I seem to be getting un-healthily obsessed with this wheel
.
Lets say you had .002" run-out in the top wheel, and the same in the bottom. As you wheel a piece of sheet back and forth, the top and bottom wheels are behaving like very slight eccentrics (not unlike me then).
So, as the metal goes through, at some points it may be that the two tolerances complement each other, giving no effective run-out. At other points R/O might not be too bad, and at the worst point, you'd have an extra 'squeeze' of 004". This doesn't sound too bad, but thats a big percentage of, say, 24 gauge sheet (which is only .022" itself).
The other issue is that once introduced to the wheels, the panel doesn't slip. Therefore if you have a loose part of the rotation in one place, and a tight part in another, you'll repeat that all the way along the panel. This starts to produce a bulge in the panel. The bigger the panel, the more apparent the distortion.
This is one of those rare cases where there is no minimum tolerance really: one thou seems to be the point where even on a stiff wheel, problems are unlikely. Trouble is, wheeling is not overly scientific, i.e. a lot comes down to feel. If you know you have run-out, it's going to be there at the back of your mind ready to pounce as soon as you get a problem panel. Is it me? Is it the metal? Of course, there's always that run-out I could never get rid of ....
You get the jist.
Anyway, with the above in mind, and as hinted at earlier, I decided the best way of getting round this was to turn the whole English Wheel into it's own lathe. Rather than using a cutting tool as such, I'd re-use the diamond lap from before, but in a machined holder that would fit where the bottom wheel lives.
I routered out a big MDF pulley (which was a real success, very useful to know if you ever need to quickly gear a motor right down for a one-off job), and set up as shown below. My little lap holder had a small reservior built in for the suds (Fairly Liquid and water), and has a jack at the rear for holding the lap steay on the wheel.
The motor and gearing gave around 120 rpm, which was perfect: fast enough to do the job, but not so fast to produce a rooster-tail of grubby lube over Steve's 911.
Anyway, to a point, the setup worked very well indeed. Straight away, and with the motor in action, it became apparent that the main shaft was badly bent, hence the problem at the 'nose' of the shaft where the wheel mounts. With this in mind, it's just as well I didn't go for a new (hardened) wheel, as there's no way I'd have been able to lap that out. With my setup though, the bent shaft becomes irrelevant as the whole assembly is being turned as one (which was the whole point of course).
Quite quickly I halved the problem down to one thou, so technically within acceptable limits. I wanted more though (don't we all). Bottom line was that the wheel didn't want to give any more. I reckon I ran the rig for twelve hours in total (I said I was obsessed), moving the lap position a tiny bit left to right every few seconds (otherwise it just started rubbing: you could hear the difference quite easily).
In the end, it just wouldn't go down past the 001" whatever I did. I concluded that the new problem was that the ball bearing race in the front of the frame probably had a little corrosion / wear / chatter in the races and balls. I'm guessing that this could easily jiggle the top wheel by plenty enough to give this error.
New bearing now on way (new old stock), the orginal BTW was dated by the bearing supplier as being 'probably pre-second world war' because if the maker's name, which changed just after the war
. As I said to Steve, I think that being the case, I can forgive it a slight error. If the new one lasts another 65-70 years, it'll be good value.
All of this messing about did, I'm afraid, mean no progress on the Special itself this week I'm afraid, but hopefully will catch up this coming weekend.
BTW, New ball race due in a day or so, and I'm hoping to pop up one evening to try it out
.
