CONTROLLED COUPLING HYDRA-MATIC
In 1952, engineers P.J. Rhoads and Kenneth Gage of the Detroit Transmission Division began work on the second-generation Hydra-Matic, a project that ultimately cost some $35 million. The new transmission, which went on sale for the 1956 model year, was known by a variety of trade names — Oldsmobile called it Jetaway Hydra-Matic while Pontiac christened it Strato-Flight and AMC dubbed it Flashaway — although all were functionally identical. The patent application, filed in November 1953 by engineer Walter Herndon (who had been part of Earl Thompson’s original Hydra-Matic development team), called it a “controlled coupling” transmission; it is sometimes known as the 315 Hydra-Matic.
The controlled-coupling transmission was a thorough redesign of the original Hydra-Matic. Like its predecessor, it had three planetary gearsets, giving four forward speeds plus reverse. A fluid coupling took the place of a conventional clutch; the engine drove the coupling’s impeller through the front planetary gearset, to reduce creep at idle. There were two oil pumps and an external oil cooler to prevent overheating. With a cast iron case (except the front torus cover, which was aluminum), the complete transmission was fairly heavy, at 240 lb (109 kg), and quite bulky.
Most of the changes to the new Hydra-Matic were designed to smooth its notoriously jerky shift quality. The first major change was replacing the troublesome brake bands with sprag (one-way) clutches for both the front and rear gearsets. (A rear band was retained, but it was used only in Low range.) The second and most dramatic change was the addition of a second, smaller fluid coupling between the front and rear planetary gearsets.
The new coupling, which was about 25% smaller than the main coupling, replaced the original Hydra-Matic’s front clutch pack. The coupling’s impeller was driven by the ring gear of the front planetary gearset, while the turbine was connected to the front planetary’s sun gear. Since a fluid coupling can’t be mechanically disengaged, it incorporated valves that would allow the oil in the torus housing to be drained or refilled in only 0.4 seconds. In first and third, the coupling was empty and thus effectively disengaged. In second and fourth, it was full, driving the sun gear of the front planetary at the same speed as the ring gear and putting the front gearset in direct drive.
These changes made Hydra-Matic substantially smoother than its predecessor. The 1-2 shift, which was achieved by filling the front coupling, was almost imperceptible. The 2-3 shift, which still involved the engagement of the rear clutch pack, could still elicit a modest thud, but it was far less pronounced than before. (When the new transmission was first shown to the press, engineers admitted they would have preferred to replace the rear clutch pack with a fluid coupling as well, but cost and space considerations had ruled out that possibility.) The redesign also eliminated the need for periodic band adjustments, which most owners had neglected anyway.
Like the earlier Dual-Range version, the controlled-coupling Hydra-Matic sought to provide a gear for every occasion. It retained the Dual-Range H-M’s second (D3 or Super) Drive range, which would keep the transmission from shifting to fourth gear at speeds under about 75 mph (120 km/h); Low range kept the transmission in second up to about 40 mph (64 km/h). The new Hydra-Matic also offered both part-throttle and full-throttle kickdowns for passing.
Owners soon discovered that the new Hydra-Matic was not quite as efficient as its predecessor was, nor was it as durable. Aggressive driving could destroy the sprag clutches, operation in extreme temperatures could be erratic, and the aluminum torus cover was prone to hairline cracks. Nonetheless, most buyers considered the new Hydra-Matic a welcome improvement.
Thanks to its daunting production costs, the new Hydra-Matic found fewer users than its predecessor did. It was used by Cadillac, Oldsmobile, and Pontiac and was sold to AMC for use in some 1956-1957 Nash and Hudson models, but GMC and Chevrolet trucks stayed with the earlier Dual-Range transmission until the early sixties, as did Rolls-Royce, which built the Dual-Range H-M under license.
Over the next few years, GM made many minor modifications to the dual-coupling transmission, most aimed at improving its durability and cold-weather performance. The most noticeable change came in the 1960 model year, when the case was redesigned to reduce its considerable bulk. None of these revisions altered the controlled-coupling Hydra-Matic’s basic operation, nor did they make it any cheaper to build. It remained one of the most expensive transmissions of this era and only its ample production volume kept the price within reason.
TWIN TURBINE DYNAFLOW
At the same time the corporate Engineering Staff was developing the controlled-coupling Hydra-Matic, they were also hard at work refining the torque converter concept. Although Chevrolet had backed away from the pure torque converter drive approach, Buick remained firmly committed to Dynaflow and was looking for ways to reduce its inherent limitations.
The first result of these efforts was the Twin-Turbine Dynaflow, introduced for the 1953 model year. Like the first Dynaflow, its basic concepts were developed by Oliver Kelley’s corporate engineering team while the production version was overseen by Buick staff engineer Rudolf Gorsky. As its name implied, the new Dynaflow featured a redesigned torque converter with two turbines along with a single impeller and a single stator. The primary turbine’s output shaft now drove the ring gear of a planetary gearset. The gearset’s sun gear was carried on an overrunning clutch. At low speeds, the sun gear was locked in place, causing the turbine to drive the pinions and planet carrier (and through them the transmission mainshaft) at reduced speed, thus multiplying the input torque. The secondary turbine was splined directly to the transmission mainshaft. At low speeds, the impeller drove the primary turbine, but as engine speed increased, the secondary turbine would gradually take over. At cruising speeds, both the stator and the sun gear of the primary turbine freewheeled, causing the torque converter to act like a plain fluid coupling. There were still no perceptible gear changes — it remained a continuously variable transmission.
The result of these changes was greater torque multiplication, up to 2.45:1 at stall. Since the multiplication was now partly mechanical, the new converter also provided some relief from Dynaflow’s low-speed lag and slippage. Nonetheless, really brisk takeoffs still demanded the use of emergency low. The latter was particularly useful for passing at around-town speed; while the twin-turbine converter provided adequate multiplication from rest, it had no mechanism to “downshift” for a sudden burst of speed.
To address that limitation, in 1955, the twin-turbine Dynaflow gained a new variable-pitch stator. Like the variable-pitch propellers used on aircraft and boats, the stator blades could switch from a high angle, for greater torque multiplication, to a low angle, for greater efficiency at cruising speeds. Flooring the throttle would flip the stator blades back to the low position with an effect analogous to a kickdown downshift in a stepped-gear transmission. The following year, Dynaflow added a second, fixed-blade stator, increasing the converter’s maximum torque multiplication to 3.5:1. Low could now be held until just past 60 mph (say, 100 km/h), allowing magazine reviewers to record some rather racy 0-60 mph (0-97 km/h) times.
Although the Twin Turbine Dynaflow was finally approaching the performance and flexibility of contemporary stepped-gear automatics, that was not enough for Transmission Development Group chief Oliver Kelley. Kelley’s team was hard at work on the ultimate torque converter transmission: the triple turbine.
THE TRIPLE-TURBINE AUTOMATIC
If the controlled-coupling Hydra-Matic could be considered a belts-and-braces makeover of the original Hydra-Matic, the triple-turbine automatic was an ambitious extension of the Dynaflow concept. Had it worked as intended, it would have validated GM’s (and Buick’s) faith in the torque converter drive concept and probably would have revolutionized Detroit’s thinking about automatic transmissions.
In broad strokes, the triple-turbine automatic was not unlike the contemporary Borg-Warner or Chrysler TorqueFlite transmissions, which had a torque converter and two planetary gearsets. It could, in a very general sense, be deemed a three-speed transmission (which is how some sources incorrectly describe it). However, it’s more accurately described as a continuously variable transmission with three overlapping ranges; like Dynaflow, it provided seamless changes of ratio with no distinct shift points.
The triple-turbine transmission had a five-element torque converter with a single impeller/pump, a variable-pitch stator, and three turbines, the first two mounted on overrunning clutches. The first turbine drove the sun gear of the rear planetary gearset; the rear planetary’s ring gear was locked in place, driving the planet carrier and the driveshaft at reduced speed. The second turbine drove the ring gear of the front planetary, while the sun gear was locked, driving the planet carrier. (The two planetary gearsets were actually identical, but driving the ring gear rather than the sun gear provided a different ratio.) The the third turbine drove the output shaft directly with no mechanical reduction.
As with the twin-turbine Dynaflow, the three turbines each had a different blade pitch, making them effective within different ranges of speed and load. As engine speed increased, the first turbine would exceed its maximum effective speed and begin to freewheel. As did, the second turbine took over, followed by the third. The effective gear ratio at any given moment depended on the relative speeds of the three turbines. At cruising speed, the third turbine was effectively hydraulically locked to the impeller, giving direct drive like top gear in a conventional transmission. The mechanical ratio, which was infinitely variable, was multiplied by the torque multiplication of the stator, which also varied continuously based on speed and load.
Unlike the earlier Dynaflow and Powerglide, there was no Low range and no way to manually lock the transmission into a lower (higher numerical) ratio. The only manual control provided, other than Park, Neutral, and Reverse, was a “Hill” or “Grade Retarder” position, intended for descending steep hills. It locked the ring gear of the rear planetary gearset to the case, causing the driveshaft to drive the front turbine; the turbine then acted as a brake on the impeller, mimicking the effect of engine braking. (Some owners made the mistake of treating it like Low in Dynaflow or Powerglide, which was worse than useless from a performance standpoint and sometimes resulted in transmission damage.)
Although we’ve referred to the triple-turbine automatic in the singular, there were two distinct production versions: Chevrolet’s Turboglide and Buick’s Flight-Pitch Dynaflow. Both were based on a common design, developed by GM’s corporate Engineering Staff, but they differed in many details and as far as we know, they shared few if any parts. The biggest differences between the Buick and Chevrolet transmissions were gear ratios and stator design; while both used variable-pitch stators, Chevy’s had only two blade positions, while Buick’s stator was infinitely variable.
TURBOGLIDE AND FLIGHT PITCH DYNAFLOW
Turboglide went on sale in late 1956 as an option on 1957 Chevrolets, priced about 25% higher than the two-speed Powerglide. Contemporary reviewers were impressed with Turboglide’s smoothness, although they found its stepless operation somewhat disconcerting. As with a modern CVT, the relationship between engine speed and road speed was not linear, which made the transmission feel as if it were slipping badly. (There was some slippage, but not as much as subjective impressions suggested.)
As with early Dynaflows, Turboglide performed best when cruising or in steady acceleration. Although it could easily squeal the tires on quick takeoffs, its responsiveness in the real world left much to be desired. Even with the variable-pitch stator, jabbing the throttle for passing or to accelerate out of a turn would cause the engine to bog as the turbines fought to overcome their rotational inertia. Unlike Powerglide or Dynaflow, there was no longer the option of engaging Low, so drivers had little recourse beyond flooring the throttle and hoping for the best. Inevitably, fuel consumption in stop-and-go driving was profligate, although cruising economy was reasonable enough.
Turboglide’s complexity posed a variety of manufacturing challenges and early production was slow. Many of its difficulties related to its die-cast aluminum case, the first aluminum transmission case Chevrolet had ever built (Powerglide retained a cast iron case until 1962). As with the later Corvair engine and the Buick/Oldsmobile aluminum V8s, complex aluminum castings were at the bleeding edge of GM’s metallurgical capabilities and it was not uncommon for the cases of early Turboglide transmissions to split, particularly if they were overheated. There were other teething problems as well, many of which were beyond the skill and training of the average technician. Toward the end of the 1957 model year, Chevrolet quietly extended warranty coverage for owners of Turboglide cars who complained about transmission problems.
Buick’s triple-turbine Flight Pitch Dynaflow, introduced the following year, had an even less auspicious debut. Pre-production units were so buggy that it almost wasn’t ready for production in time. Buick’s increasingly harried general manager, Ed Ragsdale, became so frustrated that he ordered his engineers to do whatever it took to make it work, regardless of cost. The tooling bill eventually came to a harrowing $86 million (more than $650 million today), which didn’t include development expenses.
Flight-Pitch went on sale in the fall of 1957. Despite the last-minute fixes, it was grievously unreliable until late in the 1958 model year, doing considerable damage to Buick’s already-checkered reputation. In years past, Buick’s quality control had been among the best in the business, but the division’s explosive growth in the mid-fifties had stretched its manufacturing processes beyond their limits. The 1957 Buicks had already been notorious for poor quality and the problematic Flight Pitch (and the equally disastrous air suspension, introduced at the same time) made the ’58s even worse. Buyers, nervous about reliability, dissatisfied with Buick’s overwrought styling, and spooked by an economic downturn, stayed well away. The result was the lowest sales Buick had seen since the end of the war.
By 1959, Buick and Chevrolet had resolved most of the triple-turbine transmissions’ mechanical problems, but buyers remained understandably wary. The Powerglide and Twin Turbine (nee Variable Pitch Dynaflow) were known quantities and few customers saw the point of spending up to 30% more for the triple turbines, particularly given their checkered reputation and dubious resale potential. Sales were very low.
The commercial failure of the Flight Pitch/Triple Turbine caused a lot of finger pointing at Buick, much of it directed at Oliver Kelley, who had become the division’s chief engineer in 1957, not long before Flight Pitch debuted. Both Kelley and Ed Ragsdale were ousted in 1959, Ragsdale taking early retirement, Kelley moving on to become chief technical adviser to the head of GM’s new Defense Systems Division. Buick dropped the Triple Turbine as soon as Kelley and Rollert were gone; it disappeared after 1959.
Chevrolet stuck it out with Turboglide through the 1961 model year. The late-model Turboglide was reasonably trouble-free, but it still suffered the same basic design limitations as the early models. Some owners replaced it with Powerglide and survivors are relatively rare today. In retrospect, the triple-turbine automatics were an interesting concept that was rushed to market before it was ready. Unfortunately, public reaction was so negative that GM never revisited the idea and it became a dead end.
Even if the triple-turbine transmissions hadn’t been so flawed, the late fifties was not a good time for radical new designs. The new car market slumped in 1956, and by 1957, the U.S. was in a recession, sending buyers in search of smaller, cheaper, more economical cars. During the recession, GM began work on the X-100 project, later known as the Y-body, which would give Buick, Oldsmobile, and Pontiac their own compact models for 1961. The corporation also dictated that the 1961 full-size cars would be moderately downsized, making both shorter and narrower than the Brobdingnagian 1959-1960 models.
The new, smaller cars called for new automatic transmissions. The existing Twin Turbine and controlled-coupling Hydra-Matic were too big and too expensive for the Y-body and in any case, it made little sense to mate a 350 lb (159 kg) Olds “Rockette” engine to a 229 lb (109 kg) Hydra-Matic. Pontiac’s “rope-drive” Tempest, meanwhile, needed a rear transaxle.
The senior compacts were originally intended to be one of GM’s first true inter-divisional projects, sharing engines, suspensions, and powertrains. The ongoing rivalry between the divisions frustrated those ambitions and the three cars ended up being quite different; among other things, each had a completely different transmission.
The Pontiac Tempest’s optional automatic, dubbed TempesTorque, was a two-speed torque converter transaxle. Many of its internal components were shared with the rear-engined Corvair’s Powerglide, repackaged for their new application; the torque converter was mounted at the back of the transaxle, behind the planetary gearset and the differential. TempesTorque also differed from the Corvair transmission in its unusual split-torque top gear (see sidebar below), which was intended to reduce slippage at cruising speeds.
The Buick Special, meanwhile, had a new two-speed transmission called Dual Path Turbine Drive. Developed by Gilbert Hause of the corporate Engineering Staff, it was a compact, extremely lightweight transmission; weighing only 96 lb (44 kg), it was actually about 10 lb (5 kg) lighter than the Special’s standard three-speed manual. Dual Path had a single planetary gearset and a three-element torque converter with a single turbine and a single variable-pitch stator. Unlike the Twin Turbine unit on full-size Buicks, Dual Path started in low and shifted automatically to high; in high, it used the same split-torque system as TempesTorque. Dual Path’s greatest novelty was that it did not use a conventional reverse gear — in reverse, a clutch locked the torque converter turbine in place, allowing the freewheeling stator to drive the transmission backwards.
It would have made sense for Oldsmobile’s senior compact, the F-85, to share Buick’s Dual Path transmission, but the Detroit Transmission Division instead opted to develop a new and significantly cheaper third-generation Hydra-Matic: the Roto Hydramatic.