoubtedly of the greatest interest, and on this account certain features and some performance figures call for comment".
In this initial period, the only contemporary competitor seems to be Rolland Cross. He started slightly before Aspin, and his engines gained slightly greater acceptance. He had his share of failures and problems, but generally overcame them, nevertheless his engines were bypassed and he suffered the same fate as Aspin. In the next period there are two major newcomers to the rotary club, they are DMW and NSU, so in the next section I will look at Aspin, Cross, DMW and NSU.
This is the most interesting period historically in the tale of the Aspin engine. The first section dealt with the initial development and ideas as laid down in a series of patents, and we are able to discuss what patents were published previously and how they might have helped or influenced Aspin. This second section deals with the maturity of the Aspin design, and the competitor designs that appeared. The final section deals with the subsequent development of Aspin's ideas by others, and how the rotary design has been taken forward. I leave it to you Dear Reader to judge whether the post-1970 activity has contributed anything significant to the discussion, or not.
|First we will review what Aspin himself did to improve his engines during this period. We left the 1930's with patents covering both oil cooled rotors and solid rotors in a water cooled head. For purely practical purposes he suggested a method to make the rotor assembly and housing to be accessible and changeable from the top of the engine without the need for the cylinder head to be removed. This would have been of great interest to the fleet operators. In 1943 there are three patents concerned with replaceable or service exchange units. The initial patent suggests the idea, then one adds a sealing ring to make the assembly installation gas tight, and another to spring load the assembly as well as using a sealing ring. The former insert is water cooled (at least in the drawing), whereas the latter has double tapered roller bearings not just a ball bearing. I am not at all sure why two patents, and why the difference. Having a sealing ring the subject of a patent seems nonsense. Why radically different bearing arrangements? The two patents were applied for at the same time so it must have been deliberate.
A patent in 1944 extends the function of the service replaceable unit by making the assembly moveable up and down by means of a cam controlled by the throttle and engine vacuum. In this way the compression ratio of the engine can be varied according to load. The purpose of the patent was to improve thermal efficiency during part-load conditions. Another 1944 patent adds a cush-drive rubber insert into a split drive-dog to reduce backlash and help keep the teeth engaged. The advantages are longer component life, quieter operation, and the damping of any drive resonance. A third patent of 1944 suggests that rotors should be cooled by forced convection where fans or forced induction in aircraft is used. Quite strangely, this patent (in the drawing but not in the text) introduces an offset rotor. In the fourth patent of 1944 shows a liquid cooled rotor driven from the side by using helical gears, the rotor does not permit any service exchange, and the rotor is not co-axial with the cylinder bore. But such an engine was in the drawing of a patent granted in 1942, so what is the patent trying to say?
1944 Squish head
|In the fifth patent of 1944 the offset rotor is formally documented, with the aim of reducing gas leakage and shielding the bearing surfaces from excessive heat. the total surface area exposed to the burning gas is reduced, and a considerable squish band is introduced. The drawing shows the squish very clearly, but the text does not refer to it at all, so it is not clear that this was a deliberate design feature, of whether it was coincidental. The improved gas sealing and the addition of squish certainly make this the best design so far. I think that it must have been coincidental as a patent fully describing squish to aid combustion would have been invaluable in later years when the two-stroke technology took off, for the two-stroke a squish is paramount. If Aspin had squish fully covered then I am sure he would have pressed home his advantage.
In his sixth patent of 1944,and the last for more than a decade, Aspin consolidates many previous patents by showing a service exchangeable unit with an offset rotor using the split cylinder head introduced with the offset rotor, high squish, and improved sealing rings.
1961 split-port rotor
|After this mammoth outpouring of 1944, Frank Aspin is quiet until 1961 when he launches another good idea: that of a split charge. In this patent he uses a split port to split the charge, the earlier opening inlet port is open to atmosphere and the inducted air will partially flush the residual unburnt gas from the previous cycle, the later inlet port is carburetted to provide a relatively rich mixture for easy firing. Overall, the mixture is stoichiometric and could be described as a lean-burn engine.
|DWM Motorcycles of Segley (between Dudley and Wolverhampton) were interested in rotary valve engines for a number of years. In the 1960's they were involved with , and provided a chassis for, the rotary valve motorcycle engines developed by Alpha Bearings of Dudley, the Alpha Centuri. Back in 1950 they built their own two cylinder rotary valve engine using a conical rotor. The rotors were driven by a common central dog gear, so that the two cylinders rotated in opposite directions. Each cylinder had its own inlet port, and they shared a common front exhaust port. Their patent does not describe how the rotors will take the combustion load, presumably on the conical face thereby incurring considerable friction. The patent does not describe the lubrication mechanism nor how the excess oil will be scavenged rather than burned. I would like to see some evidence that such an engine was actually built, and ran.
|The NSU Company from Wurttemberg Germany made a serious attempt at a rotary valve engine, and was granted a patent in 1955. In contrast to the rather superficial DMW patent, NSU acknowledge the difficulties to be faced. They describe the sealing surface of contact between the rotary valve and the cylinder head as being a region of high pressure and high temperature, suffering sliding motion with metallic contact, required to resist combustion pressures, and transmit combustion temperatures to the cooled cylinder head. They claim that their design will "entirely or to a great extent" avoid these difficulties. The text does not convince the reader that this will be so. The rotors are mounted in a single ball bearing, and the whole assembly is air cooled by fan. The only convincing detail of the patent is that the port sealing is effected by an insert that is separate to the rotor and the combustion pressure is bled behind the insert to push the insert and ensure a gas tight seal during the combustion and exhaust phases.
Throughout the whole of Aspin's professional life there had been one constant competitor in the race to make
a successful and universally adopted rotary valve engine. He was Rolland Cross. He had his own ideas, and
at no time ever borrowed ideas from Aspin. Rolland Cross was located in Bath, and started as early as 1922,
so he had some ten years on Aspin to get his ideas established. His first patent was 1925.
Whereas the Aspin design was a cone spinning vertically in the cylinder head, coaxially with the cylinder,
the Cross design was a cylinder running perpendicular to the cylinder and parallel to the crankshaft. In his early work,
Cross like Aspin had some remarkable power figures, in this case he claimed 70 bhp/litre for his engines.
The location of the valve made it easy to drive the rotor directly from the crankshaft by chain, shaft or bevel gears. An early four cylinder car engine was considered to be very satisfactory, and the design allowed the rotor to be water cooled for its entire length. Cross was one of the first to identify and discuss the problems of rotor lubrication. Contrast this with the many others who seem more interested in having a patent than in actually making an engine. This thought also occurs to me in later decades. Cross thought that you could add oil in the smallest quantities to provide satisfactory lubrication, or supply oil in excess and remove it by some scraper device.
The rotor was pushed against the open port in the top of the combustion space, resting on raised lips around its edge. This raised lip was generally about two thousandths of an inch (0.002") high and 3mm wide. This was found to be sufficiently resilient to seal and resist wear, and be sufficiently flexible to take up small inaccuracies in the rotor. The head and rotor bushes were split along the centre line of the rotor and this allowed for any thermal expansion of the rotor. While this was overwhelmingly satisfactory, it was possible to overload the rotor, and the resolution to this concerned Cross for some time. These problems were resolved by the mid-1940's using a system of controlled valve loading. All the problems of thermal expansion were apparently resolved and it was possible to start the engine and drive away at full power without ill effect. The basis of this controlled loading is to use a reaction to the combustion force to apply pressure to the top of the rotor to compensate for the pressure trying to push the rotor off its seat. In addition to 250cc and 350cc motorcycle test engines, a number of light aircraft engines were built: 4 cylinder 100mm x 100mm to produce 150 bhp from 3.1L (48 bhp/L).
In 1944 there was another flurry of journalistic activity. Autocar had tested a family car with a four cylinder engine. They reported 64mpg at 30mph, 50mpg at 40mph, and 45mpg at 50mph. Also ready in 1944 was a 4.6L truck engine. There were reports in the press of a flat-4 aero engine and a 650cc twin engine. Wharfedale wrote about an aero-engine flat-4 as being complete, there is a drawing and a photograph. It was an all alloy engine with steel cylinder liners, 1.8L, 80 bhp @ 5000 rpm, 130lb weight and 83mm bore. A number of these engines were made, and most were exported to the USA where interest and purchasing power seemed to be higher. In a much later article by LJK Setright, there is a photo of the 650cc twin and a short description. It is now certain that it was a 650cc top end for the Meriden Triumph Bonneville, and this project dates from just before the Meriden collapse, so I guess 1971. In a roll of drawings from the Aspin family, I received detailed drawings of the Aspin barrel and a batch of Meriden Bonneville drawings showing the Triumph crankcases and timing side covers. This Aspin engine was to have variable speed valve timing, which was the subject of a 1971 patent application (see below). A flat-4 aero engine had been tested by the Royal Engineers. It was reported as a 1.0 litre engine and may have been the prototype for the 1.8 litre flat-4 aero engine made later. This 1.0L engine was tested for 2000 hours on a rig and then for 52,000 miles in a Morris Minor car, before being tested for fuel economy and showed itself to be significantly better than the control engine.
In the 1960's the Aspin engine was certainly much more refined now, and so it should be after 30 years of development, but the interest in the engine was not self-sustaining and Frank seems to have to fight for ever piece of publicity. Ford UK were interested enough to send two or three Cortina engines to be converted for evaluation, there was a Wolesley 6 car and a bus in Northern Ireland. In some unpublished lecture notes from the late 1960's, Frank Aspin claimed that his engines with "suitable investment" would be able to produce 300 bhp/ litre. Exaggeration or what! 300bhp/l would do well in a motorcycle GrandPrix so to claim this before 1970 is just unbelievable, and I guess that no-one did.
The Northern Ireland bus had a converted 8.6 litre Leyland truck engine. The vehicle ran successfully for many thousands of miles until oil burning
problems started. The cause was the breakup of the special metal coating on the rotors, and it/they were withdrawn from service. The Company did not
demand its repair or replacement so the test could not have been convincing. Better, yes, but maybe not worth the additional effort or special consideration
during servicing. The engine was repaired and used as a electrical generator for the Aspin works for many years, then it powered a saw mill.
This engine later became part of a job-lot involved in the renovation of a single decker bus RH206. The Aspin engine was one of several engines offered to power the bus. The engine is a petrol Aspin top end on a diesel bottom end. It is this engine that is photographed elsewhere.
In a letter dated 1966, he is planning to sell conversion kits as an after-market item though a dealer in Knutsford who had shown some interest. It seems a bit late in the day for this, and I don't think it ever took off. By this time, the original engine company had lost its patent rights due to the passage of time, and Frank was worried that someone might steal his ideas. He describes how he has now a new and simplified design with a "completely different design of rotor, a different combustion space, and operates at higher pressure". The new design was to be the subject of new patents and was "therefore very valuable". Although it was difficult to see from the inside, after 30 years of hard effort to get the engines accepted with very little success, I doubt that there were many (or any) persons waiting to take over and make their fortune.In 1978 Aspin had his last patent, and potentially his best. If he had kept the offset rotor and the high squish the it would have been near perfect IMHO, but he reverted to the cylinder and rotor having a common axis. In this patent he suggests a split charge induction so that the early air-only charge will scavenge and the later fuel-rich charge will not be contaminated. The two inlet tracts are throttled by a mechanism which can vary the contribution proportion according to engine load and effect a fuel cut-off when the engine is on overrun.
BSA in conjunction with Dr Gordon Blair of QUB fame open the decade. They are concerned about the symmetric nature of piston ported two-strokes, and that the "normal" disc-valve arrangement with carburetors on the side cannot be used with three an four cylinder in-line engines on a common crankshaft. their proposed solution was a rotary valve in the inlet tract to allow asymmetric inlet timing, very much like the way a reed valve was used by Yamaha in their engines. In the same year (1971) Henvaux of Belgium sought to simplify and improve the poppet valve by introducing a rotary valve (Cross style) driven by a Geneva-Stop mechanism. This is almost identical to the Ted Mellor rotary valve New Imperial of 1940, itself perhaps drawing on the ideas of Mueller from 1914. Ted Mellor blended in the 1921 Francis suggestion to use tapered conical rotors that were pressed into their housings by springs. Since the Geneva-Stop mechanism (GSM) produces a very intermittent motion then this allows the stationary rotors to give off much of their heat very easily. A face cam lifts the rotors off their seats by a few thousandths of an inch so that they will turn freely and without undue lubrication when the GSM rotation is due. The engine ran very freely and revved well I am told but the exposed valve gear gave off a definite chatter, enough to prohibit nearby conversations. The engine is still mechanically complete as far as I cold tell about 2000, but whether it would actually run I do not know. The Geneva-Cross mechanism works well but is very noisy, but it does have a quick action for opening or shutting the valves. In a 1971 patent, Aspin tries to gain the advantages of quick open/shut without the noise, and he does this by suggesting a means to vary the rotational speed of the rotor in an eccentric fashion so that the valve is open longer with respect to the crankshaft. It was this variable speed motion that was to be used in the 650cc Triumph Bonneville conversion, that was shortly before the Meriden collapse. In 1977 an American named Gentile suggested rotating cylinders or balls in the cylinder head to control the porting very much as BSA had done earlier in their two-stroke inlet tracts. There is no discussion of oil or friction, only that the seals are "sprung loaded" by gas pressure bled from the cylinder head. The arrangements looks very similar to the current configuration by Coates Engines. The inlet and exhaust valves are controlled by separate rotors, and all the inlet rotors are driven by a common drive shaft, as are all the exhaust rotors by their own shaft. Each cylinder has its own inlet and exhaust ball. Exactly the same arrangement is proposed by Hopkins, also an American. The RCV company of Dorset UK do (or did) produce an Aspin derivative engine (wrongly attributed as Aspen in their notes) using a rotating inner liner to the barrel. Nice idea, but Ho Hummm it has been done before (and not attributed). But RCV actually made it work, well done Lads.
What really strikes me is how rotary valve ideas are being recycled without adding anything that I can see.
What I think just awful is that most of these newly patented engines are not even built. To those who did build these reworked ideas, whether or not their alleged improvements worked, you have my respect.
With that in mind, I have selected a few examples of those that were built.
|Took ideas from Aspin (1936) Edwards (1930) Milner (1984) plus modern knowledge on sealing rings to make it work.
In principle, an excellent idea as long as friction and sealing is thought about carefully.
|Rotating inlet valve to achieve asymmetric inlet timing. BSA sought the same with QUB Dr Gordon Blair
but didnt make it work.
Original idea by DOT and made to work, but didnt work well
Good effort by DOT.
|Ford Motor Company
|Two rotary valves but combustion load taken by poppet valve. Common space dilutes charge and spoils fuel economy. Unless the incoming charge is going to be flushed by a final pulse of fresh air, this is not going to be effective. The Darracq Motor Company built a car using this scheme, it almost ruined the Company.
|Coates Spherical Valve Engine
|Rotating ball valves with holes in. Similar in concept to the Cross engine and many others except
having an individual ball valve for each valve each cylinder. No reason why it wont work quite well, they claim just less than 100bhp/litre
which is quite soft power, less than most sports street motorcycles.
OK, why dont we see every other family car using this type of engine?
In the earlier patents of Tebaldi and Doutre I have blurred the requirement slightly, since these have rotating sleeves not individual balls, but the passages are individual to each cylinder. Coates probably rely on liquid cooled castings and high temperature coatings for their spherical balls, whereas Doutre working in primitive oils and cast iron uses sprung loaded tapered sleeves to automatically keep the clearances constant.
From the patents that Frank Aspin submitted, there were two ideas that might have made his fortune had he understood them and protected them. I am not at all sure that he understood their worth, but of course, I have hindsight.
|The ideas behind this patent from 1969 can be readily seen in variable valve timing drawings for curently available family cars. Most modern family and performance engines have variable exhaust valve timing dependent on speed and load, controlled by oil pressure or computer. Aspin wanted to open the valve ports quickly and dwell for a while, then close rapidly, effectivly lengthening the valve period. The amount of extension was fixed by the geometry of the components as there was no need to vary the valve dwell at that time. The solution is the same even if the goal is different.
|This patent from 1944 is fundamental to the performance of all two-stroke motorcycle engines since Yamaha used it in their 1970's TD and YD engines. The squish fires a high velocity gas jet to massively increase turbulence and shorten the ignition burn time, and the close clearance of the piston to cylinder head at TDC will prevent early detonation.
My thanks to the Aspin family for permission to use anything I can find, including some personal letters and lecture notes which I acquired independently.
And borrowed from the archives, material published in periodicals 1930 - 1950
LJK Setright, Some Unusual Engines
I Rhodes, Velocette Book
P Irving, his autobiography
M Hunter, Rotary Valve Engines
H Ricardo, High Speed Internal Combustion Engine
European Patent Library http://ep.espacenet.com/