Six Strokes of Genius, 40% increase in power, from water...

Bruce Crower's Southern California auto-racing parts shop is a temple for racecar mechanics. Here's the flat eight-cylinder Indycar engine that won him the 1977 Louis Schwitzer Award for racecar design. There's the Mercedes five-cylinder engine he converted into a squealing supercharged two-stroke, just "to see what it would sound like," says the now half-deaf 77-year-old self-taught engineer.

Crower has spent a lifetime eking more power out of every drop of fuel to make cars go faster. Now he's using the same approach to make them go farther, with a radical six-stroke engine that tops off the familiar four-stroke internal-combustion process with two extra strokes of old-fashioned steam power.

A typical engine wastes three quarters of its energy as heat. Crower's prototype, the single-cylinder diesel eight-horsepower Steam-o-Lene engine, uses that heat to make steam and recapture some of the lost energy. It runs like a conventional four-stroke combustion engine through each of the typical up-and-down movements of the piston (intake, compression, power or combustion, exhaust). But just as the engine finishes its fourth stroke, water squirts into the cylinder, hitting surfaces as hot as 1,500°F. The water immediately evaporates into steam, generating a 1,600-fold expansion in volume and driving the piston down to create an additional power stroke. The upward sixth stroke exhausts the steam to a condenser, where it is recycled into injection water.

Crower calculates that the Steam-o-Lene boosts the work it gets from a gallon of gas by 40 percent over conventional engines. Diesels, which are already more efficient, might get another 5 percent. And his engine does it with hardware that already exists, so there's no waiting for technologies to mature, as with electric cars or fuel cells.

"Crower is an innovator who tries new ideas based on his experience and gut instincts," says John Coletti, the retired head of Ford's SVT high-performance group. "Most people won't try something new for fear of failure, but he is driven by a need to succeed." And he just might. Crower has been keeping the details of his system quiet, waiting for a response to his patent application. When he gets it, he'll pass off the development process to a larger company that can run with it, full-steam.