Designing sheet metal drivetrain…
Which configuration of drive rails is best and why?
I would guess that A is the most stiff but I’ve definitely seen pretty much every variation.
https://image.ibb.co/mxLAUw/Drive_Rails.png](https://ibb.co/gSGTGb)
2386 ran B for a couple of years and have recently switched over to A. B can work well but assembly could be a little tough when putting parts in the middle of the side section. A works really well because you have smooth edges inside where you’re putting a lot of parts into. It also allowed us to do a chain run on the outside wall for 2017, allowing us to save a lot of space. Option A is by far our favourite and I’m sure we’ll be using it in the coming season.
The three years my teams ran sheet metal drivetrains, we used “D”, with belts in between the rails and wheels cantilevered to the outside. The latter two years, we bridged the lower connection with our bellypan, and the top connection with a strip piece with additional mounting features, forming a closed box structure. It was absurdly rigid in .09" metal with heavy lightening features, we likely could’ve gone thinner. Having no interior flange makes mounting features to the inner edge of the chassis rail like gearboxes or diagonal superstructure members very easy as well.
We were very happy with this approach and would use it again, but one nice thing about some of the other options is the empty space below the flange. On our design, there is a belt directly below the flange in most places, which means we must be careful about fastener length and location, or add additional mounting features to the rails. “A” is nice if you aren’t using cantilevered wheels, as it provides them with the closest support possible, and you don’t have to worry about any components hiding under a flange within the rail.
I am also very partial to the robowranglers “Big C, Small C” rail design originally from their X009 prototype, which I believe they then used on both their 2015 and 2016 designs.
I personally prefer A (If you want, I can dig up some pictures of 1086’s frames if you can’t find any). Since you don’t have to worry about axles or wheels getting caught on flanges when you are installing it. If you want to have some extra fun, you can make the wheels live axles and then run the sprockets within the flanges so the chain can be easily removed while also being protected by the flanges above and below it. Plus, you can use the flange out for a bumper mount.
Moved to A with dead axles in 2016 & 17 and it’s been great. Being able to drop wheel modules out just by pulling one bolt has been awesome. Bearing and bolts, etc don’t get in the way of your bumpers so you never need to pocket the bumpers.
The one time we have run sheet metal we ran setup A. Setup A allows you to use 1114 style axle supports in addition to giving you easier access to putting in hardware or rivets on the rails. Other setups do not fully allow that.
We haven’t done sheet metal rails for awhile but we used to do D. We however will never use D again because it made assembly far more complicated than it needed to be. If we ever use sheet metal again we would defiantly go with a dead axle method in style A
If you are looking at going with a custom sheet metal drivetrain of the Option A style, I’d recommend a slight modification: flip the bend direction of the top flange on the inner rails, so that these are a Z bend instead of a C bend.
— — — —
| | | |
— — — —
The advantage of this over Option A is that it still allows full bottom (drop) access to a dead axle drivetrain, but now allows you to face mount extrusion (such as 1x1 or 1x2) to the inner rails for building a super structure.
We used D for a while. We hated it. A large part of that was that it was terrible for maintenance and construction, ultimately costing us the quarterfinals at OCR last year. With regards to the actual construction, D was quite annoying with the bumpers, so I would suggest that you use A.
The trick with D is to make the outer rail as easy to remove/pop out as possible. Mount everything you can to the inner rail or the top structure I mentioned, and make the outer rail pop off with a couple bolts. That way, you can replace internal parts from the side, rather than from the top and having to worm your hands around the flanges. As I mentioned, it is also more appropriate for cantilevering if you choose to go that route, which leaves less to do between the rails. If you are putting everything between the rails, then yeah there are better options.
In three years of competition we only needed to open up a rail once, and were capable of a complete drive side teardown+rebuild in ~10m.
The outermost C flipping outward is very nice for defining a bumper perimeter without protruding bearings, axles, etc.
The innermost C flipping outward lets you mount superstructure outside of the drive rail and also lets you avoid breaking the flange for gearbox clearance which is a big help.
They are all basically the same or similar amounts of stiff; if you aren’t making a full tube profile it’s all going to be comparable.
So, A or C, depending on preference. That said, there’s so much more you can do with sheet metal drives if you have the expertise (custom “tubing” profiles, structural bumpers, etc) that the kind of basic C channel sheet frame is underutilitizing the resource a bit, and not necessarily better than the Kitbot.
This is kinda worthless advice without examples 
I would recommend channel in the S and C channel shapes Art recommended are preferred.
I think posting up what you plan to do with the front and back rail is just as important here. In the past I have run a front and back rail that incorporates a bellypan into one single piece. This is obviously challenging depending on the game but gave us an extremely rigid frame.
Definitely check out any of 971’s frames on Chief Delphi or on their website for some really great examples of sheet construction.
Just chiming in here - you don’t have to use sheet metal to have this discussion. Last pre-season we experimented with making a chassis from 3x1 aluminum channel. That size fits really nice into AM14U end plates. It might be easier for some teams to work with extrusion rather than sheet (for instance, if you have a mill or router, but not a brake).
We’ve been using one like this lately. Seems to work ok, but does have some issues with being able to assemble it easily.
Thanks everybody for the great replies…
I’m definitely liking the A and C styles more now (with the S). We ran parallel plates last year and I thought “two-screw removal” was super nifty - that was until we had to get the other end of the dead axle out for each wheels. We had to remove the outer plate maybe once or twice a competition.
Looking through 148’s work and some of the replies above, I see some lean towards sandwiched C design. Has anyone ever tried a design like where the wheels are embedded in the sandwich? This ideally would make slim 2015-esque drive trains possible with just sheet.
https://i.imgur.com/Am5Ld86l.png](https://i.imgur.com/Am5Ld86.png)
If I understand you correctly, yes. Brendan’s picture above (post 14), and the AM14U series are among the many cases where wheels are between the sheets. If you do this, accessing the wheels becomes more of a challenge than with cantilevered axles and wheels on the outside, and also reduces the track width relative to the chassis width.
With dead axles, the answer may be to pull the axle, then drop (or lift) the wheel out; this is how we changed some front and rear wheels on our AM14U2-based robot in 2016. Removing a wheel between the plates on a driven axle often requires removing the outer plate or another similarly involved process.
Not quite what I was asking… We did parallel plate last year and ran into many of the same issues you describe. Here is a close-up on the front of the frame I posted above.
The specific size of your outer “C” is a pretty tough shape to make with a press brake. It can be done with a large offset punch but not all sheet metal shops are equipped to handle that kind of geometry.
You might be able to do it on a manual (finger) brake but you run into other challenges there. It’s going to be pretty difficult to get the “C” to be the right dimensions so that the top and bottom legs are actually parallel.
What are the fasteners on the two (dead?) axles on the right side of the picture?
Thanks.