RETURN OF THE LOCKUP CLUTCH
GM abandoned its remaining split torque transmissions after the 1964 model year. It appears that at least part of the rationale for the wholesale switch to more conventional automatic transmission layouts was the desire to more fully exploit torque converter multiplication — not only for starting, but also to bridge the gaps between the geared ratios, especially with two-speed automatics. In the sixties and early seventies, having the torque converter available in all gears was a higher priority for most American automakers than a slight improvement in highway fuel economy.
That changed abruptly with the 1973–1974 OPEC oil embargo, which in the U.S. led to the enactment in December 1975 of the Energy Policy and Conservation Act of 1975 (EPCA). Title III of EPCA included the first-ever fuel economy requirements for U.S.-market cars. Those rules, now known as CAFE (for Corporate Average Fuel Economy), mandated a fleet average fuel economy of 18.0 mpg (13.1 L/100 km) for the 1978 model year. Since that was less than two years away, the EPCA rules sent automakers scrambling for quick fixes that wouldn’t interfere with their ability to also meet the latest federal emissions standards.
Chrysler decided to revive the lockup torque converter clutch, which at the time was still used in some truck and bus transmissions and a few European automatics, but to our knowledge hadn’t been used in a U.S.-made passenger car in about 20 years. Although the fuel economy improvements a lockup clutch offered were modest, it was a feature Chrysler could quickly add to most of its existing passenger cars without a massive retooling bill. The addition also benefited an important Chrysler client: American Motors, which purchased Chrysler TorqueFlite automatics for its own vehicles.
Chrysler’s lockup clutch, added to many (though not all) TorqueFlite transmissions for 1978, performed the same function as the old Packard Ultramatic unit, but with fewer parts. The actual friction surface was arranged in a ring around the inside front cover of the torus housing, between the flywheel and the torque converter turbine. When engaged, a hydraulic clutch piston splined to the turbine hub slid toward the flywheel, locking the piston (and thus the turbine hub) to the torus cover and forcing them to rotate together at engine speed. As with Ultramatic, clutch engagement and disengagement was controlled in the same manner as a gear change, using opposing governor and throttle valve pressures to engage the clutch piston either after or simultaneously with a shift into third gear.
Other automakers quickly followed suit. GM and Ford had added lockup clutches to all of their passenger car automatics by the 1982 model year. Many Japanese and European automatic transmissions were so equipped by the mid-eighties. Like the Chrysler and Packard units, most lockup clutches of the eighties and nineties were hydraulic, although a growing number used solenoids to open and close the hydraulic valves, allowing clutch engagement to be controlled by computer.
CENTRIFUGAL AND VISCOUS LOCKUP CLUTCHES
An alternative to the orthodox hydraulic lockup clutch was the centrifugally operated bypass clutch, which was a completely mechanical lockup not requiring any hydraulic controls. The idea had been around for decades and had previously shown up on some nonautomotive torque converters as well as a few passenger car automatics, notably the ZF unit offered in the Peugeot 404. A variety of automakers and automotive suppliers, including GM, Borg-Warner, and JATCO, returned to the concept in the seventies, and centrifugal bypass clutches appeared on a variety of production automatics throughout the eighties.
In the Borg-Warner type, developed in the seventies and used by Ford for its 1982–1986 C5 transmission and some front-wheel-drive transaxles, the actual clutch plate was permanently engaged, connecting the torque converter turbine to a transfer disc ringed with friction shoes. A one-way clutch in the transfer disc’s hub allowed the disc to drive the input shaft whenever the disc turned faster than the turbine.
The friction shoes functioned in a manner analogous to the shoes of an expanding drum brake, using the torque converter’s torus cover as the drum. Shoe position was controlled centrifugally by a series of spring-loaded weights. With the turbine stalled, spring loading held the shoes in the disengaged position. Once the turbine started rotating, the weights’ inertia would effectively “throw” the shoes progressively outward, compressing the springs and forcing the shoes against the inner circumference of the torus cover.
Under load, when the speed difference between the impeller and the turbine was large, the shoes would slip against the torus cover’s inner surface, allowing the torque converter to function normally. However, as the torque converter approached coupling stage, reducing the difference between engine and turbine speeds, the shoes would find enough purchase to wedge the transfer disc against the torus cover, forcing disc and cover to rotate together. Since the torus cover was bolted to the flywheel and always turned at engine speed, this caused the transfer disc to overrun the turbine, locking the one-way clutch separating the two and forcing the torus cover, transfer disc, turbine, and input shaft to all rotate together at engine speed.
A centrifugal lockup clutch had several advantages over the more conventional hydraulic variety. The most obvious was that it required no separate controls, making it somewhat cheaper than a hydraulically or electro-hydraulically controlled clutch. Also, because the lockup process was completely mechanical, dictated mostly by turbine speed, a centrifugal clutch worked in all forward gears, at least theoretically providing greater fuel economy benefits than lockup clutches that worked only in top gear. Furthermore, engagement was more progressive — and thus less obtrusive — than conventional hydraulic lockup clutches, which tended to engage and disengage with a noticeable thump. One tradeoff was that the friction shoes were subject to more wear, since they would slip any time there was a significant increase in load. Another compromise, at least for Borg-Warner centrifugal clutches, was they didn’t do much for engine braking, since the turbine could overrun the transfer disc.
Another unusual lockup clutch variation, used in some versions of GM’s TH440-T4 (a.k.a. 4T60) front-wheel-drive transaxle, was the viscous bypass clutch. Developed in collaboration with Eaton Corporation and first used in the 1984 Cadillac De Ville, the viscous coupling was positioned between the torque converter turbine and the flex plate.
As with a conventional plate clutch, hydraulic pressure within the converter pressed a friction surface on the outside of the viscous coupling’s housing against the torus cover, causing the coupling’s driving (input) flange to rotate at engine speed. As engine speed increased, shear within the viscous coupling’s silicone working fluid would bind the driving flanges to the driven flanges, which in turn drove the transmission input shaft. At higher speeds, pressure within the coupling would more or less lock the flanges together, causing them to drive the input shaft at close to engine speed. A control valve also allowed the viscous coupling to be completely disengaged when necessary by forcing the housing away from the torus cover surface.
Unlike a plate clutch, the viscous coupling still allowed a small amount of internal slippage, which GM and Eaton argued was balanced by the ability to lock up at road speeds as low as 25 mph (40 km/h). Just as importantly, so far as Cadillac was concerned, the viscous bypass clutch’s engagement or disengagement was progressive enough to be imperceptible. However, the viscous clutch was both less efficient and more expensive than the conventional lockup clutch used in other TH440-T4 applications.