I’ve become really interested in sheet metal designs over the years of watching FRC competitions and looking at the robots compete. This year I finally decided to start working on a CAD model for a swerve-compatible robot frame that is lighter, takes up less space (or takes up different space so we have more room inside), and more aesthetically pleasing than our current (and past) frames made of Item (similar to 80/20). At the moment we have no engineers who know much about sheet metal, but we do have CNC mills (I know they take a while) and the proper equipment to bend large quantities of metal. What I need help with is how to actually design a frame and/or components and how to attach the parts together. So far all I have been able to come up with in terms of attaching parts is this: https://docs.google.com/file/d/0B9fwpOaYSNDnaWxXYmIySldVRjQ/edit?usp=sharing
(Sorry for URL instead of direct image. Still have to figure that out.)
Attachment methods vary based on what to parts you are trying to join. The trick is to use geometry to give you strength. Imagine you are making the parts out of paper (or even mock them up in paper), and go through the thought exercise of how they will hold up to different forces.
On the picture you have attached, I would at least overlap the two top flanges and add rivets there so that you have more than just the bend resistance of the metal keeping the parts at right angles to each other. Based on the strength needed, you might add more reinforcing features, but overlapping the flanges is a cheap and easy thing to do.
JT, We just started using sheet metal last season. We have a laser cutter sponsor so it’s easier for us than it will be with you and the mill (remember to fillet your corners to the size of your end mills).
Here is the presentation I put together for the team last season when we decided to start doing this. It’s definitely not complete and I was learning at the same time so I don’t make any promises that is all correct but it will get you started.
Also go back through and read the threads on other people’s designs. Some really smart people, give really helpful advice. Here are a few of the recent threads. 1114 Inspired Chasssis, 488 Fall Project 1 and 2, Sheet Metal Drivetrain Prototype and there are many more like it.
JT, we’ve been doing sheet metal frames for 3 years now. Would be happy to send you those CAD models (Inventor) if you wanted them, along with info on assembly.
The function of the part you are needing to attach is the driving force of the attachment method in the design. Let’s list some methods to join sheet metal.
Permanent joins
Spot welding
Mig
Tig
Braze
Epoxy
Semi permanent
Pop rivet
Solid rivet
Tog-L-Loc
High bond adhesive tape
Temporary/Removable Joins
nut and screw
Screw and tap/thread
pem nuts
Pem studs
self tapping screws
drilling screws
Plastic connector
1/4 turn fasteners
Riv-nut
Inter locking joins to position or fixture and increase strength, rigidity
mixture of tab and notch and hardware
dowel pins and holes
interlocking puzzle piece joint
Dual Lock
What do you think of our list? Did I miss anything?
Thanks for all the replies! Sorry I have not responded yet, I have a habit of forgetting to post… so I will reply to all the posts in one post.
Travis Schuh: Thanks for the tip, I was actually thinking about doing just that, but I’d have to see how I could get a CNC mill to mill those tight corners.
AllenGregoryIV: Yes! I have seen your robot, the sheet metal work looks fantastic! I really admired your robot last year. Thanks for all the great links, I will look into them when I don’t have homework piled up… (tomorrow).
Sanddrag: Post taken into consideration, thanks for clarifying.
jspatz1: PM sent.
Akash Rastogi: I will send you an annoying amount of Facebook messages if the need arises, thanks for the offer.
Roystur44: I think we would most likely be using 3/16" rivets and screws. Your addition of dual lock at the end of the list intrigues me, I will look it up.
Thanks for all the great replies! I will work more on the CAD model tomorrow and post an improved corner!
It seems like your resources are multiple CNC mills + hand brakes to bend metal. Maybe it’s just me, but it sounds like those are resources better suited to tube and plate based construction with bent sheet metal for specific parts rather than a full sheet metal robot. In a lot of cases you can make good, rigid, light structure in less time and about the same weight going that route. If you have a turret punch / laser cutter sponsor and precision bending equipment, sheet metal does get more attractive.
That bend will be hard (impossible maybe?) to get.
Rather, make the inside dimension of one rail match the outside dimension of the other rail.
Also, it looks like the two flanges on the left rail coming to a sharp corner with each other is a possible issue. Someone more knowledgeable in sheet working can clarify.
Let me start with I design sheet metal parts on a daily basis. While the top bend that would connect the two flanges is rather difficult, the guys in our shop would say its doable. The caveat to that is I work with extremely skilled artisans that do this for a living. That being said I would never ask them to make this.
That “S” bend in the part is what we would call a joggle. There are special tools and techniques that make this easier. Keeping tolerances can be difficult and may take time to get right.
A more simple solution would be to rivet/bolt a gusset plate across the top two surfaces.
Rather, make the inside dimension of one rail match the outside dimension of the other rail.
This will also work, and most likely be stronger than the above mentioned method. Just be sure to leave room for tangent lines when designing the length of the inside piece. I wouldn’t design to the exact distances given by CAD. Give yourself some room.
If you decide to use this method or Adam’s method, be sure to give yourself more than a single attachment point between the two pieces. I would use 3 points or so.
Also, it looks like the two flanges on the left rail coming to a sharp corner with each other is a possible issue. Someone more knowledgeable in sheet working can clarify.
Again, this is doable but not preferred. In order to do this you would need dies (or fingers) of the exact length of the flange. This could require cutting tooling. It is also a bad idea from a stress standpoint. A sharp corner (the tangent point of two bends) should be relieved. Most software packages have this as an option in the sheet metal tools. I would put about a 1/16" or so relief radius in this part. It generally also applies to bends that are in the middle of the parts (eg the flange doesn’t run the full length of the part).
I would also recommend putting a radius or chamfer of decent size on each corner.
Do you mind if I use this presentation with some modifications to help with training our members? (I will give you guys credit at the end of the presentation)
Our team uses aluminum 1" box tubing for our frame, which we temporarily attach somehow (I dont remember?) before having our sponsor (one of their employee’s who does nothing but weld aluminum all day) weld it all up nice. This year’s frame was far too complex, but it came in at just 13lbs and is stupidly strong (could probably withstand 1,000s of lbs of force in any direction). It also is easy to work with, has nearly infinite practical durability and is very rugged. We subsidize this tubing with sheetmetal for more robot subsystems (shooter, drivetrain assemblies, ramp lowerers, innertube grabbers).
We typically use 1/8" alum (powder-coated of course) for heavy systems like drive train and climbing brackets, and 1/16"-.0404" for shooters and such. We used .0202" (basically IS paper) in 2011 on our gripper because if it got banged up, it was super easy to hand-bend back into shape.
