Seeing as the season was cut short, Summer CD starts early this year.
This is a (mostly silly) concept for a 6T 32DP gear cut into the 8mm Falcon shaft. Included are two designs: one for a 2-part pinion which has the gear teeth running the full length of the shaft, and one for a 1-part pinion with a shorter usable gear width.
The one-part shaft is similar to the Thriftybot 8mm round shafts. However, the diameter of a 32DP gear cutter necessarly means that the usable gear length has to be shorter, so that the gear cutter can clear the hub.
The two-part shaft has a separate hub piece (here, the same piece as on the base Falcon motors) so that the gear can run the full length of the shaft.
Additionally, much like Vex’s 20DP motor pinions, these gears should not use the nominal C-C distances for gears; rather, I added a large addendum, so these gears would both use the 8T center distance.
Do you know what is involved in make addendum shifted gears? Could you just use a gear cutter with the addendum shift or would you have to get into serious gear hobbing?
To my knowledge, it would require a custom gear cutter or some clever hobbing. Most gear cutters are number 1 through 8, with 1 cutting teeth for the largest gears, and 8 cutting teeth for the smallest gears. a #8 cutter nominally cuts around 12-13t gears. 6t is obviously much smaller than 12, so most COTS gear cutters wouldn’t be able to cut that gear in the first place.
Even if you did have a 6t gear cutter, it would be quite the hassle to do the addendum shift. You might be able to approximate small addenda by offsetting the cutter by the addendum, then taking multiple passes to fill out the width of the cut, but this will lead to imperfect involutes and poor efficiency, so I wouldn’t recommend it.
Yay! I’ve been talking about doing this to a CIM shaft for a bit but now it finally makes sense and somebody bothered to put in the effort to CAD it up.
Anybody wanna put a ballshifter in the Falcon shaft?
For a comparison of just how small this gets: Each of these is a 10:1 reduction, one using the 8T 20DP Falcon pinion and the other using the 6T 32DP shaft.
What ideal situation would this be used in? I can’t imagine it being used for drivetrains due to the small teeth and the entire thing being too compact. You couldn’t possible fit 2 falcons next to each other. Would this be ideal in high reduction gearboxes such as a climber. My main concern may not be a big deal but that’s that depending on how the gearbox is put together, there is no way to retain the shaft from the other side. It is 100% a cool concept, but I really don’t see how it would be used. Plus, with falcons apparently shredding aluminum 20DP gears, I’d be wary of using 32DP gears even if they are steel.
The motor shaft doesn’t need to be retained, as it’s bolted down to the rotating part of the motor; i.e. retained in the same way as the shafts on an ordinary motor. If you’re asking about the hex shaft, that’s already a solved problem - see Vex P/N 217-2733, which lets you have a bearing in the shadow of a motor face (people use this a lot in drivetrain gearboxes).
If you’re clever with idler gears, you could do a drivetrain gearbox, but I think a climber would be an effective application. Alternately it would package nicely for low-load applications (@ all the teams that used a Falcon as a control panel spinner).
You’re right about the teeth chewing each other up. With proper depthing, a healthy current limit, and some effective design, you can minimize this issue.
Since the stall torque of the Falcon is constant, the the stall load on the 1st stage teeth is also constant. Unless a concern is the 3/8" hex being unable to take the load, you can put as large of a reduction as you want on the first stage.
I love this, but have you tried running a simulation to find the max torque the gear? Or even a Lewis approximation of the tooth strength. I think this is one of those times where upgrading to an 8t would be handy. I’m worried about the core of the gear simply twisting off.
A 10:1 reduction off the 8t would still be the 8t gear and a 80t 2.5" gear, compared to 8t and 80t 4" gear. Large space savings, but less likely to simply snap off.
The current design constraint here is that the assembly (as currently put together) needs to be able to fit through the nose bearing of the Falcon motor, so the OD of the pinion had to be less than 8mm ~= 5/16". If you make it an 8t gear, the pinion is cut down a tad and looks rather wonky.
You could assemble the two-part pinion in place, with a bearing permanently retained between the shaft and hub, but that would likely make assembly difficult, especially since it looks like the nose bearing is pressed into the face plate of the motor.
I would love this for any mechanism which requires a single falcon and one reduction.
As for drive trains, on a swerve drive, this would be perfect. You are only using one falcon per module, so you don’t need to worry about the too-close-motors problem. And the way we do it, we only have a single stage gear reduction. I would buy this product for sure if it existed.
Regarding the question of if 32DP gears are strong enough for drive trains: In 2018, 775 Pro drive trains were in vogue, because there were no brushless motors yet, and they were a lot lighter/smaller than CIMs. I think the vast majority of them used a 32DP pinion on the shaft and a 32DP gear as the first stage. While there were some problems with these drive trains (burning the motors…), i don’t think that destroying the gears was one of them. I know that we ran in this setup all season and it was fine. Of course, a Falcon clearly has a lot more torque than a 775 Pro, so that’s worth analyzing further.
I would be interested in how long something like this could survive a high-power application (like drivetrain). It is probably more applicable to medium power, high reduction applications (like arms/mechanisms).
Slightly unrelated, looking at these CAD models, is there no boss to center the shaft? Just the mounting screws? We don’t have a Falcon, so I haven’t taken one apart yet. With no competitions to go to this year I haven’t seen one in person either.
(Edit, this is incorrect; see Mike Corsetto’s comment below)
While you’re correct in that the Falcon has more torque than the 775pro, it’s also using a 6T gear which increases the force going through each individual gear. Doing some quick math, a Falcon on a 6T pinion has a stall load of 354lb, while a 775pro on a 12T pinion has a stall load of 26.8lb. Since the load from the Falcon is an order of magnitude greater, I’m not certain the pinion could take it, so that’s why I feel current limiting would be so necessary.
Doing some hand calculation: Letting t (tooth width) = 0.07", h (tooth height) = 0.068", b (gear face width) = 1". I’m not using the same loads I computed earlier as they were for a lever on the exact pitch diameter, i.e. with radius 6/2/32" = 0.094", but the distance from the center axis to the outer tip of the tooth is actually 0.151". This leads to a tangential load Ft of 220lb at stall.
Plugging in our numbers, we get \sigma_b = 18 ksi. 4140 steel has a yield stress of approximately 60-100 ksi (depending on heat treatment), so it’s within load concerns here. Of course, this is a pretty rough approximation, and I would expect to see accelerated tooth wear on this pinion.
Using a value of b equal to the engaged portion of the gear, though, is much more concerning, since
Running the FEA, however, results in sub-optimal results. While the tooth does not seem to separate from the shaft, there are high stress points exceeding yield incurred across the root of the shaft and in between the teeth. These are additionally in optimal conditions, which means that it only gets worse from there on out. (Note: since SolidWorks doesn’t have a builtin material for 4140 steel, I used 4130, which isn’t quite it but pretty close)
Current limiting can be something of a saving grace here. With a 60A current limit, the tangential force on the tooth tip is 51 lb. This leads to a new approximated stress of 4.2 ksi, which is acceptable.
Re-running the FEA, using a 60A current limit on the motor:
The stress in the shaft is well within the yield stress when under a 60A current limit. Good news, then - the shaft could be usable in applications which don’t require all the torque of a Falcon 500. However, aggressive current limiting and voltage ramping would be required to avoid gear failure.
That being said, I would like to look into ideas for an 8T gear, for reasons which @asid61astutely noted.
As always, let me know if I’ve made any mistakes in my math or simulation.
