Hi CD,
For as long as I have been on NRG we have been using versachassis; however we’ve always put it together using T-gussets and #8 bolts.
Doing more reading on CD I’ve seen that rivets and the standard versachassis gussets are a popular alternative.
I can see the weight savings quite obviously. My question regards any loss in strength.
How much weaker would the chassis be in this case?
I would say “mathematically some, but not enough that it actually matters”. VEX does their holes for 5/32" rivets, though I prefer slightly larger 3/16" rivets that are significantly stronger. (220 pounds of shear strength vs. 310, which is how most tube-and-gusset setups would interact with the rivets. The 5/32" rivets are probably enough much of the time when backed by a bumper to help spread out the deceleration from a hit, but this is just one of those places where I’m playing conservative.)
We used a lot of them (probably too many) in 2016 to hold together a tube frame, and it withstood everything FIRST Stronghold could throw at it.
Conventional wisdom would be to try it on a testbed robot and beat the tar out of it to see how you like it, though this one is a pretty low risk. (Especially if you decided to go 5/32", as that shouldn’t require replacing any metal if you want to go back to #8 screws.)
We typically use 3/16 aluminum rivets. According to Fastenal these have 320lb of tensile strength and 260lb shear strength. #8 bolts are typically significantly stronger. That being said, rarely will your drive train experience forces in excess of 1000 lbs.
Fatigue stress is also a thing, however, as long as the rivets are the proper length with the right size holes I’ve never had an issue with them coming out over the season.
I’ve heard that, in shear, rivets are stronger than bolts per weight because rivets don’t have threads. Any strength loss by switching specific materials or sizes is a different issue.
That said, I’ve always used bolts because the team had them in stock, and because we could take apart our drivetrain if something went terribly wrong.
We use T gussets and found that rivets work well on the horizontal section of the T, and we put bolts on the Vertical part.
You should be fine using 5/32nd rivets as long as you drill out the intermediate holes once you line up and rivet the 1" spaced holes. If you want to be extra sure, drill the holes out for 3/16 rivets.
Rivets are definitely, absolutely, totally fine for this application. Hundreds of FRC teams put their frames together in this way. They tend to stay tighter too.
Lighter, cheaper, stronger. The mantra of product improvement teams everywhere. 
During my current team’s first trip to MSC, I walked our build mentors over to 217’s pit and introduced them to Paul. He demonstrated how the Thunderchickens perform remove-and-replace work on riveted sheet metal chassis elements. Paul’s main point, which he demonstrated very convincingly, was that drilling out an old rivet and replacing it with a new one (same size) is about 3x faster than removing and replacing a bolt-nut combination, even when the nut is easy to reach. Many chassis rivets are installed in locations where reaching a nut would be a major issue on the completely assembled robot.
The Average Joes started riveting soon afterward. Time is precious in an FRC pit.
I don’t really enjoy using rivets in the corners of the chassis. For the bellypan there’s not much option to use anything but rivets, but for the corner gussets where you only have 2-3 fasteners going into each 2x1, I prefer screws. A pair of 10-32 screws will beat a pair of 3/16" rivets.
For other mechanisms I like rivets, but I bolt them to the drivebase with 1/4"-20s. A 3/16" hex bit and a drill can remove and replace 4 1/4"-20s faster than I can rivet them without pneumatics on hand.
On 5459 Last year, we used 5/32 rivets and gussets on both sides of 90 degree tube joints on a swinging climbing linkage for a 100 lb robot with no noticeable wear or breaking all season long. It is probably worth running numbers, but they hold up well if you gusset both sides. (We also used them on a west coast drivetrain and it was extremely rigid).
To be blunt, this seems to be sort of just a gut-feeling thing not really backed up by experience - “it doesn’t seem right” sort of thing. But this is fine, this is something tons of teams do all the time. And if you are drilling out the rest of the rivet holes in your T gussets, you’re generally putting in more than 3 rivets, especially when you also count the belly pan rivets.
We’ve never done a VersaChassis, but we’ve done a few VF manipulators and I’ve done some VF chassis for roughly FTC scale robots. 98% of the time, rivets are the answer - lighter, easier to install, not too difficult to remove, and cheaper unless you pull them a lot of times. So what’s that 2%? When you build a chassis according to the usual configurations, the rivets mostly keep the gussets and the members from sliding relative to one another. This is what blind rivets are good at - handling shear stresses without much weight. They’re not as good 1:1 as threaded steel fasteners of the same size, but they’re a lot lighter and less expensive - as suggested above, double up if in doubt.
If a given fastener is going to experience significant tension (that is, a force trying to pull the rivet out), it’s back to threaded fasteners.
I have had rivets in such situations loosen up before, so your gut feeling that this is a “gut feeling” of mine is unfounded.
Belly pan rivets, as I stated, are fine. Plenty of virtual gusset there and plenty of rivets. A drawing of the versachassis gusset (link) shows that if it is installed on a 2x1, you only get two 5/32" rivets per gusset holding the long siderails to the front/back rails. I like having more than that given my experience with rivets and versaframe in particular.
We use 3/16" rivets for our corner gussets, but make our own gussets with more rivet holes. If we were using the standard T shaped Vexpro gusset, I would want to use bolts for the gusset segment that only gives you two holes to attach to the frame. If you have only a couple of rivets holding a bracket to a tube in a high stress area, they can loosen up over time or break with a large enough impact.
Here is what our corner gussets looked like last year. Once during a practice session we ran into a field element (with no bumpers on) and sheared off the 8 rivets holding the corner gussets onto the main drive tube (4 on top, 4 on bottom). Bolts might not have sheared, but it was easy to rivet back on. This wasn’t ever an issue when we had bumpers on the robot.
Once you go rivets, you won’t want to go back. Use bolts in areas were you expect to be removing & replacing items for wear, damage, etc. throughout an event. Something like your chassis can definitely be assembled using rivets - we used 5/32" with no issue on 4329, some recommend 3/16". Go bigger in areas you think will take some abuse.
Pick up one of these and you’ll be assembling robots with ease - https://www.harborfreight.com/14-in-air-hydraulic-riveter-61482.html
The thing I love about going to a gusset style of building is that any student of any experience level can do it. Clamping the gussets, transfer drilling holes, and riveting those gussets using an air riveter are all very easy operations. I’ve had brand new students with no mechanical knowledge making competition robot parts this way.
Rivets are also much lighter and are very cheap when bought in bulk. Save some weight and money and start riveting!
I like your gussets. Do you by any chance have it in CAD?
Moving on a bit.
No grade #8 screws probably are not a lot stronger than a properly done rivet.
Properly done rivets mean tight holes so the rivet takes up all the space when set. Rivets are a team player. More is better. Especially when done in a pattern like the S_forbes gusset plates.
For things that need high strength, we use structural or aircraft grade rivets. They retain and lock the anvil in place which adds substantially to the strentgh of rivet…
Quoted for emphasis. If you’re about the 3/16" life, go buy like five or ten #11 drill bits and keep them with the rivets because Those Are The Drill Bits You Use For Things That Will Get A 3/16" Rivet.
And, of course, buy rivets that match the thicknesses of the materials you’re using.
Rivet question:
If I’m riveting a 1/8" gusset onto 1/8" tube, do I use a rivet with a range from 0.1880"- 0.25", 0.251"-0.375", or will either work well?
My team uses thr .1880-.25 grip range fine. Both shpuld work. That said .125 is really thick for most gussets. Consider using .09" material if you are making them yourselves.
Not the exact model, but it’s a simple pattern:
Hi RickyBobby,
This is a question that I’ve had a bit of trouble with myself in the past.
Really it comes down to what type of poison you prefer. Since you’re asking this question I take it you understand that the sum of your material thicknesses should fall within the grip range of your chosen rivet. However, on these edge cases it can get a bit tricky. But first, the long and short of it is that for these applications here you can usually get away with using either.
Now going a little more into it just for the sake of practicing my typing skills a bit more (nephew has really gotten on about how slow I type so I figure it’s better to practice!) If you know your stock material is coming in a little undersize you can easily get away with using the smaller grip length version of the rivet. Alternatively, if you only of the longer version and you’re worried about reduced clamp load on the joint, you can throw a washer on one side of the joint to make the joint thicker where you’re riveting, that’s a fun garage trick really…
One more thing, now with the internet and catalogs its a lot easier to buy hardware but something we used to do a lot was the following hand and thumb rule: 1.5 times the rivet diameter plus the sum of the material thicknesses. That gives you the rivet length - NOT THE GRIP RANGE!
So, in this case…lets say you’re using a 1/8" rivet to join two pieces, one is 1/8 and the other is an 11 gauge aluminum bracket. That means, 1.5x.125 + .125 + .09 = .1875 + .215 = .4025 rivet length - Now we can go look at McMaster and see that a rivet with diameter 1/8 and length .4025 (97525A425 for example) and we see that the grip range contains our needs!
And another thing, this conversation seems to be talking about rivets vs bolts - hey Boltman where are ya?! I think that both have their uses but I give my vote to rivets for this use case, if this talks still going on next time I’m here I’ll give more reasons why I think so.
Merry Christmas everyone! Hope you all get the gifts you want and get to see your families. Make sure to give Santa cookies and beer…I mean milk :rolleyes: !
Best Regards,