FRC T-bone-ing and Hexagonal drive

[MichaelBick]

Originally Posted by Ty Tremblay

1678 went to Einstein with that drivetrain and pioneered the math and testing with 971. I'd hardly say they're an example of a typical FRC team.

Most teams shouldn't make this a priority. Putting, say, a week into a hexagonal drivetrain won't make your robot better than putting that same amount of time into the things you mount to your drivetrain.

You've gotta walk before you can run. If you're already running, build a hexagonal drivetrain.

I totally agree that rectangular frame perimeters are easier to do. For some teams, it might be relatively easy to build a hexagonal frame perimeter. For some teams, it may be hard. Some teams will value the extra stability highly, and others will not.

As with everything in FRC, each team needs to make their own analysis to decide what choices will give them the greatest utility.

[Ty Tremblay]

Originally Posted by asid61

Having seen 1678's robot itself up close the last couple years, I have to say that it's not their robots' build quality that makes them what they are, but rather a fantastic driver and fantastic strategy. This year, their robot seemed to be made mostly out of versatubing-like material and other COTS parts, excepting the can grabbers.
Their 2014 robot was even simpler than this year's robot IMO. Surgical tubing catapult, Vex ballshifters IIRC, dual intakes, and ball stabilizer. If somebody could get a closeup of the way they made the hex chassis, that would be interesting to see.

I don't think you understand my point. 1678 is a team that has the organization and resources to devote effort into properly testing and designing a hexagonal drivetrain.

I don't want this post to leave teams thinking that if they build a hexagonal drivetrain they'll miraculously be better robots when they still can't handle the game object efficiently.

A drivetrain won't win you an event, what you mount on top will.

[Ginger Power]

Originally Posted by Ty Tremblay

A drivetrain won't win you an event, what you mount on top will.

In general I agree 100%, but there are examples...



Especially in weaker regions, a drivetrain alone can get a team into the playoffs for most games.

[EricH]

Originally Posted by Ty Tremblay

A drivetrain won't win you an event, what you mount on top will.

Ty, would you consider PVC, speed drive, or filecards to be more important for the Beatty BEAST of '02? And I'm sure Element '06 and Element '07 would disagree--winning a regional from the last pick with nothing but a drivetrain (and good partners/defense ability) kind of puts a hole there...


That being said, a drivetrain by itself won't win an event, but a bad drivetrain will lose you that event. The three most important elements of a robot are drivetrain, drivetrain, and drivetrain (Mr. Bill Beatty, in paraphrase). However, I would consider rapid acquisition of game objects to be a close fourth--and if you can acquire them, you can presumably remove them from your robot.

tl;dr: A good robot--which is a combination of a good drivetrain, a good manipulator system ("on top of" the drivetrain), a good drive team, and a good strategy--is essential to winning an event, but different teams' mixtures of those four elements can all win.

[Ty Tremblay]

Originally Posted by EricH

Ty, would you consider PVC, speed drive, or filecards to be more important for the Beatty BEAST of '02? And I'm sure Element '06 and Element '07 would disagree--winning a regional from the last pick with nothing but a drivetrain (and good partners/defense ability) kind of puts a hole there...


That being said, a drivetrain by itself won't win an event, but a bad drivetrain will lose you that event. The three most important elements of a robot are drivetrain, drivetrain, and drivetrain (Mr. Bill Beatty, in paraphrase). However, I would consider rapid acquisition of game objects to be a close fourth--and if you can acquire them, you can presumably remove them from your robot.

tl;dr: A good robot--which is a combination of a good drivetrain, a good manipulator system ("on top of" the drivetrain), a good drive team, and a good strategy--is essential to winning an event, but different teams' mixtures of those four elements can all win.

Game breaking strategies like 2002 are not an accurate example in this era of FRC. 3rd robots aren't going to get picked for their hexagonal drivetrains. They're going to get picked for having solid, well driven drivetrains. If a team has to sacrifice either of those two in the least then they are not a team that should build a hexagonal drivetrain. My team, 319, is one of those teams.

[Lil' Lavery]

Sure, there are exceptions, but I don't think Ty's assertion is incorrect. You're generally not going to win an event because you have a stellar drive train alone. A bad drivetrain can certainly lose you an event, but a good one isn't going to win without a functioning manipulator, drive team, and strategy.

Bumper profiles and bumper fabrics are things that are important for the 90th percentile teams trying to become the 95th or 99th percentile teams. They are far less important for the 50th percentile team. Rather than spending time, money, and manpower into researching bumper shape/material, it's probably better to invest that into, say, intake shape/material.

[jkelleyrtp]

I don't understand what all the fuss is about when it comes to their complexity. Instead of going all-out like 971 in 2014, many teams can easily make something like the kleinbots' offseason CAD or 148's x009 prototype.. Both of these accomplish the same goal as a hex-chassis but with much less complexity and required knowledge.

[cadandcookies]

Originally Posted by jkelleyrtp

I don't understand what all the fuss is about when it comes to their complexity. Instead of going all-out like 971 in 2014, many teams can easily make something like the kleinbots' offseason CAD or 148's x009 prototype.. Both of these accomplish the same goal as a hex-chassis but with much less complexity and required knowledge.

Not sure I'd qualify x009 as something "easy" for many or most teams to make. Good sheet metal fabrication is hard to find in a good deal of places around the country, and the design knowledge necessary is a whole other animal.

Bottom line is hex adds complexity. In some cases it might not add much, but if you're going to get tripped up on that you should probably be doing KOP or a transition drivetrain like VersaChassis anyways.

[jman4747]

Originally Posted by cadandcookies

Not sure I'd qualify x009 as something "easy" for many or most teams to make. Good sheet metal fabrication is hard to find in a good deal of places around the country, and the design knowledge necessary is a whole other animal.

Bottom line is hex adds complexity. In some cases it might not add much, but if you're going to get tripped up on that you should probably be doing KOP or a transition drivetrain like VersaChassis anyways.

You're not wrong but I don't think it would be too hard to adapt a standard WCD or kit bot chassis to give the sides a slight hex shape but exactly how hard depends on, as you said, the machining resources of the team and how much the side sticks out. The other main problem is it makes bumpers that much more complicated...

[GeeTwo]

If you're already doing a custom built chain-in-tube WCD (or any other 6WD+ design with all live axles), is it really that much harder to do a hexagon? It looks to me like it's just moving the corner wheels to the inside of the tube/channel and adding some blocks to mount the angled bumpers. Making the bumpers with the odd angles shouldn't be that much more difficult than a rectangle; the hardest part would be cutting and sewing the cloth to give a tight fit.

This is probably not a 50th percentile team issue, but I suspect it's a good bit below 90th.

[BrendanB]

Originally Posted by Ty Tremblay

The important thing to remember here is that a hexagonal frame perimeter is an ENHANCEMENT and not a necessity. Don't sacrifice build time to design a hexagonal drivetrain unless you think that t-bones are what's holding you back.

In my opinion, only the top ~5% of FRC teams can both benefit from a hexagonal drivetrain and have the resources to build one without sacrificing elsewhere on the robot.

+1

In 2014 3467 had some extra edges up our sides which were a necessity for our shooter packaging and not added for t-bones which ended up being a benefit in a heavily defensive game.

We poured a lot of time and effort into the frame extensions, bumpers, mounting, & maintenance that didn't need to be spent there. I wouldn't advocate building another frame like that in the future and fully agree with Ty that unless you have an higher level of manpower, experience, & resources you should probably avoid trying this during build season.

To help with t-bones I would prefer a drop down omni wheel or ball caster that teams can easily remove or add depending on their needs without locking our drivebase into a specific design/layout early in the year.

Personally the bumper construction & mounting is what makes this frame style very hard to pull off easily.

Originally Posted by sanddrag

Can anyone elaborate on what exactly is happening with the plywood in the bumper in a robot like 1678 pictured above? Is it separate pieces cut and joined somehow?

Can't speak for how 1678 achieved theirs, but on 3467 we had five individual pieces of plywood with the three side pieces held together by a patiently cut down/sanded 2x4 since we didn't have access to sheet metal. It was as unpleasant as it sounded and very weak. IIRC a versa gusset or two was thrown in at some point after the plywood started to crack.

[Ari423]

I have to agree that a good drivetrain won't win you an event, but a bad drivetrain will make you lose.

In 2014, our drivetrain was a sheet metal beauty. A marvel of engineering, to quote our chassis mentor. It only weighed 25 lbs IIRC, and it was sturdy to a fault. However, we spent nowhere near enough time making a manipulator, and we ended up with a robot that couldn't control the ball. At competition, we played lights-out defense, to the point where our opponents scores were consistently 40 puts below their average. We even started keeping a tally of how many robots we disabled throughout the season (we got to 9 in 2 events IIRC). But since we couldn't do anything with the ball, we never got picked for elims and we didn't make it to DCMP.

In 2015, we only spent a few days making a quick VersaFrame and Nanotube gearbox mecanum drive chassis, and we spent a lot more time working on our manipulator. By competition, we had a 100% success rate for stacking (we didn't drop a single stack), which got us picked for elims at all of our events and advanced us to DCMP (almost CMP). However, our drivetrain didn't drive straight or fast, so we wasted a lot of time lining up with the scoring platform. Also, the unnecessary weight in the drivetrain kept us from adding the air storage tanks we needed to speed up our pneumatic elevator. If we spent a little more time on the drivetrain, we might have made it to CMP.

Tl;dr - As a medium-to-low resource team, we did pretty well when we spent more time on the manipulator, but we could have done even better if we improved our drivetrain.

[MichaelBick]

Originally Posted by BrendanB

Can't speak for how 1678 achieved theirs, but on 3467 we had five individual pieces of plywood with the three side pieces held together by a patiently cut down/sanded 2x4 since we didn't have access to sheet metal. It was as unpleasant as it sounded and very weak. IIRC a versa gusset or two was thrown in at some point after the plywood started to crack.

The best way I have seen so far is making a seperate, hexagonal, gusseted bumper frame out of 1x1 tubing. If you have the ability to make gussets it is just like making any other mechanism frame out of versaframe.

[Michael Corsetto]

Originally Posted by BrendanB

+1

In 2014 3467 had some extra edges up our sides which were a necessity for our shooter packaging and not added for t-bones which ended up being a benefit in a heavily defensive game.

Personally the bumper construction & mounting is what makes this frame style very hard to pull off easily.

Can't speak for how 1678 achieved theirs, but on 3467 we had five individual pieces of plywood with the three side pieces held together by a patiently cut down/sanded 2x4 since we didn't have access to sheet metal. It was as unpleasant as it sounded and very weak. IIRC a versa gusset or two was thrown in at some point after the plywood started to crack.

I was shocked with how similar our bots ended up looking in 2014, great minds or something I guess...

We pretty much made our bumpers the same way as you guys, except replace the 2x4 with a funky looking sheet metal piece. Sheet metal was bent in-house with a vise and hammer (we were still working out of a shipping container in 2014...)

Agreed with everyone on this bumper shape not being a priority for most teams. We worked with 971 on the science behind it, mostly because we were tired of being immobilized mid-match when we just want to score points. The research we did before build definitely paid off in a game like 2014.

-Mike

[T^2]

Originally Posted by Michael Corsetto

great minds or something I guess...

Great minds think alike, weak minds seldom differ.

The bumper sides were in three pieces of plywood, connected by a sheet metal piece. The piece was bent at the workplace of one of our mentors. At first, we maintained frame-to-bumper legality by extending little metal tabs from the drivetrain to the plywood. Turns out, those tabs were pretty weak to side-on high-speed ramming. We replaced them with C-shaped sheet metal brackets extending from the drivetrain to support the bumpers, which I made by clamping the machined piece in a vise and whacking it with a mallet.