pic: GUS Team 228's Prototype Sheet Metal 6WD Chassis

As one of our 2011 preseason activities, we updated an older prototype 6WD sheet metal chassis we had designed back in 2008, since both our 2009 and 2010 robots were 4WD.

The main chassis rails are 1/8" 5052 powder-coated aluminum, with the middle lateral supports being 3/32" 5052. Unlike the past two years where our sheet metal parts were mostly limited to our drive train, we plan on using a lot more sheet metal parts (particularly 3/32" and 1/16") on our superstructure and manipulators for the 2011 season.

The pneumatic shifting gearboxes use a variety of gears a la carte from AndyMark, and are designed for a low-gear speed of 5 ft/sec and a high-gear speed of 13 ft/sec. The center wheel is direct-driven, and the outer wheels are powered via 25p roller chain. The wheels are 4" x 1.5" Colson wheels with knurled 6061 aluminum inserts pressed into them.

With everything shown (including all hardware and simulated roller chain set to the correct weight), SolidWorks says this drivetrain weighs in at 36.76 lbs.

Probably a silly question, but what’s changed since the last design and why did you make the changes?

We went from single-speed to dog-shifting transmissions, from 5"x2" to 4"x1.5" Colson wheels, from 35p to 25p roller chain, from keyed shafts to hex shafts, changed hole pattern from #10 clearance on 3/4" spacing to alternating #10 clearance and 5/32" (for rivets) on 1/2" spacing, and overall dropped a lot of weight.

Also, the standoffs were modified to have a Delrin sleeve over them. As shown, it was designed for an exact number of even number of 25p roller chain links, so as to eliminate need for a master link. In this state the chain does not contact the standoffs, but as the chain stretches these standoffs can be moved to act as tensioner. Another benefit these standoffs have is it allows us to nest pneumatic accumulators into this (otherwise wasted) space to help keep COG as low as possible:

However, depending on whatever curve ball the game throws us, we may opt out of using these standoffs in lieu of a single sheet metal belly pan and/or sheet metal superstructure parts. If we had access to a live-tooling lathe, I’d combine the sprocket and knurled insert hub for the outer wheels into a single piece, to allow the end shafts to be dead axles.

I drew up a model similar to your 2008 model. One of the questions I had in doing that are what are the pros and cons of putting the chains on opposite sides of the driven center wheel versus having both on the same side? I would think having them on opposite sides of the center wheel would be easier, but then is there a mechanical reason for having them on the same side.

Looking at this model, why did you go with squares instead of triangles for lightening?

With everything shown (including all hardware and simulated roller chain set to the correct weight), SolidWorks says this drivetrain weighs in at 36.76 lbs What was the total weight of your robot, with the upper mechanism ?

Putting the sprockets on the same side (inside) adds almost an inch to the lateral wheelbase of the robot. This slightly improves turning by moving the angle of the wheel relative to the center of turning closer to perpendicular, which allows for a greater percentage of the wheel’s tractive force to be used in the desired direction rather than “wasted”.

Our sheet metal shop sponsor that fabricates our drive base uses turret-punch machines for their work. Slotted pockets are easy to punch out (punch two or four holes, then use straight punch to connect them). Triangular pockets require a lot of nibbling to make, which slows down the fabrication time. Since their time and work is donated, I like to keep the production time to a minimum.


There is no upper mechanism on this model, only a drivetrain and base chassis.

When sheet metal shop mentors donate their time, there are some simple steps you can take in SolidWorks to help reduce the time to form/stamp the part.

Ask what the standard tools/punches are in their turrets. This way you can design around their standard setup -you can add these to your SolidWorks Design Library under Sheet Metal folder. Then you can drag and drop these shapes from the library. There are standard sheetmetal library features there now to help.

Ask what bend radius to use. By defaut, SolidWorks uses a K factor = .5, this is the neutral axis, but in reality, the K factor might be .44 on one machine and .47 on another.

Ask about materials - what does the sheet metal shop have as scrap ahead of time. Use SolidWorks SustainabilityXpress to compare different material properties such as density, tensile strength, young modules in one big table - plus you get to see the resulting carbon footprint.

Turn on Sheet Metal tab, right click any tab in the command manager. The first feature in a sheet metal part is a Base Flange. Work in 3D, then make the Flatten state in the drawing. If you plot the drawing in the flatten state, try to size, trace it on paper, you will get an idea of what your part will look like. Cut out and score the paper at the dashed bend lines.

This tip comes from John V Neun, if you dont have access to a sheet metal shop, you can print the flatten state of the part to scale, trace with a sharpie on your sheet metal and cut out with the proper tin snips- wear safety glasses.

At 2010 you can create multi body sheet metal parts, the cut list will create a seperate representation for each body. You can also now use Mirror part in an assemby for sheet metal parts and my favorite new feature, based on the sheet metal vendor I used to work with, you can export a sheet metal body directly to .dxf or ,dwg format. From the part, right-click on the Flat Pattern feature and click Export to dwg/dxf

I used sheet metal a great deal when I was designing robots - I learned alot by visiting the sheet metal shop often and working with the sheet metal vendor. Marie

How do you tell if your designs are strong enough? Do you do FEA? Something else?

If it’s a “unique” part we do calculations and/or FEA on the part. But for the majority of parts, we use past experience and intuition, as there isn’t enough time in the FRC build season to fully analyze every part. By managing flange locations carefully (flanges strain harden the material) and properly cross-bracing sheet metal parts, you can create strong structural members with relatively thin materials.

The most important thing when working with sheet metal is that many designs may only work as a complete system. Whereas many non-sheet metal drive bases work just fine alone, many sheet metal designs would fail without the cross bracing of all the other mechanisms and structure. For example, here is the chassis for our 2010 robot:


Alone, it would never have survived a single practice round. But with all the other parts and mechanisms (swerve pods, kickers, electronics belly pan, superstructure, etc), our drivetrain/chassis was more than strong enough to survive pretty hardcore New England defense unscathed with a Regional banner. (The bearings in our swerve pods and our kickers were a different story. Live and learn. :o)

Our main sheet metal machine shop sponsor is a SolidWorks shop, and has a plugin to take the SolidWorks files, adjust the K-factors to match their machines, flatten them out and nest as many as possible in a sheet of metal. Our team took a tour of their facilities before the 2007 FRC Kickoff, and it was what helped make me a huge fan of working with sheet metal.

I love sheet metal too. Your trusses are beautiful. Marie

That pretty hardcore New England defense was sure tough in Atlanta. :stuck_out_tongue:

I think that sheet metal’s biggest “flaw” for a lower budget team is how it lacks rigidity until you build the whole robot; teams without practice robots can really only test and tweak once everything’s together, rather than system-by-system.

However this plays straight into one of my favorite strengths of sheet metal - it’s really easy to get duplicate parts, so you can build a practice robot to make up for all that lost debug time!

I’ve been a fan of your designs for quite a while. They are always clean, organized, and very well thought out. Keep up the good work!

Some questions about this chassis in particular:

How much ground clearance is there?

If the game calls for an open front, do you have a plan for joining the frame members mid way?

Would it be possible to hex broach the Colson wheels directly instead of machining knurled inserts?

For the mounting hole pattern on the frame, I’m wondering if you could make them all 1/4" holes. This would allow you to choose either 1/4-20 fasteners, or #10-24, since 1/4" is the drill size required for #10-24 PEM nuts. So if you have a spot where you want to use # 10 screws, just put in a PEM nut.

I like the way you have designed in a series of holes to allow repositioning of the standoffs to serve double duty as chain tensioners. This is much more reliable than using slots to reposition a tensioner. Do you know how much excess chain length would be taken up for each position of the standoff?


Rob- I’ll let Art handle the rest of your questions, but I think I can handle this one, as I’m a huge fan of Colson wheels as well.

Colson wheels typically have large inner diameters as you can see in this above image. The hub is a thermoplastic of some sort. My experience with Colson wheels has been nothing short of spectacular…AFTER you get them mounted to a robot. As far as wheels go, they can be a bit tricky to design with. I like Art’s knurled insert design because it simplifies the hub design greatly.

In the past I’ve made hubs that fit into the inner diameter and then have grooves cutout that grab the ribs around the hub of the wheel. They worked well, but are a relatively complex part for just being a wheel hub.

So to answer your question, hex broaching these wheels would be quite difficult knowing that the diameter of the inner hole on the wheel is already quite large. Also, the plastic hub makes for a less than ideal hex broach solution because of the material properties of plastics (shaft will strip out inside hub).


Thanks for the info on Colson wheels. I got a sample from Mcmaster a while ago, and they do seem to be nice wheels. Obviously they were never intended to be powered, which is why they are difficult to mount hubs to. The one I got has a 1/2" bore, but this is with plastic bushings. The hub itself is 7/8", which could be broached, but now that you mention it, the plastic is a bit soft, and may not hold up well under stress. I was looking for an alternative to knurling because while we do have the tools, it seems like it would be a bit time consuming to get it right on a manual lathe.
I guess with Colsons the initial effort to mount them is offset by the benefit of never having to change treads.

I was wondering if you could post a picture of the hub you are talking about?

Rob you are right on in regard to the effort to mount them. They really are NICE wheels though, and can be worth the effort. If you have the resources to make hubs, you could make something similar to this…http://sphotos.ak.fbcdn.net/hphotos-ak-ash1/hs760.ash1/165126_825763481219_1814541_45818923_1743508_n.jpg http://sphotos.ak.fbcdn.net/hphotos-ak-ash1/hs745.ash1/163748_825763366449_1814541_45818922_512013_n.jpg

This is just a very quick CAD I made from memory of what I did to mount Colsons on the 05 and 06 robots I helped with. The inner hub is converted into that round boss sticking off the hub where you could press bearings into it. Then, the notches around the outside grab a hold of the ribs you see on the Colson wheel. The hole pattern can then be drilled through the wheel so you can compress sprocket, hub, and wheel with the same bolt pattern to make entire wheel assemblies. Just an idea to get those other ideas flowing…


Can someone provide a part no. from McMaster for these wheels? I know they’re available elsewhere, but McMaster seems like it might be a more reliable source.

I haven’t been able to find any good information regarding bore diameter and overall hub diameter, but if the hub is large enough to accept the pseudo-standard 1.875" six hole circle, these should be pretty easy to mount.

Searching “rubber wheels” in mcmaster brings up the page

Colson Performa (the wheels in this thread) are on the lower half of the page titled “Performance rubber wheels”

We used the 3"x7/8ths wide ones in the 4 corners of our 6wd swerve this last year, live axle 3/8ths hex hub, and 3 bolts going through the wheel and hub.

The hub was a simple 1.25" diameter .1" ish wide flange, 3 bolt pattern with a smaller diameter that went through the wheel hub.

Not having to worry about retreading is nice.

Arthur, It’s great to see more awesome sheet metal musings from your team. You guys really nail the simplicity and efficiency aspect of your designs every year and have become one of my favorite teams to watch during the build season.

This looks like the most robust, clean, well thought out 6wd chassis I’ve seen posted up here. It’s one of those designs that makes me think “I wish I had one!”