I have been doing some research on transmissions, and it seems that many teams are using “two speed shift-on-the-fly” transmissions that use a shifting dog to engage one gear ratio or another by locking together with a counterpart. I was originally thinking of a tranny with a similar mechanism, but I was worried that the shifting “on-the-fly” would result in a very violent transition between high and low gear, in addition to excessive wear on the shifting dog, from meshing with it’s gear-mounted counterparts. Is there a trick to these transmissions, or do you just have to move the dog really really fast and hope it doesn’t grind against the dog gears? If my concernes are justified, then what kind of life expectancy do you get before you have to replace the dogs?
If you are going to design this yourselves (which is pretty involved), you never know, but apparently the Technokats never had trouble with their dog mechanism. If you’re skeptical, don’t throw in a shifter at all; we did it last year and performed well. Of course, that’s not to say you can’t or shouldn’t. Shifting is sexy. But, if you decide to make your own or fashion it after another team’s design, just to be sure to machine its dog components out of steel and make sure the track for the dog rides smoothly (and keep tolerances to a couple thousandths of an inch). If you do that, the dog shouldn’t deform at all, no matter how much normal wear it gets. It might be loud (all the better :)), but as long as it doesn’t deform, it should lock beautifully every time as a little throttle is applied to get everything turning and lined up. I think.
Jon’s description is very good. I don’t have anything to add, in that regard, but just as a precedent, teams have been making drivetrains that shift gears for a good while, at this point. A lot of them have done this by jamming one set of gears directly into another, and for the most part this has worked well, to my knowledge
If you’re worried about wear in a dog mechanism (assuming that you’re using (a) pneumatic actuator(s) to shift), you can adjust the force which the actuator exerts, and thereby reduce the force which would cause the wear, in the first place. However, if you decide to construct this mechanism, I’m almost certain that the pressure of the pneumatics system will be a fairly minor consideration. As long as you don’t use a 2" bore actuator, I’m pretty sure the shear strength of gear steel will be more than adequate.
From talking to several of the folks who roll their own shift on the fly transmissions, I have gathered that there are significant wear/breakage issues with this type of transmission.
This isn’t to say that they are not managable or that the benefits are not worth it, but it is not as simple as some folks may think.
For what it’s worth.
Joe J.
bump
the Technocat tranny’s dog is solid steel, as well as the reciving dogd milled into the gears. and it only had to work 2 minutes at a time say 30 times Thats a bit much but i imagine it would hold well never the less
We’ve been using this type of shifting (the “dog”-style) over the past two years on FIRST robots and have not had one iota of problems in that area.
This style of shifting was not invented by us… these dog engagement gears are frequently used in motorcycle racing transmissions. The benefit with this sort of gear switching is that there is no clutch needed.
Andy B.
You seem to have gotten a lot replies but few answers. 
I’m not an expert, but I think the answer is yes, there is the potential for quite a bit of shock load on the dog as it engages. There are two answers as to why the dog does not “wear,” I would bet that most teams who use this type of tranny have software that does some type of “speed matching,” so that the dog and the gear are moving at approximately the same speed when they engage, reducing the shock. The other answer is that the dog is made out of some pretty beefy steel, so it can handle the load for many cycles. Eventually it will fatigue, but the Technokats at least seem to be saying that after two years of competition/off-season competition exhibitions etc. they still haven’t broken anything.
In short the answer is avoid breakage with speed matching and beefy design.
Can some describe a “dog” system? Or point to a good white paper? Unforntunately, “dog” is a very common word and google is more than happy to rain my search with cats and dogs!
Any engineering drawings available? Sketches? More importantly, equations!
the technokat white paper is http://www.technokats.org/documents/shiftonfly.pdf
for the 2003 two motor design
Last year we were using the Generation 1 Technocats shifter that had the separate dogs mates attached to the gears. In practice we had assembled the transmission too narrow and while at full speed it shifter and ended up in high and low immediately locking up the transmission and shearing off the three screws that held the dog mate onto the gear. We took it apart and welded the gear to the dog mate and it worked for a while longer until it was assembled yet again narrow. This time, under a full power shift it again locked up the transmission but the shock went through it and sheared off the motor pinion on the bosch. We put a new motor in and used it the entire season and post season with no signs of wear on the dog or dog mates. With generation 2 having the dog mate milled into the steel gear it looks to be even stronger. My ‘non- expert’ opinion is that it is oversized for the loads that are being applied to it by our motors in the kit and for its limited use in a FIRST season it works spectacularly.
Along with the whitepapers from the Technokats’ 2002 and 2003 transmissions, here’s a picture of team 716’s implementation of this type of “dog” shifting mechanism:
http://www.chiefdelphi.com/forums/pictures.php?s=&action=single&picid=5606&direction=DESC&sort=views&perrow=4&trows=3&quiet=Verbose
Dave,
The TechnoKats/Andy Baker white paper (referenced above) is probably the best thing you will find to help explain this. If you still need any help with visualization please email me. I can try to put together some additional 3-d CAD views and/or animation of how this shifting arrangement works.
Good Luck!
We have not seen any wear after the three off-season competitions we ran with the square dogs. In fact the shifting seemed to get easier the more we ran the gearboxes. The square ‘dog’ is a nominal size, ie 1.25" square, while its mating hole is .004" larger, ie 1.254" square. This gives the needed clearance for the dog to fall into the hole. The corners do not mesh exactly, mostly clearance; the force is applied at a point similiar to a box wrench on a nut.
BTW we do not have any speed matching when we shift. The drivers pull a trigger to shift up and let go to shift down. Shifting down while running at high speed is not good for the gearbox just as it is not good for your father’s car 