The Hydra-Matic, GM’s first fully automatic transmission, was a great success, inspiring a host of rivals — including some within General Motors itself. This week, we look at the origins of Dynaflow and Powerglide, the ambitious but ill-fated Turboglide and Triple Turbine, the later controlled coupling Hydra-Matic and Roto Hydramatic, and more.
TORQUE CONVERTER DRIVE
As we saw in our first installment, Oldsmobile’s original Hydra-Matic, introduced in 1939, was the world’s first really successful fully automatic transmission. By 1952, GM had produced more than two million Hydra-Matics, which were used by Oldsmobile, Cadillac, Pontiac, and a variety of other automakers, ranging from Kaiser-Frazer to the short-lived Muntz Jet. Hydra-Matic was standard on Cadillacs by the early fifties and went into more than 80% of all Oldsmobiles and Pontiacs.
Notably missing from Hydra-Matic’s list of users were GM’s other automotive divisions, Chevrolet and Buick. Buick’s absence was particularly curious considering that the division had actually manufactured Hydra-Matic’s short-lived semi-automatic predecessor, the 1937-1939 Automatic Safety Transmission. Although corporate management had browbeaten Buick general manager Harlow Curtice and chief engineer Charles Chayne into offering the semi-automatic transmission in 1938, Buick’s engineers had never liked it and they liked Hydra-Matic even less.
There were several reasons for Buick’s disdain for Hydra-Matic, among them divisional pride. Some Buick engineers still resented the cancellation of the “Roller,” a friction-drive automatic they’d developed in the early thirties, which they felt was a superior transmission. They were also displeased with Hydra-Matic’s endemically jerky shifts. As we discussed in last week’s article, each of the H-M’s gear changes was mechanically complicated, involving a coordinated application and release of brake bands and clutch packs. Shifts were always firm, and if the transmission was not in perfect adjustment or if the bands were worn, it changed gears with a jolt. On Oldsmobiles and Cadillacs, which used open driveshafts, this could be mildly annoying. Buick and Chevrolet, however, used torque tube drive, which combined the driveshaft and axle into a single, rigid unit carried by the rear springs and connected to the transmission via a single universal joint. With a torque tube, each shift sent a pronounced thump through the frame and body, which Curtice and Chayne, in particular, considered unacceptable.
Engineer Oliver K. Kelley, who had been part of Earl Thompson’s Transmission Development Group (which had led the development of the Hydra-Matic), had similar reservations. Hydra-Matic worked, but it was not a particularly elegant design, being both complex and expensive to build. In the fall of 1939, Kelley and his colleague, George Smith, wrote a letter to Victor Olsen, general manager of the new Detroit Transmisssion Division, suggesting that it would make sense for the corporation to develop several distinct automatic transmissions, optimized for the needs of the various automotive divisions. Olsen agreed and Kelley and Smith returned to the corporate Engineering staff to begin work on a simpler, cheaper automatic transmission, originally intended for Chevrolet.
Kelley and Smith’s thinking soon turned away from fluid couplings (as used by the H-M) to torque converters. A torque converter is a type of fluid coupling, with an additional reaction member (a stator) that allows it to multiply engine torque at certain speeds, much like mechanical reduction gears. The torque converter was not new — it was first patented in 1905, at the same time as the ‘plain’ fluid coupling — and it was common in marine and industrial applications. Inventor Oscar Banker had patented a torque converter automatic design in the early thirties that had seen some use in trains and buses and GMC began using Spicer torque converter transmissions (designed by Lysholm) in its Yellow Cab coaches in 1938, but converters had yet to catch on for automobiles.
The appeal of the torque converter for Oliver Kelley was that it offered the possibility of torque multiplication without the use of conventional gears. An torque converter automatic could offer a useful range of multiplication with only a single planetary gearset — and eventually, perhaps, with no gears at all. Such a transmission would be both simpler and cheaper than Hydra-Matic, better suited for inexpensive cars like Chevrolet.
The ramp-up of military production just before America’s entry into the second world war turned the Transmission Development Group’s thinking away from passenger cars to a different application: armored fighting vehicles. Cadillac had already started adapting Hydra-Matic for use in tanks, but its torque multiplication was adequate only for lighter tanks and its gear changes still resulted in brief reductions in speed, something that could be fatal in combat. The torque converter drive offered the possibility of greater torque multiplication as well as an uninterrupted flow of power.
The first tangible result of the Transmission Development Group’s labors was Buick’s M-18 Hellcat tank destroyer, which used a modified Allison TorqMaster bus transmission with a torque converter replacing the clutch. The Hellcat proved very successful, validating many of Kelley’s ideas. It set the stage for torque converter transmissions for GM’s postwar passenger cars.
BUICK DYNAFLOW AND CHEVROLET POWERGLIDE
As the war wound down, Kelley’s group resumed work on automotive torque converter transmissions. Although their original thinking had revolved around Chevrolet, in mid-1945, Oliver Kelley demonstrated a prototype to Charlie Chayne, who was so impressed that he talked Harlow Curtice into test-driving it personally. Curtice subsequently ordered Buick’s own engineers, led by Rudolf Gorsky, to refine the corporate engineers’ ideas into a production version for Buick.
Buick’s new transmission, which went on sale in January 1948 under the name Dynaflow, took the torque converter drive concept a step further than the wartime Hellcat had. Although it used two planetary gearsets, providing two forward ratios plus reverse, Dynaflow did not shift at all in normal driving. Instead, it relied on the torque multiplication of its complex, five-element torque converter. To expand its useful range of operating speeds, the converter used two impellers and two stators, each with a different blade pitch. At stall, it provided a maximum torque multiplication of 2.25:1, not far from the 2.39:1 first gear of Buick’s standard manual gearbox. As engine speed increased, the converter multiplication gradually faded; at cruising speed, both stators freewheeled and the converter acted like a conventional fluid coupling. In effect, Dynaflow was what we would now call a continuously variable transmission, although unlike modern belt/chain/roller-driven CVTs, its operation was purely hydraulic.
Dynaflow provided unparalleled smoothness with no discernible gear changes. The consequence was lethargic low-speed throttle response. Unlike a mechanical reduction gear, the torque converter produced useful multiplication only at certain speeds. If you caught it outside that optimum range, stabbing the throttle produced more churning than forward motion, even with the 276 lb-ft (373 N-m) of Buick’s biggest straight-eight engine. Slippage was also high, which resulted in heavy fuel consumption. Buick tried to compensate for the latter by using a taller (lower numerical) axle ratio with Dynaflow, but that did nothing for performance.
There was a ready cure for Dynaflow’s mediocre acceleration: manually selecting Low range added a 1.82:1 mechanical gear ratio to the converter’s multiplication. Low could be held to about 50 mph (81 km/h), providing much snappier takeoffs and stronger passing response. Unfortunately, the transmission had no way to shift automatically into high; the driver had to move the selector back to Drive and doing that too frequently tended to damage the clutches and bands. Buick’s owner’s manual cautiously described the low gear as “emergency low.”
Despite those shortcomings, Dynaflow was well suited to the character of postwar Buicks, which emphasized refinement and comfort over performance or road manners. The average Buick buyer was not terribly concerned with fuel economy and welcomed Dynaflow’s lazy smoothness. It was too bad that Buick no longer offered limousines; Dynaflow lent itself admirably to a processional pace.
Dynaflow was followed in 1950 by Chevrolet’s Powerglide. Powerglide was a variation on the Dynaflow theme: conceptually similar, but differing in detail. Its greatest novelty was a second, smaller fluid coupling, mounted inside of the inner pump vanes of the five-element torque converter; the smaller coupling provided engine braking in downhill coasting.
Powerglide suffered the same limitations as the Buick transmission, which were more pronounced with Chevrolet’s six than with Buick’s eight. Powerglide-equipped Chevys came with a bigger, more powerful engine, but the new automatic added more than seven seconds to 0-60 mph (0-97 km/h) times and resulted in less-than-frugal fuel economy.
HYDRA-MATIC VERSUS DYNAFLOW
We may take it as a sign of GM’s considerable wealth and resources that it entered the fifties with three distinct automatic transmissions while most of its competitors were still struggling to develop even one. Until the debut of Chrysler’s two-speed PowerFlite in 1953, the only other automaker to develop its own automatic was Packard, which had introduced its torque converter Ultramatic in 1949. (GM, not convinced that the venerable independent had the resources for such a feat, later sued Packard, claiming Ultramatic’s torque converter design infringed on Buick’s patent.) Both Studebaker and Ford turned to Borg-Warner to develop their first automatics, while most of the other independents either bought Hydra-Matic or went without.
In a later era, GM’s corporate management would undoubtedly have pushed for standardization, but in the early fifties, General Motors controlled nearly half of the largest automotive market in the world and allowing the divisions to remain independent and competitive was an affordable indulgence. Just as important was the fact that GM had invested considerable capital in development and tooling for Hydra-Matic, Dynaflow, and Powerglide, which the corporation was not about to casually discard.
Moreover, the divisions’ customers were developing strong feelings about the pros and cons of each transmission, which sometimes resembled the rivalry between competing sports franchises. Hydra-Matic fans extolled its efficiency and crispness; Dynaflow supporters proclaimed the virtues of that transmission’s seamless action, and derided Hydra-Matic’s lurching shifts. Even the chief engineers of Oldsmobile and Buick exchanged good-natured jibes about the comparative merits of “Dyna-Slush” and “Hydra-Jerk,” although the latter camp had to swallow their pride for a while in 1953, when the destruction of the Livonia Hydra-Matic plant briefly forced Olds and Cadillac to use Dynaflow. The irony is that both transmissions emerged from the the same group in the corporate Engineering Staff; browsing through GM’s transmission-related patents from this period reveals many common names and evidence of similar design methodology.
Other manufacturers quickly settled on a compromise — two- or three-speed planetary gearsets combined with a torque converter. Chevrolet went that route in 1953, reengineering Powerglide to start in low and shift automatically to high; Packard adapted Ultramatic to do the same late in the 1954 model year. Instead, GM’s engineering staff spent the next decade refining the Dynaflow and Hydra-Matic concepts.