pic: How It's Made: 148 & 217 Robots

[artdutra04]

Originally Posted by Akash Rastogi

John- would it be possible for you to put on a presentation in Atlanta this year, or even a powerpoint presentation up here on CD, detailing your design process in Sheetmetal in Solidworks? This question is also directed to others who use sheet metal fabrication such as 217, 228, 1771, 1477, 1902 and any others I don't know of.

I've been learning the features in SW but I don't know what works or what doesn't in my designs when, hypothetically, I would try to get them manufactured at a fabrication site. Are there limitations using sheet metal during fabrication?

There are a few general rules that "should" be followed: all flanges should be at least 4x the sheet thickness long; all holes should be at least 2x the material thickness away from the edge of bend; if being made on a turret punch sans laser, the minimum hole size is the material thickness; and the material bend radius is generally equal to the material thickness (there are tables of standard bend radii for each thickness of commonly used sheet metal, but consult with your local shop).

Originally Posted by Akash Rastogi

What are the advantages? (other than speed of fabrication, weight, duplicates)

Other than what you listed, I'd also add their ability to adapt to really interesting geometry.

Since this is turning into a general sheet metal design thread, and since I know I'm not the only one who learns by seeing photos of things, here's a bunch of photos of various features of how 228 has used sheet metal on our last few robots:

The following sheet metal part was the first one that I ever designed (back in my senior year of high school). It allowed us to have a pivoting arm that bolted to an IFI aluminum sprocket, which enabled us to do drive by scoring in autonomous, while retaining the ability to pick up tubes from the floor. This would have been pretty much impossible with anything but sheet metal construction.



Here's a photo of it mounted on the robot: http://team228.org/media/pictures/view/3000


Another interesting feature is the ability to have square holes. While it's not the best thing on earth for stress concentrations (it's actually pretty much the worst), it can allow you to run square tubing right through the middle of parts. On 228, we only use CNC sheet metal parts where we need critical tolerances; otherwise everything else is made with mostly 1"x1"x1/16" box tubing, 1/16" polycarbonate sheet, or various sizes of aluminum angle. So interfacing with box tubing is a great advantage for us that sheet metal offers.

Here's how we used it on our 2009 robot. The 1/16" box tubing ran through the 1"x1" hole and face mounted to the inner chassis rail with bolts. The material around the outside of the box tubing acted as a corner gusset for the rear chassis rail.



Here's a photo with the box tubing super structure attached: http://team228.org/media/pictures/view/4838


Another great feature that we use a lot (as visible in the above photo), is we have .201 (#10 clearance) holes on .75 spacing along many of the important flanges on the robot. While we do model pretty much everything mechanically important on our robots before we fabricate any parts, there are occasions where we want to move something, modify a part, or add new parts onto the robot (last year, it was traction control omni wheel assemblies). Having all these preset mounting holes makes it very easy to do without without needing a drill.


Tolerances aren't an issue as long as you make it clear to the machine shop in your part drawings the critical features and the necessary tolerances. Like Paul (and now John), we use bent sheet metal gearboxes integrated into our chassis (via direct drive, live axle drive trains). Last year I wasn't brave enough to use rivets for critical features, but I'm open to them in the future (if we buy a pneumatic riveter, and as long as the GDC doesn't have inane FRAME PERIMETER rules that make rivet heads illegal). All center-to-center gear distances have .003 added. Here's a photo from our 2009 robot.