[technicalities]

[TECHNICALITIES]

GORDON JENNINGS

HISTORICALLY, THE DESIGN TREND in reciprocating, internal-combustion engines has been toward the short-stroke/ large-bore layout. Today, with bore/ stroke ratios for most recently-designed engines near 1 : 1, it is often assumed that we are seeing the trend at its leveling-off point; that things may change in other areas, but that in terms of bore and stroke, engines will remain roughly "square." There is reason to think otherwise.

Yet another trend is working to change the situation: the trend toward higher and higher specific power output. We have engines in ordinary touring motorcycles, now, giving the sort of power that would have been expected only from all-out racing engines a few years ago, and considering the enthusiasts’ seemingly insatiable thirst for more performance, it is unlikely that this trend will be halted. More performance means more power, and' that means higher engine speeds — which, in the end, leads us to further reductions in stroke. The best way of handling the matter is to reduce both bore and stroke, and multiply the total number of cylinders, but that is a costly option. You can do somewhat the same thing by simply enlarging the bore and shortening the stroke, and that is what has been happening over the years.

Where will it all end? There is no end in sight. Engines now in existence have proven that an exaggerated (by present standards) bore/stroke ratio works very well, and has marked advantages. The little English Ford passenger-car engine, for example, with an 80mm bore and a 49mm stroke, confounds everything you have been told about ultra-short stroke engines. It gives excellent low-speed pulling power, good economy, and is a marvel of power and reliability compared to its longstroke ancestors. Interestingly, this basic engine has provided the essentials for the highly-successful Cosworth-Ford racing engine, which is one of the few from the world of automobiles having a specific power output on a level with the better motorcycle racing engines. Especially interesting, is the fact that an 80mm x 49mm bore/stroke combination gives us a 246cc unit cylinder, which is right in there in a most popular motorcycle displacement class. Let’s see what we can do working with that primary combination.

It is safe to assume that we can use a maximum piston speed of 4,000 ft/min, and in applying that, the maximum crank speed becomes about 12,500 rpm. That is well above any existing 250cc “single”; yet, well within the stress levels we know to be safe with fairly ordinary materials and manufacturing methods. So, our projected “redline” will be at 12,500, and to provide a safe margin between redline and power peak, we will design for maximum power at 11,500 rpm.

For that operating speed, an inlet valve of 1 3/4" diameter is sufficient, and it (with a 1 3/8" exhaust valve) will fit into the 80mm (3.15") bore without resorting to a deep, hemispherical combustion chamber. In fact, the included angle between the valves would be (according to a quick sketch) only about 50-degrees. Thus, the combustion chamber is quite shallow and reasonably compact even with the very large bore, and there is no difficulty in getting a sufficiently high compression ratio. Finally, the narrow valve angle permits us almost complete freedom in terms of overlap in the valve timing, and very little likelihood that the valve heads would cross and foul at that inevitable moment when the engine is run past the point of valve-float.

An interesting aspect of this approach is that with such a short stroke, the cylinder-head joint can be brought down amazingly near the crankshaft, even when a reasonably long connecting rod length (relative to stroke) is provided. So, it becomes almost unnecessary to go to the bother of an overhead camshaft. Instead, it might be better to use a chain-driven camshaft high in the cylinder. In this arrangement, the timing sprockets would be only 5" between centers, placing the camshaft centerline only 1.25" below the cylinderhead. Now then, by locating the camshaft where its axis meets the intersection of the valve centerlines, we can use “General Motors” stamped cup-type rockers, pivoting (through a spherical joint) on studs in the head. Valve adjustment is thus taken care of without any attachment at the ends of the rockers, and the rockers themselves are, by their very nature, exceptionally strong and light. We can employ the ultra-light Triumphtype cam followers, to save space and weight, and as the pushrods would only be about 3.5" in length, the total valvegear mass would not be appreciably greater than in existing single overhead camshaft designs. Given the help of modern valve springs, this high camshaft engine should rev to 12,500 without valve float

— though I am not prepared to say what the long-term effects would be.

One of the knottiest problems facing anyone designing to such a radically short stroke would be a satisfactory crankshaft. The conventional layout, which has separate flywheels and a crankpin press-fitted, would not work; too much overlap between the crankpin and the mainshafts

— assuming that the pin and shafts were of adequate diameter. The answer is to use the Hirth-type crankshaft construction, like that employed by Porsche and Mercedes-Benz in their roller-bearing crankshaft racing engines. As applied here, this would entail making the flywheels with half of the crankpin on each wheel. The mating faces of the crankpin halves have radial splines, and are pulled together by means of what is called a “differentialthreaded coupling.” Each crankpin half is drilled through from end to end and threaded, one having a slightly smaller bore and coarser threads than the other. To assemble, you insert a special bolt, which also has different threads and diameters at its ends, and simply screw the bolt in. Both sets of threads engage simultaneously, but because the coarse-threaded end “gains” on the fine-threaded end, the crankpin halves move together. This is so effective that an end pressure of several tons can be applied using just a small “Allen” wrench.

(Continued on page 16)

While I have been discussing this hypothetical, ultra-short stroke engine on the basis of maximum revs and maximum power, that is not its only area of usefulness. Should you want to use such an engine for scrambles, or even trials, where low-speed power is important, this can be done by reducing port diameters and changing the valve timing. With these measures, the short-stroke engine will behave exactly like the old long-stroke singles, most people fancy to be the only type capable of delivering the “torque” needed for scrambles and trials.

To carry the matter a bit further, it is intriguing to consider that by going to an 87mm bore and a 42mm stroke, the valve sizes we mentioned earlier could be accommodated without resorting to the hemispherical combustion-chamber at all. In fact, the best arrangement then would be to make the lower surface of the head quite flat, and use the valve-clearance pockets in the top of the piston as a combustion chamber. This has been done in a few engines already, notably the latest Cosworth Formula-II racing engine, and it has worked very well. Actually, with compression ratios in the order of 12:1, we are forced to use the clearance volume for a combustion chamber whether we like it or not.

Exaggerated bore/stroke ratios will allow us to at least flatten the area above the piston and reduce the convolutions in the combustion chamber. There are signs that the more daring designers are leaning heavily in that direction, and it will be interesting to see if the trend develops further. There is a lot to be said for it. Probably, we would be seeing more evidence now, but for the conservatism of managing directors and the like. It is, of course, difficult to find much fault with such conservatism, for the average motorcycle buyer is rather conservative, and no one wants to pump money into a product the public will not buy. But, if someone does come along with some silly little ultra-short-stroke affair with nearly vertical valves and a virtually non-existent j \ linder hv all means buy it. for such i appeal'' to be the shape of the future. 1»