This is a gearbox with a 47.8:1 reduction. It uses AndyMark gears (12:50, 14:50, 14:45). It is supposed to be as light as possible yet not be too hard to fabricate. If we do fabricate, we now have a sponsor that will waterjet the plates for us :). It has plastic igus bushings instead of bearings to save weight. The spacers are delrin. Craig Hickman cheeseholed the side plates for me. He helped me with some gearbox basics too so I owe him thanks.
-Vivek
Looks nice. I would make one reccomendation-replace the bushings with bearings. You’re going to have very high frictional losses, especially in the beginning stages where the speed is extremely high. I really don’t think it’s worth it to save such a small amount of weight. Also, how are you using the FP? in an AM planetary? DeWalt? It looks like all your gearing is 20 DP and would give you ~100 rpm if you were using a CIM, so I’d assume you aren’t just mounting the motor straight to the gearbox.
I think someone else mentioned recently that they used the Igus bearings on a high speed application, and they failed because they aren’t rated for over 2000 rpm, or something.
I am using a 12 tooth 20 dp gear press fit onto the fp motor shaft. I am indeed mounting the motor straight to the gearbox.
FP’s have a free load rpm of 15700. 15700/48=327 rpm. Hmm… I came across the rpm limit in their catalog too but disregarded it.
rpm for first stage- 3768 rpm
second stage- 1055.04
third (idler)- 295.4
fourth (output)- 328.2 rpm
It looks like the first one will definitely need bearings but the other three stages will not. This gearbox is for an off season upgrade so every ounce counts (to comply with the rules anyway) We weighed in at 117.8 pounds and are also planning on adding fiberglass struts with powered rollers for the manipulator (maybe).
Thanks for the correction.
-Vivek
Are you machining this from gear stock? If not, where are you buying one with such a small bore? Also, why not keep the existing pinion and use a 32 pitch initial stage?
sanddrag,
AndyMark has gearstock that they are selling to us (I think). We are machining the holes.
I initially wanted to use the existing pinion and 32 pitch gears but I could not find gears thick enough for reasonable prices. They also did not come pre-hex broached like the AM gears.
About the rpm rating of the igus bushings:
http://www.igus.com/pdf/j.pdf
I looked and it says that their bushings have a surface speed of about 197 fpm. (19712)/(3/8pi)=2007 rpms
I picked the ultra low friction bushings from igus. Please correct me if I did my math wrong.
thanks, Vivek
Since it sounds like you are really concerned about weight, you can also just open up the FP transmission from the KOP and use the plastic gears from those for the first few reductions in your transmission.
Another way to lose weight is to use Lexan instead of aluminum for your sideplates. Just note that you won’t able to swiss cheese Lexan as much as the aluminum; you’ll have to make the ribs a little wider.
You also won’t ever get 327 rpm out of your transmission unless you’re running the motors at above 12v. Even without having the transmission connected to anything, the few percent losses of torque at each gear reduction will mean your motors won’t be running at free load.
Check your math; the circumference of a circle is π*d.
With a 3/8" diameter shaft, the maximum rpm that I would run those plastic bearings with is 1003 rpm. 197.0 ft/min * 12.00 in/ft = 2364 in/min. 0.375 in * π = 1.178 in. 2364 in/min / 1.178 = 2006 rpm. Divide that in half to factor in a 2x safety margin, and you get 1003 rpm.
art,
I realize the efficiency loss and think I can live with it.
We are planning on water jetting these side plates (well, manually drill the bushing holes to get right diameter and just have the jet mark them with a 1/8 hole) so the material would have to be aluminum.
Thanks for correcting me on my math. That was a pretty stupid mistake on my part 
This is not exactly a final version of my gearbox so I have room for error 
I also used about 1.5 times the torque that the application worked beautifully at so I should be fine.
I am probably going to switch to bearings on the first stage.
After some cheeseholing I got the weight of the whole gearbox down to 2.5 pounds w/o motors.
thanks, Vivek
I think it’d be best to have the water jet cut it like 0.010 under and then ream the hole.
Whenever we get waterjetted parts the holes are usually clearance holes or tap drill holes and their diameter is not critical. Our guy cuts them slightly undersize then we open them up with the correct size drill.
I know Pink just cuts them slightly undersized and then anything that needs a bearing in it gets opened up with a reamer in a hand drill. I’d never have guessed you could get away with using a hand drill, but it sure works for them.
Work for them it does. It is a good idea to have the jet cut the holes slightly under.
thanks for the suggestion,
-Vivek
I understand your predicament with the hex shafts. A future consideration with gearboxes though, as your stages move closer to the motor their is clearly less torque transmitting through the gears. If you were to run your stress analysis on your gears in your first ratio I’m sure you’d find that you don’t need a terribly thick gear for your ratio allowing you to use a 32 pitch gear. Usually in my transmission designs I start small and work my way up in size through the stages of your gearbox. This saves weight by using less steel, cast iron or bronze (depending on the application) in the transmission.
Disclaimer: Always check you stresses before you decide on a pitch size however, there is nothing worse than stripping teeth.
Pat,
I plan on heavily pocketing and cheesing the first few gear stages. The only gear sizes I found for 32 pitch for reasonable prices had 1/8 inch face widths. 
I am just going to stick with AM gears for this gearbox. Bright side: The 45 tooth output gear only comes in .54 inch face width so I can pocket it like I would the other gears and not worry
thanks, Vivek
This is just for future reference. Companies such as Browning or Boston gear usually work through a middle man supplier such as Motion Industries (thats the one I’ve used in the past for such things and had good luck with great lead times). Through them you can buy gears or gear stock and they sometimes do some custom broaching work depending on the case. Also some of these companies carry other things like fittings, pneumatics, bearings, etc. Other good resources are MSC and McMaster. Your design does look pretty good though.
If you really want to use 20dp gears for the first stage from AM, you can face down the ones for the first stage to probably 1/8-3/16" (run the Lewis Bending Numbers on them to be sure they are strong enough; There is an excel that is a cheat sheet for them in CD media) to save a lot of weight. Using 32dp would be ideal as you can keep the stock FP pinion for convenience, but then you have to get them hex broached somehow.
I would like to also back up what Cory said about bearings versus bushings. The efficiency losses will be terrible with bushes at those RPMs, and running them out of spec is just asking for unnecessary trouble.
I assume you’ll be cheesholing the gears, too? I mean, if you’re fretting over bearings vs. bushings, I can only think that you’re planning to do that, but didn’t see it in your CAD work.
Not knowing your end application, would it be reasonable to ask if a worm gear might be a suitable design. You can buy worm gears in ratios up to 80:1 off the shelf and they tend to lock in position… a handy thing if this is for an arm or lift.
Just a couple of thoughts… otherwise it looks great and you’re getting some good advice on this thread!
Jason
any particular reason you couldn’t just run a NBD setup in 1st gear? that’ll get you a 47:1 reduction and would be faster and easier to make.
The only difference being orientation as the NBD setup would be longer and skinnier.
edit: nevermind just read the title this is for two… whoops, but you could just run two setups or 2fp’s into one dewalt
http://www.chiefdelphi.com/media/photos/20145
http://www.chiefdelphi.com/media/photos/19245
We also have the water jet cut them slightly undersized and ream them out. I can remember one instance where I failed to do so on a part and we ended up having to put a nut on the other end due to the hole being to big to tap with the bolt that we wanted to place in there.
Pavan.
…
That is a good idea. It would be much cheaper and lighter.
thanks, Vivek
Made me realize a completely COTS solution to getting ~ 48:1 from two FPs.
Two FPs into AndyMark planetaries into a toughbox yields a 46.8~1 reduction.
I know it’s not the lightest, but It requires no machining or work besides assembly (and the FP pinion).
Adam,
I thought about that a while ago and ruled it out because it was too heavy. We are pretty concerned about weight.
How much does a prealtered dewalt gearbox assembly weigh?
If it is less than 2 pounds, it just may be worth it.
thanks, Vivek
Well, I altered the gearbox and added Idler sprocket mounts.
Here are a few pictures:
http://i200.photobucket.com/albums/aa287/vivek_chittineni/FINALRENDER.jpg
http://i200.photobucket.com/albums/aa287/vivek_chittineni/DSCN7920.jpg
http://i200.photobucket.com/albums/aa287/vivek_chittineni/DSCN7921.jpg
The first ones a render, the second two are of its construction so far.
thanks, vivek
EDIT: disregeard the messy desk behind the partially completed gearbox. 