New(ish) Drivetrain

Hi everyone!

I wanted to get some feedback/opinions on a drivetrain I’ve been designing. With the rise of teams successfully using simple ctc chain runs (with that added bonus stretch), I thought it might be cool to design one of these drivetrains for myself. I’ve been trying to challenge myself to speed up the process of manufacturing, especially on the drivetrain. For next season our team is planning on routing a standard 1" hole pattern into a ton of tube in the first couple days of build season (and in the weeks leading up to season for training/practice bot) to reduce manufacturing time once the robot is designed, so I thought I might take advantage of that pre-machined stock to speed things up a bit. The bearing blocks are routed from 1/4" aluminum, and are designed such that the wheel drop and chain stretch offsets are accounted for in the part (hole is not centered), keeping the hole pattern the standard 1" tube hole pattern we like to use everywhere. By using a few templates or just quickly drilling/drill pressing in oversized clearance holes for the shaft, these bearing blocks will bolt in to the closest holes that haven’t been drilled out and make life super easy for assembly, leaving all the hard alignment work to those bearing blocks. The next challenge was gearbox holes, and assuming I don’t do custom gearboxes next year (in which case I’ll match the pre-existing hole pattern) I’ll either mill out those two holes or use a template and drill press carefully. Bellypan is lasercut 6mm Baltic Birch, and all of the tubes are 1/8 6061 Aluminum.

The cad can be found here:
https://cad.onshape.com/documents/0d4fa499c3575d31041de583/w/0690736d523e1da52640d8b9/e/659b20324900375b21ae3c91

Screenshots/STEP file can be found here:
https://drive.google.com/open?id=0B8...lVTeVdXNTZTZ3M

I’d love to get some feedback on this type of drivetrain as we have never really run it before (kinda did in 2016 but lots of different stuff going on). Obviously this method of speeding things up really only works for my team (unless yours has the same shop, resources, etc.), however I think lots of the concepts could be used by other teams looking to speed up their drivetrain process, make it more adjustable on the fly, or just make life easier in the shop while sending plates out to sponsors.

For no reason other than the fact that I was bored, I designed this drivetrain using Onshape variables and some backend Featurescript. Essentially, that means it can be modified to suit any game’s frame perimeter rules. You can edit these variables on the left, on top of the features list in the drivetrain frame tab. None of the assembly beyond whats referencing the frame will change (for example, changing wheel size will move the bearing block over, but it won’t replace the 4" default wheel in there) (yet).

Here are the variables and what they do:

#Length: adjusts the overall length of the frame
#Width adjusts the overall width of the frame
#drop adjusts the wheel drop
#Wheel_Size adjusts the spacing of the wheels so they fit (buggy, yes I know I’m working on it)
#Gearbox adjusts the spacing of the gearbox mounting holes
#L_Length adjusts the size of the L Brackets
#Chain_offset adjusts how much added spacing there is from the ctc of the chain runs. Default should be .02, a number I got from other posts on chief and discussions with designers more experienced than I am
#GB_Slot_Width sets the width of the gearbox cutouts in the bellypan
#GB_Slot_Length sets the length of the gearbox cutouts in the bellypan

Thanks in advance for all of your feedback, I’m looking forward to seeing what everybody thinks!

-Marcus

Special thanks to Anand, Chak, and that mechanical crew for putting up with my stupid questions, as well as Michael Bick (1836 alum) for inspiration from his 2016 drivetrain and Adam Heard from 973 for pushing me and my team to get faster and better at everything we do (also answering my stupid questions yesterday).

Looks great, Marcus! You can definitely see the 973-inspired design style in there.

Really the only thing I’d change is eliminating the bearing blocks. If you’re not using sliding tension and just doing C-C, you can put your bearing holes directly into your 1/8" thick tubing (this doesn’t work well for 1/16" wall tubing, but that’s not relevant here). If you really wanted support for the full 1/4" thickness of the bearing, make your bearing plates 1/8" and combine that thickness with the tube wall thickness, though again even having the plates isn’t necessary.

This is also entirely optional, but if you have solid bumper support, you can turn the front and back pieces of tubing into 1/16" wall for weight purposes.

Keep up the great work!

Andrew,

Thanks for the quick feedback. The bearing blocks are in there, not necessarily for full bearing support, but also just to speed things up. With only one machine doing our tubes and plates, doing drive rails on that router takes lots of time away from other manufacturing priorities. By using these bearing blocks I aimed to create a more modular system where we can take the hole patterned tube already routed and use very simple, entry level tools (drill press in this case) to make an equally functional drivetrain that allows us to easily change things and prototype different configurations needed. This should free up our router to manufacture other parts, as well as allow us to put the drivetrain together much much faster.

I’d expect to see this drivetrain in a cad for a Second Locomotive robot coming soon…

-Marcus

Looks great! Is there a reason you opted to not use VersaBlocks?

Hey Nick,

I decided not to use versablocks because they are designed (per my understanding from the similar vexpro WCD bearing blocks my team typically uses) to tension, and while I’m sure they could be used without a cam/tensioning setup, those bearing blocks don’t mesh with my pre-existing hole pattern, and aren’t offset for the extra .02 ballpark one should add for 25 chain ctc. These custom bearing blocks can easily be made on my CNC router and make life super easy and quick in the current setup.

Thanks for the feedback!

-Marcus

The Versablocks that ollien was referring to are these: https://www.vexrobotics.com/bearingblocks.html

They are designed to clamp around a 1x2" tube, and are made out of aluminum. You can get enough clamping force that they won’t move out of position, even if you don’t have use the cam that Vexpro sells. The center hole is offset vertically by .0625", so if you put the middle Versablock on your drivetrain “upside down” and the ones on the ends “right side up”, you end up with a 1/8" offset of the middle wheel.

This is our drivetrain structure from the 2017 season, using Versablocks without cams, and a plywood base. It worked great! https://goo.gl/photos/kZymWaEPNcBQcRxP7

One advantage of making your own is that you will save money, but with the design you posted you might be better off just machining the bearing holes directly into the frame if you can do it precisely.

My bad, was looking at the material for the wrong part. Just edited my previous post. Either way, I still prefer this setup because they are easily made in house very quickly. As I sorta explained above, the reason I’m not machining them directly into the frame is for simplicity/speed with using my pre-existing hole patterned tube. Really don’t want to sink a bunch of time into routing drive rails when I can just make these little plates and bolt them in.

I’m not super fond of the way the versablocks bolt over/under the tube. If I use them, it will be inconvenient to mount things nearby to the top of drive rails, and I need to account for them in bellypan and other stuff like that. This custom bearing block setup just makes life easier with my team’s resources, and I’d be hesitant to switch to these vexpro bearing blocks in this drivetrain unless we have issues with the custom ones. With custom blocks I can also run good experimentation with different ctc spacing, and modify the drop depending on presence of obstacles on the field or weird drivetrain dimensions.

However, thanks for pointing me in the direction of these bearing blocks. They seem to be a really cool COTS product that I haven’t used before and I can see some great uses in other drivetrains/applications.

My only concern with this design is the amount of potential flex in the frame induced by the weight of the gearboxes. Ideally, it would be a good idea to run some supports between the gearboxes so they aren’t cantilevered like that.

I’m in agreement with Andrew that you should just drill directly into the tube – with the predriled tube that you are planning to use that is a super quick operation on the mill, and you will get better accuracy that way. Personally, I’d probably run a WCP bearing block on the gearbox axle though for peace of mind.

The 2016 drivetrain had simple bearing blocks like this so that we could adjust c-c and center drop. Because that field was so different from past years we really didn’t know what would work, so the bearing blocks gave us the freedom to quickly adjust if/when our initial dimensions didn’t work. Assuming a flat field you should know those parameters, so I think it would be safe to drill directly into the tube.

Thanks for the feedback. My team (and many others) have successfully run gearboxes like this for many years. With proper maintenance, assembly, and taking care of the gearboxes, this should not be a problem, especially on COTS flipped cim gearboxes. However, if we do see problems like what you’re describing I will add some churro shafts between the two gearboxes or something to that effect to minimize the cantilever and make sure they hold up.

Michael,

Aside from the issue being that our mill is not working, I think being able to modify and iterate on drivetrain dimensions is still important, even if the field is flat. We are trying to get our drivetrain done super early this season, and this setup allows us to change things if they need to be changed. While you know the max frame size, it’s hard to guess things like how much space you need for over bumper intakes, or if you want to put a cheval mechanism somewhere. This type of modularity allows us to very quickly get a drivetrain done and then change things if needed down the road. This setup also allows us to iterate on different ctc spacings and play with center drop in the offseason, so it makes sense to stay consistent with the setup during season.

Hope you like my stolen idea of the little markings to identify orientation/iteration, wonder where I got it :smiley:

Thanks for the feedback,

-Marcus

For anyone that would like to look but doesn’t have/want to deal with Onshape, here is a Google Drive Folder with some screenshots and the STEP file available for download:

https://drive.google.com/open?id=0B8y-Pada5qtBSlVTeVdXNTZTZ3M
Thanks everyone for the feedback so far! Keep it coming!!

I would argue it’s harder to modify an existing tube with precision holes than it is to make 1/4" bearing blocks from scratch.
I really like the offset bearing block thing- I might steal that particular idea for the future.

We plan to run bearing blocks like this pretty much forever now.

Having no critical operations on the predrilled tube saves a lot of time, leaves the option to change C-c and let’s you run 1/16 wall tube (although I’d say stick with .125).

Looks like a really great, simple drivetrain. There’s nothing extremely crazy about it, but it’s really well tuned to your team’s specific situation, and that is what makes it truly good IMO. Do you have a weight on this?

Hey Mitch,

As I said above, this drivetrain is designed so that everything (most prominently overall frame size) can change easily. I went in and changed the default to 28 by 28 per your text, and the weight came out to about 16 lbs. This is including everything properly weighed using density/material settings in Onshape except for gearboxes (I’d love it if someone at WCP could let me know the weight of this gearbox, it’s not on the site), hardware (rivets/bolts/snap rings), and chain. Obviously it’s missing some bumper solution which is on my list of additions to make.

Thanks for the kind words,

-Marcus

Very interesting design. Curious as to how many teams will use this style of drivetrain in the coming years.