Well, we have also 3 rows of holes on each side as well, so that’s ~1oz/foot.
Which I would wager ends up being a lot more than drilling a couple of holes in your drivetrain gears, which I believe I saw suggested here.
Well, we have also 3 rows of holes on each side as well, so that’s ~1oz/foot.
Three rows of holes on the sides would mean you have 4x as many rows of holes as I do (you have eight, I ran it with two), so your savings on an 18" segment would be .052 pounds. Or 0.035 pounds per foot. Or 0.56 ounces per foot.
With the calculated information: Bet.
Same tooth count, DP, face width, and bore; am-3488 is the weight-relieved version used in the EVO gearboxes.
The difference is 2.8 ounces in one gear, and in a drivetrain you’d be running at least two such gears. 5.6 ounces. (Repeat the exercise with the 40 tooth gears used in a Toughbox’s 5.95:1 ratio against its EVO counterpart, and it’s 4.28 ounces for two.)
Do some robots use more than 8-10 linear feet of 1x2 tubing? Sure! If I was overweight, would I advocate for adding a hole pattern in an at-least-two-often-four-setup process ahead of looking at those gears?* Heck no!
*For the record, I’d look at other things before even contemplating the gears. But we’re comparing two things right now.
There are a lot of excellent reasons for a standard hole pattern in tubing. The weight savings from a bunch of #9 holes is miniscule enough as to not be “worth it” for solely that reason. It’s a great idea to do it from a modularity perspective (all of my teams’ robots do this), but for weight it basically doesn’t help.
Drilling holes or turning down steel drivetrain gears reduces their weight substantially - easily 1/4 lb per gear depending on the size. In terms of weight per operation, it really can’t be beaten.
Re: aluminum gears in the drive
We saw much more frequent aluminum gear-shredding in the drivetrain 2019, probably due to a combination of the 3x brushless motor drivetrain and increased amount of defense (and this was in COTS gearboxes). We took all the teeth off several 52t aluminum gears directly after the pinions that year.
That all being said, weight of steel gears adds up. Aluminum gears can still have a place in certain locations in the drive, but you really need to either be certain you’re fine to make it work ahead of time or keep a very close eye on the wear (and have replacements ready to go). Ex. we won’t run an aluminum gear on the output stage of the drive lower than 24t after our experience driving over the defenses and shredding teeth in 2016.
Why would you not use a gear like the am-3821 which starts at 3.6oz? Or on one that size even an aluminum one.
Sometimes, it really does have to be a 50T gear. For example, a Toughbox-series gearbox can’t run a 52T gear (like am-3821 is) because there’s no matching 12T gear for the appropriate spacing.
Aluminum drivetrain gears have been discussed to death earlier in the thread. Emphasis on “death”.
Here are some of my favorite ways to keep robots light:
As mentioned previously, getting rid of this can sometimes drastically reduce total weight, but sometimes a few well placed cylinders can accomplish things lighter and better than motors.
Where ever you can, try to stick to aluminum fittings instead of brass. I can’t find exact PNs right now but they are available online.
Use an appropriately sized manifold. If you know you only need 2 cylinders, get a 3 port base instead of a higher number (1 overhead is usually good for later additions).
Don’t use those giant pressure gauges or inline pressure regulators that come in the KOP and require brass fittings. Automation Direct has some really small and light push-to-connect legal variants you can find here.
The standard FRC Viar compressor sold by AM is really good, it mounts well and is pretty robust. If you are really in the need for weight savings, there is this Thomas compressor. Be warned - it is more expensive, difficult to mount, and requires the use of bent metal tubing to be FRC legal.
This is a weird one but for the 2020 season we only used 5/32’’ tubing and fittings on components post manifold. We found over previous seasons this worked well in select applications compared to the 1/4 standard variant and provides weight reduction, which can be important if you end up having pneumatic actuators on mechanisms with long tube runs.
More of a design thing, but try to keep the battery, main breaker, and PDP close to each other. This reduces the length of large gauge wire that needs to travel between these components.
When looking around for non-specific items, look for components with non-metal cases and the right amount of ports that you need. A good example of this is this Monoprice Ethernet switch we recently switched to vs this Netgear Ethernet switch that we used back in 2016/2017 because it was the first thing we found on Amazon.
Echoing what a ton of people said earlier, don’t skimp too hard on your drive train. Make sure your gearboxes are bulletproof and that the system will be able to keep up during competition. That being said, we have successfully used 1/16’’ wall box tubing for front/back frame rails. We have also switched out colsons/pneumatic wheels in the past for AM HiGrip wheels. We love them to death, even though they can sometimes get a bad wrap, and weigh a significant less compared to other wheel alternatives.
You can see the HiGrips and the front drive rail that we swapped out for 1/16’’ tube to get some weight back at a competition.
Don’t get too crazy with it, but something we found helpful was “scalloping” the edges of plates. All this is doing is getting rid of unnecessary attachment points between the plate and what it is providing rigidity to, which can net you good weight savings.
You can get away with 3D printing pulleys in a lot of cases. This not only helps with part procurement and integrating designs together, but also allows you to use belts and pulleys in a more creative way that can allow for a lighter mechanism than just using gears. A good example of this was in 2016, where the pulleys driving our wheels were printed out of PLA. They held up all season without any problems.
People often design intakes with large rollers on them or lots of different wheels, with a giant hex shaft going through the middle. You often don’t need that shaft. Try to replace smaller wheels with a more lightweight roller supported on both sides and either a dead axle system, or a way to get rid of the middle unused portion of the hex shaft.
You can see inside the bottom transparent roller that the hex shaft driving the roller extended from the 3D printed tube insert to the bearing on the far side of the plate, retained on both sides by snap rings.
Lots of cases where bushings would be a lighter weight, better solution than bearings. Good examples are intake pivots, mechanisms that drop down once a match, and other low load rotary points.
Thunder Hex from VEX has been a real game changer, but you don’t need to use 1/2’’ hex everywhere. VEX also sells a 3/8’’ variant which is perfect for intakes, rollers, conveyors, and standoffs.
Every shaft on our 2020 intake was 3/8’’ thunderhex, would have weighed a lot more with the 1/2’’ variant.
Thanks @JackTervay for reminding me about this one.
They weigh a lot, don’t use them.
So if we add contingency stickers everywhere, will i gain infinite HP?
We used brushless motors and built an (almost) all polycarbonate robot (6672 Fusion Corps 2020 Season Recap and Engineering Notebook!) this year and it was one of the first times we did not worry about weight. In the past weight was a big factor during the design process, but we made sure to stay under when designing so that we did not have to Swiss cheese
looks at team number
What if we use two of them?
To help teach students about the importance of designing with weight in mind, I have them calculate how much weight a .5 in hole removes from different thicknesses of aluminum and then how many holes are needed to remove a pound of aluminum.
Hint: the answer is
OMG that’s too many holes.
You may THINK you want to stay up until 3 am drilling all those holes, but your parents, your school’s principal and all of your teachers, your robotics mentors, your friends, and anyone else who likes you KNOW that it’s a bad idea.
For robots with large surface area and low load plastic sheet parts, you can save entire pounds by switching to a thinner plastic. I’m a huge fan of using WCP’s 1/32" polycarbonate for sponsor panels and hopper walls. Properly mounted, these will do the job just as well as 1/8 or 1/16" thick.
I also like how it’s thin enough to cut with scissors, so you don’t need to take up valuable CNC time for intricate shapes or clearances.
It also looks hecka slick but that’s not too important
Mcmaster also sells some pretty thin PETG and HDPE films that you can use in similar applications, but I haven’t tried these for large surface area FRC parts yet.
This is an unpopular opinion around kids who have been spoiled by CNC, but using unpocketed thinner wall or lower density materials saves more weight than pocketing generally. Try not to use thicker materials unless you absolutely need the geometry. We use lots of unpocketed polycarb in places where we used to use pocketed aluminum and it’s definitely helped us stay underweight.
Supremely dumb and extreme example, but in 2011 we would have saved even more weight with using a solid thinner wall box extrusion vs this overpocketed thick wall extrusion. It also likely wouldn’t have failed as spectacularly as it did.
In 2018 we were overweight by a few pounds about a week before bag day. Our elevator was designed to use 1/2in round bearings to support the mast as it went up in down. Yes, it was overkill. The easiest solution was to take all of the 1/2in bolts out that held the bearings on, chuck them up in the lathe and drill a 5/16in hole through all of them. That along with a few strategic holesaw holes brought the robot to 119.9lbs. We did consider going to 3/8th and probably could have got away with it. From now on, it’s standard practice for us that any bolt that is larger than 5/16in gets a hole in it.
LEDs scare away opponents…that counts right?
Our team is by no means experts at making light weight robots, but there were a few things that helped us out this year.
Rather than a sponsor panels, we simply put vinyl stickers (cut out on a Cricut) on our aluminum frame.
Overall we were pleased with the look, and didn’t need to add the extra weight of sponsor panels.
We also decided this season to use more plastic in places we would have previously used aluminum. We used nylon sheet, acetal, and polycarbonate to make different structures. This saved a ton of machine time, and saved us a bit of weight.
The CNC told me it lives for that
He does. However, I’ve found it very difficult to fit a finished robot into the CNC.
It is much easier to speed-hole parts BEFORE putting them together.
Our robot this year is extremely complicated and is right up at the weight limit. Apart from the usual things like thin wall tubing, polycarbonate construction and brushless motors, we’ve implemented several tricks that are pretty specific to this year’s robot, to keep it from being overweight :
We changed our flywheel to a lighter one with a larger diameter, which gives the same angular momentum (of course both wheels were heavily pocketed so that the mass is concentrated near the edge).
On our intake, we used thin (0.8mm) hollow polycarbonate tubes with 3d printed hubs as rollers. We used a short hex shaft for each end, which is much lighter than having a long hex shaft with wheels on it.
Our Swerve modules use a small brass bevel gear and a large anodized aluminum bevel gear, instead of two steel gears.
We’ve used belt reductions instead of gears/versaplanetary wherever we possibly could.
The fact that our climb telescope assembly ended up weighing less than 1.5kg didn’t hurt either : )
I’m general, being aware of weight in every stage of the design process is the main guideline, and finding places like these, where weight can be saved without any compromise, is always good.