What metal should we use?

We are putting in metal orders for the next season and we are wondering what metal we should buy for making a chassis. And for general use.
We are thinking of using 6061-T6 aluminum 2x1 but we are unsure if we should get a 1/16” or 1/8” thickness.
We also heard that Vex’s aluminum 2x1 is not actually 6061, is this true?

I would recommend getting both 1/16" and 1/8". There are places where the thinner more lightweight 2x1will be better, and other places (like the drietrain) where the thicker 2x1 will be more useful.

According to their website the VersaFrame Tube stock is 6061 T6.

Edit: I just cant spell…

3 Likes

Vex’s 2x1 is 6061. Some other suppliers might sell you 6063, which is weaker and more annoying to machine (but potentially still usuable depending on application).

1/8 thickness is nice if you are directly supporting bearings inside the tube, as that extra wall thickness helps prevent some exotic failure modes. The downside is 1/8 wall is heavy. 1/16 wall is fine as long as you properly handle the thin walls - for example if you use an external bearing block like the VersaFrame clamping block, 1/16th can work. 1/16" wall is generally stronger and lighter than cutting weight reducing pockets into 1/8 wall, so that’s something to note.

The Vex 2x1 is offered in 0.1" wall which is a really solid compromise between the two thicknesses - this might be your best bet.

7 Likes

A couple of things about aluminum tube stock:

  1. know that Vex’s versaframe stock is not actually 1/16" (0.065) and 1/8"(0.125) wall tube. It’s actually 0.05 and 0.10 wall, which may be slightly lighter but is also thinner and thus not quite a strong as what you’ll get in comparable tube elsewhere.
  2. it’s got all those pre-drilled holes in it. When you’re building a custom frame, you need the holes where they need to be, not necessarily on the 1" centers that Vex puts them at.

Honestly, if you’re really going to build a custom frame (and we have been for a couple of years) you’re better off getting good 1x2 stock from a reliable supplier (Online Metals, Metals Supermarket, etc.) and doing all the cutting, shaping, pocketing, and drilling yourself. We use 0.125 tube for all serious load-bearing applications (main frame, lifting mechanisms that support the robot’s weight) and 0.065 for everything else, so you should definitely get a stock of both in 6061-T6. It machines well and is good for welding too (if you have the capacity for that or can get it done.)

2 Likes

There is a very very long argument/research paper in the making about this somewhere on the forum that concluded with “don’t worry about it too much”. It’s, like, probably fine even if some of the numbers look closer to 6063 than normal.

But yes, do not buy material explicitly labeled as 6063. “Architectural aluminum” is a derogatory term in robot building. Related are “hardware store aluminum” and “chewing gum with metal shavings in it”

7 Likes

I’m envisioning oriented strand board made with milling machine chips instead of wood.

6 Likes

I’m pretty sure Vex’s bent sheet metal is 5052 not 6061.

For FRC structural uses this really isn’t an issue. The T6 age after bending really isn’t worth the additional cost.

1 Like

Note that welding will knock “T6” down to about “T4”:

6 Likes

If you have budget, I’d recommend also having some 1x1 in 1/8th and 1/16th wall thicknesses, and also a modest amount of 1x1 angle (probably in 1/8th) and some 1x2 channel. These end up being helpful at times.

2 Likes

I recommend 6061 Alumiunum for pretty much anything in FRC. However for deciding between 1/8" and 1/16" I would look to your drive train. What we did was with tank drive we decided to stick to 1/8" because it would taking the most impacts from other robots but when we switched to swerve we decided to go with 1/16" because its lighter and swerve modules are heavy.

True, but of course it does recover much of that strength relatively quickly. There’s a reason that 6061-T6 is used so often for applications that are welded (bike frames, for instance) since it is still quite strong as long as you don’t try to use it immediately. We’ve welded our frames for a couple of years (we have a good MIG setup and I teach proper welding technique since I used to do it for work) and our frames have stood up to some really heavy action and hard falls (including from the traverse bar this year) without so much as tiny deviation from true. I know a lot of team are leery of welding because of the skills necessary and the expense of the setup, but it’s a valid technique if you know what you’re doing.

1 Like

Thick + windowed heavily is usually your best bet, even if you can’t do something on a cnc just have someone go at it with a hole saw.

1 Like

@JamesCH95, don’t you have the FEA to confirm the opposite?

6 Likes

We spent a whole weekend with two hole saws to cut 3 pounds out of our frame. Deburring all the holes took at least as long as cutting the holes. This was necessary for functionality and to pass inspection.

I do, and it does.

https://imgur.com/a/JgE47hY

[SPEAKING VERY GENERALLY] 1/16in walled tubing experiences 30%-60% less stress than 1/8in walled tubing that has been modified to the same linear density (same weight per length).

Often 1/8in tubing is often not required for drivetrain usage EXCEPT that our ubiquitous hex bearings have an undercut under their flange that dramatically reduces the amount of material engagement with the tube wall. More annoyingly these undercuts can vary widely lot-to-lot. .02 to .05ish is not unheard of. This is one reason why some teams mount bearings into bearing blocks and not directly into the tube wall.

Groove can be seen in the drawing, but it is undimensioned and thus generally uncontrolled.

image

6 Likes

Yeah, thin walls are definitely the preferred option when the loading is distributed evenly across the whole cross section, so I would recommend the thin walls as much as possible. But in addition to the bearing undercut, you have to be care with load concentration at bearings and fasteners, the thick walls are much stronger locally at these points. Also for anyone out there not used to thin walls, do not both through a thin wall tube without putting a spacer into the tube to keep your bolts from crushing the tube. The force required for a decent hold on your tube will crush your tube, and if your don’t tighten the bolts, your tube will move around and enbiggen your hole. You can bolt through a single wall of the tube, and if loading is a concern use washers or bolts with wide heads to distribute the load across more of the tube wall.

0.095 wall is a great compromise but not easy to find in 6061 T6, particularly in Ontario. I just ordered some from a very reputable manufacturer. The minimum order was 1000 pounds. If you’re in Ontario and want some, we’re reselling it. 12 ft lengths and pickup in Oakville only.

3 Likes

I am compelled to respond to most of the points in your post.

Agreed.

The 2x thicker material has 2x the bearing area/bearing strength. The bearing undercut makes the 1/8 material appear 3x-5x as strong vs 1/16 material. Properly-installed bolts and rivets will be transferring load through the friction of their clamping force and not ‘see’ any difference in wall thickness.

Yes. 10-32 bolts have a proof load (the tightest viable installation torque) of 660-2,380lbf depending on bolt grade. Any FRC-typical tube will be crushed or otherwise be unable to properly support one of these fasteners. Rivets are strongly recommended for fastening to tubing because they will only engage one wall of the tube and thus can develop their full clamping load without need of a spacer.

As a generality washers have a 45° load cone, that is to say a washer can transmit clamping load at a 45° angle from the edge of a bolt or nut to the material being clamped. So a washer .05in thick can made the ring of clamped material have an outer radius .05in wider. It helps, but a washer generally does not distribute clamping load to its own OD.

Fun picture:

Reference articles on this last point.

Related to @George_C 's post, we use sheet metal for essentially all of our components in part because we can pick exactly the wall thickness we want. We often use 11awg awluminum, .09in thick, in 5052H32 for drivetrain rails.

2 Likes

Wow, thanks for all your references and information. We actually sometimes counter sink standoffs and plates so that when they are bolted to a tube or sheet metal they contact on a wider footprint. We use 1/4-20s with .669” diameter heads so we can spread the load.

In an attempt to be a little more clear about what I meant when I said stronger locally, I was talking about a single fastener in the middle of a wall, on the thin wall tubing, that wall is more susceptible to planar distortions and eventually fatigue within 1/4-1/2” of the fastener. In that area it is better to think of the structure as a plate in flexure rather than a tube. Of course using many small fasteners can mitigate these effects, but in my experience, when you have to use a small number of high load attachment points the thick wall can be the superior option because you can have more material at those attachments and massively lightweight a little farther away where the loading is distributed more evenly.

Also I have drilled out too many rivets to use them except when I’m completely sure we aren’t going to have to make changes, instead we use short bolts through only one wall. Sometimes we have to drill access holes on the opposite wall to get the bolts in, but that’s not too bad.

2 Likes

We use aluminum rivets with aluminum tails. They drill out faster than bolts can be unscrewed, especially with a rivet drilling tool that stops the flange from spinning. Even faster than that, aluminum rivets can be snipped off with flush-cuts.