Sunday, September 28, 2014
Coulda hada automatic ... in 1905.
I've always been puzzled by the universality of the sliding-gear transmission when better alternatives were available first. Ford had a vastly superior planetary system in 1908, then gave it up and joined the crashbox crowd in 1928.
But wait! That's not all! Sturtevant made a car with a fully automatic 3-speed in 1905. It appears that Sturtevant was more interested in franchising its automatic than making a complete car, but neither side of the business went anywhere. Disappeared from history. Why?
There's an excellent account of the Sturtevant in The Horseless Age from 1906. See page 295 of the PDF.
I've done a half-assed animation by screwing with the actual diagram. The animation starts in neutral, goes through 1st then 2nd then stays in 3rd for a few seconds.
The flywheel (bright green, letter A) carries the centrifugal weights (letter J) around with it. At idle, none of the clutches are engaged. As motor speed increases, the weights move out, pushing the wedges forward. The wedges pinch a clutch disk for first gear (blue, letter H) against its friction surface on the flywheel. As speed increases more, the weights go out more, and pinch the second gear clutch (dark green, G) and finally the high gear clutch (orange, letter F). Each clutch has its own concentric shaft leading to its own gear train.
What happens after that is pretty much like a typical constant-mesh three speed, with one big exception. When in second, clutches 1 and 2 are both turning their gears. When in high, all three clutches are turning their gears. This would grind everything to powder, EXCEPT that the end stage of each gear train has a free-wheeling ratchet. When the final driveshaft is moving faster than the endpoint of the first-gear train, the first-gear train disengages. Same for second gear when third is engaged.
According to the writeup in Horseless Age, it worked nicely. You simply pushed on the throttle and each gear came in when it was needed.
Sturtevant even provided a D2 range for passing or hilly situations where you needed to get higher on the engine's power curve without dropping into third. GM didn't get around to D2 until the mid-50s.
= = = = =
Later: I can see one problem with the basic concept. It's a problem that may not have emerged with the relatively inflexible engines of 1905. You don't want gear selection to depend mainly on engine RPM; you want it to depend primarily on the car's road speed, weighted somewhat by throttle position. This could have been solved by adding a couple more centrifugal governors. Certainly possible with 1905 technology.
I've done a half-assed animation by screwing with the actual diagram. The animation starts in neutral, goes through 1st then 2nd then stays in 3rd for a few seconds.
The flywheel (bright green, letter A) carries the centrifugal weights (letter J) around with it. At idle, none of the clutches are engaged. As motor speed increases, the weights move out, pushing the wedges forward. The wedges pinch a clutch disk for first gear (blue, letter H) against its friction surface on the flywheel. As speed increases more, the weights go out more, and pinch the second gear clutch (dark green, G) and finally the high gear clutch (orange, letter F). Each clutch has its own concentric shaft leading to its own gear train.
What happens after that is pretty much like a typical constant-mesh three speed, with one big exception. When in second, clutches 1 and 2 are both turning their gears. When in high, all three clutches are turning their gears. This would grind everything to powder, EXCEPT that the end stage of each gear train has a free-wheeling ratchet. When the final driveshaft is moving faster than the endpoint of the first-gear train, the first-gear train disengages. Same for second gear when third is engaged.
According to the writeup in Horseless Age, it worked nicely. You simply pushed on the throttle and each gear came in when it was needed.
Sturtevant even provided a D2 range for passing or hilly situations where you needed to get higher on the engine's power curve without dropping into third. GM didn't get around to D2 until the mid-50s.
= = = = =
Later: I can see one problem with the basic concept. It's a problem that may not have emerged with the relatively inflexible engines of 1905. You don't want gear selection to depend mainly on engine RPM; you want it to depend primarily on the car's road speed, weighted somewhat by throttle position. This could have been solved by adding a couple more centrifugal governors. Certainly possible with 1905 technology.Labels: 20th century Dark Age
¶ 7:54 AM
