The Chassis Project.

Technical Technical Discussion
1

I’ve been asked to recommend or design a chassis that will be used by a new FIRST team. While I understand that FIRST made a great step forward in providing such a chassis in last season’s kit of parts, and that they will, presumably, be including it again this season, I think we can all agree that there is certainly room for improvement. For teams without access to engineering support, materials, or machinery, it’s an excellent solution. For teams without access to engineering support, but without other restriction there is freelanceFIRST and me.

So, I’m starting this thread as something of a survey. With the experience you each have now, what features would you like to have seen in a ready-made chassis design?

To start, I’ll list some of my concerns and criteria.

  • The chassis should be light. What is an acceptable chassis weight?
  • The chassis should be able to hold its own in a pushing match. Traditionally, four-wheel drive is the easiest way to attain reasonably tractive effort.
  • The chassis should be able to turn. Castors make turning easy but offer less precise control and less traction, potentially.
  • There should be ample room for attachments. Arms, baskets, booms and other devices should have easy mounting points. Electronics, battery, and wiring should have adequate space, be insulated, and be neat.
  • Parts should be common and easily cut or machined with basic tools. Repairing damage should be easy because of common parts, logical assembly, and similar accessibility.
  • The chassis needs to be stylin’ ;). I don’t do ugly.

So, what would you add or take away as a consideration? What would you change.

I’m looking for as much feedback as possible with the hope that I can develop an adequate concept that could be used as is or adapted by teams to suit their needs.

Thanks.

2

ok, i’ll give it a shot

  • weight
    as light as possible without sacrificing strength. it depends on the material too. 1/8 aluminum weighs the same as 1/4 inch plywood (i think, correct me if i’m wrong), but you would use more plywood than you would aluminum. however, you can only cut so many holes in a skeletal aluminum frame, while the number of holes you can put in a plywood chassis is usually limited only by one’s endurance and the battery of the hand drill.

  • traction
    you are correct in saying that 4-wheel drive is the easiest way to get reasonable traction - however, there is a tradeoff. the more traction you get, the harder it is to turn, and so the lower your motors must be geared in order to overpower the side traction. this problem can be solved by any number of little tricks (omniwheels, retractable casters or skids), but each of these just adds more complexity and, seeing that they lack mounds of experience, could possibly hinder a rookie team more than help.

  • turning
    manueverability is, IMHO, the most important quality of a robot. If you can supplement manueverability with lots of speed, then your golden. If you can be everywhere instantly, then you can and will dominate, no matter what the future game turns out to be (remeber 25, anyone?). as always, though its a tradeoff between manueverability, traction, and speed. It is also easy to have way to much manueverability too, though. I would at all costs stay away from permanent casters as a way of boosting manueverability. I know from my time in the scouting service that fixed casters on a robot is a sure sign of inexperience (I have heard of some teams who were pretty good anyways, but i’ve never personally seen one). Casters make it precision driving difficult, especially at high speeds, and basically cut traction to zero. all an opponent has to do is hit you from the side and your bot will spin like a top. my personal favorite design is the 6 wheel design, with the 2 center wheels lowered just a bit. it basically the traction of a 4 wheel design and gives it loads of extra turning power.

  • attachment space
    if your chassis has a solid bottom, there is rarely a lack of space for mounting things. plywood works great (you can paint it, you know, and it actually looks kind of cool).

  • repairability
    everybody disses plywood as a building material, but they really shouldn’t. It is extremely forgiving, will flex under imact, quite strong, and extremely easy to handle and repair. despite what naysayers might think, aluminum will bend long before plywood will break, and when aluminum bends it stays bent, while playwood will just spring back into place.

  • looks
    some people say that its ugly, and it certainly is if you leave it unpainted, but its easy to give a wooden chassis it a coat or two of paint before you start screwing things on to it. (although i didn’t personally agree with my teams color scheme of choice this year, i can’t tell you how many people asked me what our robot was made of - it was hard to tell that it was wood unless you really got close. look for yourself)

wow, that was a lot more than i intended to write, but i guess what is done is done.

3

I see the “ready-made chassis” as the beginning to the downfall of a part of FIRST we all love, the unique designs and technological advancements that come from the building of the machine.

See the learning process is in the making and if it is “ready-made” there will be little learning involved.

As long as there has been FIRST, there have been rookies. How did they do it a few years ago without a “ready-made chassis”. Have the games gotten harder, no just different. So, if we all did it back then, there is no reason today’s rookies can’t. If anything, they have more of an advantage already without the “ready made” chassis simply because of advancements over the years.

I am all in favor of ready made things like motor mounts because everyone has the same motors, and everyone is going to mount them. But I think chassis and gearboxes ready-made for teams is going too far. That is really where a lot of the design innovation comes from.

And as for the plywood, I guess you would call me one of the “dissers.” We bent 30x30 bosch extrusion. When I say bent, I mean bent. Plywood would have had no chance in the matches we played.

4

Tank Drive

5

I just want to put my 2 cents in here. I have been involved with two rookie teams as a mentor. In one case there was nothing but the kit of parts. In the second case there was a “ready-made” or prefab frame kit. In both cases the teams adjusted well. In the first case the students got going and made it go on their own. Total time was about 3 weeks. They learned a tremendous amount about steering, power, balance and weight distribution. In the second case total time was about 3 hours. The team learned about steering, power, balance and weight distribution. The difference was there was no initial phase of head scratching and do-overs. For the new teams it would be a great help. Maybe the idea could be that “rookie” or inexperienced teams could request this as an add-on to the kit. The more experienced teams, or those that have access to machinists, metal lathes and aluminum welders could fore-go this “luxury” and blaze new trails in design and fabrication. In both cases the teams would get what they want/need. In the first case the new teams would get past the initial frustration and build a functioning robot. The more experienced teams could get on with their building and prototyping without the problems of the extra stuff cluttering up the work area.
I realize it’s more than 2 cents worth but it is a different perspective.
Go FIRST!:slight_smile:

6

And as for the plywood, I guess you would call me one of the “dissers.” We bent 30x30 bosch extrusion. When I say bent, I mean bent. Plywood would have had no chance in the matches we played.

You bent it??? I do not even think people who build Battlebots with this stuff had problems with it bending. The only problem my team had with this stuff is that some of the pieces got wacked out of place.

I am all in favor of ready made things like motor mounts because everyone has the same motors, and everyone is going to mount them. But I think chassis and gearboxes ready-made for teams is going too far. That is really where a lot of the design innovation comes from.

The chassis never came ready-made. In fact I did not even know there was a complete chasis. I know those two huge pieces of extrusion were for the gearboxes to be mounted but what other pieces completed it.

7

*Originally posted by Adam Y. *
You bent it??? I do not even think people who build Battlebots with this stuff had problems with it bending. The only problem my team had with this stuff is that some of the pieces got wacked out of place.

In reality, everything is bendable. If another robot hit you head on, both robots at top speed, with a corner, yes even 30x30 bosch will bend. My team used 20X20 bosch, and we only bent our back bar durning regionals. 30x30 bosch is ok, but it’s too heavy and too strong. You don’t need it to be that strong. 3030 also has premade right angle joints for sale, so all you do is cut it to the right length and use it like legos. The bad part is you can’t triangulate, first because you would have to make your own joints, second it would be really heavy. With 2020 bosch, you can triangulate all you want, but you have to make all the joints because they don’t sell any for 2020. My team, just on the bottom layer of our frame, had about 20 plates holding the thing together with about 500 nuts and bolts. In the end, our entire frame was about 30 lbs and indestructible. We did have to tighten those 500 bolts every match though. My final word: if you have a mill, go for 2020. If you don’t, go for 3030.

As for drive trains, 4 wheel tank drive is the easiest to do for the performance it gives, but upgrading to treds isn’t that hard, and it gives a lot of traction and other advantages. With treds, you have constant contact to the ground with a large area. You can get some pretty high traction soft rubber treds, and since they’re treds, they get ripped up less, so you can get even softer rubber than on wheels.

But hey, it’s your robot, you decide. I’m only reccomending stuff.

8

We have access to 5 mills (and 2 NC’s) and go with 30x30 still. I was pretty sure they still made connecting hardware for 20x20. Anyway, we make out own conecting pieces (out of 1/4" plate BTW) because the Bosch ones are just too expensive. We use all right angles and have no problem. We could do a custom angle if we wanted though, we would just need new connector plates of have it welded. However, we broke a steel wels, I don’t think we’ll try our luck at aluminum (the weld was good but the force on it was much)

9

*Originally posted by Solace *
**ok, i’ll give it a shot

[list]

aluminum will bend long before plywood will break, and when aluminum bends it stays bent, while playwood will just spring back into place.

**
acctually, thats not entirelly true, when the aluminum bent on our robot, we unbent it, just fine, i mean, after 4 competitions, we replaced it, but it still held up for that long

10

I am one of those people that will never be convinced that plywood is an acceptable building material for any teams that have an alternative. Plywood is not a very sturdy base to be mouting big, heavy objects on, such as arms, which would be moving back and forth, while flexing the plywood. If you have metal, use it! It is really not that hard to work with and offers a much stronger and more secure surface for mounting stuff.

My list:

Weight: using 10/10 from 80/20 last year we made a frame that weighed less than 25 lbs. It was a simple 30"x36" with a “top deck” that was 4" taller than the bottom on it. With all motors, electronics, and battery mounted, we had a working robot that weighed less than 80 lbs. You hear about these teams that are bending massive pieces of aluminum extrusion. If you look deeper, you will see that these are teams who have tons of speed, but perhaps not so much power. They will ram into an opponent at top speed to try and dislodge them, causing bends to occur. We built a robot that had the power to move other robots just by pushing them, no ramming involved, which meant that 10/10 worked very well for us, and kept weight low (In fact, too low).

I would have to say that a four wheel drive chassis is definitely the easiest to construct for the benefits you gain. However, if you are geared for high speed, and you arent using the terrible wheels FIRST supplies, you may have trouble turning. In this case, I also like the 6 wheel design with the middle set of wheels a tad bit lower than the other 4, giving the robot a pivot to turn about. If omni wheels, or the really slick wheelchair wheels are used in this situation, turning becomes even easier.

For plenty of attachment points, just make a basic rectangular chassis. If a team wasnt going to have many attachments, you could just build onto the bottom. If you wanted to add many/larger mechanisms, you could place the motors, wiring, battery, and other random stuff in the very bottom to get a nice low center of gravity, and then make a second level of the frame on top where the mechanism is added.

for ease of use 80/20 and Bosch extrusion are tops. All you have to do is cut it to length, get the fittings, slide a nut in the groove, put a fitting over it, and screw it all in. After cutting and finishing the pieces for our frame, we put it together in less than 3 hours, and could have done it even faster had we been more familiar with 80/20. The only problem is that the connectors and such for extrusion start to be VERY expensive when you buy a lot. You can make your own, but that requires machining abilities.

Stylin’… Cant help ya there, our team has never made anything that could remotely be called styling in recent years :slight_smile:

Oh, and the whole rookies getting a premade chassis thing… they arent getting a premade chassis, they are getting two pieces that can be used to form a chassis… M is not trying to give all rookie teams a prebuilt frame. The point is to get some sort of guide that a rookie team, or any team for that matter, can refer to if they arent sure how to go about making a well built, reliable base for their robot. This guide is essentially the same as any white paper that has been written. How many people have been inspired by such white papers as the Technokat’s series of papers on their successive transmissions, and go on to create a smaller, better, lighter version? Noone calls those papers a stem of creativity, so how is this any different? Some of the greatest ideas we ever get come from seeing something else that someone has done and improving on it.

Just because older teams didnt have the luxury of getting some of the “reduced assembly” stuff from FIRST doesnt mean that noone else should get it. What anyone who says this is saying is essentially “If I cant have xxxxxxx, then noone else can have it either”, and is extremely unprofessional. Its not like teams are getting a crate with a prebuilt transmission and frame in it. They still have to figure out how to assemble all this stuff and make it work with their robot.

Is the creation of a guide to chassis building going to lessen creativity? No. Teams will do whatever they please with it, and I guarentee we will not see 900 cookie cutter robots out on the fields in March. How many teams did you all see last year that used he included frame members and had robots that looked the same? Few, if any. These things FIRST are providing are options to facilitate the building of a working robot, not required parts.

Sorry that was so long,

Cory

11

At first, thinking about a premade chassis, I might think that you would see a few hundred of the exact same robot.

But then I think to myself, maybe for some rookie teams, that wouldn’t be so bad. If it was your first or second year, I would hate to come out and have an immobile robot for every match while you see everyone else out there screaming around the field. It would go a long way to getting every team excited about the robot aspect if they could start out in the first few weeks with a mobile robot that has at least some amount of robustness to it.

My Suggestion to new teams. Us the 80/20 or Bosch rail, we used it for the past few years and it proves extremely easy to put together and take apart, it can be easily built by students, it can be painted. and it is strong. The main disadvantages are weight and cost.

What would be the next level after this? Prototype with the 80/20 and then when you have everything the way you want it, have the chassis remade out of aluminum tube welded together. That saves the weight and you have a bunch of 80/20 for next year. (Looked like 111 did this with their prototype that I saw at Buckeye) As for drive-train, we have done it all, 2WD,4WD, Swerve & Tank. I think swerve is out for a rookie team, and tank probably should be to because of cost so that leaves 4WD with chain drive between the axles to be the starting point for most teams. An 80/20 rail chassis with the First motor mounts to 4WD would be an awesome starting point for teams that haven’t got the resources for something really crazy, it would at the minimum keep them in the game.

my.02

12

In reality, everything is bendable. If another robot hit you head on, both robots at top speed, with a corner, yes even 30x30 bosch will bend. My team used 20X20 bosch, and we only bent our back bar durning regionals. 30x30 bosch is ok, but it’s too heavy and too strong. You don’t need it to be that strong

I get it now. You actually did not bend the metal bar but the corner got pushed back where the gusset is. It really is not that hard to do that.

I am one of those people that will never be convinced that plywood is an acceptable building material for any teams that have an alternative. Plywood is not a very sturdy base to be mouting big, heavy objects on, such as arms, which would be moving back and forth, while flexing the plywood. If you have metal, use it! It is really not that hard to work with and offers a much stronger and more secure surface for mounting stuff.

Plywood is not the building material we are talking about. I agree with you plywood is an really stupid idea. Something more thicker and more dense will work much better like my team used. Wood is a very good construction material for a robot as long as you are not going to send it into Battlebots.

13

Thank you everyone who’s replied. I am reading with great interest.

I’m going to attach a screenshot of my first “draft,” made before receiving your feedback. Given what you’ve each said, I’m probably going to scale back a bit, potentially switching to the use of extrusion.

I was hesitant to use extrusion in the first place because it failed where I used it for the 2002 season. But, as someone mentioned, that robot was involved in a series of high speed collisions – ultimately resulting in many sheared pieces of Bosch 30x30. Additionally, I find extrusion to be expensive and limiting, and am not thrilled with its fastening systems.

The concept that drove the creation of this original chassis was a lot of what I outlined above. There’s extra surface area on the back wheels for added potential traction across carpet and other uneven surfaces. If the playing field is, for example, entirely HDPE, the extra wheelsets can be removed. The lexan platform along the bottom serves as an electrically isolated electronics mounting surface. It also adds some structural rigidity across the beams running the length of the chassis. The aluminum top shelf is an open space for mounting additional mechanisms and adds more structural strength. It can be holed and milled out as necessary to reduce weight, etc.

The channel that comprises the main part of the structure is designed to be pseudo-modular, assuming the machining capability to add the shown holes. It will feature uniform hole patterns to accept standard bearing plates. This will allow teams to add axles whereever they’d like along the robot’s length among other things.

The front wheels are non-existant. I had intended to specify omniwheels and 4WD chain drive, but am now considering leaving it as is and letting teams decide on a solution themselves, whether it be omniwheels, additional Bead-Lok wheels, or castors. Of course, if the machining capability is available, omniwheels would be my preferred choice.

Feedback is appreciated. Thanks.

Edit: Of course, there’s every chance I might make more than one design in the end, too.

simple_chassis.jpg


simple_chassis.jpg

14

Adam Y,

Nothing got “pushed back.” We used custom 1/4" T6 plates that joined the extrusion by the top of the frame members, not the insides and we put massive torque (and lock-tite) on all fasteners. All of the corners on the machine got smashed in where the extruxion became about half it’s normal width and there was no longer an channel there.

Additionally, what I was really talking about. Was the front and rear frame rails on our robot of 30x30 extrusion. If you took it off and set it on a level plane, the center of the piece would be about an inch above that plane. It got bent. Not to mention the severely warped channel.

Cory,

The Skyway 9x2 Bead-Lok wheels really do suck UNTIL you turn them down on a lathe to have a flat profile. This gives really good traction (even though they are smooth) on carpet while still allowing enough sliding for 4wd tank style turning.

Oh, and the T-nut can be rotated in from anywhere in the channel, it does not have to be slid in from the end.

M,

As for limitations to extrusion, I haven’t found anything you can’t do with it that you can do with square tube of the same size. But there are many things you can do with extrusion that you can’t do with equivalent square tube.

The T-nuts are fairly expensive at about $0.50 apiece. The fastening gussets and the such are absurd at over $8 apiece for some.

I am totally for extrusion (30x30) but fastener pieces need to be custom made for cheaper.

Now, another interesting concept I saw at the SCRRF scrimmage was a team that used the FIRST provided chassis, but added two more helical gearboxes at the other end of the robot and had a shaft running down each side of the chassis linking the two for 4 wheel tank drive. That was really cool.

A couple pictures of the beginnings of a basic 4wd universal tank setup chassis made of 30x30 extrusion. Is uses 1/4" plate “triangles” to keep together all the joints. The upper deck is for arms. Gearbox and electronics mounting plates are not shown. Chains drive the rear wheels and chains link the front and rear wheels. There are four chains total after the output of the gearboxes. It weighs 18-19 lbs in the first picture.

BTW, that top plate was made with a drill and circular saw with a wood blade and a file.

http://www.team696.org/2003pics/team696/128.jpg
http://www.team696.org/2003pics/team696/133.jpg

15

I forgot one thing. The 1/4" joining plates my team uses are a bit heavy (and excessively strong). Next year we will be going with 3/16" but doubling up (joining top and bottom) on the four outer corners.

You all should also take a look at this thread

16

The Skyway 9x2 Bead-Lok wheels really do suck UNTIL you turn them down on a lathe to have a flat profile. This gives really good traction (even though they are smooth) on carpet while still allowing enough sliding for 4wd tank style turning.

I really would not recommend really modifying those tires much. It looks like they have a tendency to disintegrate a lot and really do not stop if they start too.

17

*Originally posted by sanddrag *
Now, another interesting concept I saw at the SCRRF scrimmage was a team that used the FIRST provided chassis, but added two more helical gearboxes at the other end of the robot and had a shaft running down each side of the chassis linking the two for 4 wheel tank drive. That was really cool.

That was Team 1135, Whitney High Shmoebotics. It took them a while to get going but they did well in the end.

18

Tank drive is the way to go, as a rookie team last year we can attest that it did wonders for us.

a) It is easy to assemble. Just attach the drill motor assemblies to a frame, add a sprocket and a few chains… voila!

b) It provides both power and precise control. Lets face it… CASTERS SUCK! How many teams last year with casters didn’t have trouble making it up the ramp? Not many :frowning:

c) Easy to add upgraded functionality. We just happened to add a servo gear shifter so we could be fast, or powerful.

Here is a pic of what our base looked like last year. Very simple square base with a two beams running in the center to hold the drills.

naked_bot.jpg
naked_bot.jpg1152×864 88.2 KB



naked_bot.jpg
naked_bot.jpg1152×864 88.2 KB

19

*Originally posted by Adam Y. *
**I really would not recommend really modifying those tires much. It looks like they have a tendency to disintegrate a lot and really do not stop if they start too. **

In my limited experience with those tires, I’ve found them to be resilient. Obviously, the wire mesh of the 2003 game was a special circumstance that was destructive to all sorts of tires.

Turned down on a lathe, those wheels seem to provide adequate, accesible traction. Also, they come in the kit and they’re keyed.

20

One more reason to go with extrusion is than you can slide the pillow blocks for chain tension adjustment. We would have been in major trouble without that. Another substitute is an adjustable or sring loaded idler sprocket to use as a tensioner if extrusion is not used.

You can also do the same thing with square tube but you woyuld have to have slots milled.