21st century dynaflow.

Tom Kasmer is an inventor. And a restless one at that. From a prototype capacitive discharge ignition system (in 1963!) that brought him to the attention of Bendix, to the rocket igniter he designed for the Lunar Excursion Module, to a concept for a one million-Watt laser for the Strategic Defense Initiative (among others), Kasrner-a confessed "gear head" with a degree in physics-has been busy inventing. His latest creation, the "Hydristor" (a combination of the words "hydraulic" and "transistor"), draws its inspiration from the ancient Buick Dynaflow automatic transmission. The Dynaflow is so old and prone to leaks that ancient cave paintings purportedly show prehistoric vehicles equipped with the device parked over pools of hydraulic :fluid, but Kasmer insist his patented (#6022201, #6527525, and #6612117) device eliminates those concerns.

Kasmer admits the Dynaflow is a strange starting point, but it's the transmission's variable-pitch torque converter that caught his attention: "The Dynaflow allowed the car to accelerate hard by varying the pitch of the torque converter." Its leaks and inefficiency put an end to it. But Kasmer persisted, and his work with IBM gave him the idea of combining a dual-pressure vane pump with the continuous stainless steel belt design found in high-speed dot matrix printers. Spring-loaded sliding vanes placed radially about the center of a rotor are contained at their outer edge by the flexible belt. This seals the vanes' edges even when the rotor is not moving. Hence, no Dynaflow download on the garage floor.

The shape of the band is determined by a set of individually adjustable pistons located about the circumference of the pump. (If you look head-on at the pump they would be located at the cardinal points of a compass.) By manipulating the opposing piston pairs from fully open to fully closed, the shape of the belt can be changed from round to elliptical. In the former configuration, a shaft attached to the central rotor moves freely because the fluid volume is balanced. Pressurizing one opposing piston pair causes the band and vanes to take on an elliptical shape that alters the volume of the adjacent chambers, and drives the rotor. "Varying the piston positions allows me to control fluid displacement between zero and the maximum amount in almost infinite increments," says Kasmer, "and have two separate hydraulic circuits that share a common fluid circuit."

He also claims that initial testing of a prototype more than five years ago (the project has moved in fits and starts in unison with investment capital) showed the Hydristor to be almost 95% efficient. "Unlike a conventional dual-chamber vane pump," he says. the vane edges don't slide against the inner surface of the pump. There's no metal-to-metal contact because the continually replenishing supply of oil being pushed through the pump creates a hydrodynamic bearing. Belt wear is kept in check because it 'walks' very slightly behind the vanes, and spreads wear over a larger area. As for the fluid containment question, Kasmer says that the mesh is so fine that even though the belt "floats" above the vanes it prevents the fluid from rushing through the seals. "And the combination of these characteristics," he adds, "means the speed of operation can be increased well beyond conventional vane pump technology, increasing the hydraulic power density available from a vane pump."

While it's possible to combine a single Hydristor with hydraulic motors to drive a vehicle down the road, Kasmer is pursuing a plan that uses the device as a high efficiency continuously variable transmission. An American automaker (Kasmer and the company in question asked us not to mention its name) is providing a large SUV that Kasmer will fit with a Hydristor drive in place of the conventional automatic transmission. And the U.S. Army is interested in the device for its efficiency, the fact that replacing conventional gearboxes with Hydristor would give it a fleet of automatic transmission vehicles, and the possibility of using it as a driven differential by shifting the belt from side-to-side. "Imagine skid-steering a Hummer like a Bobcat to avoid an ambush," Kasmer says of this idea.

According to an engineer at the American OEM investigating this technology: "The Hydristor is a simple device with a lot of flexibility that's also very compact. Theoretically," he continues, "highway fuel economy could see a dramatic improvement, and city mileage could increase by as much as 25%." And though he shies away from some of Kasmer's wilder claims (e.g. accelerating a 7,000-lb SUV from 0 to 60 mph in five seconds while returning 50 mpg on the highway), he does hold out the hope that the Hydristor will provide greater efficiency than is expected from coming transmission technologies, and at a production cost comparable to today's automatics. "By adding speeds to conventional automatics," he says, "automakers are able to reduce the torque converter's use, lower the engine speed under a variety of conditions, and get the same or better performance while using less fuel. If the Hydristor lives up to its billing, manipulating the position of the vanes and pistons would create near-infinite number of speeds, and allow us to lower the speed of the engine under most conditions."

Kasmer still has hopes to use a Hydristor derivative as a driven differential, as veil as pairing it to flywheels that can tilt n pitch and roll to keep a vehicle firmly planted on the ground. But that's a story for another time.