We had this happen, and are keeping it around as a reminder.
Yup, that’s not good!
We like to make a bracket from sheet aluminum and rivet it to the tube, whenever we need to attach something. That way there is something we can tighten a bolt onto.
Using internal spacers will help prevent the tube from collapsing.
Now, when this happens on our team, we have the student/mentor sign the part and it goes on the “Wall of Shame”.
It was at the end of a long build session, and one of our team leads didn’t notice that she didn’t have the washer on. She said she suddenly realized “this should be tight already.”
Sigh… What a waste of a perfectly centered hole…
That’s why I like rivets.
Thats what happens when you use the wrong fastener for that function.
You should use Riv-nuts/Rivet-nut/Clinch nuts in that type of situation to prevent that possibility of that happening. In industry that is called mistake-proofing because the failure is inherently by design.
Also as an FYI don’t use washers unless you have to. You can simply use flanged head bolts that have a washer face built in. This prevents accidentally omitting the washer and the possibility of losing the washer and creating foreign object damage.
This is almost exclusively what we use for this reason:
Us too. Use caution, however-- outside of FIRST, button head cap screws tend to be avoided because they aren’t designed to provide as much joint strength as socket head cap screws, for example.
“Don’t crush the tubing!” is sort of a team mantra. Unfortunately, some freshman mannaged to indent a 1/8" gusset plate about an eighth of an inch. I guess “snug” is a loose term…
I’ve been fighting to use rivet nuts for some time and out head Coach/Advisor just didn’t believe that they were robust enough. So the other day when one of the other mentors was saying how their sub-team wanted to use them but were being over-ruled I devised a little test to show that they won’t pull through.
In this blog post on our website you’ll see a picture of a properly installed steel rivet nut from Mc-Master McMaster-Carr that was subject to attempted destructive testing. It was installed properly and a bolt threaded into it. Then we placed it in the hydraulic press and proceeded to push on the bolt from the back side. The 1/8" aluminum that was over an aprox 1.5" x 1.5" hole deformed first, once it started to tilt the bolt then deformed but there was zero sign of the rivet nut wanting to push through the material.
I’m not surprised you had to convince someone unfamiliar with rivnuts of their strength. Every time we get a new mentor, even experienced ones from other teams they question us on it. I then show them some heavily abused robots that have rivnuts in high stress locations where they survived. After that most people start to think of how to utilize them. Since we use mostly 1/16" wall 1x1 we rarely have a robot that doesn’t need them somewhere.
To the Riv-Nut Proponents:
We’ve used these in some locations in the past, only to have them strip their holes and start spinning on us. This was (I believe) in some 1"x1"x1/16" aluminum tube. I’d love to use them more/again but I was somewhat put off by them causing us to miss a match when we couldn’t remove a bumper. Do you have any suggestions as to what we did wrong?
This is why I try to force our kids (and some of our mentors) to use hand tools for a bit before they’re let loose with cordless drivers or drills. It’s too easy to be overzealous with a power tool and ruin your part.
The piece of tubing in question had been repositioned a number of times as we tested the conveyor system to find the best place to position a cross member. The holes were already drilled for mounting, and we needed to move it around to find out where performance was the best, so we weren’t going to use rivets. The whole assembly is designed to be (relatively) easily and quickly removed from the robot in case we need to repair it or get access to other stuff.
It was done by hand, using a nut driver and a wrench. And by one of our most careful and sharp students. She was just paying attention to the other side (which had very little free space and needed to be fairly precisely positioned) and over tightened the bolt.
But I must say I like a student (or anyone) who owns up to a mistake and uses it to teach others not to repeat the mistake. So she earned a little more respect from me over this.
Using the knurled version like I linked to is preferred over the smooth version. The other big key is to make sure they are installed properly. Having a good tool is part of that equation, so you can ensure the proper compression is achieved, as is making sure that the hole is the correct size. If it slides into the hole easily the hole is too big.
Here is a good tool from McMaster. I’m not finding the one we use, if I do I’ll update.
I’m with Mr V on this. I’ve never had a properly set rivnut fail in Al. The only issues I can thiink of are when you try to set it in a material with a higher Rockwell hardness which prvents the knurls or splines from digging in to the parent material.
We have had issues when people who haven’t used them before don’t compress the nut all the way however, which I was reminded of last night when I realized the only person on the team other than me who installed our rivnuts for the last four years graduated last year. :o
The big thing I do not like about rivnuts is when the do fail, the corrective action can be time consuming. I know in a perfect world they will not fail, but the heat of a six week build session & multiple matches in tournament tend to cancel out the perfect world scenario. They are useful in their place & are better than tapping thin gauge Al. I second the idea of using the installation tool to set them. It has hardened face to get a consistent set.
We put PVC spacing in between anywhere where we will be tightening a lot. It’s light, durable, and a great option for any team who is without rivets.
Note to self: Get some rivets.
Last year we put spacers in the AndyMark wheels to prevent binding the bearings by over tightening the axles. I have a mechanum hub that is deformed from over tightening. Seems like about the time people are trained to be aware of the issue, they graduate. ;(
We only put the bolt through 1 wall of the tubing. We use round socket or phillips bolts, and bore a larger hole with a stepper bit (not sure of its official name) on the opposite side for access. We have a magnet in case the bolt drops. We did this on our drive train last year, and on a pretty high sheer-stress piece in prototyping this year. We haven’t had a single problem yet, and the drive trains had (have again this year) only 1/16" wall tubing. The key is torsional support.
It also saves us some weight (maybe 1lb) overall on the bot since the bolts are all 1" shorter.