No metal bot

I know that Team 58 - Riot Crew has done robots in the past where they are mostly made of wood. From what I’ve seen, they’ve worked really well.

Still have the shooter!

Jim, spot on. I’ve scanned that section as well. I actually have Behind The Design from 2006 so I also scanned the bit on 190. It’s more controls oriented but talks some about the design process.

Here’s the links to each PDF.



As an aside, I know the Behind The Design books aren’t floating around that much nowadays. I have 2015 & 2006. If anyone wants specific sections scanned let me know.

2 Likes

Thanks for sharing, Nick! A slight tangent (probably not worth a different thread), but I’d love a “behind the design” for 2020 robots, especially those that may not see a field.

2 Likes

Since it’s relevant to the topic, I recently stumbled onto this picture from our 2014 season:


(Sorry for the blurry picture, this is all I could find.)

If we’d done this as a dead-axle setup, we could probably get away with wooden drive shafts. You could swap the bearings out for nylon bushings easily enough (durability is a different question). The only structural element left to deal with would be the bolts. Could you replace them with dowels and wood glue?

5 Likes

Wow! That 'bot is giving me flashbacks to my second FRC season, with 1346, in 2005. We used the (then brand new) KitBot chassis that year, but laminated our own foam core fibreglass panels for the superstructure and built the arm out of plywood. We missed out on a few things like having a low CoG and avoiding large-diameter, long stroke pneumatics, but the wood and glass won us our first ever FRC trophy (Creativity) and set us on a long path of using baltic birch plywood in structural applications.

Thanks to the magic of the Blue Alliance, I’m even able to find 15+ year old video of the machine! Quals 88 - Greater Toronto Regional 2005 - The Blue Alliance

The chassis sort of reminds me of our 2010 drive train pic: 1346 Plywood Drivetrain and pic: 1346 Breakaway Twins

Our 2009 bot also had a wooden drive base, but we returned to aluminum tubing for the superstructure as we had a student who was excellent at welding thin walled tubing that year. pic: 1346 Side View 2009

It was never so much about avoiding metal for us, and choosing the most appropriate material, which in our shop (where we had a CNC router, but not a CNC mill) was often wood. It is inexpensive, forgiving, easy to repair with epoxy, and has excellent stiffness-to-weight when used wisely. Biodegradable and low carbon intensivity, too.

Wood is nature’s carbon fiber!

Jason
(P.S. Hope you’re still happy with that 'bot, too!)

6 Likes

oof, 2014 was rough on bots.

As far as the fasteners: Could just design it with proper joinery. There is a joint for everything. Accept the fact that the wood will move some and design around that. Make the frame out of ash or hickory and you would be fine. Hold it together with wedged tenons and a plywood panel to keep it from racking too much. Well executed it would probably do better than a lot of bots (2014 lawnchair bot anyone?)

I mean it would suck if your robot went to a more humid environment and expanded past the margins for inspection…

1 Like

Haha, yeah, I wasn’t involved in 2014 so I unfortunately can’t answer the glaring “why would you do this?” question. It seems like a terrible year to be playing with wooden robots, to me. Dovetailing the joints would be amazing, but probably beyond our level of competency. Mimicking the plate & tube construction style with wood is a reasonable compromise for the end goal of making a wooden robot, IMO.

Humidity expansion of the frame is one of many reasons not to build your robot to max dimensions :slight_smile: .

1 Like

Just have a 3D printer churning out markforged bolts, obviously.

2 Likes

Dovetails would look amazing for sure, especially with contrasting wood colors (inlayed dovetails?) but you would probably want yo use a framing joint.

The plate and tube method is good, but I am partial to the “shell” method. spread the forces out and lower the loads. At a certain point “frame” and “shields” on the robot become one.

1 Like

Maybe the move is to get a large piece of solid lumber and mill the whole thing out like a wodden Battlebot.

The guts of Duck! for reference:

9 Likes

Not playing to the strengths of the material there though…

1 Like

I didn’t realize we were trying to make a good robot here.

2 Likes

lol

legit bumper question: If I make my robot frame from plywood can the frame itself be part of the bumpers? can I make the bumpers structural to the robot? What counts as “frame” so long as the bumper is “supported”?

I see a significant loophole for robot mass exploitation. Say you had 5lbs of fabric&noodles. 10 pounds of external frame is significant.

Given the right game this could be a serious question.

3 Likes

This picture is giving off certain vibes

6 Likes

I will kill the ambiguity. I AM DEAD SERIOUS.

Hmmm…

The first sentence of the bumper rules is:

A BUMPER is a required assembly which attaches to the ROBOT frame.

You’d have to make the tautological argument that because each piece of the frame is attached to at least one other frame piece, that your FRAME is attached to your FRAME.

Each set of BUMPERS (including any fasteners and/or structures that attach them to the
ROBOT) must weigh no more than 15 lbs.

Since we’re defining the whole frame as also being the bumpers, your frame is going to have to be pretty lightweight.

BUMPERS (the entire BUMPER, not just the cover) must be designed for quick and easy
installation and removal to facilitate inspection and weighing.

This could get you in trouble, too. Luckily, they didn’t define “quick” or “easy.”

I think the most realistic way to do this would be to build a kitbot-style dead axle drivetrain, where the outer plates are also your bumper plywood. So, the bearings and frame supports would attach directly into the bumpers.

That’s kinda what I was getting at. All the important stuff about frames in the rules (“support”, “perimeter”) is handled by a thin, very light weight piece of material. A large amount of rigidity comes from the bumper plywood/backing material(s) attached to the bumper itself.

For 2767, could you directly bolt swerve modules to the bumper plywood?

Like, using them as really elaborate corner brackets?

1 Like

We use inboard rails. so there is one more degree of separation between the frame and the module than most teams.

Now, could we tie a pair of rails together with a belly pan and add rigidity with the bumpers? Quite possible.
The 2020 manual on bumper backing material:

R24. BUMPERS must be constructed as follows (see Figure 9-6):
A. be backed by ¾ in. (nominal) thick (~19mm) by 5 in. ± ½ in. (~127 mm ± 12.7 mm) tall
plywood, Oriented Strand Board (OSB) or solid wood (with the exception of balsa). Small
clearance pockets and/or access holes in the wood backing are permitted, as long as they
do not significantly affect the structural integrity of the BUMPER.

So there is a lot to interpret with the material selection, and there is no limitations on the corner joinery that hamper max rigidity. There is potentially even more to be gained with the rules regarding mounting systems. In theory I could back the whole bumper with a 1x2 tube and call it my “mounting system”. Or in the context of the 2020 2767 bot: attach the chromoly frame as built
to the bumpers instead, use the “rigid fastening system” to attach the rails and bellypan/super structure.

R24.
G. must attach to the FRAME PERIMETER of the ROBOT with a rigid fastening system to
form a tight, robust connection to the main structure/frame (e.g. not attached with hookand-loop, tape, or tie-wraps). The attachment system must be designed to withstand
vigorous game play. All removable fasteners (e.g. bolts, locking pins, pip-pins, etc.) will be
considered part of the BUMPERS.

“main structure/frame” is not defined, so I am going to interpret it by mass instead of by rigidity.