paper: 3008 Custom AM Super Shifter Plates Pocketed

Thread created automatically to discuss a document in CD-Media.

3008 Custom AM Super Shifter Plates Pocketed
by: GRT808

Symmetrical custom gearbox plates for using AM Super Shifter gears.

Weight of each plate = 0.125lb (50% lighter than the AM Super Shifter Plates)
Hex 500 Output shaft.

custom_gearbox_full_asm.stp (1.85 MB)

There’s barely any metal on this thing! Doesn’t it bend?

I imagine it’s also supposed to be for a WCD? Will it be mounted just by the standoffs or will there be other support structures? Because those things will probably twist off like Gundam parts coming off of a sprue.

Anyway, I like that you solved that stupid gear setup on the SS (Why not symmetrical, AM? Why?!?!?!?!). That’s cool.

I believe the goal was to fit into the existing profile of 5" by 5" box extrusion that the shifter come in. If you have ever opened one up, or played around with trying to modify one, you will notice there is literally no unused space.

EDIT: This was intended as an answer to why SS aren’t symetrical, but when it was posted I realized that might not be clear.

I know that, I just don’t like it. :\

Of course it will bend. But by how much, and do you really care? (Remember that the CIMs get additional support from the cylinder bracket.) Any stripped-down Super Shifter is probably more susceptible to abrasive wear issues, due to gear misalignment resulting from flexure—but how much efficiency and gear life are you losing, and with a bit of wearing in, will you achieve a new, acceptable steady state that accounts for the misalignment?

Besides, if you’re milling out the gearbox (rather than laser or waterjet cutting it), you can leave a thin wall (so the lightening is pocketed, rather than straight through). That way there is something to resist torsion in the individual ribs, if it’s actually a problem.

The asymmetry lets you pack a pair of Super Shifters into an 11.75 in × 7 in × 5 in space (plus cantilevered shafts), driving the left and right wheels of a robot. In games like 2008 and 2010, where width mattered, that was a useful attribute. This arrangement is compatible with wheels down to Ø4 in (on a very flat floor), with the gearboxes mounted low in the frame.

You really ought to increase your radii on the pockets. If this is getting waterjetted, they’re ok, but if you’re milling it you are not going to find a machinist who will want to use a .060 endmill to machine all your corners with.

If I was making this I would want to not have to use anything smaller than a 3/16" end mill, but 1/8" would be acceptable. Anything smaller and I’d tell the designer to go play in traffic.

You should also make your radii a bit larger than the radius of the cutter you intend to use. If it’s exactly the same size the machine can’t interpolate the radius-it simply comes into the corner and then abruptly moves out. This won’t affect the accuracy of your part, but you will get an improved surface finish (and tool life) if you increase the radii a bit over nominal.

Yes we have gotten it waterjetted thanks.

As for it bending we are using aircraft grade aluminum (not sure what alloy) but in our past competitions we have used the same grade and have never seen it break in any of our applications. I’m a bit more concerned about the effort it is going to take to make everything fit (such as the bearings) since the jet is so precise.

We are using 0.187" aluminum plate.

Any suggestions or concerns I should be aware about before season starts?

I think aircraft grade aluminum is a waste of money.

If you properly support that plate you will see zero benefit from using any alloy stronger than 6061.

You may actually have problems getting your bearings to fit if you waterjet it to the nominal size of the bearing. Most waterjets are only going to have repeatability of about ±.002 to .005 inches.

This isn’t going to be good enough to reliably cut a press fit for a bearing. We have waterjetted our sideplates since 2009 and have the bearing bores cut ~.015-.030 undersize so we can go back and remachine them to ensure roundness and press-fit tolerances. I know 233 waterjets them about the same amount undersize and then reams them to the proper size for the bearing.

Alloys of aluminum tend to have very similar elastic moduli, no matter their strength—in other words, if it’s just bending you’re worried about (and not stresses due to bending, which are tiny in this application), the alloy and temper don’t matter very much. (That’s why Cory is telling you to save your money.)

Strength (not stiffness) matters when you’re worried about yielding or breakage of the material. For most gearboxes, this isn’t an issue. Instead, you need some degree of stiffness to maintain alignment (e.g. when chains or wheels put loads on shafts, or when the frame of the robot flexes during use). While much of that stiffness depends mostly on the connections (tolerances, fastener types, preload, friction, etc.), the plates could play a role in it if you lighten them aggressively like that.

By the way, the kerf from waterjet cutting isn’t perpendicular to the plane of the workpiece (except on the fanciest tilting-head machines). Also, the surface finish depends on how quickly the machine is travelling—slower is better, but time is money. While it is possible to cut many gearbox plates directly on some waterjets with no secondary machining, very precise cuts probably will not be possible or feasible at the shop you’re working with.

You should experiment with bearing fits if you’re going to pursue this, and consider flanged bearings and gap filling retaining compound (see Loctite’s 600-series documentation). Also, by introducing error into bearing positions, you introduce error into your gear meshes. You may need to add a couple thousandths of an inch of extra clearance between nominal centres, to account for potential misalignment. (A little backlash is less worrisome than binding gears.)

You may need to add a couple thousandths of an inch of extra clearance between nominal centres, to account for potential misalignment.

So by this you mean try and incorporate what Cory was stating that they do?

We have waterjetted our sideplates since 2009 and have the bearing bores cut ~.015-.030 undersize so we can go back and remachine them to ensure roundness and press-fit tolerances.

flanged bearings and gap filling retaining compound (see Loctite’s 600-series documentation).

And thank you for this piece of advice. We are using all the AM Super Shifter bearings so they are all flanged, but I had heard of a gap filling retaining compound but could not recall its name. Will look into that.

Regarding the misalignment due to bending plates, do you think placing standoffs in better positions or adding more would be an adequate method to correct that?

As for the aircraft grade aluminum being a cost issue, we kind of piggy back off of scrap so it is not an issue, none of it comes out of our pocket.