For commercial engines, a manufacturer carefully designs and builds a product that delivers the torque curve needed for the defined application. In racing, the “thousand monkeys” effect quickly finds what works best.
The Needs of the Application Determine the Shape
The high-torque diesel engines found in large trucks are mirrored in motorcycling by the large displacement and low-end torque of touring-bike engines. In both cases there must be real muscle to get a heavy vehicle moving without drama or wasteful clutch slip. Because touring also requires passing and on-ramp performance to move with traffic, touring torque curves are biased to the low- and midrange, sloping downward above 3,000–3,500 revs. Why?
Again, just as with heavy trucks, top speed is determined in relation to traffic. Nobody tours at 100 mph, just as no trucking firm wants or needs performance much beyond the speed of traffic.
The Sportbike Era
The rocket ships of the sportbike era (1986-2007) were sold by races won, especially Daytona. They needed plenty of top-end power; if your company’s “stock” 600s couldn’t manage 160 mph on the Daytona banking, don’t even bother making that trip to Florida. That required pushing their peak torque up close to peak power (meaning within 1,500 revs). This made their torque curves look like those of the touring bikes, but flipped right for left. Sportbikes had it all on top but were so-so in the midrange and downright weak on the bottom. Tour jobs had it all on the bottom and mid, their torque sloping right off above that.
How did those sportbikes get going from a standing start? They had enough torque down low to be decorously rideable in the presence of policemen, but if you needed a fast start there was no alternative to revving the engine high into its torque range and slipping the clutch. Over time (and the burnout craze) those clutches became large and capable. During a maximum-effort clutch-slipping start, the clutch acts as a power divider, sending down the chain what the rear wheel can accept and dumping the rest into clutch slip, where it becomes heat. The bigger and heavier the clutch, the lower the peak temperature reached by the clutch discs off the start line.
In roadracing, getting two standing starts from a new set of clutch plates is outstanding, and service after one such start has been the norm.
Just for fun, stage a roll-on from 3,000 revs between a sportbike and a big Harley or Indian. I used to smile when I’d get a request to “put lights on my TZ750 so I can ride it on the street.” The big TZ’s torque doubled at 9,300 and peak power came 1,000 or 1,200 revs above that. That meant that it had to be ridden as a racebike to get anywhere in a hurry, so it would be left far behind by the street stuff (any boulevard bike, for instance) in a roll-on.
A more difficult problem is the heavy ADV bike, which needs good bottom torque, for getting moving or slogging through loose stuff, and high torque everywhere else to keep things moving, such as on the highway or evading third-world bandits on the way to Patagonia. This one needs a “tabletop” torque curve, and fortunately it can be had from the combination of four valves per cylinder and high valve lift over fairly short duration.
Euro 5 and the Tabletop Torque Curve
The coming of Euro 5 emissions limits is now imposing this tabletop torque curve on most streetbikes. Here’s why. In all the usual descriptions of the four-stroke cycle, we are told that the intake valves open, the piston descends on its intake stroke to bottom center, where the intakes close and the piston rises on compression. As it nears top center, a spark ignites the compressed charge, which forcibly drives the piston down on its power stroke, transmitting power to the crankshaft. Then at bottom center the exhausts open and the piston rises on its exhaust stroke, pushing out the expanded combustion gas. Repeat as necessary.
Trouble is, it takes time to open and close those valves; they can’t just snap to full lift instantly. We have to start opening the intakes a little before top center, so they can be mostly open when the piston reaches maximum speed on its intake stroke at about 76 degrees after top center (that’s where the crank throw and the con-rod are at right angles to each other). And as the piston approaches bottom center, still on its intake stroke, since the intake flow has reached a really high speed—hundreds of feet per second—it would be a shame to stop it by closing the intakes now. So we leave them open for a time after bottom center, letting that high-speed flow keep right on charging into the cylinder.
Timing and Duration
It is also a known source of power loss to wait until the piston has reached bottom center on its power stroke to begin opening the exhaust valves. Residual exhaust pressure does a really good job of pushing out exhaust, so the sooner the exhaust valves open (before bottom center) the less power is lost in making the rising piston do the work of forcing the remaining exhaust out.
Following the rising piston on its exhaust stroke, we note the intakes beginning to open a bit before top center, as described above. It’s also premature to close the exhausts right at top center, because exhaust flow, too, has useful momentum, carrying it out of the cylinder “for free.” So we leave the exhausts open until a bit after top center.
This word picture leaves us with a time, at top center at the end of the exhaust stroke, when both intakes and exhausts are open at the same time. The intakes are just beginning to open and the exhausts haven’t quite finished closing. This period is called “valve overlap” and it is of intense interest to emissions regulators. Why? Because this is “the emissions moment” when fresh charge from the intake side can get swept out of the cylinder by the last of the departing exhaust gas.
Well, why can’t we just let in pure air until the exhaust valves have closed, and only then open the fuel injectors and begin to spray? Consider that fuel reaches the cylinders in three forms: 1) as evaporated fuel vapor, 2) as unevaporated larger fuel droplets, and 3) as “wall wash”—liquid fuel sliding along the walls of the intake port. Just because the fuel injector isn’t spraying doesn’t mean there’s no fuel in the intake port, and it is this fuel, getting lost out the exhaust, that causes emissions regulators to dislike valve overlap. To meet Euro 5 limits (and the similarly tight limits of other nations), makers are having to drastically cut overlap.
To avoid serious performance loss, they then draw on what has been learned in racing about how to control valve motion when lift is high and duration (total number of degrees the valve is open) is short. For this reason, even super-performance bikes of the present era have reduced valve overlap.
More Torque-Curve Lessons From Racing
It’s interesting to compare 125s and 500s in motorcycle Grand Prix roadracing in the early FIM years, 1949 onward. If we multiply times four the power of a given year’s 125 singles from, say, Mondial and MV, we get a substantially bigger number than the MV and Gilera fours were making in that year. No surprise—that just means that 125s were tuned more sharply than 500s. In 1953 the 125 reached 17 hp and the 500 four 56 hp. Then contrast our theoretical “multiplied 125” and its 68 hp with the actual 56 hp of the 500: big difference. Why? 125s can use all the power it’s possible to give them because their torque is too small to knock them down, but the 500s often become slower as a result of being tuned more sharply. The 500′s power has to be smoother and more usable because there is so much of it. As Rob Muzzy observed years ago, “The harder you tune a four-stroke, the more it acts like a two-stroke,” meaning that making an engine more strongly dependent on exhaust and intake wave effects transforms its normally smooth torque curve into a two-stroke-like Alpine terrain of harder-to-ride peaks and valleys.
The late John Surtees praised the 500 MV (making 67 hp when he began to ride it in 1956) for its wide, smooth torque, which allowed him to steer the bike with its throttle through long smooth bends. In that same year MV’s 125 made 20 hp, which multiplies to 80 for comparison with the 500. Clearly, MV did not tune its 500 as sharply as it did its 125.
It was said that back in the late 1940s Gilera mechanics might putt to the corner store for cigarettes on one of their bland, easy-to-ride 500 racers.
Meanwhile, to keep up with the fours, Norton was having to tune its single-cylinder “Manx” 500 to such an extreme that it couldn’t accept full throttle below 5,500 rpm (due to “megaphonitis,” or negative exhaust-pipe effects), but gave peak power at 6,800. The resulting narrow powerband made it as hard to keep “on the boil” as the little 125s.
In our own era these tuning effects bled through to make powerful sportbikes weak at bottom and midrange because all their tuning was working to build high-rpm torque and peak power.
Two-Stroke GP Bikes
Even in the two-stroke Grand Prix era of 1975-2001, 125s and 250s were always tuned to greater extremes than were the 500s. Torque spikes that might put an incautious 500 rider on his head were just fine on smaller-displacement bikes making much less torque, no matter where that torque peak occurred. And so it was that at the close of the era, when 500 two-strokes were making 185–190 hp, the top 125s carried on to 55 hp (implying that 220 was possible for a 500), and the 250s to over 100 hp. Near the end, a European-designed 500 appeared that was claimed to make more than 200 hp, but it was uncompetitive with bikes that were less powerful yet easier to ride. Races are won not by how much power you have, but by how much you can use.
The tale is told that Suzuki long ago did some focus-group testing with GSX-Rs that had been given easier-to-use torque curves. People loved the combination of radical sportbike style with usable power.
Similarly, when Honda was preparing to build its CBR900RR Fireblade (introduced in 1992), there was serious concern that such a light, powerful bike might be almost unsafe for average riders. The development group then worked up a “rideability index”—a list of qualities that would make such a powerful bike easier to ride. Then they built a prototype incorporating high rideability scores. Not surprisingly, focus groups of average riders loved the bike. What the engineers did not expect, however, was that when professional racers tested the high rideability bike, they went faster on it than they did on purpose-built racers of similar displacement.
What this revealed was that motorcyclists, and some engineers, had come to believe that really fast bikes had to be hard to ride. But in fact racebike torque curves had not been sufficiently engineered to make them usable. Honda used this knowledge to dominate the first two years of four-stroke MotoGP, from 2002-2003, just by making very rideable bikes. At the time, retired GP pilot Randy Mamola commented that the new four-strokes had smoother power delivery than most production bikes.
A later and inevitable outcome of this work has been the development of electronic rider aids, which allow very high states of engine tune to be engineered to deliver high rideability. As Loris Capirossi told me during his time on Suzuki’s V-4 MotoGP bike, “If you switch off the electronics, this bike becomes unrideable—worse than a 500 [two-stroke].”