Most anything is possible:
John, did your hub have a short section that was cylindrical? If Tom has the same hub that I'm thinking of...it is curved over the entire length. I talked to motorized bicycle Monkey last night, and his is cylindrical.
I didn't want to advertise that with some modification my adapter could be made to work. If it doesn't then I have to refund a part that will go into the scrap bin @ 35 cents per pound.
The CAD graphics below show the potential pitfalls of an improper installation.
The installed adapter has a bolt circle of 2.956" for small flange wheels, (I have another one for wheels with a larger flange, such as the OCC). The sprocket has a bolt circle of 2.956". The bolt holes in the sprocket are only a few thousandths larger than the bolts. If the adapter is not installed as intended, (graphic #2), the sprocket will not bolt up. This was the very reason I had to make sprockets for the adapters. The Chinese bolt circles vary. The bolt circles are not concentric to the teeth either.
The adapter is made to work, by boring the through hole to the same dimension as the wheel hub. By mounting the adapter to a cylindrical hub, the adapter and sprocket assembly can slide along the hub for a perfect chain alignment, and then final tightened. If installed on a hub that has been machined to an elliptical, or parabolic, shape the adapter will not make full contact, (graphic #4). Obviously the ability to slide the adapter for chain alignment will be lost also.
There are two ways to make it work, somewhat, for this type of hub.
1) Digitize the hub so that the curve can be plotted and input into the computer. 2) Machine the adapter in the mill, leaving the bore small. 3) Fixture the hub in the CNC lathe, and create in inverse geometry inside the hub. The ability to adjust the adapter will be lost so it will be necessary to take the measurements for sprocket alignment and make the mount accordingly.
1) Digitize the hub so that the curve can be plotted and input into the computer. 2) Machine the adapter in the mill making the bore .030" larger than the smallest diameter of the hub.
2) Using the digitized geometry from the hub, create tangent lines and determine the angle. Create chamfers in a manual lathe, after making a fixture, according to this angle. With this method, you would have to cut and test fit several times in order to sneak up on it, and get the correct bolt circle.
Certainly I can't stop anyone from trying to make it work, but I don't want it back when it doesn't.