View Full Version : Knurling to a diameter

02-09-2012, 08:07 AM
Hi all,
I need to knurl some 8mm bolts, and the section I need to knurl needs to remain 8mm in diameter. Is there a good way to estimate how much I should turn them down before I knurl? Thanks!

02-09-2012, 08:09 AM
I think knurling raises the diameter...

02-09-2012, 08:47 AM
Try knurling a sample at the full dia and measure the difference

02-09-2012, 08:49 AM
I'd do some samples first.

Depending on the dia of your part and the teeth per inch of the knurl you're using, the pattern may not work perfectly.

02-09-2012, 10:05 AM
I think knurling raises the diameter...

Knurling does indeed increase the diameter, as long as it is an OD knurl. ID knurls reduce the diameter, as in knurling valve guides.

The OP stated that he wanted to turn the diameter down so that the knurl would be same as original diameter.

I would just use trial and error. It might cost an extra 8mm bolt or two, but that shouldn't be much of a cost. Or knurl it and then turn the knurl down to the original diameter. There would still be some knurl left.


02-09-2012, 10:15 AM
Dorian has a simple X-cel spread sheet that will help you work it out. It tells you how deep you are supposed to knurl, your start diameter, and what your finish diameter is going to be. Gives feeds and speeds too, I find it pretty handy and pretty accurate. Free download on their site.

02-09-2012, 01:34 PM
do you have alink? dont see any^thing on their site.

02-09-2012, 01:46 PM
The diameter increased depends upon the knurl, but for a rough approximation, I believe you can count on about half of the distance fed into the piece to be displaced outward.

02-09-2012, 01:49 PM
do you have alink? dont see any^thing on their site.


In red about half way down the page.

Paul Alciatore
02-09-2012, 04:33 PM
You do not specify what kind of 8mm bolts you are working with. I would assume you are going to use one of the lower grade bolts as knurling a grade 8.8 bolt may be problematic.

OK, you do say it needs to retain the 8mm diameter. In the above context, that could mean several things. Most commercial bolts have their threads formed by rolling. Thus, the diameter of the threaded section before rolling has to be less then 8mm. Now an 8mm bolt may have an un-threaded section that is 8mm or it may have that un-threaded section at the same lesser diameter that the threaded section was before the thread was rolled on. And in either case the tolerances will be somewhat loose: in fact, the un-threaded section may not even be completely round. So, do you want an 8mm knurled section or do you want to preserve the actual diameter, whatever it is?

If your bolts have a reduced diameter un-threaded section, you may not even be able to increase it to a full 8mm by knurling because the knurling process raises the OD by a lesser amount than thread rolling will. But you do not even say what your tolerances are so the amount of increase you get may be enough even it doesn't actually reach 8mm.

If your bolts have a nominal 8mm diameter at the un-threaded area, even that will likely not be a full 8mm. So you may be able to use a partial depth knurl on it and stay at 8mm. Another approach would be to do a full depth knurl and then turning it down to 8mm after. If that does not work for you, then you need to decrease the diameter first. Others have pointed to sources for calculating the amount.

Video Man
02-09-2012, 05:28 PM
It depends on what pitch knurl you have, depth of knurl, etc. FWIW a program I have predicts: Enco medium knurl 21.3 teeth per inch, converted to metric: turn work to 7.35mm, produces an even 19 points with a final diameter of 7.64mm. This is the closest number of full points that the program predicts for the closest finished diameter. (A larger turned diameter would prouduce, theoretically at least, a partial point which would make bad knurl. This program picks the largest integral number of points to make a finished diameter closest to the target size.)

If you know your knurl pitch and want to try it, post the info and I'll run it again...

02-09-2012, 06:51 PM
Home Shop Machinist had a project awhile ago that cuts knurls. It doesn't form them and the OD would be unchanged as long as you didn't go too deep.

It's probably more project than you want for a few 8mm bolts though.

I have yet to get a really good knurl consistently, which is why I am in the process of making the knurl cutter.

As others have mentioned, trial and error will be what you need for the results you want.

02-09-2012, 07:30 PM
I have yet to get a really good knurl consistently,

Until recently I was a knurling hack, bad hack, horrible consistency, just crap.

I did some reading, put that to use and just ran off 61 parts where I had to knurl 8.5" of 1" diameter 6061. I got 60 good knurls. Ran 140 of these 6 months ago and I had about a 15% failure of good knurls, good enough for the customer, but not perfect. I also did a 380 part run in 12L14 last month, I was about 5% bad knurls.

Anyways, the two things I found that make it work pretty consistently. First, insanely slow speed when engaging, literally as slow as you can turn your lathe. I was engaging at about 40 rpms on those aluminum pieces.

Getting the tool into the part, you want to get to full depth in less that one revolution. I was doing this on a CNC, but basically you want to slam that thing in there as hard and fast as you can. Maybe a hard stop on the cross slide would work and just slam that handle.

Once engaged, crank the revs up and just feed her along. At the end where it came off a shoulder, I actually tapered out a bit so it didn't muck up the shoulder.

I was psyched, 510 inches of good knurls, only 8.5" of crap knurls. Thats more inches of good knurling than I've made in my entire life, maybe not, but it was the best success I've ever had.

I was also making sure I was getting as close to a whole number of teeth on the OD as I could. That Dorian spread sheet I linked to up there calculates all that out too.

02-09-2012, 07:32 PM
Thanks. I'll give it a shot with trial end error, I think that will be easier than calculating it.