But it looks like you have some problems with your chain interfering with your gearbox standoff and possibly your belly pan too. I would suggest you downsize to 16T sprockets instead of the larger ones you’re using.
Have you considered how your corner gussets will stand up with significant torsional loads? Or are you using a bellypan? Also, how will you make sure that the two sides of the drives are parallel? Rivets holes usually have enough play in them that locating of them doesn’t work super well. Are rivets the only things holding the frame together, or is the tubing welded too?
I only ask because I’ve recently been working through some of the same design problems myself. It’s interesting to see how different people have worked through the same problems.
By the way, I really like those bumper extension tubing thingys that surround the center wheel.
Thanks! Nice observation, that’s on the change list for the next revision.
Right now it’s just those gussets, but I’ve been considering adding a bellypan (for both strength and electronics). Would a bellypan also help with keeping the sides parallel or would the play in the rivet holes still be an issue?
I would much prefer welding the tubing but my team has never had success welding aluminum. We have access to a Miller Syncrowave 200 TIG and a Snap-On MIG 300, but no one with much experience welding. Would it be worth having some members learn and practice welding in the off season?
Bellypans do add a lot of strength (especially torsional, which was partially why I mentioned it to begin with). If you’re designing a WCD, you’ve probably seen some well done bellypans before, but even just for inspiration, take a look at 254’s or 1538’s. They do a really great job with the elex layout. In terms of keeping the sides parallel, it will definitely help. There will be a lot of rivet holes far apart on a bellypan. Because they’ll be separated by a longer distance, small amounts of wiggle in the rivet holes will change the relative angle of the two frame pieces you’re joining less than it would witj your gussets right now, where the rivet holes are relatively close together. (If this made no sense, consider the effect of changing the length of one leg of a right triangle a with either a small or large hypotenuse. With a longer hypotenuse, changes in the leg will have less effect on the opposite angle.)
Welding is both a lot of fun and very useful in FRC. I highly suggest that people learn it, especially TIG, during the summer. I’ve recently taken up TIG myself, and can personally attest to how much results are driven by experience and practice. TIG is generally considered to be better for thin aluminuim (like in a frame for FRC), but if MIG’s more your cup of tea, do that, it works fine too. I don’t know about your machines in specific, but Miller tends to make nice equipment.
I’ve changed my mind several times on this, but my current thoughts on welding vs riveting are these: do both. Welding is very strong, but tends to warp frames (and a lot!) if it isn’t done with proper care. Even careful tack welding, shimming, and fixturing can leave you with a slightly warped frame. Riveting, on the other hand, is much faster, and of course is less permanent. However, as I touched upon earlier, it isn’t quite as strong. My advice at this point is to make your gussets be fixtures for welding. Set up the frame nice and square, rivet it all together, and see how it looks and performs. Then, clamp it to a welding table, and, tack welding first, etc. etc, weld it together. You ideally get the strength of welding with the ease and straightness of a good riveted frame.
As goes with all of my advice on CD, try it yourself and make up your own mind. I’m just suggesting things based on experience. See what works for your team. Follow my suggestions to the letter, mix and match them, or even throw them completely out the window. It’s all about you and your team.
Rivets and gussets are definitely both rigid and adequately strong for this application. I don’t understand where you’re getting rivets with enough play that you can’t hold a frame together with them.
We used overly thick 1/8" gussets and a 1/16" solid belly pan this year instead of welding. Chassis was just as rigid, really. No noticeable performance or strength difference at all. Welding is not mandatory for this kind of drive. And there is definitely no need to do both permanently.
Here’s a picture of our setup. The thickness of the gusset and the number of rivets used were both excessive, but rivets are almost “free” in terms of weight.
What hasn’t been mentioned yet is that belly pans are almost critical for resisting shear – never mind torsion. My bosses at work call shear the forgotten force, and for good reason; it’s more difficult to simulate than simple point/line loads or torsional loads. Something as thin as .025" sheet aluminum, even with additional lightening holes, will make your whole frame much stiffer when it experiences shear forces.
Our 2011 and 2012 robots were entirely riveted, no welding of the frames. Our 2011 WCD was not noticably different from the others in terms of strength or rigidity.
I hate anecdotes, but I’ll be a hypocrite; Rivets hold airplanes together.
Also, the bellypan doesn’t have to be exotic like 233/254 does. We made two of those and while they are cool, the resource cost is so high. We’ve run both garolite (G-10/FR4) and a high quality 6mm plywood with great results.
We avoid welding on the drive base because of the warpage and also because of the additional time and energy. For us welding means outsourcing, which means we’re without the robot for a few days. Gussets are plenty strong and keep the build process completely in house, so we like them.
I was less thinking of rigidity after the rivets are installed, and more before. When a rivet is first placed in a hole (but not installed) the two pieces being joined together can still move in respect to each other a small amount. 3/16" nominal rivets are designed to go into #10 (.194") holes. While the OD of the rivet is probably larger than .1875" exactly, there is still some play between the rivet and the hole when it hasn’t been crushed yet.
Rivets are rigid enough to hold drivetrains together, no problem, just as you said. We riveted (and didn’t weld) parts of our drivetrain together this year and the year before. After the rivets were installed, the drive was quite rigid. However, I definitely noticed a significant amount of play before the rivets were installed. I do wish I had taken more care to make sure the drive was square, as rivets aren’t perfect for locating frame members together. A bellypan certainly would have helped with getting everything square before the rivets were installed.
Riveting with a bellypan is probably perfectly square, strong and rigid. But, if the OP has TIG and MIG welders in house, why not learn to use them in the offseason?
Just like I said before, it’s perfectly fine to completely disregard my (or anyone else’s) advice on CD. Personal experience and experimenting in the offseason should always trump what someone says on this forum.
Check out Cleco fasteners. They work as temporary fasteners to hold things together. My team used them when we got our sheet metal back to make sure everything was square and all the holes line up, before we riveted. They need a special tool to install and remove, but they take no more than a second.
They’re known as “hole-grip clamps” on McMaster-Carr
I would not recommend this. Smaller sprockets load your chain more, and make your drive less tolerant to variable tension and alignment inperfections. We did a similar drive with #25 16t sprockets, and were throwing chains left and right. There should be plenty of room on a drive like this for bigger sprockets.
Agreed. We try to stick to (if #25 chain) at least a 22t sprocket on the drive. Some rough tests I’ve done have shown this to be the smallest tolerable size. And when I say tolerable I mean it works pretty darn well.
However, if you’re aligned perfectly the smallest of sprockets won’t give you an issue. But are you really going to be perfect?
We traditionally have run 22T sprockets for 3.5-4.5" wheels, and have a lot of runtime on that. It’s bulletproof.
We ran 16T sprockets for 4.3" wheels for the start of the season, was bulletproof.
We then ran 16T sprockets w/ 6.3" wheels, and threw the same chain 3 times. I assume some sort of minor misalignment existed that was exaggerated by the very high chain load. This was late in the season and only on the comp bot (practice was fine with massively more runtime) so we never investigated the issue.
I’d say 16T sprockets are fine if your chains are all lined up properly for even 6" wheels, and are bulletproof for 4" wheels.
I know this exceeds the rated working load of the chain depending on how you look at it, but whether or not that situation actually arises is unknown.
I can’t say I’ve had any similar experiences. We ran 16T sprockets with untensioned chain and it lasted us two regionals and champs with absolutely no fails.
You are correct about the chain loading and for that reason I would say that it’s best to run the largest sprocket you can on your drive. However on this drivetrain it looks like the 22T sprockets OP is using are too large and the next smallest size is 16T, that is the reason I suggested using 16T sprockets. Also, I’m pretty sure 16T is standard for WCD’s, and I’ve never heard of WCDs popping chains.
Thanks for all the input. I’m planning on replacing the 22T gears with 16T. Since I dropped the middle wheel 1/8’’ rather than raising the corner wheels 1/8’’ or doing 1/16’’ on both, the 22T sprockets make the chain run into the bellypan in the current config.
I’ll also a 1/16’’ aluminum bellypan, and I’m going to keep to gussets and rivets for now, although I’ll try to get some people interested in welding since that can still be useful. Are there any other suggestions on improving the drive?