Do any other aluminum gears in frc have endurance testing published?
1690 did some aluminum and other materials on swerve. 1690 Swerve Drive 2019 off season project
I mean, they still sell the steel version, so if you have a bad experience with the aluminum ones, you can always just swap them out for those. Imo, the weight savings are significant enough that I’d run the aluminum ones with regular replacement over the steel ones regardless; 0.111 lbs vs. 0.323 lbs, nearly a pound (0.848 lbs) of weight cut on a full four-module drive is pretty wild, ngl. Doesn’t look like they’re coated though, dunno how that will impact their wear characteristics.
Our aluminum gears held up incredibly well - two years in and not a scratch. However, they are 60t and were made out of 7075 aluminum with a Teflon-infused coating, and we drive them with a brass gear, which really does contribute to their longevity (very low friction and we don’t need to grease them at all). I would assume WCP have put this through the some rigorous testing before releasing it so I wouldn’t worry about it too much.
(I’m guessing that there’s a reason why the small gear isn’t aluminum - probably didn’t hold up in testing, just like our tests)
Nope, and that’s a problem IMO. It’s hard enough to scrape together enough money to build an FRC robot in the first place, only to find out later that the slick gears you just bought will (depending on application) have to be replaced in less than a season. It’s false advertising by omission.
I’m happy that most vendors now offer at least both steel and aluminum options, I just wish they were more forthcoming on their storefronts about why teams would select each.
I have a history of calling out FRC vendors for insufficient documentation, but I’d say “false advertising” is a bit too strong an accusation here. In recent years, I’ve found WCP has actually been solid/improving in terms of documentation for their non-VEXPro products.
I’d love it if more load rating testing and guides (similar to the VersaPlanetary load rating) were available for power transmission components, but at the same time providing blanket guidance for bare gears might be a tall order, as gear wearing will differ greatly based on fabrication tolerances, motor power, and field terrain (among other things).
Overall thoughts here:
After seeing 1690 successfully field robots with aluminum bevel gears I started to look at plastic and aluminum bevel gears. We worked with our gear shop to make a few samples and have done some testing. Our baseline was it would be used for the competition robot only which for us gets about ~ 100/matches year @ 2 min/match ~ 200 min or 3.5 hours of practice. So far we’ve hit more than enough run time to feel good about using it in season on 1323. My global thought for products is if I’d use it on my frc team then other teams could use as well if they want to.
We have 4 teams testing the new swerve modules with all aluminum bevel gears. By the time the new swerve modules release we’ll have enough “long term” testing to say when/where to use these gears. All swerve kits will come with steel gears and the aluminum is an optional buy.
The tradeoff for aluminum vs steel gears has always been weight vs longevity/breakage. Overall if teams have doubts then they should just purchase the steel version as we offer all major gears from 6T-64T in steel along with all bevels in steel. We have put a big effort into making a manual for everything and will soon roll out example cad/assemblies for products. I’m unsure if a steel vs aluminum gear manual is in our pipeline as its too hard to predict all use cases.
I find that FRC vendors are often held to an insane standard that’s not paralleled in industry. Most of these product ecosystems have more documentation than even McMaster stuff.
WCP does a fine job in product documentation.
I also find it extremely hard to believe that WCP will sell a high profile item that they would not be confident enough to have their own WCP sponsored teams use.
Shoutout to the McM ball bearings that fall apart if you look at them wrong.
They’ve got a decent amount of garbage hiding in there, it helps when a lot of stuff is “name brand” or you kind of know what it is already.
McM might have 5 lines on a “high strength, braided sheath V belt”, but I know that’s a Gates Super HC and will work perfectly. Same with a “red and orange, comfort grip 1000v insulated screwdriver” being a Wera Kraftform. Bolts have near universal grades and thread class, etc.
The standards we hold FRC suppliers to aren’t out of the question compared to industrial applications (I’ve been on both of those phone calls), but they’re certainly disproportionate.
Perhaps with justification from the team side? Getting $50 for your FRC team is a lot harder than even $5000 at work, and I’ve thrown a trash can with a couple grand of bad parts into the dumpster countless times.
While I would also like longevity tests, I find it impractical to show longevity
As you stated, it all depends on the mechanism and usage. Gears in an elevator will probably last longer than ones in a drivetrain. Also, if you use 1/16" thick aluminum as the material to support the gears vs 1/8" or 1/4", that will also affect the wear.
I remember when we first used a 1/16" sheet metal drivetrain - we didn’t put extra support for the gearboxes and the gears turned to powder. Now that we put an extra 1/8" plate to help support the gears, we haven’t had an issue with wear
For something like gears, I think it’s really difficult and not very practical to list the longevity
It’s not difficult? It’s very common for gear manufacturers (if not all retailers) to list the limiting torque for gear tooth bending strength and surfaces durability (for some defined cycle life). These can be derived from formula, but consumers don’t know all the details about geometry, material and surface treatment. We have to trust the vendor.
I appreciate that WCP is doing their diligence to test these gears on a robot (thanks for mentioning that @R.C ) But I don’t think it’s unreasonable for them (or any other vendor) to run an endurance test on half a dozen samples, apply whatever conservative margin they like, and then publish the result. Especially when the limits are being pushed (as they are with aluminum gears).
How would you simulate an FRC endurance test? The loads vary wildly between teams, games, and configurations.
I can run through the ANSI gear calculations but it probably won’t be very accurate. I would like to see how much the additional undercut on FRC gears affects strength, but that’s really just to satisfy my curiosity.
You (they) don’t have to simulate an FRC season any more than Boston Gear has to simulate every mechanism ever built with their product. Just test whatever your published load limit is. After that it’s up to teams to ensure they design below that limit, but now they can do that armed with actual data.
Can you please walk us through how you would do that for a drivetrain with Boston gears?
How you would calculate the loads on a gear in a drive train and compare them to published load ratings? Have I understood your question properly? I don’t want to hijack this thread any more, but here you go:
Approach 1) Start with the worst case motor input - the stall torque. Increase that by the gear ratio of every stage before the gear in question. Add some safety factor (start high, maybe x3 or x4, work down with experience, or look up some standard load factors in the AGMA gear formulas). Done.
Approach 2) start with the maximum output wheel torque at slip (robot weight * cof * wheel radius / number of wheels touching the ground). Be extra conservative on all those numbers. Reduce that value by the gear ratio of every stage between your gear of interest and the wheel. Add some safety factor. Done.
The load factors will change with gear speed (first stage gears wear faster), reversing loads, impacts, and crappy shaft alignment. But if the ratings came with published assumptions for those things I’d be happy enough. Knowing exactly how much longer steel gears will last compared to aluminum for a given “model loading condition” would be very useful information when teams are deciding what to do with their money.
I believe this product was meant for use on their COTS swerve module, so there really isn’t much variation between configurations(robot weight would be the main variable and you can simply load it all the way) . It seems that they are already testing them on actual robots which, for FRC teams’ use is probably more than enough
That assumption is wide enough to easily prove aluminum is valid or invalid.
This is where the problem is. Manufacturers can rate all kinds of things, but they’re useless at the end of the day because of all the unknowns.
A student could do lewis gear calculations and calculate the force on the gear teeth from a stalled motor or the robot slipping on the ground. If they use that to make their drivetrain, it probably won’t work, depending on a hundred factors that are difficult or impossible to quantify. The knowledge would hurt more than it would help, ultimately, for all but a few people who understand the whole problem.
That picture pretty clearly looks hard anodized (likely with Teflon impregnation) to me, but I can’t speak for RC.