I have a bicycle brake disk that I wish to mount on a (bicycle) hub. The disk is RH threaded and is meant to mount on the left side of the hub, so when the brake is applied it screws down onto a shoulder on the hub. My application means I need to mount this disk on the right side.
So what I think I'll do is make a shouldered collar threaded RH on the outside, and LH on the inside, and cut LH threads on the hub. The disk will run up to the shoulder on the collar and the collar will run down to the shoulder on the hub.
How thin can I make the collar? - there's a limit to how far I can reduce the hub diameter. The disk is threaded 1.375 x 24TPI; I'll probably make the collar from either brass or mild steel; and the hub is some kind of aluminium alloy.
And I presume the shoulder would be on the outside of the disk? That should work OK, since the tendency will be for the disk to screw to the shoulder, and the LH threads inside will tend to screw the collar inboard.
The hub is probably the limiting element for strength of a threaded section. You probably won't like the machining characteristics of it, but you'll be better off with stainless for the collar. Otherwise, you'll get nasty dissimilar metals corrosion in the union, almost right away.
You didn't say how thick was the brake disk. That plays into how small a difference you can live with between the hub o.d. and the disk hole i.d.
I have no idea how much force a bike wheel exerts on the disk, so please accept my cornball estimates as just a way to solve the problem.
Figure a reasonable working strength on the stainless similar to mild steel - say 40kpsi to keep it conservative. If the disk were, say, 3/16" thick, then the 'adapter' portion of the collar would be 3/16" wide, shoulder to end - 0.1875 inches. Say the disk exerts a torque of 1000 inch-lb at the hub (ymmv). Then you need 1000/40000 of a square inch of metal in the cross-section of the adapter - 1/40th square inches. At .1875 wide, that's 0.1333 inches thick (between thread valleys). Add the thread depths of both the inside and outside threads to that to get the major o.d. and i.d.
Likely, the threads would fail before the collar squirmed out of the hole. One rule of thumb says at least five buried threads to maintain the characteristic strength of the metal. A 3/16" thick disk doesn't give you that many buried threads at 24tpi, unless it has a thicker hub.
Actually, he'd have to make a new disc and a new hub. Still, that would probably be a better solution than the collar idea.
When it comes to brakes, you don't want to mess around and risk failure. Even on a bicycle (imagine having it fail at a busy traffic intersection at the bottom of a grade).
The makers of the commercial brake have invested considerable time and money to insure that their design works as intended. Will the OP be able to say the same thing about his collar?
If he copies their design, using the same materials, dimensions, and tolerances, but simply reversing the thread direction, he can be fairly sure he'll have no problems. But a makeshift collar will be an unknown factor. Not to mention that it introduces additional points for out-of-roundness, out-of-plane, out-of-balance, etc. All the things a person wouldn't want in a disc brake.
Of course, the other option is to solve the problem that's preventing the OP from mounting the brake on the left side of the hub as intended.
If I understand it correctly, the threads are taking all the torque load when the caliper clamps onto the rotor. So as long as you remove the unthreading action (from reversing the RH thread on R side), nothing will have changed as far as safety and function is concerned.
So, howzabout pinning the hub, rotor and lock collar in place? The pin can be relatively small diameter since its only function will be to prevent the collar from backing off. Can you match drill after assembly and pin or even thread and install a small bolt??
Any decent bike will have aluminum hubs & rims, stainless spokes and brass nipples. All three of the materials are in direct contact with each of the other. Aluminum crank arms are in contact with chromoly or Ti pedal and bottom bracket spindles. The list goes on. Granted, the aluminum is probably anodized and the brass is usually plated.
Also, the OP only said he's using bike parts, not making a bicycle so it may or may not be an "outside thing". It may not even get wet.