Knurling

[NormParticipant @norm]

Can anybody help me to produce clean diamond knurls?

I am using a doule headed knurling tool with 2 daimond knurling forms.

It seems to me that the dia of the shaft to be knurled has to be a multipule of one diamond pattern. At present I get mangled diamonds and fine slivers of metal. It must be simple?

[NormParticipant @norm]

[Steve GarnettParticipant @stevegarnett62550]

I’d be the first to admit that I don’t do a staggering amount of knurling, but what I have done (and a bit of investigating) led me to the following conclusion:

 

Yes, the circumference of the piece you are knurling has to be exactly divisible by the spacing between the individual knurls, or you get multiple cuts, as you are describing. My experience with doing this so far on circumferences that are very slightly out is that if the wheel is ‘almost’ there, it will slide into the correct cut – but this is very limited in its extent, and only seems to work if there was a decent level of cut on the first pass.

[mgjParticipant @mgj]

I always thought that too – but it doesn’t seem to be so, because I have one of the Hemingway scissor (clamp) type double knurls and it works perfectly whatever the diameter.

 

Assuming the knurls are of good quality the trick, I’m told is to go very slowly and add lots and lots of coolant. So, I run the lathes at about 50 RPM and add lots of coolant.

 

Lots of coolant washes the chips away and ensures you get nice clean professional looking knurls. Also with that clamp type knurling tool you can apply quite a bit of pressure without loading the lathe bearings.

 

I used to use a single knurl – straight in type. Again that worked very well at low speeds with a lot of coolant, but I prefer not to load up the lathe bearings. 

 

I’m afraid I have no experience of knurling without coolant. I know people say their brush and dab systems are marvellous, and I’m sure they are, but somehow when I use tiddly amounts of coolant, I find myself turning the tap up sharpish, so forgive me if I have my doubts.

[Tony Pratt 1Participant @tonypratt1]

I have done a fair bit of krurling and it doesn’t seem to matter what the work piece or  the krurl diameter is. As has been said before use good quality sharp knurl wheels, slowish speed and plenty of coolant. I only use a scissor type knurling tool as I don’t like the idea of the lathe bearings having to resist the sideways knurling force if using only one wheel.

I start with the wheels overhanging the end of the job and tighten the scissor nut to start cutting, once it gets cutting properly the wheels seem to find their own tracking and a decent krurl appears, you can then traverse slowly to get the krurl length required.

To reiterate the krurls must be sharp and use plenty of coolant, I have tried the dab and brush method but it tends to get a bit hairy when the brush gets trapped in the krurl wheels!!

[Steve GarnettParticipant @stevegarnett62550]

I must admit that my knurling experience has been with a pair of wheels with relatively wide pattern spacing, and these are the ones that double-cut. I recently acquired (but haven’t used) a much smaller scissor-type knurler, and looking at the cutter spacing on the wheels with this, they appear to be much closer together – to the point where the cutters could hardly avoid sliding into a previous cut even if they wanted to. And I’d also presume that using a copious amount of coolant/lube would help that process quite a bit – hence more consistent results.

 

So I’d guess that that the OP might be using a relatively widely-spaced pattern wheel – in which case I’d still say that double-cuts and slivering can easily happen.

[ZigFireParticipant @zigfire]

Hi,

Don’t do a lot of knurling but an old timer showed me his way which works well.

Start the knurl with the wheels overhanging the job, say 30% on and 70% off the job.

Also don’t have the knurling tool perpendicular to the work, have it a little opened up (say 1 degree).

Start the job and slowly feed tool into work before engaging travel, Slow speed and plenty of lube and result should be okay. No need to have a specific dia to match the tooling. As usual trial on some scrap first.

Cheers

Michael

Edited By ZigFire on 02/02/2010 03:23:31

[Ramon WilsonParticipant @ramonwilson3]

Just like to confirm that Tony and Zigfire have it spot on. Diameter isn’t important just slow speed and feed and – if possible – lot’s of coolant to clear the swarf. The small approach angle is very beneficial too especially on long knurling.

Like Tony lot’s of knurling has been carried out however we always would set the knurl (scissor type) to the diameter first by closing the knurls to the workpiece then backing off, applying the depth (one/two flats on the top nut) winding the tool back onto the workpiece and checking for depth. This was a much gentler way than crushing the knurl into the diameter using the nut and gives a slower ‘uptake’ at the start of the knurl. If the depth needs increasing do this on a second pass.

Where possible knurling on a raised ‘land’ is much easier as well, kissing the edges with a 45 tool giving a nice finishing touch.

 

This is one op where the home machinist can struggle because of that coolant requirement, I know I do as my set up is very much a brush and jar one – I do have some very nice very short haired brushes though! A ‘squeezy’ bottle with a small spout to get a good jet is okay too just use as much as you can. I wouldn’t recommend using the thicker high pressure cutting lubricants however because they tend to help the swarf stick in the knurls.

 

Keep on knurling

 

Ramon

[Steve GarnettParticipant @stevegarnett62550]

I should also like to point out that if the circumference/pattern don’t divide equally, and slippage doesn’t occur, then you will get double/multiple cuts – or knurling has somehow managed to defy the laws of physics/mechanics! I do recognise though that there’s a slight buggeration factor in this, inasmuch as the circumference will increase slightly during the process as the surface material inevitably gets deformed, and obviously predicting what might happen – well, basically it’s a non-starter.Maybe that’s why it always seems to work?

 

So to me, all the advice sounds good, especially the stuff about coolant, but it seems reasonable that the smaller the diameter you knurl, the closer spaced the knurl pattern on your wheels should be, because that is where the chances of multiple cutting increase.

 

Most of the things I knurl tend to be very narrow adjustment rings, just for a bit of grip. They are a lot narrower than the width of the knurling cutters, so feeding them in is a bit of a non-starter for me. And nobody particularly cares about double-cutting, because these things aren’t on show in the slightest and only need adjusting once every several years, as a rule, so it’s not really an issue.

[John Stevenson 1Participant @johnstevenson1]

I have proven that diameter isn’t a factor.

I took a bar of steel and machined it into equal steps at random diameters, each diameter a few thou less than the previous one and ran a knurl along the whole length.

 

The whole length came out perfect regardless of diameter.

 

I do a lot of knurling that has to be accurate as regards no twist and even.

 

 

 

This is a 30mm steel bar knurled in one pass at 500 rpm and flood coolant.

 

John S.

[Steve GarnettParticipant @stevegarnett62550]

Posted by John Stevenson on 02/02/2010 09:42:21:

I have proven that diameter isn’t a factor.

 Well, at least two of us have experience that is at odds with that, although I think that there may be reasons (explained in previous posts) as to why…

 

There’s a document I found (you can read it here) which seems to indicate in tip no.1 that the slippage thing is indeed critical. Norm, perhaps you should read it? It may help…

Edited By Steve Garnett on 02/02/2010 11:03:18

[NormParticipant @norm]

John S, how would it work with a diamond cut? I expect it would be ok with 2 opposing diagonal knurls but not using a preformed diamond set of knurls.

I have just read my posting again. Whoops, sorry about the spelling. Dyslexia or something I’m afraid!  But thank you everyone for your quick and informative responses.

Happy knurling!

Norm

[mgjParticipant @mgj]

Can you get preformed diamonds? I’ve never seen them, but then I have never looked.

 

I did quite a bit with straight knurls – (Dore Westbury, Quorn, GHT VDH , rotary table etc) Somehow that always managed to sort itself out and defy the laws of physics and circumferences. Only thing there was always to use tailstock support, obviously.

 

More recently I’ve always done diamonds using fine opposing diagonal wheels. Again, the laws of physics don’t seem to apply.

[chris stephensParticipant @chrisstephens63393]

Hi Guys,

What some of you seem to be forgetting is that knurling is NOT a cutting action, it is a deforming action, where the metal “flows” and is therefore more akin to metal spinning than standard lathe cutting. It is because of this deformation that diameter does not play much of a part in normal knurling.

A little thought for those who think diameter is critical,  do you pick the dia, at the peak or the trough, of the knurl, for your calculations?

chriStephens 

[mgjParticipant @mgj]

Or maybe like a gear- in the middle.

[Steve GarnettParticipant @stevegarnett62550]

Posted by chris stephens on 02/02/2010 18:11:19:A little thought for those who think diameter is critical,  do you pick the dia, at the peak or the trough, of the knurl, for your calculations

 That’s the point – you can’t calculate it, because it depends entirely on the deforming force, etc. And none of this knurling that works defies the laws of physics – the knurl wheel slips or rolls into the correct position if there’s plenty of lubricant and a suitable surface for it to slide down. I’ve only had it go wrong on relatively small diameters with one particular set of knurl wheels, and that has very widely spaced and deep deformers (if this is what we call them now?), and with a relatively shallow cut, which was almost certainly a mistake. It seems to be that if you accidentally start with a double knurl, the wheels find it very easy to continue with it!

[John Stevenson 1Participant @johnstevenson1]

Any idea why my picture has disappeared again ?

 

Been there all day now gone?

 

John S.

[mgjParticipant @mgj]

Steve- double knurls – that’s true.

 

You have to go for it, and that’s why I like the scissors knurls,  – you set the pressure good and tight, 50rpm and flood coolant and 10 secs later you have your knurled finish.

 

With apologies to the laws of physics, but some know that I am probably a lot more mathematical and rules of physics abiding than most. I am sure you are right – a modicum of slippage is the answer..

[chris stephensParticipant @chrisstephens63393]

I think Meyrick has it right, lots of pressure and do it quickly. This applies especially to materials that work harden, dither for too long and your chance of success drops markedly, and the chance of making more of those  nasty slivers of swarf increases. If those slivers are not washed away as soon as they are produced they will try to embed themselves in the job, ruining the appearance. Again Meyrick is right about flood coolant, don’t use neat oil (applied with a brush), it only makes the slivers stick. 

  

For those new to knurling, let me explain the reason you have both straight knurls and diamond ones. Straight knurls are intended for knobs that only have to turn, diamond ones are for knobs that have to be pulled as well as turned. Well in principle this holds, but you are entitled to do what what ever you want, the Engineering Police will not come a knocking if you get it wrong or only have one type of knurl.

chriStephens 

[russellParticipant @russell]

thanks everyone for contributing to this discussion.

 

there is something i am confused about. when i want to knurl over a length longer than the width of the rollers, should the rollers be square to the work (in which case, it seems i am trying to force the ‘leading edge’ of the roller into the work, resulting in a mess) or at a slight angle, so the knurl is developed in depth as the (longitudinal) feed progresses, in which case, the leading edge will not be as deep as the rest (assuming i cant feed right off the knurled section, which is likely as the head or tailstock will be in the way)

 

or is there something fundamental i am not understanding?

 

thanks…

 

russell

[Ramon WilsonParticipant @ramonwilson3]

Hi Russell

It is much easier on the movement of the knurl if there is a slight lead-in angle but this does have the effect you describe if you cannot  move the knurls well past, but preferably not off, the edge of the knurling area. Ideally it is more practical to try to arrange this situation but of course that is not always possible.

 

Depending on the type of knurling tool an angle is beneficial, whether push or scissor type – the  rather long armed Jones and Shipman scissor tool for instance would flex sideways considerably, the cutting resistance bending the tool away from the direction of travel so having a lead angle helped overcome this to a degree  and prevent the opposite happening ie the leading edge digging in. This is less of a problem when reversing the pass as the knurl was already formed. Once the travel stops this tool would then ‘pull’ itself along the knurl a little for the extent of the movement in it’s arms and due allowance would need to be allowed. Not good when it’s rather close to the jaws! but if quick enough the lathe could be run in reverse to overcome this effect!!

If the problem you describe of a thinner pattern  was encountered the tool was pulled back, the angle reversed and the tool slowly engaged again picking up the knurl. Not ideal but acceptable.

 

There is also the possibilty of course of making the area to be knurled slightly wider and then machining the ‘thin’ area away afterwards.

 

As most of the previous postings show ‘pressure’ knurling is not exactly a science, it can at times be be rather hit and miss, even in a works environment. Though some advocate – and are successfull – in using higher speeds I have always found the lower speeds give a more controlled approach especially when starting the knurl the, higher speeds tending to produce a fine crossed pattern and the higher speeds definitely needing the ability to get the chips away quicker as well.

 

The one thing I’m fairly certain of is that this operation proves far more thought provoking than most, especially to the novice. Like many things it is best to practice on some scrap first and gain some experience rather than wait until that perfectly turned new part is ready for that ‘experimental’ op to finish!

 

Hope this helps -Ramon

[Gordon WParticipant @gordonw]

I’ve never managed a good knurl, and this thread is very interesting. I’ve got a cheapo scisser tool, I get the impression that knurling under power feed results in more mangled patterns, and hand feed is better, might this be because hand feeding lets the knurls find their own thread, so to speak?

[Steve GarnettParticipant @stevegarnett62550]

I tried an experimental longer knurl under a very slow power feed, and it worked fine – if anything slightly better than a hand feed, I thought. (I’m normally forced to do them entirely by hand because of the size of them). I strongly suspect that if you do what Meyrick et al suggest and run your lathe as slow as it will go, use loads of coolant, a very slight angle (about 1 degree or thereabouts) and the slowest feed rate you have, they’d be okay. I don’t think that it’s how you feed the knurls in that’s really the issue – it’s whether the knurl can follow the deforming process accurately. And as long as there’s enough of the combination of coolant and force onto the job, how you progress it (within reason!) doesn’t actually matter.

 

Incidentally, what looked better on my  test was the overall finish. I suspect that this is because the material deformation at any given point was essentially the same, more so than I could ever have got it feeding that slowly by hand. But since everything – materials, speeds, coolant, tooling, etc – about this process will be slightly different for everybody, it’s very much a case of Your Mileage May Vary…

Edited By Steve Garnett on 03/02/2010 10:35:16

[John HaineParticipant @johnhaine32865]

If the circumferential pitch of a knurl is of the order of say 1 mm, the perimeter of the work will be an integral number of knurl cycles long as long as the diameter is an integral multiple of  “1/pi” mm, or about 0.3 mm.  If the depth you form is at least this much then surely the metal will deform to match the “integral number of teeth” criterion whatever the diameter of the work. (Adjust these parameters to suit the actual knurl pitch you’re using.)

 

This seems to correspond to what many people are reporting.

 

I have also read in MEW of a type of knurl which has a sharp face on the side and, mounted rigidly in the toolpost, is fed in from the end so that it actually cuts rather than forms the pattern – can’t remember the name offhand.  I suspect that with this type the diameter and in-feed will be critical.

[Steve GarnettParticipant @stevegarnett62550]

FWIW, the circumferential pitch of the wheels I get the double-knurls with is just a tad under 2mm. 0.63mm of deformation – ooh, that’s quite a lot…

[Author]

Posts