Help out rookie/less experienced teams

Alright. I’ve seen some people ask how to make drivetrains, how to organize electronics, etc, so I think the CD community as a whole should help them out.

First off:

Building a Non-Kitbot chassis:

To build a very robust, non-kitbot chassis, you want the outside of the frame to be the largest part of your robot, then you want to make sure that there are plenty of supports and good fasteners. 675 uses 3/8ths inch Grade 8 bolts with a shoulder on them, and a washer on each side of the bolt. Gloria Machina, 675’s bot, is layed out like this.

http://www.lotechdesigns.com/host/images/8628GloriaChassis.jpg

The thin black lines are vertically-mounted peices of quarter-inch thick, 3 inch tall aluminium.
The maroon rectangles are the wheels.
The dark blue strips are the same type of aluminium, but bolted on top of our gearboxes to support the chassis and gearboxes.
The thick black strips are some other sort of material, and hold the stump of our arm, along with our ramp bracket and battery.

Not to scale.

Building a good Drivetrain:

First off, unless you want insane traction and a hard time turning with tank drive, steer clear of the KOP wheels. They’re very grippy and to lose traction, teams have had to cover them in plastic or zipties.

There are several different kinds of drivetrains you can do. You can do a direct-drive system with all wheels powered by their own motor (4 or 6 wheels works best for this), a chain-drive system with all wheels powered with their own motor/gearbox (again, 4-6 wheels), chain-drive for each side (2 motors generally, 4/6 wheels total), or some other interesting systems (like a 6-CIM drive system, or a 12-wheel system, not recommended for rookie or less experienced teams).

There are also several different types of wheels. Traction, Omniwheel, and Mecanum are the three most well known. Traction wheels have a grippy tread that gives, as the name implies, traction. The design on the tread can be all sorts of things, from a crisscross pattern, to a pseudo-car-tire design, to the kind of stuff that comes on IFI Traction wheels.

IFI Traction V2 wheels:


Not the two different types of tread. The kind on the left is called “Roughtop”, and the type on the right is called “wedge”.

Omnidrive wheels are cool. They are normal wheels in the sense that they are round and give you traction, but only forward and backwards. A common design for omniwheel useage is traction wheels in the center of a 6-wheel chassis, and omniwheels at the corners. This improves turning speed greatly over an all-traction wheel chassis.

AndyMark Omnidrive Wheels, 6in Diameter:
http://www.lotechdesigns.com/host/images/2026yhst-33833170891817_1942_2834515.jpg

Mecanum wheels are probably the coolest, however. The wheel design is similar to Omnidrive, however different in such a manner that the robot using them can move in any direction and rotate easily, allowing it to slide around much like a hockey puck. There is a special way of mounting them, however, and that is so that the 4 wheels form an X when looked at from the top down.

Mecanum Wheel, 8in diameter:
http://www.lotechdesigns.com/host/images/8556yhst-33833170891817_1937_1422730.jpg

Notice the staggered rollers.

Alright, well, that’s about all I can post about, as that’s about all I know how to do. If anybody who has experience with Pneumatics or Electronics would like to speak up, please do.

Thanks to AndyMark.biz and InnovationFIRST.com for the pictures, and making good prodoucts.

EDIT:

And to anybody who’s interested, I hope to find pictures or acquire pictures of our robot’s chassis and drivetrain. I know that a simple MSPaint diagram of a bot isn’t as helpful as one might make it out to be, but hey, it’s a start.

I also might make more diagrams later.

Hereis Team 1345’s 2007 Chassis. Kids wanted to have a low ground clearence chassis and 6wd. We used 4.5"x4.5" box aluminum tubes for 2 modules (on the side).

In this case, we were inspired by team 1114’s prototype chassis. All we needed was some 4.5"x4.5"s. It is 37 inches long and 27 inches wide. The C channels you see are 4.5"x4.5" cut in half with a notch. There are L brackets that connect the C channels to the modules. One side of the L (the side that is connected to the module) is riveted on with .25" steel rivets, the other side is bolted on with quarter inch bolts. We ran a single #35 chain through each module and used a vertical tensioning system. The final product weighed 53 pounds (including electronics).

Here are some links to pictures of the modules and the chassis. If you have any questions, or would like to discuss, feel free to send me a message.




http://www.flickr.com/photos/7634503@N04/444303255/ (last minute cheeseholes).

does this look familiar? from Thursday morning at Davis…





Why yes, that DOES look familiar. =]

I’ll take some higher-quality photos at Nationals and post-season, though. Post-season will also give me a chance to take the ramp off and focus on the chassis and drivetrain designs.

I’ll talk to our Electrical guy and see if he can write up some tips.

By the way, how do I create a “whitepaper”?

I just made a white paper (on Center of Gravity)…it’s pretty easy just go to CD-Media and click upload on the center white paper menu. I entered the text in the description box. Get your paper in good shape (complete, proofread, etc) before you upload it though…

A few things I want to clear up;

There is nothing wrong with the kit wheels and there is no such thing as “too much traction”. The kit wheels have even less traction than the IFI traction wheels that are recommended above.

High traction wheels can make turning difficult if the robot is not designed correctly. A 4 wheeled robot, with high traction wheels on all four corners, will not turn well… It probably won’t even turn.

the best advice I can give to rookies on drive trains isn’t to necessarily ditch the KOP chassis, but to go to a 6-wheel drive. The KOP frame is amazingly useful when you have limited resources.

A good thing for rookies with limited resources to do would be to design a 6 wheel drive (remember to lower the center wheel for turning, the KOP frame conveniently has this built in) that uses the KOP frame and an off the shelf gearbox (not banebots!). The easiest two would be the AndyMark single speed and the AndyMark 2-speed shifters.

A rookie team would get an amazing drivetrain with little advanced fabrication required, and would then be allowed to work on more ambitious things without worrying about a competitive drivetrain.

Also, unless you have the resources… don’t get too ambitious with the drive train; If the drivetrain fails, it makes everything else useless as well.

I’ve found that teams have had trouble turning at a reasonable speed with KOP wheels. IFI Traction wheels seem to break traction more readily when turning, and 675 uses the Blue tread (the really tough stuff) and we haven’t had problems turning.

I’m not suggesting an overly-advanced drivetrain, just a good, effecient one.

The BaneBots Gearboxes are just fine as long as you don’t get the 2:1 motor adapter and you get the Hardened Carrier Plates. 675 is running 4 56mm 12:1 gearboxes (one per wheel), and the only drive issues we had was a loose PWM cable once, and re-greasing one of the gearboxes. And that was after two regionals.

Also, make sure the Carrier pins are the correct length. I’ve read about BaneBots shipping the short ones and robot drivetrains failing. Not good.

That all being said, I’ve seen bots with KOP wheels have a harder time turning than IFI Traction V2 wheeled robots.

Most of what I was writing about, however, was intended for non-Kitbots. I’ve already stated this, as I do not know how to build a Kitbot (This is the first year I’ve been really involved in building, and also the year that we’ve had money for raw metal for a chassis).

That being said, perhaps somebody else could make a detailed post on how to build a good Kitbot Chassis/Drivetrain?

EDIT:

Oh, and Squirrel, thanks for telling me that. After I make some revisions, I’ll probably upload it.

Just want to reiterate.

The Kit wheels have a lower coefficient of friction than the IFI traction wheels. Any robot that has trouble turning with KOP wheels, would have more trouble with the IFI wheels.

The way to improve turning is in the drivetrain design (lowering the center wheel on a 6 wheel drive, etc…).

In general, it is best to use the maximum possible coefficient of friction for the wheels (I’m going to get some flack for that). Sure, high coefficient of friction wheels will make it harder to turn (compensate for that in design…) but will increase the robots traction.

I would reccomend checking out the following teams drives (in no specific order); 1345, 217, 1114, 254/968, 330, 233, 45, 190, 234, 179, 1889, 177, 121 (excellent example of KOP frame use), 116, 1717, 125 and a whole bunch more… All are good examples of 6 wheel drives (of varying complexity), which I believe to be the most resource effective drive for most FIRST applications.

Shameless plug, but 294’s drive is pretty good this year.

Hmm.

Well, when I get some free time (probably after Nationals), I’ll whip out some (very) basic 6 wheel chassis designs involving both omniwheels and rocker chassis (but not at the same time), and perhaps a few more 4 wheel Chassis designs (or rather, more detailed ones)

Also, any and all Diagrams that I have/will make are yours to manipulate, use, or otherwise take in any fashion, as long as you don’t claim them as yours. I do not require that you say that I made them, however. Just say “I got them from a buddy” or “I got them from somebody on CD” or something to that effect.

Alright, well, I can’t edit my original post, So I’ll have to update it some other way.

Here’s a new diagram I made. It’s mostly self explanatory.

http://www.lotechdesigns.com/host/images/2501Chassisintersection.jpg

Just want to clear this up.

You can have to much traction (depending on the gearing of your drive train) All teams should (after the drive train is complete) should bring the robot up to proper wieght (using paper wieghts or what not to load down the robot) and then push against the wall. If your wheels begin to slip you are good. However if your wheels do not move, thus forceing the motors to stall a course of action must be taken to keep the motors from stalling.

There are a number of ways, the most common is reducing the coefficent of friction of the wheels.

we used the KOP chassis and the 6-wheel drive(middle lowered and the KOP wheels) during our rookie year. It worked VERY well for us. We used the banebot tranny with a supporting bearing at the end of the shaft. We had zero problems with it. Turning was easy (sometimes too easy). That’s something i’m having the programmers work on this summer.

Here are a few pics of the drive train before it was finished. It’s rough, but you get the general idea. If you want to see it in action, check out the VCU match #50 on SOAP. We did well until be broke the arm :frowning:

Hmm. KOP Wheels + Rocker Chassis… not something I had thought of before.

Kudos to you and your team for coming up with that.

I’m thinking I might just go ahead and find some graph paper and draw up the other diagrams then scan them – it’d probably be less legible that way (photoshop could fix that) but more precise.

that’s what comes in the kit…although you need to get two more wheels…so it’s not something secret :slight_smile:

Remember, I’ve stated that I’ve not built a KOP Chassis before. '05 was 675’s last year of doing so (for Triple Play). For both Aim High '06 and this year, we’ve taken aluminium stock (3x0.25 in) and turned it into a chassis.

We used fiberglass for the 1726 chassis last year and this year. But we did spend some time looking at the kit chassis, and noticed the lowered center holes.

The fiberglass works very well as a chassis material…it’s tough, easy to work with if you don’t have a machine shop, and the shapes we use work well for robots. The side rails which hold the wheels are I beams turned on their side, the ends are channel, and it’s all held together with bolted on aluminum angles (similar to those you’ve shown above). We learned about this material from our mentor team 842. We use a different frame layout than they do, though.

wow what a thread to find! Being on a less experienced team myself and being one of the few that wish to prototype sweet design, we have decided to try a crab drive (to keep up with you vetren-like). I don’t mean to be rude or anything and I understand this is a thread to posted ideas, not ask questions but is there any chance someone could ethier post or PM me on the simpliest way to start a crab drive. And before you ask I do understand it is complicated to do, thats why I am starting now :smiley:

Feel free to ask them to post it here!

This isn’t just a thread for more experienced teams to help out less experienced ones, but a place where less experienced teams can come to ask for help. Remember, we’re all gracious professionals here.

That being said, I don’t know anything about Crab or Swerve drive. Sorry.

EDIT:

Oh, but I did find this article on Wiki Books…

http://en.wikibooks.org/wiki/Robotics:_Design_Basics:_Physical_Design

Enjoy. =]

Yup, you’re right.

…Sometimes you get so focused on something (traction affecting turning), you forget basic facts…

sorry about that.

…Woah. Fiberglass?

I’ll have to mess around with that after I get a job. I’d love to see it’s potential.