11-05-2010 13:59

Jeff 801

Re: pic: 3 Hour Chassis + 2 Hours

If you can I would get the extended output shaft that AM sells and make the center wheel direct drive.

Also how do you plan to tension the chains that your using and what size chain are you looking at?

Also the chain runs on the bottom look like they will run into the square tube... In inventor with design accelerator you can generate chain runs so that you can check for clearance. If you in solidworks I think there is a way to do the same thing.

11-05-2010 14:05

548swimmer

Re: pic: 3 Hour Chassis + 2 Hours

It's looking a ton better!

You'll need some way to attach the tube supports to your outer plate. We use 1.5"x3/4"x3/4" press-fit stock tapped with 1/4-20 to do this.

The tubes should have material underneath them. There's no constraint tool in real life, so spacers have a hard time just hanging there. Think about the assembly process when you're designing robots, that helps me the most. You can usually just pound them through with a gentle persuasion tool if the hole is a tad small.

If at all possible, the design would be a lot stronger if you move the tube supports so that they are not all along the same line.

Are you running both chains on the same side?

11-05-2010 14:34

Andrew Schreiber

Re: pic: 3 Hour Chassis + 2 Hours

I agree, get the tubes to at LEAST be in 2 planes.

Perhaps also consider mounting your battery laying down on the bottom of your robot. That way you don't have to reach through anything.

I would also only have the cross supports go to the inner side plates. Use the standoffs and some wood blocks to support the outer ones.

11-05-2010 14:40

JamesCH95

Re: pic: 3 Hour Chassis + 2 Hours

I agree with the direct-drive on the center wheels, that will get all of the heaviest objects directly over your primary drive axle, and close to the center of rotation.

It's hard to tell from your drawing if you have done this or not, but if you chose to, you could eliminate the face plate of the toughbox by machining your 1/8" plates with the proper bearing and shaft holes.

11-05-2010 15:32

Ty Tremblay

Re: pic: 3 Hour Chassis + 2 Hours

I intend to weld the cross beams to the inner plates. Our team has a CNC mill, but it doesn't have a coolant system built in, thus milling operations can be quite cumbersome. By designing it this way, the only CNC operation necessary is the drilling of the holes so everything lines up square. The rest of the dimensions are more or less non-critical and can be cut with some careful band sawing.

11-05-2010 16:12

JamesCH95

Re: pic: 3 Hour Chassis + 2 Hours

I do a fair amount of aluminum machining dry without too much argument... what sort of tooling are you using? Are you making nice aggressive cuts to keep heat in the chip?

I've had good luck with carbide tooling and heavy feeds on 6, 2, and 7 series alloys.

I assume because you're planning on welding it you're going to use a 5 series alloy, which has a "poor" machinability rating, which may well be your problem.

11-05-2010 16:34

Ty Tremblay

Re: pic: 3 Hour Chassis + 2 Hours

I'm also on 190, which has plenty of Haas CNC mills for me to do my worst with. My alma mater team, 319, has the CNC I've mentioned, but I've never seen it run. I was intending to use 6061, but should I choose something else due to my intention to weld? Also, is it difficult to weld two different alloys together?

11-05-2010 16:58

Chris is me

Re: pic: 3 Hour Chassis + 2 Hours

There's a chain reduction, so direct driving the center wheel isn't possible without a gearing change.

11-05-2010 17:03

548swimmer

Re: pic: 3 Hour Chassis + 2 Hours

Could you use a smallish size end mill and cut like this:

______
|xxxxxx|
|xxxxxx|
|_____|

so the center piece falls out.

If you can use a 1/16 or 1/4 bit you can cut past the needed length/width just enough so that you end up with squared corners.

11-05-2010 17:16

JamesCH95

Re: pic: 3 Hour Chassis + 2 Hours

6061 loses its temper when it's welded, so your 6061T6 becomes 6061T0, which according to wikipedia is an 80% loss in strength, which is what I've noticed/experienced with the welding and mechanical testing I've done.

Welding two aluminum alloy's together hasn't been an issue for me. Most, if not all, aluminum alloys will weld fine with the same filler material, so there shouldn't be much of an issue welding different alloys.

Haas CNCs are pretty good machines, I used one in college. Tooling, though, shouldn't be confused with the mill itself. You could have the best CNC on the planet, but if a low-quality HSS endmill is chucked up it'll run like crap. Look into solid carbide tooling, if you're CNC'ing plates you could easily get away with a 1/4" end mill, or maybe smaller. McMaster: 8829A19 about $24 for an end mill that will make you many a robot frame. If you have good compressed air for chip clearing you could get a 4-flute which is even less expensive.

IMHO, if you want to weld the chassis together (which I heartily approve of) you should CNC your side plates from 5052 (which has no temper and won't lose strength when welded) and then use AndyMark/KOP extrusions for your cross-bracing. It'll be a little less expensive, and you'll get tons of 1/4" mounting holes. My team's most recent chassis was made from CNC'ed and bent 5052 TIG welded to the KOP frame. I did the welding and found the 1/8" 5052 to be a very forgiving material to TIG. We dropped 3-4lbs in fasteners IIRC.

11-05-2010 17:19

JamesCH95

Re: pic: 3 Hour Chassis + 2 Hours

Good point, but in the previous thread on this topic the OP mentioned making custom gearboxes.

11-05-2010 17:44

Jeff 801

Re: pic: 3 Hour Chassis + 2 Hours

From the looks of it the small sprocket on the toughbox looks to be about 16 tooth and the sprockets on the wheels look around 48 (Very ROUGH numbers) and assuming its a standard toughbox the overall reduction is 28.3:1 and those look to be 6" wheels so the top speed would be 3.05ft/s

Or on the back side of that wheel the sprocket is smaller than all the others and after looking at the AndyMark website the smallest sprocket (#25 I am assuming) is 32 tooth and the sprocket on the toughbox output shaft is still 16 that means the overall reduction is 25.5:1 and the top speed is 4.57ft/s

From what I know your average single speed robot is ~9ft/s which would cause me to thing there is some sort of difference in the toughbox or the extra chain reduction can be removed and supplemented with one of the many gearing options AndyMark has for sale.

After going back to the old thread I see that your wanting to get ~12ft/s which you can do by using the Toughbox with the optional 9.87:1 and direct drive which would give you ~11.82 ft/s

11-05-2010 18:28

sgreco

Re: pic: 3 Hour Chassis + 2 Hours

You could consider a 1x1 bar of 80/20 insead of what you have for the cross pieces. The ends can be tapped for a 1/4 20 bolt and you can just bolt it to the outer plate. This also leaves you a lot of flexibilty on the inside as far as mounting other components.

I think it would work nicely in a prototype like this. My team actually used a very similar frame to this this year, aluminum side plates, cross beams extending to the outer plate, toughbox transmission etc...Anyway the 80/20 worked nicely for us on the cross braces.

11-05-2010 18:51

548swimmer

Re: pic: 3 Hour Chassis + 2 Hours

The only potential issue with that is 80/20 in comparison to hollow tube, it's very heavy. For an off-season drive train, that would be an excellent idea, though I would recommend against it for a competition bot.

11-05-2010 18:57

sgreco

Re: pic: 3 Hour Chassis + 2 Hours

Yes this is true...to be honest I'm not a fan of 80/20 for its weight reasons(even though my team used it with out problems this year. We didn't make our whole frame out of it though, we only used it for crosses)...

80/20 is unneccessary if your team CADs well, because you can plan all of your cuts ahead of time. 80/20 can be nice if your team likes to make it up as you go and constantly be taking off and adding parts. (not that you can't do it with other materials, just 80/20 makes it easy).

I would still recommend it for a prototype though.

11-05-2010 19:05

Ty Tremblay

Re: pic: 3 Hour Chassis + 2 Hours

I've done no calculations regarding gear ratios, chain reductions, or overall speeds yet. I felt that there were more pressing structural concerns in my design that I needed to adress before conquering the gearing.

That 9.87:1 Toughbox seems interesting, but if we're going to buy more Toughboxes, we might as well just make our own (lighter) gearboxes. I just threw the Toughboxes in the CAD in the event that the team wanted to save money and use the Toughboxes we already have.

11-05-2010 19:10

sgreco

Re: pic: 3 Hour Chassis + 2 Hours

If you like the gearing in the Toughbox, there are lot of weighs to reduce weight. You can mill out some of the material in the side plates, get aluminum gears instead of steel, or just cut out material in the gear sand sprockets like this http://www.chiefdelphi.com/media/photos/30614

I'm all for custom gearboxes, but if you want to keep it simple, you have a lot of options.

11-05-2010 19:33

Chris is me

Re: pic: 3 Hour Chassis + 2 Hours

Not really, 80/20 weighs almost exactly the same as 1/8" wall 1" square tubing.

11-05-2010 19:37

EricH

Re: pic: 3 Hour Chassis + 2 Hours

And 1/16" wall 1" square tubing is called for.

For the hollow tube in general, you're right. But for the specific hollow tube called for, 80/20 will be heavier.

Comment on the cross-braces: I'm not quite sure I'd trust a 1/16" wall for that application in that pattern. Change the pattern, and you'll probably be fine--but run the analysis just to make sure.

11-05-2010 19:40

sgreco

Re: pic: 3 Hour Chassis + 2 Hours

It's close, a foot of 1x1 80/20 weighs .53 pounds and a foot of 1x1 1/8 wall weighs about .512 pounds. This is assuming 6061. They're so close its negligible.

(These numbers are what SolidWorks is telling me).

You can pocket 1x1 tubing, but you can't pocket 80/20 so that could be included in the weight tradeoff.

11-05-2010 19:41

Chris is me

Re: pic: 3 Hour Chassis + 2 Hours

Running 12fps without shifting and using roughtop tread will lead to a lot of current issues as well as iffy acceleration.

11-05-2010 21:35

Hawiian Cadder

Re: pic: 3 Hour Chassis + 2 Hours

i would recommend the use of C channel instead of square pipe, it can be just as strong if oriented correctly, and it is lighter as well as easier to attached.