I've had some challenges making smallish cylindrical shaft surfaces for my radial engine, kind of learning as I go. Compared to the valves which are turned down from SS stock, the rocker axles seem dead simple. They are made from O1 tool steel as they need to be hardened. Tool arrives looking nice & shiny but its actually about plus minus 0.0005 oversize. The issue I have found is its actually elliptical in shape, not circular like a gauge pin. You can verify this my taking measurements at right angles on the same portion of shaft. Correcting both the diameter (within tenths) and shape simultaneously sounds like a job for lapping. I have made a few tools like this squeeze clamp style with replaceable brass or aluminum cartridges. It works but lapping is messy business. And you cant get a lap on & off unless the the shaft is the same diameter. So for example turning in a lathe where you have the enlarged end knobby in place related to tail stock support.
So I gave this method a try fully thinking it will be yet another way of how NOT to do it. But I'm actually very impressed. In about 15 minutes I've made myself some nice sections of 'true' 5mm shaft stock that are within a tenth of target dimension & desired finish. And its a very controlled process. I got the idea by (carefully) mic-ing various wet/dry papers in the 600, 1000, 1200# range. I was impressed with how consistent they are across the sheet & across other sheets of the same vendor. For example the 600# I have measure 0.0075". Amazingly for some reason my 1000# thickness was quite close to 600#, within tenths, otherwise I would make a separate lap for each grit. So I worked this dimension in as an allowance, both annular (surrounding the stock) and between each lap face. Hopefully the sketch makes sense.
So I started by milling a block of aluminum, cut in half so it gave 2 symmetric pieces. Clamped them together face to face in the vise & drilled the oversize hole (0.0075 annular = 0.015" diametric). It would have been much better to ream the hole for a better finish but I didn't have the right size. So I chose the next closest numbered drill I had. There will be some buffering with the paper itself I guess. Then separate the 2 lap halves, position in vise & simultaneously mill off the face allowance from both. make a small transition radius on the hole corners The idea was to place 2 strips of abrasive paper clam-shelling the stock. If everything is dimension-ed reasonably close, by the time you fully squeezing the assembly together, it should be getting close to target dimension. You could spin the stock in the lathe, or as I did, just grip it in an electric drill. The benefit is the 'tool' never changes shape or wears out by lapping action. Fresh abrasive is a postage stamp area of wet/dry paper.
Well it worked out way better than I expected. You can actually feel a bit of vibration initially, which are the eccentric bumps of the shape. I used 600#, then 1000# mic-ing every so often. A longer section of shaft (3-4") means you are probably lapping more consistently across any one area. I figure 5 seconds of lap-sanding was about 0.0001" so it was actually quite easy to sneak up on the target dimension & finish simultaneously. I tried 2 strips of paper & also a single wrap just to get a sense of how much of the paper was in contact & cutting.
This is not a good way way of removing too much stock. Its kind of like lapping but without all the mess. Digital photo's tend to exaggerate machining marks so hopefully you can see the contrast of the finished rod & the aluminum tool. Drill holes are nasty looking surfaces under magnification once you split them open.
So I gave this method a try fully thinking it will be yet another way of how NOT to do it. But I'm actually very impressed. In about 15 minutes I've made myself some nice sections of 'true' 5mm shaft stock that are within a tenth of target dimension & desired finish. And its a very controlled process. I got the idea by (carefully) mic-ing various wet/dry papers in the 600, 1000, 1200# range. I was impressed with how consistent they are across the sheet & across other sheets of the same vendor. For example the 600# I have measure 0.0075". Amazingly for some reason my 1000# thickness was quite close to 600#, within tenths, otherwise I would make a separate lap for each grit. So I worked this dimension in as an allowance, both annular (surrounding the stock) and between each lap face. Hopefully the sketch makes sense.
So I started by milling a block of aluminum, cut in half so it gave 2 symmetric pieces. Clamped them together face to face in the vise & drilled the oversize hole (0.0075 annular = 0.015" diametric). It would have been much better to ream the hole for a better finish but I didn't have the right size. So I chose the next closest numbered drill I had. There will be some buffering with the paper itself I guess. Then separate the 2 lap halves, position in vise & simultaneously mill off the face allowance from both. make a small transition radius on the hole corners The idea was to place 2 strips of abrasive paper clam-shelling the stock. If everything is dimension-ed reasonably close, by the time you fully squeezing the assembly together, it should be getting close to target dimension. You could spin the stock in the lathe, or as I did, just grip it in an electric drill. The benefit is the 'tool' never changes shape or wears out by lapping action. Fresh abrasive is a postage stamp area of wet/dry paper.
Well it worked out way better than I expected. You can actually feel a bit of vibration initially, which are the eccentric bumps of the shape. I used 600#, then 1000# mic-ing every so often. A longer section of shaft (3-4") means you are probably lapping more consistently across any one area. I figure 5 seconds of lap-sanding was about 0.0001" so it was actually quite easy to sneak up on the target dimension & finish simultaneously. I tried 2 strips of paper & also a single wrap just to get a sense of how much of the paper was in contact & cutting.
This is not a good way way of removing too much stock. Its kind of like lapping but without all the mess. Digital photo's tend to exaggerate machining marks so hopefully you can see the contrast of the finished rod & the aluminum tool. Drill holes are nasty looking surfaces under magnification once you split them open.
