Sibling Rivalry: Horsepower vs. Torque

Listening to gearheads argue about whether it's better to have an engine with a broad torque curve or a tall horsepower peak, you may feel as if you've stepped into Bill Cosby's old routine where God tells Noah the dimensions of the Ark: "Build it 40 cubits by 30 cubits," says God. "Right," says Noah. "What's a cubit?"

So, before we go any further: "Torque" is a turning or twisting force that produces rotation. "Horsepower" was coined by James Watt to describe a steam engine's power-after watching a horse plod around a water pump to power it (a job the steam engine would soon take over), Watt devised a complex formula of which we offer this radically simplified version: horsepower equals rpm multiplied by torque, then divided by 5252.

Without motion, all the torque in the world isn't useful, and without torque, there's no horsepower. Thus, since their inception, horsepower and torque have been interrelated.

The horsepower and torque figures car magazines quote are determined when automakers measure an engine's output on an energy-absorbing device called a dynamometer, many of which still employ a version of a water pump to simulate the resistance of pushing a car down the road. Test engines are broken in and configured to represent normal installation conditions, with either actual production intake and exhaust systems, or setups that offer the same restrictions. General Motors, for example, tests only a single, "representative" production engine per family; procedures vary among the other makers.

The numbers produced, however, aren't a surprise. From the moment the vehicle requirements are set, the powertrain engineers have a good idea how powerful the engine should be-which is why many estimated engine outputs are rounded off. Over the development process, numerous engines are built and tested. So, when the final test engine is bolted to the dyno, the engineers are nearly certain what it will do. GM says for its Northstar DOHC V-8s, output of all production engines falls within two percent of the advertised ratings.

Because manufacturers are often secretive about the full dyno figures, we're left to publish "peak" horsepower and torque. It's like quoting the highest elevation for an unseen mountain range: What you don't know is whether the range starts far away and gradually rises to a modest height, or whether it languishes along a wide beach before rocketing to dizzying altitudes.

An example of broad buttes versus tall, thin peaks is seen in two versions of the 1.8-liter DOHC powerplant offered in the Acura Integra. At first glance, the base 142-horse powerplant falls far short of the GS-R's 170-horsepower engine. But look at where those peaks occur: 6300 rpm and 7600 rpm, respectively. And if that doesn't give you a clue, the torque peaks should: 127 foot-pounds at 5200 rpm for the base engine versus 128 foot-pounds at 6200 rpm for the GS-R. This means that until you really twist its tail, the GS-R offers little, if any, more accelerative force than the base model. Run it to the redline, however, and the GS-R's extra power comes to the fore.