Building the Structure of the Robot?
 

Hello all.

I believe many beginner teams (including us) use aluminum extrusion such as 80/20 or T-Slots, as it's very simple to use, intuitive, and cost effective.

However, as we progress throughout our FIRST life, we're seeing more and more cons to extrusion.

Extrusion is considerably heavy in comparison to alternatives such as custom machined lightweight aluminum.

It's also not nearly as flexible for construction, given it's only a straight 1"x1" rod really.

My question to you all is, how would you suggest/think (or if you've personally started by using 80/20) wane off the use of extrusion and transition onto more advanced structural materials?

To add to that, in what way can you accomplish this given low-moderate funding ($7,000/season), and pretty much no specialty tools (unless the tool in question is affordable). I'm saying we only have a Miter Saw, Drill Press, and Bench Grinder. Scroll Saw & Band Saw are not out of the question.

Thanks!

Re: Building the Structure of the Robot?
 

2815's past two robots have largely been made of 1" square tubing (1/16" wall) and flat and angle to gusset it. If you don't have the resources for a full-on sheetmetal design, this is hard to top.

Re: Building the Structure of the Robot?
 

Look for local sources for affordable pricing of 1x1 and 1x2 aluminum tubing. If you can find a local source who is willing to donate material, that is even better. If not, you can find great deals at www.onlinemetals.com

Tubing can be connected using gussets or by welding. A great example of the gusset method is Gus team 228 (no pun intended)

http://www.team228.org/media/pictures/view/4896

Check out the way they connect their corners using what I think is 1/16" thick 6061 aluminum sheet. I personally love this method because it is easy for teams to do without much (if any) precision machining. You can CAD up a design for your brackets, print out 1:1 scale engineering drawings, and just use the print out as templates for pieces to cut on the bandsaw and make a hole pattern with a drill press. A lot of teams use this method along with rivets but you can also use screws and bolts. Don't use something bulky like 1/4-20 though. Rivets are recommended.

Hope that helps!

Re: Building the Structure of the Robot?
 

I have some thoughts of my own on the subject, but I think I'll leave you with this instead (from CD user BJC from Team 33)

Re: Building the Structure of the Robot?
 

If you can get a welding sponsor welded tube is one of the easiest ways to go (at least for my team welding is faster than rivets and gussets). We have built robots on budgets of ~$2000-$3500 for everything (spares, prototyping, tools, and so on) using the tube. If you use 1" square 1/16" wall aluminum tubing you can build your robot structurally in much the same way you would with the extrusions but with a much lower weight and cost. Considering our machining resources usually are about the same as what you described transitioning to welded frames of thin-wall tube should be relatively easy (if you can get a welding sponsor, otherwise do gussets and rivets/bolts).

Re: Building the Structure of the Robot?
 

I also think 973 does this well. Also one thing to point out for lower resource teams is that if you get the gussets cut accurately(with a waterjet sponsor namely, which are really easy to find) you can match drill your whole robot frame

Re: Building the Structure of the Robot?
 

Yes, our entire 2012 robot (and various parts of older robots) were held together with gussets, not welding.

We made our own cnc router, so the ability to cut arbitrary gussets in house on zero notice was helpful.

Your suggestion of having a sponsor cut gussets is totally valid, a waterjet or laser will fly through 1/16" plate.

Re: Building the Structure of the Robot?
 

Just to add in, even if you don't have a waterjetting/CNC/laser cutting sponsor, you still can do the work yourself at a local hackerspace. We're planning on cutting our gussets and bellypan with a hackerspace's waterjet, rather than milling them in our own shop (which took forever).

In our experience, gussets are a great way to save weight (because you can use thinner wall tubing), but rigidity is compromised. I'm hoping to in house weld our drive base again for the sake of rigidity, but I'd concider doing riveted gussets on our superstructure again, even though our tower was a but wobbly last year.

As a final note, don't have freshman make all your frame pieces in a drill press and expect all the holes to line up. Take the time to CNC the frame members, to have experienced members do the drilling, or be prepared to have these same freshmen do a lot of filling...

Re: Building the Structure of the Robot?
 

We recently switched over to a box tube + gusset bolt together construction method: http://www.chiefdelphi.com/media/photos/38147

It's a great way to put together a stiff and strong frame that is also light, our entire frame was under 20lbs.
Using rivets instead of bolts will help a lot with weight as well.

Re: Building the Structure of the Robot?
 

I assume you're referring to Techshop.

Unless you've managed to get them to sponsor your team, or your baseplate is vastly simpler than ours, it would be extremely expensive to jet one there. Probably on the order of $500-750.

Re: Building the Structure of the Robot?
 

The current price is 2$ a minute of pump time for members at SJ. 1323's bellypan last year was about 45-50 minutes a piece on a similar sized waterjet, so it would probably come out to about $200 per. Not super cheap, but I've had laser shops charge more than that just for setup even with a discount.

Re: Building the Structure of the Robot?
 

Ditch the bolts and as much 1/8" as possible, we've never had a frame over 10 lbs ;)

EDIT: I'm referring to drive frame and superstructure frame separately.

Re: Building the Structure of the Robot?
 

The tower is all 1/16th but the drive mostly 1/8th. We are planning on switching over to 10# hardware and 5/16 rivets. :cool:

Re: Building the Structure of the Robot?
 

3/16" Right :P

-RC

Re: Building the Structure of the Robot?
 

Maybe...... :p

Re: Building the Structure of the Robot?
 

5/32" maybe? that's what we usually use

If you can't manage to get aluminum welded, you might try using steel, it's easy to weld....there's probably someone on your team who can do it. We got a new freshman to weld our steel frame together this year.

Re: Building the Structure of the Robot?
 

I couldn't agree more with Billfred on this one. The switch over to 1/16th wall box will reduce costs and weight as long as you spend some time designing your layout out instead of relying on sliding parts.

This should be an easy switch as it is still 1x1 construction. Some friends on 3302 will pre-drill their square tubing with holes on the center line every 1" (or maybe 3/4 or 1/2"). The small holes allow the box material to be easy construction similar to the "kit" extrusions. this allows allows for some kids that are not interested in design to start making stuff right away while the others design to this constraint. this is a neat option, but you do loose some flexibility of dimension. 1x1 is also really common and in a pinch can be found at most Home Depots.

If weight is a higher priority, switching to round tubing for upper structures is a great weight savings, but significantly harder to work with. Round versus square is a 25% savings for the same wall thickness. Round also has lots more wall thicknesses and material compounds to choose from. This too can add to weight savings. Make sure you get some good pictures of how 469 builds their robots. They are (IMO) masters of the thin wall no weld construction.

***********************

All this being said, 3 of my favorite and most competitive robots last year used 80/20 or star bar (25, 341, and 1023). These 3 were able to do it all. They were good shooters, good at collecting, and able to do the bridge well.

Re: Building the Structure of the Robot?
 

Or you do what we did: craft the shooter on the fly, realize you had some parts wrong, drill out the rivets, place them in the right place, re-rivet...

...and then putty over the holes before painting. ;)

Re: Building the Structure of the Robot?
 

So..

Our time doesn't really want to weld, nor do we want to allocate resources to buy a good welder.

As far as gussets go, where would we get them/how would we make them? Is it just any shape plate with bolts..?

Re: Building the Structure of the Robot?
 

2815's gussets have been 1/16" flat and angle aluminum from the hardware section of Lowe's, riveted using 1/8" aluminum rivets (usually 2-4 on each end). Cheap, easy to install, easy to repair.

I'd love to see us get to a bent sheetmetal gusset like some of the ones shared here...but the results are suitable for competition.

Re: Building the Structure of the Robot?
 

The best way imo is to make them in house with a cnc router or with a sponsor who has a waterjet. Either way it will be very little time investmet. Just make sure to make the gussets as thin as possible(1/16 is likely fine fo many applications). By making the gussets thinner you don't have to lighten them. In addition because you don't have to run a lightening pattern and because the parts are thinner they can be machined extremely quickly.

Re: Building the Structure of the Robot?
 

Uh, we have access to neither of those.

Re: Building the Structure of the Robot?
 

We use a lot of 1/16th box extrusion with gussets and ESTO connectors, one of the best investments we have made as far as hand tools go is the ratcheting rivet gun.

Re: Building the Structure of the Robot?
 

Cutting triangles out of square pieces is very easy with a bandsaw. That's how we made all the gussets for our after-season stinger addition. It also has a huge benefit: it's fast and you don't have to wait for anyone. If you want to match drill them - make all your gussets in advance, clamp them together, then drill through them in one shot with a drill press. Your gussets may not be be exactly the same, but the hole pattern will be.

Re: Building the Structure of the Robot?
 

Do you have access to a hacksaw, metal file, and a drill?

Congratulations, you now have all the tools to make gussets the way 2815 does. :)

There's not a lot of magic to it; if you're worried, get a stick of 1" square tubing and a short (say, 2') length of 1/16"-wall, 1" angle aluminum. Make a box out of the tubing, riveting it with the angle used as gussets. Kick it around the shop a bit and observe its properties in the context of what a robot would encounter (usually with bumpers, remember). You'll be pleased, most likely.

Re: Building the Structure of the Robot?
 

Drill all the rivet holes in the gusset, then clamp the gusset to the bigger parts. Then drill the first hole all the way thru, and pop a rivet in it. Continue drilling and riveting one hole at a time, until there are two rivets in all the parts. then you can drill all the rest of the holes, and put in the rest of the rivets.

We've been doing this for building our "structures" for years....it's time consuming, but cheap and easy and doesn't take much equipment.

Re: Building the Structure of the Robot?
 

Thanks for the quick replies guys, I'll forward all of this to our Mechanical Captain.

While I'm one of the more mechanically savvy guys on the team, the final call is to the captain :)

Also, another question/concern. With 80/20, we had the flexibility of just putting a drivetrain together with the parts, and then when we needed to change something or add something, it was a matter off adding a connector or taking this off or adding this. Very simple, with little modification.

I imagine that by using aluminum sq tubing w/ gussets and all, we'll have to plan our drivetrain out much more thoroughly than we did with 80/20 (unless we plan to have a lot of random holes from taking gussets off and on xD)

Is this correct, and how do you cope w/ this?

Re: Building the Structure of the Robot?
 

We've been known to add a few extra holes to our robot through bad planning. In 2011, we even took a piece of our arm structure, rotated it around, and drilled a new hole for lack of another piece.

It all comes down to what kinds of forces your pieces will see and how picky you are about appearance. When we botched parts on this year's robot, we took the metal that was still usable in-place and redrilled new holes and placed putty into the holes left behind to smooth it out. Metal that was too weakened from the holes for our liking was cut down and reused in other places when possible. Very few pieces were junked entirely.

But, of course, you can always plan it out and avoid all this. :)

Re: Building the Structure of the Robot?
 

CAD, CAD, CAD. As you can tell I'm a CADder. In all seriousness though, the best way to get better really quickly is to start CADding. It allows you to build robots quicker and cheaper. Also in response to not having access to a waterjet, I reccomend that you find a waterjetting sponsor. Any 2d work is really easy to find a sponsor willing to do work for you, because the time commitment is a lot less for them that of a full blown machine shop.

Re: Building the Structure of the Robot?
 

No, but you have access to the most powerful machine tool of all; a telephone.

We had zero machining sponsors, now we've probably had over 30 unique shops run parts for us; All we did was ask.

Re: Building the Structure of the Robot?
 

CAD

As Adam said, if you can't figure out how to make this stuff with manual tools (which you really should be able to) then get sponsors and learn how to CAD.

Call these guys. http://www.metalsupplyfl.com/

They are in your town, they waterjet cut, weld, and can do a lot of other fabrication.

Simple google searches and a phone is all you need to find sponsors. Just get up and do it.

Again though, I see no reason why you are not capable of making such parts with your current machines.

Re: Building the Structure of the Robot?
 

OK.

Say we obtain a sponsor that can do various fabrication, waterjet cutting, CNC machining, etc.

At that point I have another question. Our team organization is based around the fact that everyone works with what we have and is contributing to an assembly.

With most of the design/assembly itself being done in CAD and then sent off, what does the mechanical team do? In essence, if the drivetrain is being made in CAD and made my a sponsor, what does the actual mechanical team do while the design team is CADing it and sending it off?

Re: Building the Structure of the Robot?
 

Using what you currently have effectively is great, 973 prides itself on it, but that doesn't mean you can not strive to do more. No one suggests that you should get the entire robot outsourced (it's pretty much impossible to do so anyway), but intelligently using a sponsor or two can really improve your products.

Are you saying that CADing designs somehow takes away from the work of mechanical people? I'm unsure how to read that.

The mechanical people should be involved with designing the robot in CAD, and if the guys that just love putting stuff together refuse to be a part of the design process, they can spend time prototoyping over and over to end up with a killer manipulator.

Re: Building the Structure of the Robot?
 

Nobody ever said you NEED to use these places. The fact is that these resources are out there for when your team decides to transition into it and create more advanced designs that require precision fabrication. I posted this to simply show you that machining sponsors are easy to get.

Even if you do send parts out, you don't have to send out everything to get cut by a sponsor. Getting brackets cut by a sponsor or in-house is up to you.

Re: Building the Structure of the Robot?
 

If you still plan to do in house work(as Adam said it can be very hard to not do any in house work) then they can start machining things. However there should still pe a primary emphasis on CAD and prototyping during the first few weeks of build.

Re: Building the Structure of the Robot?
 

We do all fabrication except for welding in house. If you can get a welding sponsor you can do very well using COTS parts and simple brackets for everything else. Even if you don't have welding you can get by with simple gusset brackets made on a bandsaw and aluminum angle. All you really need is a drill press (or hand drill) and band saw (or jigsaw). We do almost all our work with hand-held power tools, and we do not have access to a lathe, mill, or router at our build site so we do not design to use them.

My point is that you don't need to use expensive computer controlled equipment to build a fully custom and competitive robot, we used to build out of a garage and we have always done custom. You should always do CAD no matter how you plan on having parts made.

Re: Building the Structure of the Robot?
 

You really shouldn't write off box extrusion simply because it's less "flexible" - there's a lot that you can use it for. In fact, its lack of flexibility makes it a prime candidate for building robot frames. Pretty much every robot my team has built uses box extrusion as the primary structural member. We've only recently started using gussets, but in the past we used both 80-20 triangles and L brackets to join pieces together. For flexibility, we just cut it up - a saw and drill are all you really need to manipulate it however you want. It just requires more caution and imagination during the actual fabrication process.

Besides, there isn't really such a thing as "more advanced structural materials". Most people use some form of extrusion, be it 1"x1" or 1"x3". There are teams like 148 and 217, yes, that have practically completely eschewed the use of materials like extrusion because their resources afford them that advantage, but that doesn't mean that sheet metal is better. You could even argue that a lot of the work they do with sheet metal is spent making it mimick extrusion - just look at their arm last year.

The point, in short, is that you're limited more by how you use a material than what that material is. Simbotics went to 2011 Galileo semis with a hockey stick.

Re: Building the Structure of the Robot?
 

[quote=slijin;1194467]
Besides, there isn't really such a thing as "more advanced structural materials". Most people use some form of extrusion, be it 1"x1" or 1"x3". There are teams like 148 and 217, yes, that have practically completely eschewed the use of materials like extrusion because their resources afford them that advantage, but that doesn't mean that sheet metal is better. You could even argue that a lot of the work they do with sheet metal is spent making it mimick extrusion - just look at their arm last year.
[quote]

Do you mean 1" x 2" extrusion? I don't know a lot of teams that use 1" x 3" as their primary frame extrusion.

Re: Building the Structure of the Robot?
 

I want teams to read this. Then read it again, and again.

Re: Building the Structure of the Robot?
 

Thanks for all of the feedback guys.

I'd obviously like to make use of nearby machine shops and possible sponsors, it's just that it's hard for me to wrap my head around the team structure of doing this.

I'm not saying it's difficult to come up with a structure or anything, either. I'm just trying to lay it out logistically in my head and would like your input.

We're going to have a team of about 30 members. The vast majority, about 22 or so, are part of the "Mechanical Team" while the other 8 are part of "Controls" which is electrical and software.

The way I see it, would we have mechanical members CADing? Making rough designs out of t slots to refine in CAD using other materials? Prototyping manipulators and mechanisms with t slots?

What I see happening is a lot of prototyping using T-Slots, 80/20, and other materials, then applying the concept in CAD but using different materials.

I doubt I'm making this clear, because I guess I don't know how to say it very clearly.

For our team, which never really had CAD guys or an emphasis on CAD, it's just hard for me to wrap my head around the idea of making it a centerpoint in our platform. (Please don't bash me or provide pointed feedback for leaving out CAD, I just wanted to get by our first year).

If you could give me some kind of outline or idea of how you make use of actual construction and CADing simultaneously, I'd appreciate it greatly. I'd just like some team structure or flow to go off of.

Re: Building the Structure of the Robot?
 

Week 1: Preliminary Design, Prototype, have a CAD for the drivetrain done by at least the end of this week(preferably earlier) so that you can start to/send out parts to be machined at the end of the week.

Week 2-3: CAD and machine parts as you finalize the CAD

Week 4: CAD is done. Finish up machining by the end of the week so that you can assemble the robot during the weekend.

Week 5-6: Program and ITERATE!

Feel free to adjust this schedule to your team, but you should definitely have at least a week before you bag the robot to program and iterate.

Re: Building the Structure of the Robot?
 

I'll take a crack at it this one:

Linking you to our bots:

http://www.chiefdelphi.com/media/photos/38030 (Not built by an army of mentors or anything)

Not Very Successful:

http://www.chiefdelphi.com/media/photos/20464
http://www.chiefdelphi.com/media/photos/27094

You can read through the specs there and see the progression that your team could go through. We have a group of lets say 25 kids on a good day and more or less a group of 10-15 on average with about 2-3 local mentors & 2-3 remote mentors.

We currently CAD about 90-95% (I wish we could finish the entire thing) of our robot. Our team directly benefited from Adam's advice above. We had zero machine shop sponsors in 2008 and never cad'd a single part. We took some extrusion, t-slot, whatever else we could fine and built a robot. While it was fun and everyone seemed involved, the robot was never great and the team was meh at best (4-5 kids showed up on average, we showed up when we felt like and built whatever we wanted).

As we got more serious:

-We started to CAD every part (minus wiring/pneumatic lines)
-Called up every shop in town(literally and 3/4 of the town shops sponsor us and some in various other towns)
-Kids started to show up and involvement got even higher.

Our goal is to:

-Be the best we can
-Teach kids how to use CAD/CNC's/Welding/Wiring/etc... We want them to learn as many skills as possible.
-Be a competitive team
-Teach our kids how to win at everything they do. <- Very important to me
-And one day to get onto Einstein.

Your issue:

-Don't feel CAD is important and/or think it will make it so the mechanical kids are less involved.
-Shouldn't be the cenerpoint?
-Why send off parts to a sponsor?

What we/Other teams do:

-Using CAD has helped us get more kids involved and save a ton of money/time.
-We use the CAD model in a way so everyone can put their two cents in, its hard for a kid to come up to a person in charge while the robot is being built and go "Change the ENTIRE design" or "Change this piece". When its in a harmless model, its much much much easier to do so. We encourage each kid to help us with design. Some of the best parts of our robot (imo), a sophomore/junior thought of it. Not a mentor or some genius.
-CAD allows you play with multiple ideas and check if everything is going to work.
-When we finish conceptual design (paper/whiteboard etc..). We have the kids that are not CAD'n, our hands on lets get building people work on:
--Actual Field Elements
--Prototypes: We made an actual wood 4Bar and working shooter
--Intake Concepts
--Crossing the barrier mods
--Testing ball jams

-Then we regroup during the week off an on and have people put their two cents in and get moving towards a finished product.

Sponsors:

-We send off 10-20 percent of our robot out in terms of making parts. Why? There is nothing wrong with sending our more or less btw.
-We feel we are getting more (outside) people involved with our program.
-Some parts are impossible for us to make, 27x37 belly pan will not fit on our 20x30 router no matter how hard we try! Large laser parts with bends we send out as well as we can't bend accurately.
-Repetition, our machines are not top of the line. If we are making 50 bearing blocks, a sponsor would have no problem doing them. We send out a lot of the parts that a sponsor can set up once or twice and run a ton of that part for us.
-This process has not only helped kids not get burnt out while making parts but has also gotten us higher community involvement and awareness.

TL;DR. Change is hard. CAD is important. Sponsors are money and I'm glad we have 20+ shops and 100+ donors. If you don't want them, its your loss (We had 10 donors and zero shops in 2008). And most of all there are plenty of jobs/stuff that needs to be done. Just gotta look a little harder and try a little harder. Adam's advice above is seriously awesome, we have DIRECTLY benefited.

-RC

Re: Building the Structure of the Robot?
 

I agree completely with what RC said. We too had never CADded a robot or had machining sponsors before this year. Before this year, we had never made it to eliminations. Once we started CADding and looking for sponsors this year, everything changed. We build by far the best robot we have ever built, finally made it to eliminations, and even were regional finalists at the LA regional. Furthermore we were the 5th seed alliance captains at the Fall Classic offseason competition recently. All of this success is directly correlated to the fact that we started to CAD our robots and look for machining sponsors.

Re: Building the Structure of the Robot?
 

Michael, having CADed the 1836 robot in '10 and helping Ryan in '09, I can say that 1836 has CADed in every year except for 2011.

RC worded everything perfectly. CADing a robot does not detract from the mechanical team members and their roles within the team.

The the value of the ability to plan and design a robot visually cannot be overstated. I like to think of it this way:

"Imagine the ability to build a product, run tests on it, and scrap it up. Only to do it again, and again, and again, until you are perfectly content with the result. All before cutting your first piece of metal"

Team 1515 CADs approximately 98% of the robot. While we go above and beyond by implementing electrical and pneumatic paths, this is not necessary. You should be designing at least 90% of the robot in a 3D modeling software such as Autodesk Inventor or SolidWorks. Both of these are freely and readily available to you.

If you need any help through the learning curve (which I admit is significant), there is an army of CD users eager to help you.

Good luck!

Re: Building the Structure of the Robot?
 

What material did you use to cut your gussets? Given that you used a router, I'd assume that you used polycarbonate or some other plastic.