Hello CD,
Simple question: Would it be viable to 3D print drivetrain gearboxes?
That is all,
18sampson
It’s possimpible.*
It is viable to 3D print gears for novelty toys
What else did you mean?
Is it viable for FRC?
Not for any reasonable drivetrain, but easily viable for some mechanisms, like say a gear ejecter in 2017.
So, 3D printed gearboxes work for light loads.
What do you mean when you say viable? universally viable? wut?
And I assumed regular FDM printing of PLA that is most commonly available to teams. I didn’t even go into other types of printing!
The question is too broad.
Yes.
We 3D printed gearbox plates in 2013, 2014 and 2015. In 2014 and 2015 I can say at least, it did NOT GO WELL. We had problems almost every match, spending much of 2014 sitting immobile across the truss. They simply flexed too much and the tolerances didn’t hold well, causing grinding.
If you have the machining resources, build metal gearboxes. If you don’t, buy COTS. $75 for a Toughbox is cheaper than $5000 for a regional.
With the current lower cost plastic filament printers the materials have limited strength.
Almost all manufacturing companies are in some stage of developing parts with Metal Additive Manufacturing. Once there is sufficient process standardization, and consistent material properties (strength, thermal, fatigue, etc.) this will really take off.
I’m sure that within a few years this Metal Additive Manufacturing technology eventually filtering down to a lower cost “consumer” models. At that point printing custom spur or bevel gears, and custom shift mechanisms, will be easy.
On a hobbyist level printer with the traditional filaments like ABS or PLA? Probably not, though I’m sure someone could manage it if they put enough effort into stiffening it up and supporting it, but it wouldn’t be worth the effort by that point.
But on a higher grade printer with some of the more structurally feasible filaments like nylon, Ultem, or polycarb? Very feasible. It’s still not a drop-in replacement for aluminum but it could certainly be done, and I think we’ll see a lot more of it in the near future.
I agree wholeheartedly. But I think “a few years” is a bit optimistic. I think we’ve still got a ways till mid-grade metal additive manufacturing is widely available.
Along those lines, how do think an injection molded gearbox would hold up?
Pretty well, considering half of the KOP gearbox is injection molded. 
Agreed on the question being too broad. Somebody with an SLS machine that could do metal would be able to print gearboxes no problem. (Possibly entirely in one go…) SLS being Selective Laser Sintering. Any team that has access to one of those would be on par with a team that has access to a waterjet as far as “resource-rich”…
Many FDM printers can also handle ABS. ABS is stronger, and with the proper settings one might be able to get away with that. Once you get into the higher-end printers that can handle stronger materials (or use a different process altogether) you get more chance for success. Stratasys machines, for example, have a decent chance.
One can also use FDM to print a negative to build a mold around…
Or one could use extra-thick plates to add extra strength. Not quite as easy as it might sound.
The problem with 3D printing a gearbox is orienting the layers of the parts so they aren’t likely to split when you put very high loads on them. There are ways to do that, naturally. But if you’re printing flat gearbox plates, you could probably get the plates off faster with a CNC router, and they’d be stronger.
With a Markforged inlaying carbon fiber or fiberglass it I’d say it’s very doable. We’ve looked into it some and you can get similar strength gearboxes for similar cost to vex pro ones.
With a PLA or ABS printer it will shatter or strip for sure, I’ve seen some people 3D print pulleys though out of standard materials.
I mean, VexPro likes them well enough. I have little experience with injection molding though, and I’ve only used it to make custom bushings for a suspension system, so nothing structural per se. So I’m sure it could work, but I see little reason to pursue injection molding to make gearboxes.
I think it is reasonably easy to design the first stage of drivetrain gearbox from 3D Printed material if you are using 775Pro motors as your drive motors (and you should be, imho). The farther you go up the geartrain the harder it is going to be to use 3D Printed materials.
Of course, if you have access to MarkForged parts or to SLS parts made from GF Nylon powder (or course SLS from metal) that makes the problem that much easier to solve but still you need to pay attention to the loading conditions.
When it comes to putting my money where my mouth is, last year Team 88 used 3D Printed first stage with 4 775Pros going into the Vex Pro 3 Cim Shifter. We were happy enough with it but will probably go a different route in 2018 – focusing on shaving off some size and weight.
YMMV.
Dr. Joe J.
We ran a version of the Andymanrk 90 degree bevel box https://www.andymark.com/product-p/am-2622.htm and customized it to add a second CIM. It worked perfectly. Just as good as aluminum. We have a Dimension Stratasys printer. Its was rock solid and not the slightest hint of flex, wear or issues.
I’ve seen a team use 3-D printed gears for low strain applications and have it work reasonably well, but I would definitely recommend against using 3-D printed gears for high strain or critical applications. However if a large number of custom gears are needed, a viable method would possibly be to 3-D print an original, make a silicone mold of it, and cast gears out of polyurethane resin (smooth-cast 300 is a good resin for this). By no means would this be a replacement for hand-cut or machined steel, but it would provide a considerable step up from 3-D printing if CNC is not available.
We printed a set of gears from Stratasys ABS, but their only purpose was carry an encoder. The teeth didn’t wear much, and we only destroyed one when we slammed a swerve module hard enough to permanently bend .125 aluminum.
I’d be comfortably using plastic gears on a very few low impact uses. It’s really the shock loading that would cause intralayer failure on an FFF/FDM component, or the torque being more than the lewis bending numbers say you can hold.
Some of use may remember the Fisher Price gear boxes from the kiddie cars. Plastic housing and plastic gears. Only metal gear was a pressed on pinion. They are wonderful gear boxes. Manny teams had problems with them because of the way they were modified for an output shaft. A gear box housing is very doable with a good hobby 3d printer. The gears are the problem, But a well dialed in printer with a good nylon filament can print good gears. They need to be much wider. The design is the key. The designer must design for the materials they are using. A classic plate gearbox printed probably will be a disaster. Recently I came across 3d printed cycloidal nema 17 size gear box.I am going to modify and print one of these for future positioning application. This summer we printed a worm gear and housing for a 9012 motor. Had to use a steel worm. Surprisingly durable. So I would say that yes a drive train gear box is possible but the design is the key. Or fund raise and just buy what you need.
This guy designed a 3D printed planetary gearbox that can lift a couple hundred pounds on a 4" winch, which is more than I expected was possible. His used almost 2 lb worth of filament though, which must have taken days to print.
We’ve been using 3D printed nylon gears on our (almost completely) 3D printed swerve drive. So far, they seem to be holding up really well! We designed them with a fairly low DP so as to not tax our printer and to help the strength of our gears.
I’d recommend taking a look at what you can print and try to print nylon. We’ve had success with Taulman’s Nylon 910 printed on a bed coated with PVA glue to get the nylon to stick.
What’s the goal here? “just to see if I can do it”, “somehow automate a production process”, or what?
Reliability will highly depend on the printing technology, and whether if you’re limiting yourself to just the case, or shooting for making the gears et al as well…
Printing JUST the case is one thing, but printing the gears, shafts, bearing, et al are a whole 'nuther issue.
A couple of years back FRC 470 printed their Choo-Choo catapult ball launcher’s gearbox’s CASE out of PLA to allow them to use one of the kit’s other motors (vs requiring a CIM. They wished to use the CIMs elsewhere.)
They simply re-engineered the provided AM gearbox case for a different kit motor, and printed it in PLA with the 3D printer they got from FIRST Choice. They bought a single matching pinion gear for it, then scavenged ALL of the REST of the gears, shafts, and bearings from one of the stock basic AM gearboxes included in the kit. That reduced the problem IMMENSELY, and created a VERY reliable launching windup motor-gearbox.
It did require a few generations of printing to get the case structure hardened enough in the right places to handle the extreme loads, but in the end it worked BEAUTIFULLY. It even won a tech award for a “partially biodegradable gearbox”. The final version was actually stronger than AM’s CIM version. In fact, intermediate generations broke -exactly- like AM’s version when overloaded beyond spec… (Interesting…). The final version never broke at all, for it was thickened up in the right places.
Anyway… The 3D printing tech chosen always has a strong influence on reliability. Basic plastic FDM tech (Fused Deposition Modeling - aka “squirting hot plastic toothpaste in layers”) can be too fragile between the layers for high load devices. Sure, a heated build chamber often reduces that problem by inducing a better bond between layers (and different plastics have different characteristics), but even with some cases if the forces are large enough I might still consider a true 3D metal printing tech. (You can only thicken the plastic so much…)
Now when it comes to creating the MAJOR force bearing parts (gears, shafts, et al), If you’re really interested in printing them I’d probably go with either a true 3D metal printing tech (ie laser sintering), or use a 3D lost wax/plastic print tech to generate cores for DIY metal casting in a two-step process.
But honestly, IMHO you should just consider automating the fab of high force parts another way (wire EDM, machining centers, CNC lathes/mills, etc) The forces on the shafts, gears et al are LARGE. If you don’t have to print them (ie you have worked up CNC EDM, lathe and/or milling solutions, or heck… just BUY the parts to stuff the case) then don’t do it! There are MANY better ways to automate the fab of those parts than to print them.
Bottom line: Always Use The Appropriate Tech, to fab the various parts in an assembly. Never expect just ONE fab tech to do the ENTIRE job for you, because EACH part of an assembly has DIFFERENT design goals.
Sure, it seems cool to “print up an entire gearbox in one swoop” as an exercise. But in practice, having a number of different automated tools crank out a SET of parts for you (while you kick back) to assemble into a RELIABLE one is (IMHO) even cooler! 
I hope this helps!
- Keith
“In Theory there is no difference between Theory and Practice. In Practice, there is…”