125x125 banner
Stylish Motorcycle Riding Gears

Reader Comments: Engine Torque

Kevin Cameron has been writing about motorcycles for nearly 50 years, first for <em>Cycle magazine</em> and, since 1992, for <em>Cycle World</em>.

Kevin Cameron has been writing about motorcycles for nearly 50 years, first for <em>Cycle magazine</em> and, since 1992, for <em>Cycle World</em>. (Robert Martin/)

In response to my recent “Bore, Stroke, and Engine Performance” story, Simon Felix wants to know if there are any oversquare engines (bore greater than stroke) that produce lots of low-end torque.

The classic example often given was built by Ford of Britain years ago. Combining oversquare bore and stroke with quite small valves and ports, it was remarkable for its bottom-end and midrange torque. There are other such engines as well.

Engine torque is not entirely dependent on an engine’s stroke.

Engine torque is not entirely dependent on an engine’s stroke. (KTM/)

Let’s compare two cylinders having the same displacement—one a stock Harley 114′s undersquare bore and stroke of 4.0 x 4.5 inches, and the other having the radically oversquare dimensions of 5.0 x 2.9 inches. Dividing the stock H-D’s 4.5-inch stroke by the rad example’s 2.9-inch figure tells us the Harley’s stroke is 1.55 times longer than the oversquare engine’s. Does that mean it will make 55 percent more torque? No. This is because the short-stroke cylinder’s bigger piston has more area for combustion gas to push against. When we do the arithmetic, we find that the big-bore, short-stroke cylinder has (wait for it) 1.55 times more piston area. The result, given equal cylinder-filling and combustion pressure, is that the two exactly cancel, and produce the same torque. This tells us that engine torque is not created by stroke length alone.

Intake Port Size and Torque

As builders have known for many decades, if you “hog out” an engine’s ports, what you’ll get is loss of bottom-end and midrange torque, resulting in a top-end-only “light switch” engine. How does opening up the ports kill torque? Once an engine’s intake stroke is about half completed, the inrushing air-fuel mixture is moving really fast. It is the kinetic energy associated with this velocity (K.E. = 1/2 M x velocity, squared, where M equals the mass of what’s in motion) that keeps the intake process going even though the piston is decelerating toward bottom center.

Making the intake valve and port bigger reduces mixture velocity, causing cylinder filling to be less complete at low- and mid-rpm engine speeds—returning the hard-to-ride “light switch” torque curve.

It works the other way too. The usual reason we build oversquare engines is to allow them to rev higher, so it’s normal to fill as much of the cylinder head’s area with the biggest possible valves. This was how British single-cylinder race engines evolved between the wars (1919–1939), with gradual reductions of stroke, increases of bore, and provision for larger and larger valves.

Building a Tractor

But what if we’re not trying to build a race engine? What if we’re an automaker needing a family-car powerplant with a wide torque range, yet all we have in production is an oversquare engine? Just put smallish valves and ports into a new cylinder head and—presto—now we have a tractor. Those small valves and ports don’t “know” how big the bore is or how short the stroke. All they know is that something is creating a pressure difference across the port. By half-stroke, intake velocity is very high, so as the piston approaches BDC at middling rpm, air keeps right on rushing into the cylinder, filling it well, resulting in wide torque.

Another less subtle example: Back in the 1970s both Ferrari and Alfa Romeo built flat-twelve F1 engines with large bores and short strokes. Ferrari’s 312T was quite successful, the Alfa less so. Now pull a cylinder head from each and invert it so we can see the combustion chambers. Alfa’s head has the biggest valves that will fit, but Ferrari’s has smaller valves with a lot more room around them.

Why? We know that if we make intake valves and ports too big, an engine can generate cylinder-filling high intake velocity only at the very top of its rev range. Ferrari chose smaller valves because that gave them stronger acceleration.

How Fast Is Too Fast?

We also know that if intake velocity is pushed too high by really small ports, air friction and even sonic shock formation put the brakes on flow, making the flow unable to keep up with the falling piston’s demand. This is what happens to stock Harley Big Twins as they approach 5,000 rpm—their short VW-like cam timing (very close to 180 degrees) and smallish valves and ports do a fine job of packing the cylinders with high-velocity mixture off the bottom and into the midrange (peak torque near 3,000 rpm), but as the pistons move even faster, intake air velocity rises into the friction-loss zone, so cylinder filling weakens. The engine is all done at 5,000 rpm. The very same thing would happen even if we changed the Harley’s bore and stroke to a radical 5 x 2.9 inches.

It is valve and port size that determine where in its rev range an engine will breathe its deepest. We associate long stroke and small bore with good low- and midrange torque because small valves and ports are all that will fit into a small bore. But the physics behind cylinder filling has to do with having the right intake velocity—not too low, not too high—in the rpm range where you need torque the most.

About the Oval-Piston Honda NR500

In the comments after my “Curtain Area” article, reader WR300R also asks an interesting question: “Was valve shrouding by its neighbors any part of the downfall of Honda’s NR500?”

The NR500 of the late ‘70s/early ‘80s was Honda’s attempt to build a four-stroke V-4 that could beat the 500cc two-stroke GP bikes fielded by Yamaha and Suzuki. To make power at very high revs the NR used oval pistons, each joined to the crankshaft by a pair of con-rods. Each cylinder also had eight valves—four intakes on one side of the oval, and four exhausts on the other—in a classic pent roof chamber. Although intended to give peak power at 23,000 rpm it eventually made 136 hp at 19,000 rpm, and never won a single GP point. Honda now regards it as the starting point for its family of round-cylinder four-valve V-4 bikes, including the RC213V racer which has carried Marc Márquez to six MotoGP world championships.

Honda’s NR500 had <i>eight</i> valves per cylinder.

Honda’s NR500 had <i>eight</i> valves per cylinder. (Honda/)

Those who have worked with conventional five-valve engines such as Yamaha’s FZR750 have noted that if only the two outer intake valves are operating, the flow through them is outstanding. But when the center intake operates with the other two, something happens that can result in less-than-expected flow. Do the flow streams interfere with each other? Did something similar happen, as our reader suggests, with the NR500 and its four intakes all in a row?

The Real Culprit

Cylinder filling and combustion in the NR seem to have been fairly good, at about 93 percent of what is regarded as outstanding, so I suspect Honda’s lack of results came more from being unable to reach the hoped-for peak-power revs of 23,000. Chronic piston cracking held them back, and when you calculate the NR’s peak piston acceleration, you find they were trying to make their pistons reliable at a level 30 percent higher than what was usual then in Formula 1.

Honda worked very hard on the NR, refining it through multiple designs. Had they succeeded in making reliable power at 23,000 rpm, then 165 hp might have been possible. The two-strokes reached that level in 1991, and six years later went on to produce 190 horses. To make that power from a four-stroke would have required reliability at 25,000 revs.

I heard the NR run at a few GPs in 1981. During warmup it made a deep booming sound, like the 1,000cc Superbikes of that time. But at speed, on the track, it was hard to tell from the two-strokes.

View full post on Cycle World