PDA

View Full Version : Knurling Questions



KEJR
09-03-2011, 09:44 PM
Hello,

Like many here I read the knurling article in HSM and went into my shop to try some knurling of my own. I had some 5/8" 6061 Al that I put in a collet and made a clean cut. I was having terrible problems getting a good knurl pattern to form and it was frustrating. I was applying oil with a brush and quite a bit of powdery Aluminum slurry was being generated.

So here are some questions I had regarding knurling:

- Is it OK to advance the cut by hand as long as you go slow?
(I don't have power feed at the moment).
- Can you cut back and forth, or are you better to make cuts in one direction only?
- How do you know how deep to go? Once the knurl engages for me Its covered with crap and I can't see the knurl well enough.


Many of my cuts seemed to be much finer than the knurl pitch so to speak and not well lined up. Sometimes I would get rectangular looking knurls.

I am using a phase II quick change adapter in the "AXA" size toolpost with the stock supplied cutters. One thing to note is that there is significant play in the wheels from side to side. I don't know if this is the type of thing where the knurls start tracking but when I advance from side to side one of them shifts and it screws up the cut?

Lets ignore for now the load on the headstock or crosslide bearings. I'd like to use this tool if it is going to work since I will probably knurl twice a year so it shouldn't wear out my half worn out lathe much more than it already is.

Thanks,
KEJR

gwilson
09-03-2011, 11:05 PM
You are probably taking too long to make the knurl. You are supposed to complete the knurl in just several revolutions. You need to bump the knurl in pretty good to start out.

I don't do much diamond knurling myself,as I find it unattractive. I use old style convex microscope type knurls,and other special knurls. Fine knurls work better on brass than coarser types because it displaces less metal.

You might try a finer pitch of diamond knurl on your aluminum,too.

I'm not too coherent after no power for several days,and generators running all night !!!:)

GadgetBuilder
09-03-2011, 11:15 PM
I have a 7x12 fitted with a Phase II AXA and I always feed by hand at low spindle speed, about 30 RPM.

Knurling is a cold extrusion process rather than a cutting process so the forces involved are quite high.

I found the knurls that came with the Phase II were too coarse for most of my use so I got fine diamond knurls. The force required for knurling seems less for fine knurls. The original knurls produced a surface that was uncomfortable to grip if fully formed. File and/or sandpaper fixes this.

The force required with the bump knurler was more than my 7x12 could handle so I built Martin Cleeve's clamp knurler and found it made the process much easier... on me and the machine.

Knurls sometimes fail to sync and may then double, producing a finer knurl than normal. I'm from the "more force" school rather than the "proper diameter" school but this subject is best avoided in this group :D

It is generally best to complete knurling in a single pass. Multiple incremental passes cold work the metal repeatedly and can lead to chunks of the knurl peeling off on later passes in some materials, particularly common when the knurl doubles.

I oil the knurls and the work heavily prior to starting, get the knurl synced on the first 1/8" or so, get the depth I want set (stop the lathe to clean and inspect if needed) and then slowly crank along the work for the desired length. It doesn't seem to make much difference if you stop at that point and release the knurl or wind it back the other way and then release.

Side movement of the knurls on their shaft shouldn't affect sync. Once the knurls sync they are effectively geared to the work and will do what's needed to maintain sync. That's why it's important to establish sync on the first 1/8" - you can turn this off and try again but if you knurl the whole length it may be a problem. Side movement can leave a little area at the end of the knurl with only one knurl impression, easily removed by adding a shallow groove etc.

It's important to clean the knurls thoroughly after use. Any knurling swarf left on the knurl could be pressed into the next work item and it can cause corrosion if it is a different type of metal.

wooleybooger
09-03-2011, 11:33 PM
are your cutters clean? i use a dental pick and a brass "toothbrush" to clean mine before using. dont back the tool out until your finished,if you back the tool out and try another pass youll have a mashed-up mess. always go left to right. dont run the tool off the end. stop and reverse carriage feed direction when the tool just begins to go off the end of the piece. work piece spring will give bad results as well as too slow a feed. feeding by hand is fine for a very short length but is hard to maintain a consistent speed over a longer length.

mf205i
09-04-2011, 03:47 AM
Machineryís Handbook 27th edition, page 1240. Basically, part diameter should be an even multiple of 1/64 inch for 64 and 128 DP, and a multiple of 1/32 inch for 96 and 160 DP. I think that this is the reason that some foreign knurls give problems as they do not conform to the ANSI standard and therefore require experimenting to get the correct, probably metric, diameter.
The other way is to set the tool at a slight angle to the work and jam it into the work hard and fast. This method works on a heavy lathe but, unless the part diameter conforms to the above, almost never works on a light lathe. This is especially true on hard materials. Using knurls with a narrow face will greatly improve your success rate as they take less pressure to get that critical heavy upset on the first revolution. You can easily narrow the knurls face with a grinder, but donít let them get hot.
Also have a look at this http://www.formrolldie.com/new_page_1.htm
Best of luck, Mike

John Stevenson
09-04-2011, 07:19 AM
So what happens if you do not have the ANSI standards or have read P 1240 in MH ?

Does that mean you will never be able to knurl correctly ?

KiddZimaHater
09-04-2011, 11:18 AM
I do alot of knurling in aluminum. One trick I learned was to bump the part, let the knurl start tracking, then withdraw, and bump into it again a litte harder the second time. Then engage the feed (Or feed by hand in your case).
And try using WD-40 instead of cutting oil on aluminum.
Don't ask me why this works, but it does most of the time.
When I get a double-pattern, or an ugly knurl, I'll file it down and try again.
Aluminum is very forgiving.
Pactice practice practice.
A knurl doesn't have to look like a million bucks, it's just there for grip.

DATo
09-04-2011, 11:48 AM
I'll go along with what mf205i said about setting the knurling tool on a slight angle. Also, I always blow the workpiece with compressed air while the knurling is in progress to keep any chips generated from galling to the finished knurl surface.

Edit: Yes, it is OK to knurl short lengths by hand. I do it often and with good results.

J Tiers
09-04-2011, 01:47 PM
So what happens if you do not have the ANSI standards or have read P 1240 in MH ?

Does that mean you will never be able to knurl correctly ?


Seems like an angled/diamond knurl will work on any size part. The knurlroller can slip a bit to allow a slightly different pitch.

A straight knurl has less "slop" factor, and might need to be closer to ideal. Or the resulting knurl might not be completely formed to the depth of the teeth on the roller.....or the knurling process brings the size to a "good" size for the pitch.

Carld
09-04-2011, 02:40 PM
The first mistake was putting it in a collet. The collet jaws may only support the work for about 1/2" to 5/8" while a 3 or 4 jaw will have an 1" or more support. Since the knurling pressure is great the force of knurling will move the shaft in the collet.

Use a chuck, it works better that way.

JoeLee
09-04-2011, 03:33 PM
By now most of us have foiund out that knurling on our small tool room lathes isn't as easy as we thought it would be, even with softer materials as everyone seems to get less than desirable results. Myself included. There is no alignment needed for diamond knurl pattern as the helix will cross somewhere as long as your pattern doesn't start to overlap or crawl, in which case you'll end up with a chewed up looking mess. Most of us don't really have lathes big enough to exert the kind of inward force to get a good pattern in a couple revoloutions. If you knurler is an Aloris type which is center off set you might find your tool post starts to move, so we use less force and morte revoloutions which can also lead to a chewed up mess as the material breaks down and filings become lodged in the knurls (flaking). What I've found that works best is to angle the knurls so you are only using a corner of the wheel which takes much less force. The sissors type works best as it takes most of the load off the cross slide screw. If you think diamond knurls are hard to do then try straight tooth knurling.

JL................................

mf205i
09-04-2011, 03:36 PM
So what happens if you do not have the ANSI standards or have read P 1240 in MH ?

Does that mean you will never be able to knurl correctly ?

Do as you like, but utilizing the standards is the difference between knurling a part and hoping it will be good and knurling as many parts as you wish and knowing that they will be good.
Mike

Forrest Addy
09-04-2011, 03:38 PM
Knurling is pretty simple but you have a few hoops to jump.

One is pitch. Mike the knurl, count the teeth. Divide the diameter by the tooth count. This gives you a "pitch" figure you use to calculate the knurled diameter. The knurled diameter is some multiple of the knurl "pitch". Miss by very far and the knurls won't track properly most likely you'll split the knurl and get extra fine knurls or some odd messy looking pattern. There are those who say you don't have to jump through this hoop of calculating. When I take the time do it I never miss making a neat, strongly patterned knurl on the first few revolutions.

Another is commitment. If you're going to knurl, mash the knurl into the work. You can't sneak up on it. This means a lot of force against the work so it has to be supported. Either the work has to be short and strongly gripped in a chuck or if longer supported with a center. If you have a lot of detail work or the part finishes with thin walls knurl it early in the game.

Yes, roller knurling is a drastic treatment where metal is plasticly deformed. When metal is repeatedly deformed it fatigues. Thus it's not surprizing that knurling generates "mud" - finely divided metal flakes; fatigue products. Use plenty of oil to flush them away. Roll that icky mud back onto the knurled surface and the appearance suffers.

By the way, always use oil for knurling. The rollers are hard and the pin they bear against is hard. Water based coolants are poor lubricants. Use cutting oil of almost any type for knurling.

You will see the pointed crests of the knurled surface rise up hollow and close like barnacles. These crests are often crumbly concoctions of pressure welded fatigued metal, weak and unstable. Dress them off with a fine file to clean metal and knurl some more to raise the knurl to near points.

How much or how little feed depends on a number of variables. Make sure the knurls present square to the work surface and you get a complete annular start before you begin to feed. A feed af about 1/3 the knurl width works pretty well for me. More for soft materials and less for harder or tougher materials. If there's doubt, you might experiment a little on the material sometime when roughing. Run a knurl on a roughed surface to establish parameters then machine it off and continue manufacture.

Jaakko Fagerlund
09-04-2011, 04:03 PM
The first mistake was putting it in a collet. The collet jaws may only support the work for about 1/2" to 5/8" while a 3 or 4 jaw will have an 1" or more support. Since the knurling pressure is great the force of knurling will move the shaft in the collet.

Use a chuck, it works better that way.
Strange collets you have there, if they only grip so little. At work I have a collet chuck that grips about 75 mm (3") and that is plenty more than a 3-jaw chuck. Besides, a proper collet (chuck) has more surface area contacting the part than a 3-jaw chuck.

Carld
09-04-2011, 04:54 PM
Ok, my bad Jaakko, my R8 and 5C collets have about 1" of support. I doubt most average lathes uses a collet that has much more support length than 1". A collet will not hold work as well as a chuck. Many times I have to use a collar clamped on the work to keep it from working it's way into the collet.

I still maintain that collets are a poor choice for knurling or any heavy work. When we did boat prop shafts at the last place I worked and we used large collets in a lathe but we also had to use a collar that fit the shaft to keep it from working into the collet as the taper was cut and then threaded for a nut.

KEJR
09-04-2011, 06:19 PM
I practiced some more today and got a good looking knurl. I started with some new stock, fed the knurl in fast, and slowly went from left to right.

It worked fine in my collet. I'm more concerned with the part flexing than my collets. My Hardinge spindle has a 5C taper so unless there is a special case I use 5C collets for everything. Alot of this has to do with the fact that I haven't made a backplate adapter for my chuck yet! I have access to other lathes at work so I'm going to make one up out of cast or C1018.

Anyhow, I appreciate the advise here. Angling the knurl to get a little lead angle doesn't sound like a bad idea. If nothing else it makes sure you have some potitive lead instead of negative. I'm also glad to hear that others have had success with starting the knurl and then stopping the machine to check the knurl after cleaning.

Even though I got a decent looking knurl I noticed that it could have been a tad deeper and it seemed like one wheel was engaged more, creating a somewhat rectangular diamond. This just needs to be set better, obviously.

I'll also try to make sure the part is in 1/32 increments. Doesn't seem too hard to do!

Thanks,
KEJR

Carld
09-04-2011, 10:38 PM
UHHH, yeah, the part is flexing or moving in the collet, the collet sure the hell isn't moving in the collet holder. In time it will wear out your collet.

I repeat, collets are not designed for high side loads but they are your collets and your money.

Jaakko Fagerlund
09-05-2011, 01:33 PM
UHHH, yeah, the part is flexing or moving in the collet, the collet sure the hell isn't moving in the collet holder. In time it will wear out your collet.

I repeat, collets are not designed for high side loads but they are your collets and your money.
Could you cite where this "collets are bad" information comes from? If you know what tolerances the collet can securely hold and use material that has a good quality surface on it, the collets won't slip or let the work move towards the collet. And they will hold side loads well, easily noticed when milling with a big cutter mounted in a collet and when parting of stuff in CNC lathe that has bar feeder and makes automated production with a collet chuck.

Tony Ennis
09-05-2011, 01:51 PM
Re: the exact perfect diameter... I'm pretty sure John S posted a picture of a beautiful knurl he cut... on a tapered part.

Carld
09-05-2011, 03:26 PM
Jaakko, it comes from working in shops with collet chucks on lathes and trying different things with them. We had one machinist that was stronger than a bear and he couldn't tighten a collet chuck on his lathe tight enough to keep a SS boat shaft from working it's way into the collet while cutting the taper or the thread while using a live center in the end of the shaft. The only way we could keep the shaft from moving in the collet was to make sleeves for each shaft size to clamp on the shaft to hold it in place. We didn't baby the cuts with little bitty cuts, we had to get them done as fast as possible.

I tried once to knurl a shaft in his lathe with a collet with a live center in the end and the pressure of knurling pushed against the live center hard enough to force the shaft into the collet. I never had that problem with a 3 or 4 jaw chuck.

If it works for you keep doing it and it may have worked for me if I took light cuts but knurling, I doubt it would ever work. I think the collet set we used there went up to about 2", maybe 2 1/4" I can't remember for sure. I believe it was a 16" South Bend lathe that he used and we had two of them and about 8 other lathes and other machines.

ftownroe
09-05-2011, 04:28 PM
The following link from the folks at Little Machine Shop has a formula for an Excel spreadsh
eet that will calculate the diameter for you and save time. Scroll to the bottom of the page.

http://www.littlemachineshop.com/Reference/Knurler.php

KEJR
09-05-2011, 04:58 PM
I did some more tooling around and found that some of my issues were tracking problems due to not being fed enough. I also found that my compound tilted while I was advancing the tool even though everything was tightened down fairly well.

So it looks like at some point I might take the plunge into a scissor type tool. Just about everyone says that this is the way to go.

Thanks,
KEJR

John Stevenson
09-05-2011, 05:12 PM
Re: the exact perfect diameter... I'm pretty sure John S posted a picture of a beautiful knurl he cut... on a tapered part.

Black art, smoke and mirrors, doesn't work, can't work because I haven't read that book or got that spreadsheet. :rolleyes:

Magicniner
09-05-2011, 06:23 PM
He's a Witch! Burn Him!
;-)

MotorradMike
09-05-2011, 06:35 PM
I remember the picture of Sir John's knurl on a stepped shaft but I can't find it.
Here (http://bbs.homeshopmachinist.net/showthread.php?t=38568) is another post of his that is equally impressive(to some of us anyway).

John Stevenson
09-05-2011, 07:25 PM
Here's the pictures.

http://www.stevenson-engineers.co.uk/files/stepknurl1.jpg

Scrap piece of bar, no idea what size it is but a series of steps cut on it and marked with felt tip marker.

http://www.stevenson-engineers.co.uk/files/stepknurl2.jpg

Knurl started at tailstock end, just pushed in until it came out nice then power feed all the way along.

You can see the original felt tip marks under the knurl.

No books or spreadsheets were harmed in the making of this knurl.

hareng
09-05-2011, 08:15 PM
Quite agree John but you have a robust machine there:D

Likewise i do a lot of round knurling 'end on' Now all you book readers find that.
The old saying springs to mind a little knowledge is dangerous.

Lew Hartswick
09-05-2011, 10:44 PM
I tried once to knurl a shaft in his lathe with a collet with a live center in the end and the pressure of knurling pushed against the live center hard enough to force the shaft into the collet. I never had that problem with a 3 or 4 jaw chuck. .

??? If you're knurling towards the live center (I ASSUME in the tail stock)
how in the world does the work get pushed into the collet (again
assuming the collet is in the headstock) ?????
...lew... We always have the students knurl toward the tail stock it
doesn't move.

Carld
09-05-2011, 10:49 PM
I knurl towards the headstock then back towards the tailstock back and forth until the knurl is done. There's no problem going toward the tailstock, just toward the headstock.

Maybe new collets in a new collet chuck would hold well enough to knurl without slipping but on a two different machines with collet chucks we had trouble with work slipping back into the collet.

Never tried to figure out why it did, just worked around it.

noah katz
09-06-2011, 01:29 AM
What prevents the knurls from being made in a different spot every revolution, unless you luck out and the knurl pitch divides evenly into the stock diameter?

J Tiers
09-06-2011, 01:35 AM
I've calculated it, and I have just "done" it.... both seem to work, but I had considerably more hassle getting a goodie with the 'try it" method. Dunno why, might not have had any relevance.

But I was usually able to get a good knurl and no splits with the calc..... got splits with the "any size and cram it in" more often

If there is a specified OD, is that WITH knurl, or BEFORE knurl......? because the knurl changes the diameter, at least the overall OD.

And that may be why it works.... if you go in until you get a "good knurl", does it end up on a target dimension?

Or does some of the shaft end up shaved off and in the chip pan, with what is left at the "right size" for that knurl, and the final OD a bit of a question mark?

Ernie
09-06-2011, 04:28 PM
I followed the above link to littlemachineshop and played with the spread sheet. Because his formula rounds off to the next smaller size, it can be misleading depending on the diameter of the stock you're about to use.

For instance, if your stock measures .762", it wants you to turn it down to .752". In fact, any diameter from .762" down to .753" should be reduced to .752". If your stock happens to already be .752" and you plug it into the formula, now it wants you to turn it down to .743" but it already says .752" was the perfect size???

Ernie