Fading knurls

Glenn, Turn your knurling tool square and plunge to proper depth on the existing knurl. Once you establish the pattern to depth, it's virtually impossible to split. Be certain to lubricate well, and run at a moderate speed, not too fast, not too slow. Coarse feed is preferred to fine feed.

Moving the carriage back and forth rapidly when starting a knurl often will prevent the knurl from splitting. As you move the carriage, rapidly plunge the cross slide into the work, trying to go to depth quickly so the knurl formed forces the rolls to track properly. That is the secret to preventing splitting. If you have more difficulty getting your knurl to come to full depth at the end, there's nothing wrong with tilting the knurling tool slightly deeper on that end to get the depth you desire. Try to not back track your entire knurl if possible. If you have difficulty blending the rework, allow the carriage to feed into the old knurl and very slowly back out the cross slide. That will blend the old with the new and you'll not be able to see where they came together.

I'm assuming you're talking about a diamond knurl, not a straight knurl. You'll notice that the looseness of the knurling tool allows the rollers to move about, always leaving a perfect knurl if you start right. In theory the diameter is rarely proper for correct , but the wandering knurl, (side to side) makes up for the error in spacing, leaving what appears to be a perfect knurl. In truth, it's not in a straight line, but you can't tell when looking at the knurl.

If you're knurling stainless, you risk work hardening the material, making is progressively more difficult to knurl, so be certain to get it right as quickly as possible. You can't allow the knurling tool to run endlessly.

Good luck, let us know if it doesn't work.

Harold

In principle, yes. In practice, no.

1) the tolerance on that is basically the crest-to-crest distance between the ridges of the knurl. Really small. If you get close the knurl will pick up and track. 2) knurling actually changes the diameter of the workpiece. If it's not exactly right when you start, it will be in about ten seconds after the tool meets the part.

Jim

================================================== please reply to: JRR(zero) at yktvmv (dot) vnet (dot) ibm (dot) com ==================================================

You would probably just have to find the depth that puts you back in rthym with the knurling and reknurl it with the tool square to the work.

I have found that if you intentionally skew the knurls a few degrees so that the leading edge is slightly closer to the work it is often easier to get a good knurl. I think it is because the knurl points are loaded less or slightly differently at the point where most of the work is being done.

I usually knurl down to the finish point, stop the motor, then reverse it off the work.

I doubt if there's a book, but I think there are basic instructions in the old Atlas lathe manual, and probably something in the South Bend lathe manual.

You're right about the pitch never being quite perfect on paper. What happens, basically, is that you plunge the tool into the work until it finds a diameter it likes, and that produces a uniform knurl. It doesn't take much fooling around to find one, either. You usually wind up with a couple of choices, depending on how deep you want the knurl and how sharp you want the points. As Harold says in another message in this thread, a diamond knurl also requires some looseness in the tool so it will follow a uniform multiple of the knurl pitch and thus produce a clean knurl.

Until you try it, you won't believe it. Give it a try if you have the machine and the tool. I was amazed the first time I tried it. Then I started knurling everything in sight. This is an early stage in learning to machine. You can tell such a person by the knurled handles everywhere in his shop. You should see the knob on my tablesaw miter fence.

Ed Huntress

That is what I found. The leading edge turned a beautiful knurl but the back side faded out. Being an oversized hand turned turnbuckle body the diameter of the part was not critical so I turned the ends down to where the pattern is clean and the knurl stands proud. After a short spell in the electropolish pot it looks like it was meant to be.

Sometimes you can make silk purses out of sows' ears. :-)

From yours and Harold's descriptions, it sounds like the knurling wheel slides left and right on its axel (axial play) as it cuts...is this correct? Does it jump, or does it oscillate in a sinusoidal pattern (like a pendulum swinging back and forth)?

How much axial play should the wheel have? (I would eventually like to make my own knurling wheel holder.)

Thanks, Michael

What kind of setup were you using. I was all afraid of knurling before I tried it. The first time I did it came out PERFECT! I've never gotten it right since. No idea what I'm doing wrong or what I did right.

I'll have to measure the OD of the piece that came out right, I still have it, and check to see if the math works out.

John

Please note that my return address is wrong due to the amount of junk email I get. So please respond to this message through the newsgroup.

Hey Glenn, you got photos and/or description of your electropolish pot? This is something I'll probably be getting into sometime in the not too distant future. How does the finish come out?

John

Please note that my return address is wrong due to the amount of junk email I get. So please respond to this message through the newsgroup.

Hmm. I just re-read Harold's post, and I don't think I'm following him. And I may be making an erroneous assumption here. It's been a while since I've used the things and I may not have been paying attention. I can't really say

*what* the role of the endplay is. It just looks like clearance to me.

My diamond knurling wheels have around 0.010" of end-play. Now I have to go run them some time soon to take a better look at what's happening. I'll have to look around the shop for something that isn't knurled.

My point is, they work, and it's easier than it ought to be.

Ed Huntress

Usually if the leading edge was right the whole knurl looked that was after I walked the tool on and then back off. I think I would usually apply a little more pressure to the infeed after I reversed it.

I gave up and accepted some pretty light knurling on some stuff I made out of tool steels like W-1 and O-1. And it was in the annealed state.

You bet, my Dad always used to say "Even a blind pig occasionally finds an acorn." Usually when I had unexpectedly gotten something right.

I've seen quite a few responses, but one thing which seems to not have been covered is what style of knurling tool you have. I'll list the ones which come to my mind, and my opinions of them and what to watch out for.

1) The original "bump" knurlers. First was simply a pair of knurling rollers in a tool holder to fit a lantern style toolpost.

Later ones had turrets with three grades of knurls in a similar holder

For the quick-change toolpost (in the starter sets) , there is usually a single pair of knurling rollers in one end of a tool holder, with the other end used for a facing or boring bit.

All of these share the same problem to some degree or other. That is that they put a large stress on the carriage, cross-feed, and compound. The forces are high enough so unless the compound is locked down firmly, it may rotate (or the toolpost may rotate) allowing it to back away from the workpiece as it travels.

Also -- the workpiece can bow away from the knurler if it is long enough to allow that much flexing

2) "Scissors" style tools. The two knurls are mounted on hinged arms, and closed onto the top and bottom of the workpiece by a bolt and nut connecting the two sides. Some actually are close to scissors -- and form a big 'X' with the nut out where you can get to it easily to tighten it. Others (such as the one which I have) are set up with the bolt and nut between the pivot point(s) and the knurling rollers.

Mine is by Eagle Rock -- though that was not clear from the MSC listing, until I got it.

These have the major advantage that most of the force required is provided by the assembly, and not passed on to the lathe's moving components.

3) Similar to scissors is the fancy knurling tool from Aloris. It has the two knurling rollers on arms which travel on ways on the front of the toolholder. A feedscrew is mounted vertically, one end RH-thread, the other LH-thread, so as you rotate it the arms move closer together or farther apart, but maintain centering around the workpiece. I got one of these from eBay, and now use it in preference to the scissors knurler. 4) T-head knurlers -- similar to scissors in that they apply the knurling force from opposite sides of the workpiece, but these mount in a turret in the lathe, and are fed on from the end.

The 'T' is actually more of a 'Y' with flat bars on the tops bored to hold the following:

The knurls are mounted in the ends of round rods which are grooved to be guided by pointed setscrews. There are provisions for adjusting the rods in to the proper position. The grooves allow you to take straight knurls and set them for either straight knurling, diagonal knurling (if you have a need for that) or diamond knurling -- with choices as to 30 degree or 45 degree angles.

These are very nice for knurling short distances onto the workpiece, but awkward for knurling the middle of a long workpiece -- or just knurling a long workpiece, unless you have a turret without the central lock pin which mine has, and nothing in the opposite station. :-)

5) Cut style knurlers. These don't emboss the workpiece, but rather remove metal to form the knurls, thus the size of the workpiece does not grow, and you can wind up with very sharp knurling The knurls are mounted at an angle to the axis of the workpiece.

I don't have one of these -- they keep going for more than I can consider acceptable even on eBay, let alone in the catalogs at new prices. :-)

6) Turret cut style knurlers. I must confess to having one of these, and expecting to use it the next run. Think of a 3-jaw chuck, with the knurls mounted on the tops of the jaws, with the tops angled to set the proper angle for the knurling cutters. The scroll plate for this is turned by a knurled ring on the OD of the body, and is calibrated as to the amount of infeed. Since there are only three jaws, you have to have two right-hand cutters and one left (or is it one right-hand and two left?) It remains to be seen what this will look like when I use it.

Now -- another thing which can help with the knurling of tougher materials. You can get knurls which are crowned, so only the center is doing the maximum penetration. I've used these with scissors style knurler on 416 Stainless. I don't think that work-hardens as badly as some of the other alloys. The knurl produced tapers off cleanly and quickly with this style of knurler.

Of course -- lube the rollers as well as the workpiece. Knurl aggressively.

I've never found a need to tailor the pitch of the knurl to the diameter of the workpiece -- as long as I was sufficiently aggressive in starting the knurling.

So -- these are my thoughts.

Enjoy, DoN.

Those work real nice, but the only objection to that design is, they make the holders for the rollers really solid and thick - so as a result you cannot knurl work very close to the spindle at all. The armstrong turret three-way one I have at home does better than the aloris (at work) in that regard. Usually this is no big deal, but I have found that the key to getting good knurling is to keep the setup as rigid as possible, and the extra inch or so of stock required to hang out of the chuck or collet does not help much, when using that tool.

Another point I would make is when knurling, I've found it helps to a) uses a live center in the end of the part, and b) apply most force towards the *tailstock* side, because this reduces the tendency for the workpiece to slide into or out of the work holder, especially when using collets.

Knurling a part held only in a collet will often lead to some interesting results. :(

Correct technique for applying sulfurized cutting oil in casa rozen:

1) apply liberal coating of oil to workpiece before starting.

2) begin knurling, aggressively.

3) once knurls begin to form correctly, pause the carriage traverse, and dip acid brush back into pot of oil.

4) apply acid brush to rotating knurls, forgetting that the upper one is rotating the 'other' way.

5) watch, stupified, as the acid brush gets sucked into the rollers and work.

6) smack self on head. Profanity used appropriately.

7) knock over container of sulfurized cutting oil while attempting to undo the damage. On a bad day this splashes all down your shirt. Cuss some more.

8) "What's that *smell*??" when you emerge from the shop at the end of the night.

Jim

================================================== please reply to: JRR(zero) at yktvmv (dot) vnet (dot) ibm (dot) com ==================================================

Hmm ... but since the pressure is applied between the top and bottom of the workpiece, the stickout problem is minimized -- until you get far enough out to where you should have the tailstock center helping things.

A good point.

[ ... ]

Done this -- except for neglecting to knock over the oil. I was using Molly-Dee at the time -- a rather expensive way to knurl. :-)

Molly-Dee smells better.

I'm waiting to see what using the latest acquisition of cutting fluid (an oil so full of flowers of sulphur that it looks like mustard) does on kurling. :-) IIRC, the name of the stuff is Sul-Flo, and it took a group purchase, thanks to the minimum quantity that the vendor was willing to sell. It works wonderfully on turning and milling, but it

*does* smell. :-)

Enjoy, DoN.

I'm not too well equipped in my head to describe what happens, but there is no diameter that can't be diamond knurled perfectly, and to a sharp V. You can turn a part in increments of .005", reducing it continually until you have no part left that will accept a knurl, and start a new, full knurl every time. That tells me that there has to be something else going on. I have taken notice that knurls do not necessarily press in a "straight" line as the part rotates. The slop in the tool allows the rollers to move side to side, with a rhythm of sorts.

pendulum swinging back and forth)?

Yep, like that! I think that the side motion makes up for the slight discrepancy in the theoretical diameter of the part, yielding a perfect knurl. If you terminate the knurl mid shaft, it slowly cold forms the metal until it is in a straight line, but it's evident that something changes, because the knurl has a slightly different appearance when forced to end that way.

In building tooling, one of the things that I often faced was a knurled knob that had a particular diameter specified. You start out somewhat undersized, knurl, measure, turn the OD, knurl again, measure, etc. With practice one can end up with a knurled diameter exactly as specified. Granted, inspection might overlook something like that, but in my field of work, I took nothing for granted. I had a reputation to uphold, and I did so.

I'm of the opinion that it would be more difficult to achieve good results with a knurling tool that has limited (side)clearance than with one that has a fair amount. Personally, I never have any trouble starting a knurl and getting it to a full V form. That's for a diamond knurl. When it comes to a straight knurl, that's a different story. That one can fail, which indicates to me that the side wandering plays a big role in diamond knurls.

The secret to success in knurling with a typical knurling tool (not a scissor type) is to not be timid. Don't start the rolls with very little depth of feed. Go for depth quickly, which prevents the rolls from splitting. Works for me.

Harold

There's a pretty good description of the process here:

Greg

Any designer who delivers a print where a knurled diameter is specified to +/- 0.0002 inch should be taken out in the back yard and be pelted with possum carcasses. Relentlessly.

IMO.

Jim

================================================== please reply to: JRR(zero) at yktvmv (dot) vnet (dot) ibm (dot) com ==================================================

circumference

Hi Michael

This is a post that I wrote back in 99. I believed it to be valid then and still do.

Search Result 2 From: darby ( snipped-for-privacy@sympatico.ca) Subject: Re: Knurls View: Complete Thread (28 articles) Original Format Newsgroups: rec.crafts.metalworking Date: 1999/01/28

When it comes to knurls, I'd go easy on the math because it (the math) all goes out the window as soon as you tighten the knurl a few thou and it starts making an "impression"

If you start to knurl a 1/2 inch center punch (shop 101), what value should you use for your equation? 1/2 inch. I think not. With the knurl just touching it would be 1/2 but now you will want to wind the knurl in enough to get the result you want. If you advance or withdraw the knurl .010, the radius of the punch will change by .010, the diameter will change by .020, the circumference will change by .064

As you are well aware a change in circumference of .064 is more then enough to destroy the values derived from this type of calculation.

To each his own, but I subscribe to the notion that , when you first bed your knurl into the work you should get a good deep impression. (You can do this by rocking the work back and forth by hand a couple of times.) Having done that you will have created a few perfectly formed impressions on the work. These impressions are pyramid shaped with a sharply pointed top, like a tooth on a gear.

When you start to knurl, on completion of the very first revolution the perfectly formed teeth of the knurl will arrive at the perfectly formed "impression" you created. When these two sets "of teeth" meet they will mesh (they are forced to) and depending how they meet it is just a question of which will slip ahead or behind the tip of the other.

Of course there is no law that demands that you "bed in" a knurl when you start because everything may just work out perfectly and you could produce a perfect job a high percentage of the time anyway. But if you do not bed the knurl well, those first few "impressions" will be ill formed and have a flat top (as a mater of fact before they are fully formed they appear to have a con caved dimple on the top) This flat top allows the knurl the opportunity to start a new path of impressions. (bad)

Regards Bill Darby

Thanks Ed, I'll print this out and try it on the lathe.

John

Please note that my return address is wrong due to the amount of junk email I get. So please respond to this message through the newsgroup.