This is the latest drawing of the prototype. 1530 lite rails, not light and over designed as this is a learning process, with custom saddles for the dead axles on the front and rear. AndyMark.biz Supershifter /w servo shift is attached via a custom saddle with a bearing on the opposite side. Actual attachment is via two through bolts on the AM supershifter. AM FIRST 2007 wheels will be driven by #25 ANSI roller chain. Chain / sprockets / universal hub for center wheels not included in design. Also not pictured, not drawn yet, are the front/rear & side bumpers. Front/Rear bumpers will be simple bolt on type. The side bumpers will need 2" stand-offs in the middle section.
Very interesting. What is your sprocket setup (input/output)? I would probably recommend going with #35 chain in a drive situation. #25 is more suited for arms and other lower-stress mechanisms. Have you considered traction wheels or are you concerned about budget. Obviously you don’t want to break the bank. How far is the center wheel lowered?
The center axle is 0.125" lower and not adjustable. The outside axles have 1" slots which will allow for chain tensioning. The chain will be on the wheel side as there is not room between the transmission and frame rail. A universal hub will drive the center wheel and sprockets attached to the wheel or a second hub will drive the outer wheels via attached sprockets. The outer axles are dead axles which saves me bearing mounts on the saddles. The use of #25 chain is due to weight considerations. I’m told by other posters in a seperate thread that teams have been successful with #25 as long as tension is dealt with appropriately. If I can get some 1020 in the shop within a week or so, I will adjust the saddle widths to the smaller profile which will drop a few pounds and then we won’t have to worry about weight as much and can use #35 chain.
#25 chain has been used successfully by many many teams for driving applications (including both of my own teams), and greatly reduces the weight. If you correctly tension the chain, and everything is aligned right, they pose no issues. It should be just fine in this setup
From what I can see in picture, you’re just bolting the angle iron on to the extrusion. I’d suggest putting some gussets or triangles in there in that setup to keep the frame from becoming a parallelogram.
At the very least, I think you need to reconsider using 1x1x1/8" angle as cross members. Even in pairs, they’re not going to give you the rigidity you’re looking for and you stand a good chance of knocking the chassis out of square, if not, worse, the first time you clip your corner on something at a reasonable speed. You can probably stand to use lighter stock on the side rails and move some of the weight and strength to the cross members.
I looked into using 80/20 extrusion as the basis of a drive train about a month and a half ago as we iterated our existing design and tried to eliminate as much manufacturing time from it as possible. Ultimately, I decided to do things a different way, but here’s a look at what I’d come up with:
We do all of our machining work and so I design for our skill set – and that means no multi-op milling. If you’re expecting to have a shop do some work for you, you could easily have them create bearing blocks from blocks of aluminum that replace the function of the individual plates depicted here.
We have the axles sitting above (or below, really) the 80/20 in an effort to eliminate milling operations on our side rails. If you’re planning on milling slots into the 80/20 to allow axle clearance for tension and you’re expecting to machine bearing blocks for your axles, why use 80/20 at all? The advantage in 80/20 is that it allows for tensioning easily, but since you’re milling the side rails anyway, it’d be pretty easy to add some slots for tensioning.
I really like this design Madison. This would be a very friendly prototype layout since it would allow you to use however many wheels you want.
As for machining the slots, all that results in is a loss of movement (and you really only need an inch to tension the chains). The real purpose of using 80/20 is that it allows you to tighten the bolts anywhere along the track, in other words, no need for a tensioning screw or other such contraption.
Will these be using AM Gen 2’s? Perhaps you should start another thread:rolleyes:
 One thing you may not have considered (and I did not consider it the first time I designed my drivetrain) was tightening the bolts on the bearing blocks. From the looks of your design it would appear the wheel prevents people from reaching the outside bolts. 
Madison, that is a really cool design. What is the purpose of having movable gearboxes? Tensioning? Either way, its a very neat modular design. Truely innovative.
I uploaded the image to the gallery: http://www.chiefdelphi.com/media/photos/29458?
This never got as far as having motors implemented, really, because it was primarily meant to examine construction methods. A lot of things are placeholders.
The advantage to using 80/20, as I see it, is that the stuff that requires a mill when implementing a sliding wheel tensioning system – the slotting – is preformed into the your frame rail. If that’s the case, milling a clearance slot for the axle to slide in eliminates any time or resource savings obtained by using 80/20 in the first place. If you’re going to be milling anyway, just use rectangular or square tubing and mill slots for the axle/bearing block clearance as well as through bolts that hold the bearings in line with one another.
I would be highly interested in a sort of synthesis of these two designs. I like Madison’s design for the lightness of the frame components, but I think there’s a lot of advantage to the style of bearing blocks the Andrew has designed. As I understand it, the main purpose of 8020 is to simplify chain tensioning, and Andrew’s block only need three screws on the top for adjustments. Madison’s, while somewhat lighter, would take require loosening 2 screws outside the frame and 2 screws inside the frame, which seems more troublesome to me. Admittedly, the loading on Andrew’s blocks wouldn’t be quite what you’d like, but I don’t think it would really present a problem. If Madison’s using 1515 Lite instead of 1010 on those side rails you could probably bolt them straight on.
I like the design you have presented here, with it’s effective use of extruded aluminum rails. However, as a pretty experienced user of 8020, I’ll make several comments:
The 1010 material you are using is actually fairly flexy in torsion. Add that to the bending stiffness and your deflections start to add up. With a cantilevered setup like this, those middle wheels may induce enough flex in the rail to either throw a chain (#25) or negate any drop you had to begin with. Anything you can do to add a x-plate or similar will stiffen up that center axle tremendously.
The 8020 fasteners are very prone to coming loose. Locktite does work, but there aren’t many threads in those types of slide-in fasteners. I would make sure that you have access to the outer bearing block screws either through the spokes of your wheels or with a quick removal of the wheel.
There is quick a bit of “slop” in the fit of those slide-in / drop-in fasteners. You can easily get +/- 1/16" loss of accuracy in the placement of those blocks. If you machine a flat recess into the plates, you can ensure the proper alignment based on the fitment of the plates, and not the mounting holes and hardware. (though I see this doesn’t work with your existing bearing placement…)
Best of luck with this design. I like seeing more and more teams come up with innovative ways to work with this material. If only 8020 ponied up and became a bigger sponsor and/or supplier for the KOP.