Nowadays frc became standardized with imperial units but as a team working in a metric country we primarily have access to metric gears, keys, fasteners etc. Is it feasible to design a gearbox without any imperial parts and by only using keyways instead of hex features ?
1 Like
From google dictionary, “feasible” means “possible to do easily or conveniently”, so i would say the feasibility of a metric gearbox depends on your access to metric COTS parts. If you have access to them, as you say you do, and have the capability to deal with keyways and non- hex bearings, then sure, it is feasible and that is probably what you should do.
As a team that is fully dependent on the 1/2 hex COTS gear and 1/2 hex shaft, it is not feasible for us to design metric gearboxes. That does not mean it is not feasible for you.
It sure is! In fact a few common FRC components actually use metric keyways (CIMs have an 8mm shaft with a 2mm key, PG71s use a 10mm shaft with 4mm key and BAG motors iirc use a 4mm shaft). But before you go down this path it’s worth considering how you’re going to use your gearbox - with COTS parts from FRC suppliers? (which will be in imperial) Or will you manufacture your own hubs, sprockets, wheels etc? It’s worth considering how to make your gearbox work in the context of the rest of your robot before going down the path of making your own.
If you decide that you need to use all the convenient imperial things from typical FRC suppliers, then it’s worth purchasing some gears and/or gearboxes from them. This will make your life easier in the vast majority of cases - having to broach and mill your own keyways will be time consuming and much of this time can be used to make your robot better.
If you decide that you must make your own gearboxes (and everything else, seeing as you are using metric) then get yourself a set of metric broaches. If you are using a keyway, 3mm or 4mm on a 10mm - 14mm shaft will work well for most high-torque situations. Alternatively, a hex broach can work - they do tend to be more expensive but it is more convenient.
However, all things considered it’s probably worth asking yourself if it is better to just get stuff shipped to your location. It might be a bit costly with interational shipping and troublesome with customs, but that money and effort could well be offset by the cost of the tooling needed and the cost of making your own everything. Luckily Vex and Andymark have a large presence with us here in Australia, but if you give the nice folks at those companies an email, I’m sure they’ll be able to work something out for you.
In the meantime, let’s hope the Americans learn to use intelligent units soon XP
1 Like
As with anything metric, the math will probably be the same – converting from standard to metric is just multiplication by constants, after all – just in incomprehensible units.
Completely possible. If I recall, @Paul_Copioli designed the 2005 kitbot gearbox all in metric.
While it might be possible, i really wouldn’t recommend it. Inches aren’t very fun to deal with, but when designing a gearbox you could keep it down to the minimum of using Imperial gears, bearings and shafts. Being able to buy those components from Vex is just wildly easier and more convenient than sourcing metric components, and they come with a lot of features that make them much easier to work with in FRC.
For example, you can’t really find flanged bearings out in the wild, those aren’t used much outside of FRC, yet they make life so much simpler. Vex’s hex shafts also come slightly undersized- this is to make them easier to fit in hex bearings. Shafts you’ll source locally will be press fit in their appropriate bearings, making everything harder to assemble. Also, keyways are a really unnecessary thing to deal with when you can just use hex, and you’ll have a hard time finding metric hex, let alone hex bearings. Not to mention, anything you source locally will probably be made of steel and not aluminum.
Basically what i’m trying to say is that while having all those things in Metric sounds more convenient, it’s really much more of a hassle than it’s worth. You’d have to get into some industry practices a lot more than you’d want in FRC, and even though it might be cheaper, it really isn’t worth it
1 Like
I find these statements both bordering on incorrect. Now, you ARE in Israel, so I expect that you’re a little more up on sourcing of metric gears etc. On the other hand, it took me a very short time to find a source for metric gear sizes here in the States–and metric isn’t common. Steel, yes… but in some cases steel can be the preferred option. If you have mill or lathe availability, you can remove a lot of weight from a steel gear.
Taking a quick shot at the keyways, I can count how many cases I or my coworkers have used hex in the real world on both hands, but I’ll have fingers left over. I can’t say the same for keyways as I’m out of toes and going through hair strands. (I also have a tendency to use something that is neither a keyway or a hex, but that’s expensive and heavy compared to a keyway–and you really want a lathe to get it right.)
But that statement on flanged bearings in the wild? I took a look at two suppliers, and at ONE of them I found close to 200 flanged bearings… in metric… and that was just scratching the surface of their stock. The other had about half of that, mostly in Imperial. I’d be surprised if there wasn’t a better selection in areas that use less Imperial units.
I don’t think using industry practices in FRC is a bad thing, per se. But, you do need to know that you’re working with those going in. It may be worth looking at doing a metric gearbox in the offseason just to familiarize yourselves with industry practices and what can you do with local parts. Given the industry standard fits, you’ll need a press (hydraulic preferred), a few broaches for keyways of various sizes, and you’ll need to either have a mill available or be able to source pre-keyed shafts. The press is for both making keyways and for making sure that bearings go onto shafts (and if you want a slip fit, you’ll also want a lathe and plenty of sandpaper/files).
Like i said, it would probably be cheaper, my main point was about how they are harder to work with because of them being steel and requiring keyways and presses etc. Probably could’ve phrased that better
Hmmmmm i don’t know about you, but we usually have a hard time not reaching the weight limit, and lightening gears sounds like a lot of time better spent elsewhere.
Interesting, maybe that’s just about Israel’s industry practices but we couldn’t find any flanged bearings in Israeli industry, and our mentors who work in the industry tell us they aren’t a thing.
This is my main issue with this concept, if you wanna do this to learn industry practices and for fun go ahead but doing this stuff in a season just demands you to manufacture a lot of stuff you could’ve just bought from Vex in-house, all just so you could avoid using Inches? That’s an aweful lot more trouble than it really is worth. I just think having the occasional Imperial dimension in the CAD really is better than having to broach and press and lighten steel gears all season long, and with all this equipment you really won’t save that much money, maybe in the long term but it’s still not worth the time in my opinion.
1 Like
I can’t speak for OP but my understanding is that they are in a situation where metric is their only option because they cannot source imperial gears locally and Vex may not be a very easy option.
That is incorrect, I know of at least 3 suppliers in Israel that sell metric flanged bearings. We used them all over our intake this year since they were more readily available than the VEX ones
Really? Which ones? I might be wrong about that, obviously it’s better to use metric when possible but i still think it’s a mistake to only use metric
1 Like
It not so much flange bearings but hex bearings that are unusual for outside of FRC. Bearings these days are mostly native metric with ID and OD sized to suit the market. Thunderhex is a mish mash because the shaft is 1/2 hex rounded to take an odd size ID metric bearing with a standard 1-1/8" OD. It could just as easily been a odd size imperial ID.
It comes down to supply chain. If you have a good source of gears sized to suit FRC applications and are willing to not use 1/2 hex to transmit torque then you can use whatever units that suit you.
Odd, yes. 35/64" is the nearest binary fraction to 13.75 mm, and its fit on a Thunderhex shaft would be too loose.
The bearing is a custom size. Vex could have sized the extrusion slightly differently and sized the custom order bearing to a decimal inch equivalent. I have to admit I always size my shafts to available bearing sizes rather than order custom bearings. It might well be the the supply chain says the 13.75MM is an less expensive order. I am by no means suggest they should.
What is really weird is you can buy metric pipe which corresponds to NPS pipe which is loosely based on the old IPS standard which is based on some long lost brain fart.
There’s not really any difference in designing gearboxes around metric components versus imperial. All you do is type in different values into CAD. Everything at VEX internally is designed in nominal metric dimensions (albeit with specific standard interfaces, SAE threads, etc to match existing “standards”). Calculations are also way easier when you stick to pure SI units.
There’s a ton of “non-standard standard” stuff in FRC. Some of these items are like vestigial tails, left over from random early COTS or Kit of Parts components that were initially one-off products 20ish years ago, but became defacto standards due to becoming widely adopted by FRC teams. Others were the result of what was simply easy/affordable to manufacture in school machine shops in the era before widely available COTS parts. And since FRC was almost exclusively a US affair back then, most of these “non-standard standards” are all nominal imperial.
In “the real world”, hex is rarely used for torque transfer in gearboxes or transmissions. The reason why is because it’s actually pretty difficult to manufacture precision products that have an accurately sized hexagon hole that is also accurately positioned to be concentric with the gear tooth profile. This is why “real world” products use shaft interfaces like press fits, keys (aka mechanical fuse), and splines, depending on the use case, as these are all significantly easier to guarantee low runout.
4 Likes
I was going to bring this one up next. @benjierex, I’m going to take you down memory lane, if that’s all right.
Imagine a world where VEX*, AndyMark, CTRE, REV, WCP, and the smaller FRC suppliers don’t exist. Your Kit of Parts is a bunch of motors of various types–and that’s your motor limit as well. The Kit drivebase is a drill gearbox, with motor, per side, that has to be put onto the robot (possibly through a secondary gearbox). And you need a different ratio than the KOP box for some other task. My first year on a team, my team ran into that for an arm.
So we sourced gears (cast iron/steel–by our first event they were all steel because we’d broken the teeth off the cast iron one), got the side plates machined by a sponsor, got the gears lightweighted by the same sponsor, broached the hubs where necessary and assembled the gearbox. All over the country (and Canada) other teams were doing the same thing. Steel sprockets? Light-weight 'em. (I have access to a robot of that period, once the pandemic subsides–that one we had to replace the drive sprockets in the offseason because we lightweighted them too much and they failed. They’re now solid steel and the robot’s overweight.)
A few years later, along came AndyMark, along came the IFI KOP drivetrain… and within 4 seasons, FRC would never be the same as COTS revolutionized the competition. Aluminum gears**, “standard” hole patterns, and the removal of the “no COTS components custom for the game” rule changed the entire game. You still see custom gearboxes, but usually it’s because the COTS ones for some reason can’t do the job (and then you know that within a year there’ll be a COTS one).
I’d tend to agree that just ordering from [insert FRC supplier here] for parts can do the job better, but I wouldn’t say that making your own from near-scratch is a bad thing. Just might be a bit more of a headache than you want in a given year. That’s why I think it’s a good offseason project if you have the resources.
*In this particular case… as we know them. We’ll call them IFI, and they only do FRC-type control systems. No gearboxes, chassis stuff, etc. If it isn’t a Victor 884, radio, Operator Interface, or Robot Controller, then they don’t do it. YET.
**I once looked into using aluminum gears in a work project, picking on VEX stock. I quickly determined that they’d be aluminum ragged-edged discs within a short time period and decided against that. But I also didn’t have the available numbers for non-steel gear tooth strength; had to do that math myself.
Ahh the Fisher Price motors and gear boxes.
QFT preach!
This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.