Experience with 3D Printed Gears

[InFlight]

Has anyone used 3D printed gears in transmission this year with good success?
I was primarily wondering about the viability of 3D printed gears for driving an absolute encoder, as nylon Hex Shaft gears aren't plentiful.

Please include the pitch and gear thickness if possible as well.

[thefro526]

In 2013, I printed a couple of gears to drive encoders and they were a non issue. If anything, they made a lot of issues go away. Specs on those were 20dp, 60T, material was ABS.

One of the tricks I picked up was to print the gears a bit oversized, something like .002/.003" over and run them against a metal gear. The plastic gear has a tendency to wear into shape when driven by an aluminum or steel gear, and you suddenly have a relatively low backlash interface.

For the most part, if you're printing gears to interface with a sensor, you'll be fine. If you're putting enough power into the gear to break it, odds are you've destroyed your sensor in the process.

[SteveGarward]

We were going to do this exactly for our winch position pot this year, but we ran out of space for it and found a different way to mount the pot instead.

Most filaments will have a stated shrinkage amount (or you can try some and work it out), and most readily available ABS I've come across is about 3%. If you use a slicer that will let you scale the model, you can bump it up a small percentage, if you need to for your filament, without altering your model.

I've found that the Form Futura EasyFil ABS pretty much didn't shrink at all, and didn't warp during the print as most ABS gears I've printed have.

Viability: They should be fine - little to no load on it shouldn't wear it down. We had a 3D printed winch pulley for our lift this year, which stood up to a lot of (ab)use with not much wear.

[protoserge]

We used them for encoders this year on our swerve modules and they worked great. There were some PLA and some ABS. Gears were about 1/8" thickness 32T.

Don't expect a plastic gear to work for power transmission of a 120lb robot

[cad321]

We used PLA 3d printed gears on our drive transmissions hooked up to an encoder for position feedback. They worked absolutely great. We took the cad files of a vex 24t, 20dp gear, modified it to fit a 1/8th square shaft and made it about 2% larger for printing. We were concerned with durability but found no wear issues at all. We ran them the whole season (greased) and never once needed to replace them. Would highly recommend to teams looking to shed some weight.

[steelerborn]

Although I haven't tried this on an FRC bot but one application could be for rotating a turret.

If you had a fairly long Face Width on the gears with a good ABS plastic you will probably be okay. If you wanted just a limited range on the turret you could just print out part of the large gear to drive the turret (90 degrees or something).

[lovelj]

Originally Posted by steelerborn

Although I haven't tried this on an FRC bot but one application could be for rotating a turret.

If you had a fairly long Face Width on the gears with a good ABS plastic you will probably be okay. If you wanted just a limited range on the turret you could just print out part of the large gear to drive the turret (90 degrees or something).

This is a great application. 3824 prints ALOT. For Rebound Rumble they needed a large lazy Suzan for passing their ball thru a turret. They printed a large v-groove bearing (10" ID) that had a belt pulley integrated into the same part (e.g. Bearing and pulley were combined). It was secured by 3 mating bearings space 120 degrees apart.

As for gears and pulleys, they work great but I'd suggest using Andy Mark hubs if you are using the gears for power. You design your gear and have a hole in the center large enough for the hub. Use bolts to secure the hub to your pulley or gear then transmit power to a keyed shaft thru the hub. They've also used the same approach for couplings to window motors and fisher price gear boxes.

I'll try to dig up pictures. If you have a printer and you use belt drive, you should never have to order a belt hubs. Print out whatever pitch, diameter, width.. You need.

[notmattlythgoe]

We actually used 3D printed gears to drive the encoders on both of our elevators and the azimuth of our pickup arm, never had a problem since they never take a high load.

Here is a picture where you can see 2 of the 3 sets of gears.

[Boe]

We used 3d printed gears on to run our drive encoders this year and they worked out great in our opinion. They were 17t 20dp and .375 face width. I believe we were using ABS M30.

Heres a picture where you can see it on the middle wheel http://frc2175.smugmug.com/2015/Nort...MG_6620-XL.jpg

[Kevin Kolodziej]

We experimented with printed gears prior to the 2013 season but ended up not using them due to strength concerns after testing. These were intended for drive gearbox main gears. We did have a few polycarbonate gears printed by a sponsor and those held up just fine during testing but opted for a different ratio in the end. These were 20dp gears, 1/2" face width.

We successfully used 3D printed gear box plates for the entire 2013 season though. Those were 1/4 thick, solid ABS from a Dimension 1200 BST printer.

In 2014 we used printed ABS PULLEYS (not gears) for our roller claw intake without any issue the entire season. The prints were done as two halves with steel dowel pins (1/16" diameter) between the two halves with press fit holes. The pulleys were impacted during rough play and even held up just fine when a screw on the roller claw wheel backed out - and the mechanism proceeded to bend the screw past the mounting point and continue to operate. No ill effects on the pulleys at all. The files were printed directly from the Vex CAD files for the pulleys (I believe they were 32t, HTD profile). The only modification we made was adding a 1/2" hex bore. Bore was oversized by .001" or .002" and we ran a 1/2" broach through the bore after completion to smooth out any deficiencies in the corners.

In 2015 we lost access to the Dimension machine but were awarded one of the Ekocycle printers. We printed some Vex EDR sprockets with 1/2" hex bores to run chain from our tote elevator and our can arms to encoders and potentiometers. This printer isn't quite as fine as the Dimension printer so we oversized the hex hole by .005" but took the tooth profile directly from the Vex CAD files and they worked great all year.

[Dylan179]

The only 3D printed gears that 179 used where in our gearboxes for encoders and even then they occasionally became unreliable if you don't load them right.

[Mschmeh144]

In our experience with 3d printed gears, they will for the most part work fine. You can't substitute a metal load bearing gear with one that has been 3d printed. It will work for a little bit then start stripping teeth after repeated use. We have a makerbot 2 and standard plastic abs plastic.

[BenGuy]

3641 tested 3D printed gears last year in a prototype. We found that they work great for low stress situations and they are very effective in those situations because of how customize able they are. In high stress situations, I don't recommend using them because they are much less strong than metal obviously.

[rsegrest]

While we did not print gears we did print brackets for limit switches this year which held up really well in the ABS. BUT...

After build season we invested in some carbon fiber filament that we are testing with now to determine its viability in terms of strength and cost vs. the ABS/PLA. Will let you know how it turns out

[edwinsander]

That is also my dream, to use some 3D Printed Gears.