Question: Interesting article, but certainly this isn't the whole story. I beams and H beams demonstrate how certain forms of ribbing can add strength on a beam or support of a given weight, correct? And, specifically to motorcycles, the swingarm on the KTM 790 Duke comes to mind. —Ryan Haag

Answer: Haag presents engine connecting rods and structural I beams as examples of successful parts with "ribs." I would add such things as Ducati closing levers, which also have I section. In these cases, the flanges—which have the job of carrying the bending, tensile, and compressive loads—must be located far from one another to place maximum material where it can carry bending load. The flanges are prevented from buckling by being joined to each other by a web.

Such I-section members are poor at resisting torsion, which is why early engines (some Wright inlines, the Offenhauser racing engines, etc.) had connecting rods with tubular shanks, which are ideal for resisting torsion. Some years ago, I ran into a valve-spring engineer and asked him why no one makes valve springs from hollow wire (material near the center of the wire makes almost no contribution to torsional stiffness). He replied there is no way to guarantee that the interior surface would be free of stress raisers. I therefore suspect tubular con-rods fell out of use at least partly for this reason.

Hollow con-rod shanks were produced in two ways: 1) by assembling the rod from separate pieces (early vertical Wright engines); or 2) by drilling from the big end (Offenhauser and some others).

Yet the usual objections to stress concentration at angles and rapid changes of cross-section remain. I was shown one of the two special maraging steel con-rods made in the 1960s by Albert Gunter and associates for the Matchless G50. It had failed from cracks that propagated along the angular junctions between flanges and web. The most likely sort of stress to have caused this is torsional.

How would torsional stress arise in an engine’s connecting rods? I asked myself that question as I read about torsional rod failures in early development of the World War II-era Allison V-12 aircraft engine. The answer is that as a crankshaft undergoes torsional vibration its deformation causes its crankpins to cyclically go slightly out of parallel with the crank axis. This oscillating angular movement is transmitted through the rod to the piston.

Because the Allison rod failures were near the small end, the flanges there were made somewhat thicker and the problem did not recur. In the case of valve-train finger followers and Ducati closing levers, there is no such source of torsional loading.

In a conversation with John Wittner (who built and tuned successful Battle of the Twins Moto Guzzis) about connecting-rod design, I asked him why the difference in shape between, say, a Norton Manx con-rod, whose beam widens as it approaches the big end, and the rods in Japanese engines, which have simple shanks of constant cross-section. He replied that working with dynamic Finite Element Analysis (FEA) had showed him a rod-beam cross-section able to transmit the tensile and compression loads was sufficient to also handle the bending loads.