West Coast Drive Design Help

My team has wanted to make a west coast drive for about a year now. Recently, I designed one to try and get this to happen during the off season. Yesterday, the team met with our machining sponsor to do a design review. They advised 1. that we reorient the bearing block or change it for something else and 2. add a support through the middle of the chassis that attaches to the 2x1 tubes and the belly pan to add more strength.

I know that what I designed is far from perfect (And oversized. We’re replacing an old chassis, so design constraints), but we all have to start somewhere. I wanted to know what the FRC community thought about what our sponsor said and the following questions.

Did I set up the bearing block wrong?
How thick should the ribs be for all the supports in the bellypan and the 2x1 tubes? 1/8 in? 1/4 in?
What would be a better way to hold the batteries?
Is the center drop enough?
Does it need the support in the middle and how would this be done?

Google Drive Link to CAD files: https://drive.google.com/drive/folders/0B8OmTfKXwOeMWG4wNnUwZWNKWXM?usp=sharing

Yes, I know about a lot of the resources that teach people how to make a west coast drives like the Greybots RAMP program and things like that. But I want to know what you guys think. As always any other input would be appreciated. Thanks for helping.

I’m not somewhere where I can take a look at the CAD right now. But if you really want some good help with your WCD come to our shop and look at some of ours. I offered that a couple weeks ago as well to someone else on your team. You’ll get good suggestions off of CD but I think you might be best served if you come to our shop with your questions and we show you some of our WCDs.

PM if you’re interested we’re happy to help you guys out! :smiley:

I can’t load solidworks files, but I will try to speak to their feedback. Usually our industry sponsors’ concerns come from a perspective of needing higher strength over the long-term, so they can be valuable.

If you think the bearing block is setup correctly for FRC, see what they think about the VEX clamping versablocks. Four 1/4-20 bolts through one set of those held up our 140-lb robot through around 180 climbs this year, no matter what torsional stresses we put on them. They’re pretty tough.

Adding a 2x1 extrusion across the middle of the robot is pretty common. It provides a lot of support in the middle of the robot for everything from a battery to elevator rails. 1678’s 2016 bot did in a clever way that still allowed the WCD chains to go where they needed.

A few questions and suggestions:

  1. Why did you go with pocketed 1/8" wall tubing instead of unpocketed 1/16" (especially since you’re using bearing blocks)
  2. Why did you go for a pocketed sheet metal bellypan instead of one made from plywood or lexan or another plastic or even thin unpocketed sheet metal?
  3. What are the bearings in the drive rails next to the gearboxes for? There doesn’t seem to be a hole in the tube for them.
  4. Maybe you want to make the gussets more of a triangle shape instead of a T. That should help improve torsional rigidity
  5. If you do want to stick with a pocketed sheet metal bellypan, you should pocket it so there are torsional members going all the way across the diagonals from one corner to the other (i.e. make sure the linear pattern has the same number of patters in both directions)
  6. How are you mounting bumpers?
  7. Why do you have shaft collars on both sides of the wheel? For that matter, why are you using shaft collars at all for the wheels instead of drilling and tapping a hole in the end of the shaft and constraining it with a screw and washer?
  8. It seems that the outside mounting holes for the gearboxes aren’t properly lines up in the CAD model. You probably want to check that before you send them off for production.
  9. Why do you have one motor with a 12t pinion and one with a 14t pinion in each gearbox?
  10. Any thoughts on where an encoder is going?

Do not use 1/16’’ tubing on a drive chassis. Speaking from experience, it will not hold up.

That’s definitely not a universal truth… we ran 1/16" wall only on our drive this year.

For a west coast drive for 95% of teams I’d recommend 1/8" all around though.

We use 1/16" for crossmembers in WCDs, and have debated doing 1/16" with siderails since our bearing blocks facemount and we have no contact w/ the edge of the tube wall.

I’ve been experimenting with thin metal plate inserts to add more wall thickness where bearing blocks and gearbox mounts are. I’ve yet to do this in real life (will be testing this December) but I think it is a viable solution.

Ok, then at least use 0.1" wall VersaFrame. You’re using 2x1" tubing with holes at 1" spacing, which is one of the VersaFrame sizes. According to your CAD model, unpocketed 0.1" wall tubing is basically the same weight as your 0.125" pocketed tube. And you’ll save a ton of machining time drilling holes and milling pockets.

Oh, I’m not suggesting pocketed tube (as it’s super inefficient).

I’m talking about 1/16 versus 1/8" wall solid.

I always forget about the .1 wall because we buy from an aluminum supplier, but it’s definitely a nice compromise that saves some weight (about .7-1 lbs for typical FRC drive siderails).

We ran 1/16 last year and it survived just fine; the bumper rail was isolated from the drive rails and absorbed all the hits. We will probably use 1/8 this year though.

I am not able to open the CAD, so just a comment on your original question.

A cross rail is a good idea. There is a lot of weight hanging on the inside of the gearbox mount rails, a cross will help. A cross will also strengthen the whole chassis to absorb hits and limit the potential damage.

Like I said it’s far from perfect or complete for that matter. As the sole person working on this without a mentor to help is hard for anyone (I do still ask my team on their opinions. It’s just no one’s splitting the work with me). I will go back and check everything. I just wanted to get a gauge of what’s missing and what’s bad.

  1. From what I saw at champ, not a lot of people chose 1/16" wall if they had such intense pocketing. We were thinking that we wanted to do something like what 3005 had for the 2017 season. We had recently got the machining sponsor mentioned before and wanted to use it to up the quality and the cool factor of our work. It seems that Adam and you recommend no pocketing 1/16" or 1/10" wall and just getting the Bellypan lasered or water-jetted.
  2. Well If the bellypan was made of lexan, there would be a giant dip in the middle no? We would have to put supports through the whole thing or something like what 1285 did with their robot in 2017. For an over sized off season chassis we felt it wasn’t worth it.
  3. I know that the direct driven wheel needs the gearbox specific bearing block so probably a mistake.
  4. True, after looking at the 1678 robot, it seems like a must have.
  5. So basically I’ll have to make it square? Is that why people do diamonds? I couldn’t get diamonds to work for the life of me so some insight on that would be helpful.
  6. No bumpers, it’s the off season and I guarantee that no one is going to be within 10 ft of it unless it’s off. (Everyone wants to catch shirts, not inspect the robot). Either way we probably should make some and there are mounting points on the Upper frame that we’re recycling (Our t-shirt cannon chassis is really broken).
  7. True, that’s for sure changing. With all the other things I thought this was a minor detail that doesn’t need the CAD for us to make the decision to change it. I did my research, but because it’s my first time, I didn’t know what I was really doing.
  8. Yes, will get changed.
  9. The gearbox was pulled from the 3847 CAD library. It’s useful so I use it. All we need to know is that there is a gearbox there. And I’m pretty sure they pulled that model from West Coast Products, so either way the model was going to be wrong, I think…
  10. Good question, my understanding was that the inside of the gearboxes there is a place to access the output shaft where the encoders fit. Is that correct?

I’m not 100% sure about Lexan because I haven’t used it before, but I know that plenty of teams use plywood bellypans unsupported in the middle. As long as you’re not putting anything really heavy on it (e.g. battery), it shouldn’t deflect much as long as it’s supported around all of the edges. I imagine Laxan, Garolite, and any other kind of plastic I’ve heard of teams using would respond in basically the same way at the correct thicknesses. This thread has some examples, and you can search CD for “bellypan” and you’ll find a number of good threads about how successful teams do theirs.

Andrew Lawrence made this tutorial on CADing bellypan lightening patterns. It’s similar to what you did, but in the end you end up with a diagonal across the whole frame.

If you’re also using this project to teach new members how to build a robot, it’s a good idea to build bumpers even if they aren’t needed. That way the first set they build won’t be the ones going on the competition bot.

The end of the output shaft is drilled out to 1/4". I believe either Vex or WCP sells a kit to attach a Grayhill encoder to there. It’s also possible to attach it to one of the front or back wheels if you’d rather save space. Something like this usually works. It’s not that important to put in now, but it’s something you should consider as the design progresses.

Build a “bling” set–because this isn’t for competition, you can use non-standard colors and use decorations. You don’t have to put your team number on them (though you may want to)–but you could put stripes, tiedye, team logos, school logos…

Then you just have to remind everybody that “We do these like we did these others, but in red/blue and with only the team number on them”.

Well, I suggested this to them about 2 weeks ago just to see the reaction. And let’s just say not everyone has the same standards when it comes to the words bling and bumpers.

It was shut down before I said the second sentence.

Trust me we would if we could.

You have to consider the tradeoff between saving a little weight and playing with fire in drivetrain design.

First, I wouldn’t pocket drive tubes at all - either use thinner tube or leave it unpocketed. It’s just eating tons of machine time for relatively little weight savings for all the strength loss.

Personally, I would rather just use 1/8" wall tube throughout the drive, even the crossmembers. The problems with 1/16" for the drive rails have been documented, particularly if you have your bearing or bearing block contacting the wall thickness of the tube via a pocket or hole. That said I don’t even really like it for crossmembers. I’ve seen teams have 1/16" crossmembers bend and buckle in games with lots of high-speed ramming (2014), and I just am not comfortable taking that increase in risk in the most important part of the robot to save optimistically 2 pounds.

Until you’re in the top tier of competitiveness, where missing functionality is really the sole limiting factor in robot success (as opposed to tuning, drive practice, your existing mechanisms actually working to full potential, etc), the weight limit isn’t such a brick wall that you have to shave every ounce out of everything no matter the risk. Obviously you want to build as light as practical whenever you can, but there are plenty of less risky places to take weight out of a west coast style FRC robot - pocketing the drive gears, 1/16" wall tubing for mechanisms in the frame perimeter, less use of thick plate in general, etc. It honestly takes a fairly complicated and feature dense robot to meet the 120lb limit if you build simple, light mechanisms, and so I think you need to carefully decide where it is “right” to save weight versus leaving the extra strength in the design.

So here’s the progress so far. https://drive.google.com/open?id=1LsUu_t8H0YynTpokM9vDiG4nsZQ19YF2

The last few changes will include:

Chain Hole in the cross supports
Removing some of the pocketing?

Just curious, does this type of pocketing really mess with the strength that much? From what I saw last year a lot of teams had minimal damage to their chassis’ other than 3005 which had a giant dent in the tube, but that was only because there was no bumper covering it.

We wanted to do this because one of our sponsors give us free metal. The machining time seems worth it, but I would agree that 3.5+ hours per side times 4 sides and 4 pieces does seem like a long time. Also $25 dollars per versaframe tube feels kinda tempting.