Alternatives to Welding?

[randomunchies]

My team is seriously considering making the change from 8020 to tubing for next season. However, we don't have the resources to weld, and some of our mentors don't feel too great that once we weld something, we won't be able to change it.

I've read around that some teams have made tubing robots without welding, so my main question is:

What are the main methods you guys use to join tubing without welding?

-Kat

[Tom Bottiglieri]

We use gusset plates (typically 1/16" alum) and rivets to hold our tubes together. We cut ours on a waterjet, but you could probably make them on a mill if needed.

[randomunchies]

Is there a better way to use the gussets other than riveting? Would it be alright if we threaded the holes and used screws?

Thanks Tom!

-Kat

[Retired Starman]

Threading works fine, provided the tube wall thickness is great enough to get at least 3-4 full threads in. Four threads in 1/8 in. thick aluminum means 32 threads per inch, so a 6-32, 8-32 or 10-32 machine screw should work. Gussets are usually attached with a bunch of fasteners, and all that tapping will not be fun. Installing rivets just requires drilling matching holes in gussets and tubes, then installing the rivets.

Many teams, however, use 1/16 in. wall thickness, in which case you don't have enough metal to do the required threads. At this point, pop rivets work best.

If you need to make a change, parts can easily be disassembled by drilling out the rivets. Takes very little time or effort, and you can reassemble them with more pop rivets.

The teams I have worked with love using rivets. Also, they really save on weight over conventional bolting techniques.

Dr. Bob

[MagiChau]

If I recall correctly, our frame is actually bolted/screwed together. I would need to ask one of the frame CAD designers for more information.

[Wayne TenBrink]

If the material is too thin for threads, or if the fastener needs to be removed repeatedly and the space doesn't allow room for a nut and a wrench, you can use a floating nut plate, which can be riveted in place.

Another "nutplate" option is to drill one or two holes into an 8020 sliding nut and rivet that in place. Its kind of a pain in the neck to do, but it works well.

[Hawiian Cadder]

we looked at many solutions for our tube and gusset frame before deciding on rivets. we used 1 by 1 by 1/8 tube with rivets and loved it. we considered bolts, but decided against them because they weighed much more, and also because we wanted the frame to stay absolutely perfectly square once it was put together. large holes tend to have more slop than small holes (at least in our shop) we decided against screws because we didn't want to tap. our 2010 robot had 700 machine screws, and that was something we didn't want to repeat. pop rivets were suggested by one of our mentors and they were excellent. they are easy to drill out if you want to change something, and they are very strong. i think our frame weighed 28lbs. (heavy, but part of the lift support was included and as our first year with tubes and gussets it was way way tougher than needed) i would recommend rivets for sure.

[sanddrag]

You could also use rivet nuts

[IndySam]

Rivets and gussets are easy to do and surprisingly strong. We have also machined framed connectors and epoxied a frame together.

[jwfoss]

Since you already have a good understanding of 80/20s T-Slot products perhaps looking into their other products would be a good first step. Using this style connection should allow you to keep a square frame, I would suggest bolting or adding gussets to sections you expect to see more load.

They sell a product called Quick Frame which has many options for connectors and comes in at about half the weight of T-Slot.

80/20 Quick Frame:
http://www.8020.net/Quick-Frame-5.asp

Another Company:
http://www.brunnerent.com/Tools/Port...y=2&strMetaTag

I'm sure there are other options out there as well.

[DonRotolo]

By tubing, I am guessing you mean square extrusion? Much of the following applies to round tubing as well.

Gusset plates can be made with a hacksaw and drill. If the tube is in danger of crushing (from nut & bolt) use hard wood (we use oak & maple), pounded into the tube, as an internal support.

Don't be so hard on blind rivets. They are faster to remove than a nut & bolt (want to race me?) and at least as strong. Never ever had one fail.

In some places we leave two of the 4 tube walls intact as 'tongues' we use for fastening. We've also used 3/4" square as an internal splice for 1" square (1/8" walls), as well as 1-1/4" square as an external splice.

We use 80/20 for prototyping, since it's easily adjustable, but move to extrusion for the final robot where it makes sense.

[akoscielski3]

Originally Posted by DonRotolo

By tubing, I am guessing you mean square extrusion? Much of the following applies to round tubing as well.

I pretty sure he means tubing like 1" x 1" with side walls of 1/8. That is what we used this year for our robot, though we did have it welded at my dads shop (biggest sponsor). But you could easily use gussets but it would be heavier than welding them.

[Mike Schreiber]

397 has done many bent aluminum robot upper structures, although it is not recommended for a chassis because of accuracy and deformation, it is definitely a cool (easyt) way to form a robot.

http://www.chiefdelphi.com/media/photos/30277

[randomunchies]

How much extra weight would gussets and rivets add to the chassis, as opposed to welding?

[Billfred]

Originally Posted by randomunchies

How much extra weight would gussets and rivets add to the chassis, as opposed to welding?

Just short of negligible. You're talking about a small amount of 1/16" to 1/8" flat plate, strategically applied, and a part of a series of aluminum rivets. Your unit of measurement would probably still be ounces for most chassis configurations.

Originally Posted by DonRotolo

Don't be so hard on blind rivets. They are faster to remove than a nut & bolt (want to race me?) and at least as strong. Never ever had one fail.

I've had two fail that I can remember:

At Brunswick Eruption 2007, 1618's robot tipped and sheared the two 3/16" rivets holding together the tower on one side.

At Peachtree 2011, 2815's roller claw was run into the alliance station wall, causing the rivets holding the lower part of the roller claw to become loose enough to become an issue. We decided to scrap the lower roller in favor of a flat plate.