I propose this as a fair test for all new cars equipped with an automatic transmission.

  • Start the engine and warm to normal operating temperature

  • Place car on a level surface 50 feet in front of a steel pole

  • Shift the transmission lever into Drive

  • Take all feet from the pedals

  • Steer car, if it moves, directly into pole

  • If car is still operational and/or salable 60 seconds later, it passes test

Well, I don't know about your car, but my car would fail miserably. Let it be known to all that many accidents have been caused by the innate tendency of cars with automatics to "creep", or go ahead on their own, with no input from the driver. It should be the characteristic of all cars, on a level surface, not to move when all pedals are released.

You can easily demonstrate just how much torque is transferred to the driving wheels with the engine at idle and in gear. Jack up the vehicle so the two driving wheels are in the air. For vehicles not equipped with "Positraction®" or equivalent, block one wheel so the differential gear will only drive the opposite wheel. The other wheel will turn with considerable torque. This torque is the cause of the undesirable creep. Caution: Do not try to stop the turning with your hands; it is impossible and could be dangerous if tried. It is even dangerous to stop it by sliding a block under the wheel because the torque may be enough to drive the car off the jack. You have been warned!!

People have been injured and even killed by creep. Usually, it is a case of contributory negligence where a driver leaves his car in neutral, brake off, and gets out of the car to do something like shut the garage door. A child shifts the transmission into gear and the car moves. It ends when the car crashes into a fixed object, hopefully before the child panics and jumps out, risking getting run over. Sometimes people have been crushed or run over by their own cars. In traffic, sometimes a car will creep into the back end of another, if the driver is inattentive or relaxes brake pressure. It is brake pressure that controls creep in ordinary driving.

The earliest transmissions, notably the Hydramatic® of 1938 and a number of years following surprisingly had little or no creep. The reason was their use of single or dual fluid couplings. A fluid coupling is illustrated at right. Fluid couplings are quite inefficient at engine idle and did not transfer significant torque. In later years, beginning in the early '50s, torque converters replaced fluid couplings in many makes. The torque converter made for a cheaper and more efficient transmission because it eliminated the need for one planetary gear system. However, it had the undesirable trait of transmitting torque even at idle, causing notorious creep. Packard realized the undesirability and possible dangers of creep and developed a rather crude method of controlling it. Studebaker-Packard automatics utilized an internal brake mechanism to absorb the torque before it could reach the output shaft. Other makes ignored creep as did the public. We now seem to think it is normal and natural for all our cars to want to go on their own whether or not we want them to, when the gear selector is in a drive position or reverse.

How can today's transmissions overcome creep? I think the most elegant solution lies in the design of a variable pitch stator. The stator is the third element which fluid couplings do not have. The stator redirects fluid flow between the drive member (pump) and the driven member (turbine). A present day torque converter is shown schematically here. The fixed stator as used in all present designs is always in the helping direction; that is, in torque multiplication. Thus, torque is always transmitted to the converter at all engine speeds. What is needed is a variable pitch stator whose blade angles can be changed by a hydraulic servo to a torque neutral position at idle, with feet off the pedals. It the car begins to roll backwards, a sensor can detect wheel rotation and the blade angles can be increased to provide just enough torque to prevent backwards vehicle movement. When the driver pressed on the throttle pedal, the blades would be adjusted to normal torque multiplication and the car would accelerate normally. The transition could be made seamless, with no engine runaway tendencies. An added benefit would be improved engine-assisted braking from road speed. Drivers of conventional transmission equipped cars often downshift through the gears to use engine compression to assist braking. Often service brakes don't have to be used at speeds above 30 mph. Such is never the case with automatics; they are notorious for being very poor performers in this regard. A variable stator could be self adjusting to maximize torque transfer from turbine to pump during deceleration providing better engine braking and saving the brakes from wear and possible overheating.

The manufacturers have prior knowledge of the extreme hazard of creep and have been put on notice to expedite the solution. Buyers of expensive makes should test drive before purchase and insist that this inherent defect be remedied as a condition of continued patronage.