This is a new chassis design for my team. It is the first time we made a 6 wheel drive, the first time we had directly driven wheels, and the first time we cantilevered axles. I am using dead axles except in the middle. This is made of 2x1 C-channel. One of the goals we had in making this was to be fairly easy to build and to not have to use precise machining because our fabricatioin sponsor has a slow turn around time. I am open to suggestions to improve the design and will try to answer your questions.
Nice CAD! Just curious, why did you put the chain sprockets for the wheels on the outside? Wouldn’t they be safer and more effective on the inside?
Your drive is going to be way too fast ( > 16.5 fps) with a cimple box direct driving a 6" wheel. You’ll probably have issues with current draw as well due to under-gearing.
Aside from what Dustin said, I can’t quite tell, but how are your drive axles supported?
You will want to take another look at your gearing.
Also, going along with Akash’s remark, with the C-channel it looks like you are only supporting your wheel axles at one point. You will want at least two. For this and other structural reasons I would suggest box tube in place of the c-channel. It comes in various sizes (1"x2", 1"x1", etc.) and wall thicknesses (no less than 1/8" for a drive base usually) and can be found at (perhaps donated by) your local metal shop. Quite a few teams (including my own) use it structurally, and you can either weld or rivet (with gusset plates) the joints.
It looks like a Toughbox Mini.
It’s a toughbox mini, the max speed is just under 9 fps I think.
Hopefully these two pictures shows how the axles are supported a little better. There is a piece of c-channel nested in the main one and slotted on the top and bottom so we can slide it to move the chain. The axle is supported in two places.
We used c-channel instead of 2x1 box because it seemed to us easier to bolt things on to c-channel than box. With a box frame, you wouldn’t be able to get a nut on a bolt. Do other teams usually use rivets to avoid this problem?
I put chains on the outside on the middle to the back wheels, and the chains on the inside on the middle and front wheels. Maybe I’m wrong, but at the time I did it, it seemed like that would take less space and the wheel wouldn’t have to be cantilevered as far. In any case, I believe the bumpers should adequately protect the chain, won’t they?
*also, sorry for huge pictures, I uploaded them to Tinypic and didn’t know how to make them smaller here.
Most teams that use box tubing just go all the way through both sides of the tubing and put a nut on the bottom.
I put chains on the outside on the middle to the back wheels, and the chains on the inside on the middle and front wheels. Maybe I’m wrong, but at the time I did it, it seemed like that would take less space and the wheel wouldn’t have to be cantilevered as far. In any case, I believe the bumpers should adequately protect the chain, won’t they?
I would want the chain inside of the wheels. Less cantilever to the sprocket means less chance for the chain to jump the sprocket.
I wouldn’t worry about the bumpers protecting the chain so much as the chain rubbing the bumpers…
You could also try a trick that a number of teams using WCD pull–they run their chains inside the frame outline.
That would take pretty long bolts though, so I don’t know if it is the best option.
I would want the chain inside of the wheels. Less cantilever to the sprocket means less chance for the chain to jump the sprocket.
I wouldn’t worry about the bumpers protecting the chain so much as the chain rubbing the bumpers…
You could also try a trick that a number of teams using WCD pull–they run their chains inside the frame outline.
I didn’t think of some of those issues with the chain. I might move that chain to the other side of the wheel if we have room.
I’m not sure what you meant by unning the chains inside the frame outline. If you meant using live axles and having the sprockets on the inside, we are using dead axes so that wouldn’t work for us.
Other teams use welding to secure their frame.
Why are 2" or 3" bolts hard to come by? And with box tube, you would no longer need the extra piece of c-channel to support the axle in a second spot, so you don’t need those particular bolts any more anyway, AND you’ve gotten rid of 30 components (1 bracket, 2 bolts, 2 nuts per wheel) in the entire drive base, making the whole thing much simpler and less prone to failure.
Also, I would probably only position nuts/bolt heads in tight spaces like that as a last resort. That’s harder to get at than you might think, and you want that sort of thing to go fast.
Other than that, CAD looks good, and with Toughbox Nanos you should be fine (the orthogonal render made them look more like CIMple boxes…). I might prefer a bit faster, but 9fps is safe, and you’ll have little trouble pushing people around. You might want to see if you can get it to the traction limit.
If they are frugal with weight and keep it at about 100 pounds w/ battery and bumpers like we did in '11, the Cimple boxes aren’t a bad choice. The reduced weight makes it much easier to accelerate, and thus easier on the motors. The speed can be harder to control in tight spaces (probably not the best choice for most of LogoMotion), but it really helps when cruising down the field. Cross-field runs occur in about 4-5 seconds with it from stop, which can really save time
Do you have a capable engineering mentor on your team or in your area that can sit with you and review your work? You’ll get a lot more from that than you will from this forum, unfortunately.
There are a lot of problems here – and that’s okay. We all learn by trying new things and making mistakes. In drive design, the devil is in the details a lot of the time and your modeling isn’t very detailed, so forgive me if I’ve made incorrect assumptions in the advice I offer.
1 – It looks like you’re using Pro/E. You have my sympathy.
2 – Do not use C-channel to build your frame or otherwise expect to add a VERY rigid (i.e. thick) piece of wood or aluminum plate to the inside to add rigidity. You’ll otherwise have a lot of flex in the frame – I know this from experience. The problems you’re anticipating with using rectangular or square tubing aren’t real.
3 – The Toughbox Mini with long output shaft driving the center wheel is supported in two places in the Toughbox housing. That puts the load on that shaft at 1" (channel leg height) + spacer + sprocket + 1/2 wheel width away from the bearing. That’s probably 2.5" – too far. There’s all sorts of detail missing with respect to how you’re going to position the wheel on the shaft – is it retaining clips? spacers? Retaining clips weaken the shaft. Spacers will exert an axial load from the wheel onto whatever surface (presumably the channel) the spacer works off of. How will you handle that thrust loading?
4 – You’ve said the C-channel is 2x1 – presumably by 1/8" thick. That means the blue parts you show as your dead-axle support are 2.25 x 1.125 x .125". That size isn’t commercially available to my knowledge. How will you manufacture those parts?
5 – It doesn’t matter because they won’t work. You’ve got a hole in blue channel and a slot in the frame channel along side slots for mounting the blue channel. I presume those mounting slots are for chain tensioning. Think about what happens when you slide the blue channel away from the center when the chain is under tension. Your axles will not remain parallel to the center axle/toughbox output and you will throw chain and add a lot of drag to the drive.
6 – Is each Toughbox attached with only two screws? That’s going to exacerbate the flex in your frame. Follow the loading – weight of the robot passes through center wheels, bending toughbox output upward. Flex in output shaft tries to rotate the entire gearbox housing about its mounting point. This is hard to describe, so you have my apologies.
7 – The location of the sprockets doesn’t matter. All of the loading is passing through bearings / hubs in your wheels. The sprockets are floating around the shafts and your biggest concern isn’t torque applied by the chain bending your axles in a horizontal plane, but torque applied by the weight of the robot bending them in a vertical plane.
None of us get things right the first time – or the second, third, or fourth. After looking at a design for a long time, it’s hard to be objective about its strengths and weaknesses, so asking for advice from new sets of eyes is a good thing. If some of the things I’ve written don’t make sense, I apologize, but I’m in a bit of a rush this morning – I’ve got presentations to give this morning on drive design 
Assuming you used 6" wheels (which is what the OP has), did you do an analysis to see how much faster cross-field runs from a dead stop would be* if you had more than just the CIMple box’s 4.67:1 reduction?
- and less current draw, and quicker acceleration and easier control in tight spaces
**
I was rushed to finish this design, but will have a chance to talk with some mentors about it soon.
There are a lot of problems here – and that’s okay. We all learn by trying new things and making mistakes. In drive design, the devil is in the details a lot of the time and your modeling isn’t very detailed, so forgive me if I’ve made incorrect assumptions in the advice I offer.
1 – It looks like you’re using Pro/E. You have my sympathy.
That is correct. Although it is hard to learn, especially because I had to teach myself, I haven’t used any other programs so I cant really compare it to others.
2 – Do not use C-channel to build your frame or otherwise expect to add a VERY rigid (i.e. thick) piece of wood or aluminum plate to the inside to add rigidity. You’ll otherwise have a lot of flex in the frame – I know this from experience. The problems you’re anticipating with using rectangular or square tubing aren’t real.
We have built non-cantilevered, 4 wheel frames out of 4x1 c-channel inthe past, hence my tendency to use the same shape here. I am a little worried about the flex so we will probably add a couple pieces across the middle or use plywood for the electroncs board in the middle.
3 – The Toughbox Mini with long output shaft driving the center wheel is supported in two places in the Toughbox housing. That puts the load on that shaft at 1" (channel leg height) + spacer + sprocket + 1/2 wheel width away from the bearing. That’s probably 2.5" – too far. There’s all sorts of detail missing with respect to how you’re going to position the wheel on the shaft – is it retaining clips? spacers? Retaining clips weaken the shaft. Spacers will exert an axial load from the wheel onto whatever surface (presumably the channel) the spacer works off of. How will you handle that thrust loading?
I’m not sure why you say that is 2.5" too far. the total distance from the edge of the gearbox to the outside of the far sprocket is only 3.5 inches. The shaft is supported with one of the blue pieces of channel (with a bearing) about an inch away from the gearbox, so the total cantilevered distance is under 2.5". Did I explain that setup ok? I was planning on putting a spacer on the axle; the spacer would rub against a bearing in that blue piece. Would the thrust load cause a problem here?
4 – You’ve said the C-channel is 2x1 – presumably by 1/8" thick. That means the blue parts you show as your dead-axle support are 2.25 x 1.125 x .125". That size isn’t commercially available to my knowledge. How will you manufacture those parts?
You are mistaken. The blue pieces are cut from the same stock as the side pieces. They nest together with the legs of the blue piece on top of the legs of the frame piece. One of the pictures I posted kind of shows this.
5 – It doesn’t matter because they won’t work. You’ve got a hole in blue channel and a slot in the frame channel along side slots for mounting the blue channel. I presume those mounting slots are for chain tensioning. Think about what happens when you slide the blue channel away from the center when the chain is under tension. Your axles will not remain parallel to the center axle/toughbox output and you will throw chain and add a lot of drag to the drive.
You are correct, that is what the slots are for. Would there be any way to keep the axles from pivoting as you described? Would putting a washer on the axle between the nut and frame (on the non-wheel side) possibly minimize this?
6 – Is each Toughbox attached with only two screws? That’s going to exacerbate the flex in your frame. Follow the loading – weight of the robot passes through center wheels, bending toughbox output upward. Flex in output shaft tries to rotate the entire gearbox housing about its mounting point. This is hard to describe, so you have my apologies.
I think I understand what you are saying. I only attached the gearbox with 2 bolts because the gearbox’s mounting holes are too far apart to fit more of them in the frame. How do other teams manage this (other than using custom gearboxes)?
7 – The location of the sprockets doesn’t matter. All of the loading is passing through bearings / hubs in your wheels. The sprockets are floating around the shafts and your biggest concern isn’t torque applied by the chain bending your axles in a horizontal plane, but torque applied by the weight of the robot bending them in a vertical plane.
That is what I thought. The current placement of the chain was designed to try to minimize the cantilever distance with dead axles.
None of us get things right the first time – or the second, third, or fourth. After looking at a design for a long time, it’s hard to be objective about its strengths and weaknesses, so asking for advice from new sets of eyes is a good thing. If some of the things I’ve written don’t make sense, I apologize, but I’m in a bit of a rush this morning – I’ve got presentations to give this morning on drive design
Thank you your advice on all of this. I appreciate it.
Your Pro/E work looks good - considering you are self taught it’s quite good actually.
C-Channel frames can work - it depends on how they are loaded.
I believe the concern with the long shaft is regarding how far the wheel is cantilvered. In general you want the support bearing as close to the wheel as possible.
That wheel tensioner isn’t going to work very well. 2 tightened locknuts holding a tightened drivetrain chain in tension only by keeping the bolthead and nut in compression with the hub is not going to work for any reasonable amount of time. You really need some kind of cam device to keep your chains from falling off.
But if I eliminated the axle brackets, I wouldn’t be able to slide the axles to tension the chain. That was the idea behind using them as I did. Would box tube really make it much more rigid?
Do you mean a cam that pushes the chain down, or do you mean a cam that rotates to push the axle brackets farther out? Assuming you meant the second thing, then I was wondering why that keeps it tight. Couln’t the force just cause the cam to rotate an pull the brackets together as you described before? (I’m not saying it doesn’t work, just that I don’t understand why it does)
Can’t you slot both sides of the box tube like you did in one side of c-channel?
Yes. At 1/8" wall it’s less likely, but C-Channel will still collapse on the open side much much sooner than box tube. Think about just the profile of each of them and which would hold up better. Closed rectangle > open c.
You can do either. The principle is that of the inclined plane. The angle of the cam and the friction between it and whatever it moves are such that no force exerted back onto the cam will turn it, but you can still turn the cam. Thinking of just a block on an inclined plane, given a downward force, the static friction force <= mu*tan(theta)mg, so if mutan(theta) is greater than 1, then the frictional force will always be sufficient to keep the block in place, no matter what force you exert straight down. However, if you push sideways (up the ramp), you don’t get the same force of friction and can move the block farther up (or farther down) at will. The block moves when you want it to, and not at all otherwise.