Technicalities

TECHNICALITIES

GORDON H. JENNINGS

IN THE FUTURE, you will probably hear a lot about "short-rod" engines. Tuners are becoming increasingly interested in this trick and it is one of the few speedtuning techniques not already fully developed. Unfortunately, while the basic idea behind the short-rod technique is sound, it is not without its dangers and disadvantages.

The technique consists of installing a connecting rod, or rods, having a shorter than standard length (from the piston-pin center to the crank-pin center). This may be accomplished by making a new rod, or by shortening the standard rod.

Benefits, if any, are derived because with the shorter rod, there is a greater displacement of the piston per degree of crank rotation at the ends of the stroke. This gives better volumetric efficiency (the theory goes) because the more rapid rate of piston displacement as the crank swings down from top center will give the column of air in the intake tract a sharp tug, and start it spilling into the cylinder more rapidly than would be the case with a longer connecting rod. Also, it may be that some improvement in power output is realized because the more rapid initial displacement of the piston expands the super-heated combustion gases more rapidly and thus uses the energy represented by the heat before it can be lost to the metal surrounding the combustion chamber. In practice, some engines have benefited from short-rodding, so there is at least some evidence to support the theory.

Before everyone rushes out to build the shortest of short-rod engines, they should know that there are some disadvantages to the scheme. It should be obvious that if there is an increase in piston movement at top center, relative to crank rotation, then the piston and everything connected with it is going to be subjected to greater inertia loads. Particularly, the piston and rings bear the brunt of the load and will be first to fail if the loads are too high.

An English research scientist, Paul Dykes, conducted an extensive investigation into the effects of inertia loads on pistons and piston rings. Dykes found that piston rings would cease to seal if acceleration was sufficient to overcome gas pressure and they lifted from the floor of their groove. It was also discovered that piston acceleration in the order of 100,000 ft/sec2 was likely to cause piston failure. Ring problems were found to be related quite closely to ring width. Wide rings are heavier, and their inertia will overcome gas pressure at a lower rate of piston acceleration than narrow rings. Very thin rings will continue to function up to circa 100,000 ft/sec2. Above that, only the Dykes-pattern ring (developed as a result of the research program and described in last month's “Technicalities”) performs with any degree of reliability.

(Continued on page 26)

With all of my reports and technical papers at hand, I know that there are sharp limits to how short the connecting rod in any given engine can be made. Yet, there are examples that kick theory right in its lofty backside and get away with it. My good friend and sometime co-conspirator, Jerry Branch, has constructed a HarleyDavidson KR side-valve 45-incher that defies all the rules. This engine’s 3.8125" stroke pumps its pistons up and down a trifle too vigorously with rods of any length, and Jerry has cut a full inch out of them, leaving a center to center distance of 6.437". The resulting stroke/rodlength ratio would not be too bad if the KR engine were being run at the unhurried 5500-6000 rpm to which its lengthy stroke is entitled. Unfortunately, maximum power is up around 7000 rpm, and the gearing is selected to give a maximum of 7500 rpm at the end of a flat track's short straightaways. With the stroke and rod length involved, piston acceleration is slightly above 110.000 ft/sec2 at 7000 rpm, exceeding 120,000 ft/sec2 briefly at maximum revs. Piston failure should occur, but it does not. Ring problems are another matter. With the standard ,0625"-width rings, Dykes’ research data indicates that ring flutter should be experienced much above 80,000 ft/sec2, and there is some indication that this is, in fact, just what happens.. However, the blow-by that accompanies ring-flutter does not become too severe, and while ring failure does occur after a time, the rings will last long enough to make the set-up practical. Most certainly, the engine is strong; Branch’s KR, with the talented Jim Nicholson at the tiller, has shown a capability for zapping the opposition down the straights that is nothing less than astonishing.

Astonishing, too, is the fact that Branch successfully shortened the rods by chopping an inch from the rod shank and welding the pieces back together. Ordinarily, this would not work; a fracture would surely occur at the weld. It works in this instance because the cuts were made diagonally, across and down the rod, so stresses are spread over a long welded joint. Also, the welds were made by a particularly talented craftsman, using heli-arc, and the rods were X-rayed for flaws after each individual bead was laid on. After this multistage welding process had been completed, the joints were machined to recover the original I-beam section, and repolished. In finished form, it is impossible to see the joint, and except for the unavoidable interruption of the forging “grain,” the joint for all practical purposes no longer exists. Two sets of these rods were made and both sets have been in service, in racing engines, long enough to establish that they are completely reliable.

It is also possible, and probably more desirable, to machine special rods from solid stock. Harley-Davidson’s “knife-andfork” connecting rods are a bit complicated for this sort of thing, but conventional rods are not too difficult to carve from the solid. Obviously, they should be made of a tough, heat-treatable alloy steel, like 4130 chrome-moly, and as you will probably be carving them from rolled plate, cut the blanks so that the plate’s natural grain is oriented down the length of the rod; not across. Best of all would be forged rods, but the price of forging dies being what it is, this will not be an attractive alternative for many tuners.

Those who do make the switch to shortened connecting rods may find that the increased inertia loads have uncovered a weakness down at the crankpin bearing. When that happens, it may be time to start thinking in terms of different bearings. This would be particularly true when roller bearings are being used.

Rollers have advantages: high load capacity for any given width being one; very low running friction being another. Indeed, when dealing with low shaft speeds and very high loads, nothing else works as well as the roller bearing. However, they do have one major failing as applied to a crankpin: cyclic speed variations make them skid, and when they skid, terrible things happen to the rollers and their races.

Such skidding will not normally occur in a constant speed application. But, the connecting rod's swinging from side to side, as the crank rotates, causes the rod bearing’s outer race to (in effect) slow as the crankpin moves around past bottom center and move ahead at top center. Thus, the rollers must also change speed, and their inertia makes them unwilling to do so. And, too, there is the fact that they must drag the cage with them, which introduces yet another difficulty. You will appreciate that shortening the rod causes an increase in rod angularity and that may be the final straw in a marginal situation.

Sometimes, it will be possible to substitute a lighter cage, or to lighten the stock cage. Usually, the cage will be a simple ring of aluminum, with rectangular holes broached through to hold the rollers. Often, some of the material in the middle of the ring can be machined away, to give it a “waisted” appearance, and that will make it lighter. An even lighter cage can be made of titanium, but materials that would not usually be found in bush-type bearings should be tin-plated. Even aluminum cages benefit from having a flash of tin over their surface. This ring will bear against either the inner or outer race (I prefer to support it with the inner race; the rubbing speed is lower), and the coating of tin will prevent the cage from wearing away the inner race. In this connection, I should mention that the tin coating should be deep enough (.010") to absorb micro-particles of grit. Uncoated aluminum cages tend to collect particles stuck to their surfaces, and this creates a disastrously effective abrasive lap.

(Continued on page 28)

Still another measure is to abandon the rollers entirely and install a plain bushing. You can simply substitute a tin-coated aluminum ring, grooved and drilled to spread oil, for the roller-bearing assembly. The Enfield 500cc single uses an arrangement of this kind and it is extremely effective. Of course, the bushing will require at least double the quantity of oil provided for the roller bearing, so the oil pump must be modified. The bushing will have a higher running friction, incurring slight loss in net horsepower, but it is my opinion that it will be more reliable at high crank speeds, in a big single, than a roller bearing.

It is also my opinion that all engines, and particularly racing engines, should be fitted with oil filters. The racing engine, usually running without an air-cleaner, will gather a surprising amount of grit in its oil supply. It arrives with the fresh mixture, from the carburetor, and is deposited in the oil on the cylinder walls. The oilscraper ring will remove the combined oil and grit from the cylinder and drop it into the crankcase, where it inevitably finds its way to the bearings. Honda’s centrifugal grit extractor does a good job of keeping oil clean; nearly everyone else relies on hope and frequent oil changes. A few use wire-mesh filters, but these are not particularly effective. ■