pic: Wood Coast Drive V2



A redesign of the wood west-coast style drivetrain found here.

Improvements include:
-Integrated bumper mount with brackets and bumper supports
-Removable custom 2-stage gearbox for a free speed of ~12.5 ft/s
-No precision machining is required for any metal components
-Single piece belly pan and top plate (made by flipping the plywood sheet between two separate laser operations)
-Custom plywood wheels designed by s_forbes (though any wheel with a 1/2" hex bore can be used)
-Equation-driven CAD for faster modification
-Incorporates new Bit Buckets CAD Library components
-Stained black for superior performance

Game allowing, we hope to use this drivetrain for the 2016 season. CAD will be released before Kickoff.

I’m happy to answer any questions and welcome suggestions for improvements.

I love the laser-cut plywood bases you guys build. I hope this works well for you, because I want to try it myself some year!

What size bed do you need for this one? ~3 ft squareish?

Very cool, however one point, everyone knows that you need to make it red to make it faster :wink:

Interesting design. Really like it. :slight_smile:

Just one thing, why did you choose wood over aluminum for the drive chassis material? Is it for the weight benefits? I always thought that wood for a drive chassis is not preferable because when, say, you get hit by another robot, then the wood may break, or something like that. Really not sure of this, since I’ve never done so much woodwork in the past.

Really digging the black. Are you just staining black, or will anything be green?
What made you go with 12 fps single speed over something faster like 16-17fps or a shifting gearbox?
What is the overall weight for this model, minus electronics?
That battery mount is gorgeous.
Cound you post screenshots of the plywood wheel in detail? That in particular looks really interesting.

Looking good, Cal!

It’s mostly for convenience. We have a Trotec Speedy 300 (~29" x 17") laser cutter in our workspace that can cut up to 1/4" plywood. Since we have no CNC metal-working tools, the laser cutter is by far the fastest method to turn an idea into reality.

It’s an excellent material, though, and surprisingly cheap. There are drawbacks (widening of side-loaded bearing and bolt holes, chipping, attachment method, etc) but we feel the benefits outweigh the issues. We may experiment with wooden gussets and aluminum tubing for the interface between the drive base and mechanisms, as well as the mechanisms themselves. While laser cut plywood is great, we like to stress that it’s not always the right material for the job.

Our laser cutter bed is 29x17", but parts up to 29x34" can be cut in two operations, one for each end, by opening the machine’s door so the material can stick out.

-Our final color scheme is TBD.
-Gearbox ratios can be easily changed for free speeds from ~9 ft/s to >20 ft/s, depending on the distances normally traveled in the game. The chassis could easily be modified to use ball shifters, mounted like this.
-Excluding electronics and hardware not shown in the model and including motors and belts, the drivetrain weighs ~28.4 lbs.
-Images of the wheel can be found here. The only difference is that 6 dowel pins are used in each wheel in the album, as opposed to 3 in the robot’s wheels.

When you decide to use plywood, do you do any calculations to guide your design? Do you use any of the formulas that civil engineers use for plywood (such as http://www.canply.org/pdf/main/plywood_designfund.pdf), do you use rules of thumb you can share, or do you just iterate until it works?

Are the captured nuts the only way you fasten the sheets, or do you glue them as well? Do you find the structure fails at the fastener typically, or some other way?

Thanks again for sharing

Does the plywood ever warp over time?

Mostly we iterate :slight_smile: Last year was the first with primarily plywood, so there was little worry about strength (no defense). We’ll see if this year is different. Typically we just try to use good judgement on the placement of sheets. We didn’t glue much on our last robot, but we’re looking at gluing for tight spaces this year where we can’t fit a captive nut. We have seen a few breaks near the slots in a recent side project… better to learn that now than in 2 months.

It does, our space doesn’t have much protection from the environment, so we get occasional warping after a few weeks. We’re going to develop a jig to flatten the sheets in the cutter before build season. On the robot it doesn’t matter much - if it’s designed right, the warping disappears when constructed.

How well do these two operations line up? It must be virtually impossible to get a perfect seam with this, but I am guessing you can get them good enough?

I would not be concerned so much with X/Y lineup as much as Z rotation when re-jigging the part. Could lead to some wonky wheel misalignment

This looks great. Looks much cleaner and more open than last year’s. Will there be a top plate or any cross beams? Also, about how many individual parts are there here? It looks like a lot less.

Are you planning to build one before build season for practice?

EDIT:

I can’t speak for the team, seeing as I graduated last year, but I would assume that the flattening jig that Kevin mentioned would be a big part of making sure they line up. Additionally, plywood sheets are cheap, so if they can always cut another.

If a CIM costs more than all of your combined drivetrain raw material, I guess you can afford to make mistakes. :slight_smile:

Everybody should know by now that red robots are unlucky, unless fielded by 1114.

Amen to that!

How does attaching stuff to the robot work? It’s probably a naive question, but I was wondering if using pop rivets is as easy as fastening them to metal? (I ask because I’ve never seen pop rivets used on wood before)

Yes pop rivets can be used with wood successfully. That is the way that all the angle aluminum, handles and latches are attached to the road cases that transport the fields and materials are made. Now when we made them for the PNW district we put a back up washer on them but the ones that FIRST made that I have looked at closely did not use back up washers.

Of course nuts, bolts and washers are also a option.

For the potential problem of bearing holes wanting to wallow out VEX makes these. http://content.vexrobotics.com/vexpro/pdf/217-4183-Drawing-20141111.PDF Cut the hole for the bearing place a bearing in this mount and into the hole and then rivet then match drill and rivet the plate in place.

In 2005 we built a wooden arm for our robot… we kept a spare in the pit so that when the judges came by they could jump on it. Nobody with aluminum or steel arms was inviting judges to jump on their arms. They were either too valuable, or too flexible to demo this way. We had the arm in a three point bend test with the two largest judges bouncing up and down on it… and came away with the first of several design awards for working with wood.

In 2010 we went “all in” with a complete plywood chassis. No laser cutter, just a tablesaw, drill press and some baltic birch plywood. We tested it against some 1" aluminum tubing we had sitting around the shop, and all I’ve got to say is that the aluminum robots were lucky that we all had (wooden) bumpers on!

The teams with the lasers have since gone well above and beyond anything we did, achieving both lighter weights and more rigid frames. I believe it was 1899, from Seattle, that had a spare chassis in their pit one year and would invite judges to stand on it, taking our “just try and destroy it” schtick to a new level.

One way of looking at it is that wood is nature’s carbon fiber. The lower density allows for great cross sectional structures and enhanced stiffness. As structures often fail by buckling or bending, the extra strength of metals is offset by the enhanced specific stiffness of wood. Add to that the low cost and ease of working with wood, and it can solve problems that metals cannot.

Oh… and should wood need repair, rather than running to the machine shop to get something welded, you can just grab some fiberglass cloth and 5 minute epoxy and have it rebuilt stronger than it was before your next match.

Please don’t interpret this as saying that aluminum is no good… there are plenty of beautiful aluminum, steel, and composite drive trains and components out there… but wood is a very, very competitive material choice for teams that want to consider it.

Jason

Realy nice design!

Can you do a predication of the weight without the battery?
Keep on the good work guys :slight_smile: .

No top plate or cross beams should be needed.

There are 27 plywood parts in the chassis, 5 in each gearbox and 7 in each wheel.

Not sure; we’ll see how much time we have. We will at least be prototyping the wooden wheels and experimenting with laser-cutting larger parts.

With electronics, no battery, ~34.7lbs. The wood chassis, including gearbox structure and bumper mounts, weighs ~10.2lbs.

I added the superstructure mount. It weighs only ~3 lbs and can be tilted by removing two quick-release pins. I also moved the bumper mounts to the superstructure mount to facilitate access.

Images