Construction with gussets and bolts

[Brandon_L]

Congrats on doing some off-season research! Selecting a set construction method for your team is immensely important and I see it often overlooked resulting in some 'slapped together' style robots that tend to fail often.

Having done both gussets+bolts and gussets+rivets each for a full season, rivets are pretty awesome. We used 1/4-20 bolts in 2013, and in 2014 replaced them with 1/4" rivets. Our hole patterns and everything were treated just as if we were using bolts, but instead we dropped rivets in the holes. And if worst comes to worst and you break a rivet for whatever reason (which has not happened yet) you can drop one of the million 1/4-20's you have laying around through the hole in a pinch*. Some would argue the weight loss is a plus as well. We've used 1/8 tubes with 1/8 plate for gussets but you could go for a thinner gusset plate and probably tube wall for sure.

2013
2014

Clarification #2: We have used a lot of rivets in the past, and do like them for certain applications; however, I'm specifically interested in trying out bolts on the critical parts of the frame- primarily drivetrain- where it can see a lot of stress due to collisions or hard driving. These would be parts that we shouldn't ever have to disassemble or replace between matches (fingers crossed) so I'm not really concerned about the speed of installing/removing them.

With the right kind of rivets, I guarantee you wont see them break. Here's what we've used successfully for a couple years now. They might even be overkill, but we opted to go for slightly stronger because of a bit of paranoia just to guarantee no issue. I have not used smaller rivets before, but I've also seen teams use smaller sizes (I believe 3/16"?) with a more aggressive pattern.

*Our kids also rigged up a system using the old KOP air compressor to allow us to rivet in our pit, which was pretty nice.

[asid61]

We use both rivets and socket head cap scews. Hex head bolts we never use. I prefer to use rivets wherever possible, generally 1/8" or 3/16" excepting the things like the gussets on the drivebase which use 1/4".
This year we used a lot more bolts because of a want to make assemblies removable, odd loading conditions that are bad for rivets, and ascetics in some places. In any situation where you're using bolts, do a weight calculation of locknuts versus 1/4" plate; surprisingly we often find that a 1/4" tapped plate is actually lighter than nuts, and that's without any sort of lightening.
Locktite is nice, but lockwashers can get the job done just fine. Our elevator back support was held on by bolts through 1x1 to 1/4" tapped plates and it never came loose due to lockwashers.
Buttonhead screws going into tapped plates also looks very good, due to the lack of a cylinder-like head and nut. I also like the increased holding area. However, for sizes 8-32 and smaller, I like to use socket head screws because the tiny size socket of button heads make it easy to strip out a screw if it's repeatedly removed and replaced.

[Chris is me]

Originally Posted by Joseph Smith

We've used versaframe extensively, especially this last season. I'm looking into construction methods that don't rely on rivets as the primary means of fastening, and that utilizes our machining resources rather than the cash in our pockets. We have a laser-cutting sponsor, so designing gussets rather than buying them would make sense for our team.

None of your machining constraints really have anything to do with bolting versus riveting. You can totally build riveted frames using those same resources. To better answer your question, I'd like to know - what problems have you had with rivets?

If it's strength - stepping up to 3/16" rivets doesn't cost a ton of weight and can be a lot stronger. If it's repairability / removability - rivet nuts combined with bolts are useful for mechanisms that have to attach and detach themselves a lot.

Bolts in gussets and tube can be a bit problematic - bolts going through both sides of the tube can crush it if you're not careful, bolts have more slop leaving a less rigid frame in some cases, etc. It's still quite doable though. You can even carefully size holes to accept both #10 clearance and 3/16 rivets, or #8 clearance and 5/32 rivets.

[D.Allred]

Originally Posted by Joseph Smith

I've been investigating some new design and construction methods that my team hasn't explored in the past, and I really want to try assembling the frame and important structures with machined gussets and bolts. I have a few questions for teams who have used this construction method. In this scenario, I'd be using two gussets, top and bottom, to create a 90 degree joint between two lengths of rectangular aluminum tubing, all 1/8" wall thickness.
1. Is it better to put a long bolt all the way through and use a nut on the other side, or to tap the gusset and tubing and use shorter bolts?
2. If a bolt all the way through is used, are large washers needed to spread the load to prevent the tube from being crushed?
3. If the parts are tapped and a short bolt is used, would it be preferable to tap both the gusset and the tubing or to drill clearance on the gusset and only tap the tubing?
4. Is locktite a necessity with the tapping method?
5. What has been your overall experience with this method of construction? Pros/Cons?
Thanks!

I should clarify, with the pretapped method, we would be using machine screws or something of the sort.

Clarification #2: We have used a lot of rivets in the past, and do like them for certain applications; however, I'm specifically interested in trying out bolts on the critical parts of the frame- primarily drivetrain- where it can see a lot of stress due to collisions or hard driving. These would be parts that we shouldn't ever have to disassemble or replace between matches (fingers crossed) so I'm not really concerned about the speed of installing/removing them.

We've used this construction method for the past two years on our west coast drive base and on some superstructures. Here's specific answers to your questions.
1. Mostly tapped the 1/8" tubing and used short screws. Our build standard is based on M6 socket button head screws.
2. We use bolt through or rivet nuts for mounting the superstructure to the base. We typically spread the load with washers for heavy structures like our tower this year.
3. We used tight clearance holes on the gussets and only threaded the tubing.
4. Loctite is necessary.
5. Pros / Cons: This method is slightly heavier. Not much of an issue for the drive base, but superstructures with 1/8" tubing can add up even with pocketing. We used this method because we had the tooling not necessarily for reliability. But it does hold up to competition abuse.

One disclaimer - We are buying the tooling for 3/16" rivets next year. Just never got around to before.

Let me know if you have any other questions.

David

[Mr V]

Originally Posted by GeeTwo

Tapping 1/8" aluminum for machine screws is an exercise in futility - for any thing heftier than holding an encoder in place, they're almost certain to be stripped out, or rattle loose, or both. Always end in steel, or 1/4"+ aluminum. Self-tapping or sheet metal screws in 1/8" aluminum are not too bad, but can still be stripped out more easily than I'm comfortable with for the FRC team.

When we've used bolts and gussets with tubing or c-channel, we've usually gone all the way through, and used gussets on both sides to help distribute the load (minimize deformation of the tubing). Never use a single hex nut for any joint under load or vibration. Either get nylon locking nuts, or jam two nuts together to lock them in place. Sprocket-type lock nuts work when they're new, and being pushed against steel (they often mangle aluminum rather than grip). They don't work as well after they've been flattened - and how many teens will notice the difference?

Originally Posted by sanddrag

I have to respectfully disagree. I've tapped 1/8 aluminum quite successfully in every thread from #4-40 up to 1/4-20 and it works just fine if you're careful about it.

I have to agree with GeeToo that tapping 1/8" aluminum is not a good choice. The general rule of thumb is that for a hard material like steel the thickness of the tapped material should equal the diameter of the fastener. For soft materials with a hard bolt the thickness of the tapped material should be 2x the diameter.

So for a #10 bolt you would want the tapped material to be ~3/16" if it is steel and 3/8" if you are tapping aluminum and you want the full strength of the bolt.

If you look at standard hardware that you'll find at your local store a 1/4" nut will be 1/4" high, a 1/2" nut will be 1/2" high ect. Note that is for a standard non lock nut, nylocks, stover nuts, kep nuts ect will be slightly taller to account of the locking portion. This holds true whether the thread is coars or fine and regardless of the strength/material of the fasteners since the assumption is that you will use the same strength/material for the nut and bolt.

Now that does not mean that you are guaranteed failure is the tapped material does not follow those guidelines, just that the ultimate strength of the combination will not be realized.

In regards to the original question I prefer a bolt and nut through just the gusset and one wall of the tube. That does mean that the order of assembly must be considered and held to since if you are joining two tubes at right angles you will block access to one of the tubes with the other tube. This method will result in the ultimate strength since you'll have double the fasteners and often with less weight. Yes it is more difficult to assemble.

The other option is to use a rivnut. But for many applications I prefer a high strength pop rivet, as others have mentioned.

[Mike Schreiber]

We do that. More on the upper than the chassis, but we have some removable panels there as well. We specifically used bolts on our stacker this year rather than rivet because of concern for fatigue / rivets getting looser over time. Not sure I'd have that same concern with a chassis though.

[cpiravi]

Have you considered TIG welding the tubing? If you have access to a welder, it would be a good idea to weld your drive base together (as much as possible ) because it will be very durable, especially for gussets.

[Joseph Smith]

Originally Posted by Chris is me

None of your machining constraints really have anything to do with bolting versus riveting. You can totally build riveted frames using those same resources. To better answer your question, I'd like to know - what problems have you had with rivets?

If it's strength - stepping up to 3/16" rivets doesn't cost a ton of weight and can be a lot stronger. If it's repairability / removability - rivet nuts combined with bolts are useful for mechanisms that have to attach and detach themselves a lot.

Bolts in gussets and tube can be a bit problematic - bolts going through both sides of the tube can crush it if you're not careful, bolts have more slop leaving a less rigid frame in some cases, etc. It's still quite doable though. You can even carefully size holes to accept both #10 clearance and 3/16 rivets, or #8 clearance and 5/32 rivets.

We haven't had many major issues with rivets- mostly just loosening over time in certain areas and needing replacement occasionally. I think didn't use the right rivets this last season though- we primarily used these when in the past we normally used something like this.

[Joseph Smith]

Originally Posted by cpiravi

Have you considered TIG welding the tubing? If you have access to a welder, it would be a good idea to weld your drive base together (as much as possible ) because it will be very durable, especially for gussets.

We've welded in the past, and while I really like the durability and rigidity it gives us, we had a bad experience when we broke a part on our robot and were unable to replace it during the competition because we didn't have access to a welder. Since then we've moved away from welding anything except modular assemblies that can be swapped easily.

[Joseph Smith]

Originally Posted by Mike Schreiber

We do that. More on the upper than the chassis, but we have some removable panels there as well. We specifically used bolts on our stacker this year rather than rivet because of concern for fatigue / rivets getting looser over time. Not sure I'd have that same concern with a chassis though.

If I understand correctly, this allows you to bolt thin sheet metal together without tapping?

[Nick Lawrence]

One way that you can prevent the tube from being crushed is to press wood blocks into the center of thin wall tubing. This probably is not necessary for 1/8" wall tubing, but could be useful to make robust connections in 1/16" wall or thinner. I've attached a quick cross section below.

-Nick

[Nato]

Originally Posted by sanddrag

I have to respectfully disagree. I've tapped 1/8 aluminum quite successfully in every thread from #4-40 up to 1/4-20 and it works just fine if you're careful about it.

Seconded. 3476 has been tapping 1/8 aluminum for a few years now. If your bolts are rattling loose, use threadlocker (given the vibrations that normal FRC machines experience, we use it on almost every bolt on our robots).

Now stripping out the threads is a different story. One of the intake pivot mounts on our 2013 machine was connected with bolts to the 1/8 tapped aluminum frame. The consistent shock loads from quickly raising/lowering the intake and normal robot/field contact was enough to strip out one of the taps. In situations like that, we found it would be better to run a bolt entirely through the frame rather than simply tap one side. (It wasn't possible on that particular mount, for reasons I won't get into )

[philso]

Originally Posted by Mike Schreiber

We do that. More on the upper than the chassis, but we have some removable panels there as well. We specifically used bolts on our stacker this year rather than rivet because of concern for fatigue / rivets getting looser over time. Not sure I'd have that same concern with a chassis though.

We used some captive nuts like that to hold on the cover for our electronics panel last year. It would have gone more smoothly if we had pressed them into the chassis before installing all the other components. We also installed extras. These came into play when, over time, one or two of the captive nuts worked their way out and were lost. Most likely, someone pushed on the screw as it was being removed and pushed the captive nut out of the aluminum.

[OccamzRazor]

I am surprised there has not been more support on this thread for using 80/20 or Bosch Profile T-Slot framing components. Yes, we love and use rivets, but the question was more about which option to use if you are bolting gussets to extrusion. I would say neither if you can afford to go the route of T-slot extrusion which will allow you to bolt a gusset on either side of the extrusion with fewer protruding bolt heads.

The T-Slot fasteners take a LOT of tooling time away from the construction of the robot parts and you get the plus side of loosening the fasteners and adjusting the position of the components without drilling any holes most of the time. We use this for prototyping every year before we transform to a lighter design. (probably with rivets)

Check out an 80/20 catalog to see what they have, they even offer a lightweight version for robot construction and some suppliers offer discounts to FRC teams through the company. We even got about $1000 in parts and extrusion donated this last year for our pit.

[Chris is me]

Originally Posted by Joseph Smith

We haven't had many major issues with rivets- mostly just loosening over time in certain areas and needing replacement occasionally. I think didn't use the right rivets this last season though- we primarily used these when in the past we normally used something like this.

If you end up using rivets again, for critical areas like a drive chassis, where you don't expect to ever be removing the rivets, high strength aluminum rivets stay very firm over time and are just as light as regular rivets. They aren't that difficult to remove if you have to - you just need to center punch the rivet hole as the mandrel tends to break flush with the top of the rivet, rather than slightly inside the rivet.

The other benefit of these pricier high strength rivets is that they are rated to work in holes going all the way up to .201". That way, you can cut holes to .196" or so and use a tight #10 bolt clearance hole that doubles as a rivet hole.

The extra cost of these rivets isn't insignificant though - http://www.mcmaster.com/#aluminum-rivets/=x9kpy0