I just finished up resawing when the Enco sale email showed up (one day 20% off). I had the parts sitting in the shopping cart so that was an easy way to drop $150. The two knurls were $76; AccuTrak convex (crowned) knurls, lots more expensive than the $12 bevel knurls but boy do they work well. Strangely enough, AccuTrak sells the knurls for quite a bit less than Enco ($30 vs $38), the Enco catalog even cribs from the AccuTrak website.
I also ordered some case hardening compound; heat your part to cherry red, apply powder, reheat, cool in some way. No precautions listed on the container, no MSDS, but geezus, the heated power gives off copious amounts of something very ammonia like. Bet that scrubbed away some brain cells. Kinda PITA to clean up, messy but not too bad. I was hoping for a bit of expansion (which apparently happens when parts are commercially case hardened) as I don’t have the reamers to make a slightly undersized hole but I think the parts may have actually lost material (due to scaling). The compound does work but don’t expect miracles (ie I was somewhat underwhelmed).
The plans call for ¼” dowel pins (which are very slightly oversize) for knurl axles, I needed two, didn’t want to buy 100 so I bought some ¼” HHS round tool blanks for $5 (probably should have gotten carbide) and chopped off some chunks with the die grinder and abrasive wheel. They measure every so slightly under size so they are a slip fit in a ¼” reamed hole (they need to be a slight press fit because that is the only thing holding the axle in place). And case hardening didn’t make the hole smaller. So I used a center punch to make a half circle of dents around the hole, which moved enough metal to give the desired fit.
Choosing the correct diameter for your knurl. Man, it seems like everybody has their method. Basically, the problem is you need to make sure the distance from the last dent to the first dent is the same as all the other dents so the knurl rolls into the first dent after a revolution or things get mushed or double stamped. You can get diametrical knurls that work on [diameters of] multiples of 1/64″ or 1/32″, you can use formulas with Correction Factors, or without, mash until you like the result, yada, yada, yada. I used the “whatever” method for the videos, turns out that I was only one thou out, ie not a problem. I’ve added a video from Tom Lipton with a nice simple formula backed by tons of experience; I use slightly different math (same result) because it suits the way I think.
One pass or multi pass? This tool just rocks because there is no [side] load on the cross slide or part. So just set the depth and go. The “push the knurl into the part” just bites; the cross slide hates it, the part hates it, I hates it, the part is flexing all over the place, you have to take multiple passes to get to depth, it just sucks. With this tool, I’m not even sure the lathe noticed it was doing anything.
Enough yack, some videos: Tom’s Techniques videos on the knurling tool.
Assembling the tool:
In action knurling ½” cold rolled mild steel rod. Notice how far I have the rod sticking out of the collet, can’t do that with a conventional tool.
Tom Lipton’s take on calculating diameters for knurling: