20-06-2010 22:51

Chris is me

Re: pic: sheet metal 6WD

Looks familiar.

Any reason not to make the bottom a full belly pan for electronics, etc?

20-06-2010 23:05

MWB

Re: pic: sheet metal 6WD

Is there any way to rotate the shifters so that the motors line up, for ascetics and very insignificant COG changes?

20-06-2010 23:39

apalrd

Re: pic: sheet metal 6WD

1. I would recommend a center divider on one side of the motors. It will add strength and also a place to mount an electronics tray (e.g. from the back to the middle with the front open, or something like that)

2. Electronics tray could be perforated aluminum. Just cut it to size, rivet it in, and zip tie the electronics. Nice and easy.

3. I like the chain tensioning method. Super simple. If you want, you could put a 1" tall by 1/4" thick block with a hole for the axle, then tap the end for a 10-24 (or 10-32 if you want) and just tighten the screw to tension it, then tighten the axle bolts to hold everything secure.

4. Are you using servos for the SS's? If you are using pneumatics elsewhere on the robot (or even if you aren't), pneumatics will shift easier then the servos. I do not know how much better they are, but they provide a lot more force to shift then the servo does. We always shift with pneumatics when using AM shifters.

5. Rivets. If you are waterjetting or otherwise CNC'ing the sheet metal plates, just add a few holes per connection and pop rivet it. Easier to assembly, and not much more work in CAD (this is especially true with a pneumatic riveter, those are great)

6. Have you done any math on the speed and gear ratios?

7. .125" seems really really thick. We use something around .050 in our chassis, with a plate around the axles (a little circle with an axle hole and 4 rivet holes, around 1.5" in diameter), or the above mentioned tensioning block with some reinforcement.

21-06-2010 00:10

548swimmer

Re: pic: sheet metal 6WD

.125 sheet is way to thick, especially with flanges. You can easily go down to .09 (a fairly standard size) wihout too many design changes. Any thinner (.05) and I'd make larger flanges just to be on the safe side.

Is there any reason you're lightening with circles instead of triangles?

Is there any reason you rendered it as an illustration?

21-06-2010 00:40

Akash Rastogi

Re: pic: sheet metal 6WD

Depending on what manufacturing process they have available, circles are great to use.

Check out pictures of 228's recent robots.

21-06-2010 00:50

548swimmer

Re: pic: sheet metal 6WD

True, I tend to forget about punches since we don't have access to one.

What about flipping the inner plate around so the flange is on the wheel side. It looks like the flange is small enough such that this wouldn't limit access too drastically, and having a flange along the entire length will be muh stronger.

21-06-2010 01:01

Akash Rastogi

Re: pic: sheet metal 6WD

If you have access to a punch, kajeevan, I would recommend checking out Art Dutra's white papers here on CD. He posted his CAD files and he has some awesome designs in there.

21-06-2010 01:12

Chris is me

Re: pic: sheet metal 6WD

How much is "much"?

21-06-2010 01:34

artdutra04

Re: pic: sheet metal 6WD

Thanks!

I'm sure sure of the rational behind their choice of sheet metal direction flanges, but since the flange profiles look identical to the ones on our 4WD/6WD sheet metal designs, here's the rationality behind ours. The black flanges are our chosen sheet metal flange profile.



By putting the flanges where they are to avoid geometry conflicts, the necessary space between two chassis rails is kept to a minimum. These aren't the most optimal flange configurations for chassis strength, but they work good enough for our applications.

21-06-2010 23:49

548swimmer

Re: pic: sheet metal 6WD

Fair enough. I usually think of a flange's strength benefits instead of its mounting benefits. Although, with roughly 5 inches of pool noodle and 1.5 inches of plywood between colliding robots, the bottom flange may not even be necessary.

22-06-2010 01:06

AdamHeard

Re: pic: sheet metal 6WD

The moment of intertia of the cross section is directly related to it's stiffness.

If you google, you'll see for a rectangle, the moment of inertia is 1/12bh^3 (base and height). Base being the side parallel to the axis the beam is being bent around.

So, a 5" tall .050" thick plate has an I = 1/12(5in)(.050in)^3 ~ 5x10^-5.

The same beam with two 1" flanges is that I plus the I of the two flanges. I = 1/12(.050)(1)^3= .004

So the beam with the two flanges has a total I of .004+.004+5x10^-5 ~.084.

.084 / (5x10^-5) ~150:1.

Even if I made a mistake in the math somewhere, this demonstrates the difference flanges make in design. I could go on for pages elaborating this, but it'd better to google and look yourself (key terms being beam, cross section, moment of inertia, etc...). You'll also then understand why I-beams are shaped the way they are.

22-06-2010 01:15

Chris is me

Re: pic: sheet metal 6WD

I meant how much of a strength difference does cutting part of the flange out for the gearbox make? A poster said that it would be much stronger, and I wanted to know how much.

22-06-2010 01:22

AdamHeard

Re: pic: sheet metal 6WD

I totally misread your question then.

Similar concepts apply, just inverse. The beam is substantially weaker due to that flange removal. Depending on how the gearbox is attached, the gearbox itself could add a lot of support to the beam where material is removed.

22-06-2010 01:51

Chris is me

Re: pic: sheet metal 6WD

delete please.

22-06-2010 01:52

548swimmer

Re: pic: sheet metal 6WD

Great explanation

The loss of a continual flange will not only decrease the second moment of area, but will also "chase" the stresses to the ends of the flange.

22-06-2010 01:58

Chris is me

Re: pic: sheet metal 6WD

If that's the case, what's your basis for saying it'd be okay to remove the bottom flange entirely?

22-06-2010 02:07

548swimmer

Re: pic: sheet metal 6WD

Different design. The flange I said would be okay to remove was the bottom external flange. The top external flange used to mount bumpers will working in tandem with a properly mounted plate should, depending on material thickness, provide more than adequate structural integrity. That's why I said you could most likely remove that flange.

22-06-2010 02:37

Chris is me

Re: pic: sheet metal 6WD

Ah. Did you use any kind of analysis to come to this conclusion? I'm trying to learn all I can here.

22-06-2010 09:15

548swimmer

Re: pic: sheet metal 6WD

Not on this specific piece, though I re-designs our kicker plate (under roughly 300 lbf) to make it lighter and not deform. This was one of the things I frequently noticed.


This is with a continual .5 inch flange on the top and bottom, 100lbf, fixed constraints at the bolt holes.
http://picasaweb.google.com/11295163... 6928451693122

This with a 5 inch break in the top and bottom flanges, same load and constraints.
http://picasaweb.google.com/11295163... 6931950242546

22-06-2010 15:06

AdamHeard

Re: pic: sheet metal 6WD

Another way to strengthen a frame such as this is adequate attachment between the inner and outer drive plates. This combines them into a much larger, and much stronger beam (a ~3-4" tall I-Beam).

If you understand the basic concepts, you'll quickly see that rather trivial and easy changes in design can cause appreciable increases in strength.

Kajeeven, this isn't a criticism of your design, I would definitely classify your attachment as adequate. It's just a convenient place to mention such concepts... Too much what on this forum and not enough why.

I do highly recommend adding a single baseplate that attaches to all members of the frame, or at least the full length of the front/back crossmembers and inner drive rails. This will greatly increase the rigidity of your frame. Think of it as an infinite amount of crossupports, providing strength in whatever direction is needed at the moment. Also provides a very low CG friendly electronics mount.

22-06-2010 15:33

Andrew Schreiber

Re: pic: sheet metal 6WD

The op may be looking at doing something different and lighter to mount their electronics (http://www.chiefdelphi.com/media/photos/27670) It would be possible to use cables under tension to mount your electronics to and use your frame (which should be pretty rigid in most cases) to ensure tension. Not only that but there is no fear of shorting out if you use a plastic coated cable.

Just saying, there are alternate ways of mounting electronics that many teams haven't looked into, all of these have the benefits and their drawbacks which must be considered.

22-06-2010 20:45

kajeevan

Re: pic: sheet metal 6WD

To answer a few questions,

I went with 1/8in Al trying to avoid having to put plates around the axles for strength and having to deal with smaller pieces. Also I was hoping to have the side plates tapped so I might not need nuts and just adjust the clutch on the drill to not strip the treads when inserting bolts.

The chain is tensioned with a bent C-channel with extra flanges for strength that hug the axle and is pulled by one 5/16'' bolt. The chain can be tensioned up to 1''. If you look closely enough you can see it.

I don't believe it's possible to have the AM SS in line with each other.
I am using pneumatic shifting.
I positioned them sideways to have easy access to the encoder and shifter.
Speeds are 5 and 18 feet per second.

There is no base plate yet but will be once the function is decided and its worked around it. But the front and back plates connect the entire base together and act like a small base plate at the triangles preventing racking.

Sorry for the late response been busy for a while.

23-06-2010 10:20

Rob Stehlik

Re: pic: sheet metal 6WD

Overall, this is a solid looking design. I agree with previous recommendations that you don't cut away flanges for the gearboxes. It looks like you may have room to bend the flange towards the wheels. Also, depending on the manipulator required, you'll likely want an opening at the front of the frame.

I agree that using thicker material can be simpler than reinforcing thinner material. I'm not sure you should put threads into .125 aluminum though. It'll strip really easily. You should look into PEM nuts. They press into holes in your sheet metal and provide much stronger steel threads. We use them all the time. The correct way to install them is with a press, but we find just cranking them tight with a cap screw will seat the teeth into the aluminum just fine.

I like this design a lot! It looks like a great way to tension the chain. My only suggestion would be to put flanges at the front of the C-channel where the pull bolt goes through. Otherwise this area would bend quite easily. Also, why a 5/16" bolt? A 1/4" or even #10 bolt would be fine.

Triangles are good

Rob