Robots with giant arms are going to be extremely common this season. With a 48" extension limit teams are going to be extending themselves further than we’ve maybe ever seen in FRC history.
There are a number of things that will come into play as a result of this:
1. Robot contact becomes more damaging: With a longer lever, opposing robots (or field elements) will impart a significantly greater torque on your robot’s pivot points. Teams will need to be very careful to avoid this type of contact, plan for certain failure points that enable quick fixes, or build their mechanisms to be durable via either flexibility or rigidity
2. Field Side Volunteers better keep their head on a swivel: the side of the field will be a dangerous place this season.
3. Greater Reach = More Tipped Robots: Teams should do everything they can to optimize for a low center of gravity. It will be a bad year for tippy robots.
4. Contact inside frame perimeter becomes easier: Many robots will be tall which will make it very easy to enable contact inside frame perimeter for a large extended arm. It will also be significantly easier for referees to call this season since the contact will take place in more visually obvious areas than in 2022.
Strategies to mitigate the issues above include:
Being concious about when you extend your arms (sticking to safe zones is a good idea)
Building robust mechanisms with planned failure modes for situations that involve robot contact
Don’t build long arms and instead strategically target other objectives
To expand on this post, if you are planning to build some sort of arm to tackle this seasons challenges, be careful on how you design said arm.
1/2 Hex shaft can and will bend or shear under high torque loads that a long arm will see. Switching to steel hex shaft won’t really help.
Some robust arm building techniques:
Use a dead axel setup where the arm rotates around a static shaft instead of a live axel setup where you rotate the axel
Use larger diameter aluminum round tubing (3/4" OD or larger) or something like the REV Max Extrusion for the axel
You can buy oil embedded flanged bushings and tube nuts from McMaster for various OD tube
7/8" OD tube and bushings play nicely with COTS sptockets
Have a chain for the final reduction, the chain will have a substantially higher % of tooth engaged compared to a gear where only a couple are engaged (commonly referred to a wrap), the chain will also tolerate shock loads
Backlash is the enemy of a long arm, consider ways to reduce this throughout your system such as shimming hex gears in a gearbox or tensioning a chain
Consider ways to counterbalance your arm to reduce the load on the motor
Do not try to use a weaker planetary gearbox like a VersaPlanetary to drive a high torque arm, the small internal gears and the thin casing will shear/tear out (ask me how I know )
Beefier Planetary gearboxes like the MaxPlanetary and AM CIMsport are probably fine
Reinforcing the tube near the axel with machined or 3D printed inserts will stiffen up that joint and prevent the tube from buckling
Is anyone aware of a past robot (or openalliance) cad example of a high torque-handling arm pivot that applies the principles and recommendations here? We have a lot of inexperienced folks (me included) & while I’m grok’ing most of this, a design to show and walk through to explain the what’s and why’s would help. Thanks.
You can probably poke around the decades of 330 single jointed arms; one thing Kremer sort of glossed there is making your own mounting holes connecting the sprocket to the arm, rather than relying on the fairly small pattern they usually come with. Adjustable chain tensioners on large reduction arms are fairly worthless (never seem to be strong enough), though the “snail” cam type on a bearing block for the sprocket can work.
Most of this stuff is “common sense” but it often requires pictures of an FRC implementation for things to click.
971 2018 is honestly a terrible example for most teams given the CF tube construction.
Also note that this is a cycling game, where smaller frame perimeters are encouraged for the end-game. Keeping your COG low will greatly improve your robots handling and stability while driving. You don’t have to build a giant A-frame arm or single stage elevator to the height limit.
A low COG tank drive will drive circles around a high COG swerve drive all day. Acceleration / ability to change direction quickly is the most important aspect to be able to cycle quickly through defence.
This is saying that if you adhere to these ratios, it should be impossible to break the gearbox because the motor will be incapable of applying enough torque to cause that to happen. If more torque than that is applied on the arm, the motor will just back-drive.
You are likely going to be hard pressed to find much CAD for arms this large. Especially if you are considering level 3 scoring. It’s been many years since FRC has built arms at this scale and going back that far is often not helpful since a lot has changed in 15 years of FRC.
It also depends on what resources you have available to you. For example, 971’s 2018 robot is great but many high resource teams would struggle to build that. I consider most of what 971 builds to be aspirational Maybe someday we can build robots like that but not this year.
Given your statement that you don’t have much experience, consider some of the vendor solutions like ION from REV. The have some great documentation to get you started.
If you have the resources to build something custom, follow the advice above and you’ll be in good shape. Here is an example from our 2019 robot that was a pretty high torque arm. It was built to lift the robot as well as game pieces. We never had the time to work all the backlash out, but it certainly put out torque and was reliable.
Could just one or the other of these be followed or are both needed? Could a typical 1/2 inch hex shaft be the dead axle, and that’s enough mitigation with sprocket + bearings/bushings spinning on the axle? Or do you need both the larger round tubing and a dead axle set up? I think you mean larger round tubing for a live axle (+ bushings probably)… I seem to remember someone saying smaller axles are ok (or even better) with a dead axle setup. Thanks.
I think 7461’s robot applies all of them in a way that’s COTS approachable, though I don’t recommend trying to pull off some of the weird packaging constraints (not biased)
Requested the sprockets and 7/8” bushings be ordered this evening, along with a backup axle (we had a length of 7/8 already). Good to learn this - thanks!
)
Maybe someday we can build robots like that but not this year.
