Courtesy of Ducati

Fundamentals Of Unsprung Weight

Wheels and engine valves face similar performance challenges

"Unsprung weight" is a phrase from back when people were excited about sports cars from MG and Austin-Healey. It has been upstaged in the present era by concern over how we'll look in nightclub selfies, yet it has as much meaning and importance as ever.

Unsprung weight is that part of a vehicle that rises and falls over every irregularity in the road. On the front of a motorcycle, the unsprung weight is the wheel, tire, brake discs, calipers, axle, and lower fork sliders. The sprung weight is everything supported by the springs—engine, chassis, fuel, and rider.

The importance of unsprung weight is what it can tell us about suspension’s ability to keep the tires against the pavement.

An oxcart has no suspension, so its entire weight is unsprung; the whole cart must rise and fall over every bump, giving a hard ride. If the oxen get a move on and the cart’s wheels hit a big enough bump, the wheels may momentarily leave the road, with only the acceleration of gravity to pull them back down.

When this wheel-leaving-the-road-after-a-bump occurs on a motorcycle in the middle of a bumpy turn, the bike jumps sideways, giving us a fright. Our grip is lost completely.

There’s an analogy to be made here with valve float. Engine intake and exhaust valves are opened by rotating cams, and the cam lobe is just a very refined sort of “bump in the road.” The valve spring is made stiff enough to normally hold the valve mechanism in contact with the cam lobe, but if the engine is accidentally over-revved, the cam may turn so fast that the valve train “floats” off the cam profile for an instant. This is valve float, and follows the same laws as do the sprung wheels of vehicles passing over bumps.

In the case of engine valve trains, engineers seek to make the moving parts as light as possible so that the valve spring can keep it all in contact with the cam. The very same is done with the unsprung weight of a motorcycle—the wheel is made as light as possible by making it of aluminum or even magnesium. The overweight brake discs of 1972—7mm thick—have been replaced by much lighter and narrower discs 5.5mm thick. Or, in the case of MotoGP, metal is replaced by ultra-light carbon-carbon discs. The first-try caliper designs of the early 1970s (some weighing as much as 4-1/2 pounds apiece) have been refined by stress analysis to weigh as little as one pound, yet have the stiffness to give excellent brake lever feel. The bias-ply tires of the 1970s weighed as much as 10 pounds each, with the 3-pound weight of an inner tube on top of that. The coming of cast wheels in the 1980s banished the inner tube, and the post-1984 switch from bias-ply to semi-radial tire construction chopped another 3 pounds out of the tire itself. Heavy yet overly flexible solid wheel axles have been replaced on many bikes by lighter yet stiffer tubular axles.

ultra-light components on wheels
Using ultra-light components on wheels helps reduce unsprung weight.Courtesy of Yamaha Motor Racing

The valve train analog to this lightening process was the switch from pushrods-and-rockers to the minimum-weight valve trains of overhead cam (OHC) engines. In the old system, the cam lifted a tappet or cam follower, which in turn lifted a pushrod. The pushrod tilted the rocker arm, whose far end, bearing against the end of the valve stem, lifted the valve.

But in an overhead cam valve train the extra weight of pushrod and rocker is gone, and the cam lifts only a light tappet plus the valve itself, needing only half as much valve spring pressure to prevent valve float and keep the valve train in constant contact with the cam lobe.

So it is on the road. By lightening the motorcycle’s unsprung weight, it takes more speed to “float the tires” over a given rough road, and when a tire does momentarily leave the road, it returns to full contact much sooner than could the heavier unsprung weights of 40 years ago.

How much sooner? The lightest unsprung weight I’ve measured on a motorcycle is slightly more than 30 pounds, and the heaviest is essentially twice that. Acceleration in Gs is the weight of the object being accelerated, divided by the force that is accelerating it. That force comes from the suspension spring(s). If the wheel is carrying a 300-pound load and unsprung weight is 30 pounds, then if the wheel “floats” after a bump, the acceleration driving it back to full pavement contact is 300/30 = 10 G. But in the case of the 60-pound unsprung weight, that acceleration is 300/60 = 5G—half as much.

This makes it clear that our low-unsprung-weight vehicle can go much faster through a bumpy corner—without “wheel float”—than can the one with heavier wheels. It’s just like the difference between a heavier pushrod-and-rocker valve train and a lighter overhead cam valve train. And if the lighter wheel does hit a bump suddenly enough to toss it into the air momentarily, it will return to pavement contact much sooner.

This also explains why motorcycles with some lateral flexibility designed into their chassis can have higher corner apex speeds (at high lean angle) than can bikes with more rigid chassis. A rigid chassis, because it cannot flex, is tossed upward as a whole by bumps, losing contact with the pavement and jumping sideways as it does so. A more flexible chassis allows fork and steering-head flex or swingarm flex to move with the bump instead of the whole vehicle being lifted by it. That chassis flex lets the tires follow small bumps better and minimizes the “air time” of the tires over larger bumps.

Providing just the right amount of such chassis flex hasn't been easy. Honda's first "flex bike," the 1997 NSR250 (possibly influenced by engineer Shuhei Nakamoto), was too flexible, making it unstable in a straight line and destroying the confidence of its rider, Max Biaggi. With a chassis stiffened by carbon fiber Max became champion that year—but by only two points.

Why didn’t the normal suspension handle all this? With modern tires allowing lean angles over 60 degrees, the normal suspension moves in quite a different direction from the bumps.

Can’t computers figure this out? Computers model complex physical systems such as flexible motorcycle chassis well, but they are helpless to model rider confidence. Factory test riders are told, “The engineering boys tell us this ought to work.” They roll out on the latest chassis to find chatter or instability or steering delay instead of the hoped-for improved grip in fast corners. For that reason this problem continues to defy formal analysis, requiring expensive and time-consuming physical testing with living, breathing human riders.