Making knurling wheels

Hi

I have a job that I need to make some custom pattern knurling wheels.

They are for form knurling some 8mm 420 stainless rods up to 130mm along the length. I have tried a few methods so far without good results.

I am going to make a large heavy duty scissor type knurler to hold the knurling wheels. The wheels will be ~80mm diameter and 130mm wide (full width). The style of the knurled pattern means that I am unable to make a small wheel and traverse. It has to knurl the full width at once. This means I need a relatively heavy duty tool. The wheels will be supported at both ends by arms made from 100m x 20mm steel bars.

My question is regarding the knurling wheels themselves. I am planning on using D2 as it has been recommended for this, but I am unsure about what hardness to aim for. I have a pretty accurate kiln but I haven't found any info on what hardness to give the best strength/toughness combination.

Any suggestions appreciated

Thanks

Sincerely doubt you can do this unless the stainless is at yellow to white heat & lots of water to cool the knurls ..maybe hydraulic actuation of your 'scissor knurling tool'? Consult a steel rolling mill for knurl material.

Well, a quick 10 second read on my favorite knurl manufacturer's website says that their wheels are made of "Hi-Cobalt" steel (obviously they don't want to tell you which alloy) that is hardened and tempered twice to a final hardness of 63C.

Grinding a knurl at an angle on a peripheral is tricky because the grind is on an arc or radius to the part.

Can you put enough torque into an 8mm rod to knurl across 130mm in one go?

Is there any way to do cut knurls? They'll be easier and look better.

I use a lot of 3/4" knurls made from T15 about 62-63 rc

Maybe have a look at thread rolling machines for some inspiration and an idea of the force you are likely going to need.

Screwmachine said:

Can you put enough torque into an 8mm rod to knurl across 130mm in one go?

Click to expand...

What he said! ^^^

Powered rolls would do it fine. When we roll threads using an axial roller the jaw pressure must be high or the blank will spin in the jaws.

How many thousand do you plan to do?

Hi

Thanks for the replies.

So job is around 2000. To be honest, the torque thing has been the main thing I've been unsure about. I was think of bringing the 2 rollers down slowly but I have a fear of the driven end of the bar chewing itself up.

This is a rough idea of what I was thinking. The scissor is pivoting from a boring bar holder. I have a dickson QCTP.


I don't have an easy way to power the rollers. If it came to it, I could hook up a motor and worm drive but that's a bigger project. I'd also need to allow the parts to freely rotate in the chuck and just be using the lathe's X axis really.

The main attraction of making the rollers is the per part speed as once the process is set, it will only be a few seconds.

I have been researching using cut knurling tools so far and have had mixed success. The knurl pattern is a cross hatch pattern but not diamond shape, it is axial and radially aligned grooves. I can cut knurl both grooves separately, but the issue is that the second cut is flexing out of the way. Once the first cut is made, a steady cannot be used without damaging the first cut. Also there is a bit of a burr being raised that requires a repeat pass of each cutters.

Another sort of back up idea is to mount the 2 cut knurling tools at different angles so they are inline in the Z plane. Then in front towards the chuck I can have a travelling steady to keep things in place. I'd need to have the tools running on their own axes to engage and disengage. This method would mean the cutting forces are much smaller but would take a little longer as the tool would need to traverse the length of the cut. And the tool would be a little more complex.

Below is the tool I have made to produce the radial grooves. It works well for shorter lengths, but just the longer ones are trickier. It works kind of like a pinion gear cutting it's rack.


Anyway, the responses are pushing me in the direction of the cut knurling tool. Maybe if it's a lot of repeat work I might make a driven form tool.

Thanks

Apologies for my previous flippant reply .. so you want a square knurl pattern. I think cut knurl is the way to go, two passes with different tools or one pass ganged tools. A steady on a sacrificial zone? I think both need to be canted wheels, diagonal pattern to get the cutting action especially for the axial grooves .. not sure about the peripheral grooves. You need to FLOOD coolant to flush chips. Maybe possible to use an end feed tool with hollow mount? & a leading support bush? Mixed wheels? Now, what was I drinking?

Yeah I think I need to do both cuts simultaneously as there is too much deflection without a travelling steady. Once the first cut is made with the steady, I can't use it without damaging the grooves from the previous cut. I need to mount the 2 tools parallel in the Z axis plane then use the steady ahead of the cut area.

I like the idea of a guide bush. The material is centreless ground beforehand, so pretty tight tolerance for a bush rather than steady rollers. Also if there are 2 tools at different angles, the cutting forces will be different from the 2 travelling steady fingers and a bush would take care of it wherever it wants to deflect.

Thanks

When you finally decide that you have to use driven rolls, an idea that might be worth exploring might be to mount spur gears on the headstock end of the shafts supporting the rolls. Have a mating gear on the collet that holds the shaft to be knurled. The stainless shaft slides axially into position and the three gears engage. Strong clamping pressure is applied and the collet/gear/ shaft begin to rotate in sync—-no double tracking, no tearing up or twisting off stainless round stock. Once the shaft is knurled in just a few rotations, the clamping pressure is removed, round stock advances and is cut off.

Most likely the work of making such a setup is too great for a 2000 part run. But, I have to say it seems cool. Maybe this will lead to a more practical idea? Yes, I do think the gearing would tolerate the slight mesh variance needed. Making a cut-knurl device to do the do the job might involve a comparable amount of fabrication and subsequent tuning.

Denis

Thanks for the idea Denis.

I think for now I will press on with the cut knurling plan as I have a knurl that is about 95% of the way there. Just need to build a jig to hold everything together.

I like the headstock mounted gear idea. I was thinking of using an encoder on the spindle and driving the rollers through a work gearbox with a stepper. A little like Tom Lipton shows here Dogmeat Helical Milling - YouTube. I thought it would be easy to tune the gearing if it needed to be, but your idea is a whole lot simpler to construct.

The customer reckons it is a repeat order (often said....) so perhaps I might get involved with it next time. The form rollers would definitely be a lot quicker than traversing the length, so it is very tempting.

Anyway, I will post when I have progressed

Thanks

I think this job of yours would be a no-fuss routine thing for a FLAT knurling machine but if the knurling is decorative rather than purely functional that 420 stainless might demand more than usual fine tuning. In case you're unfamiliar, two parallel platens, one fixed, one moving with workpiece in between .. rather like rolling a plasticine sausage between your palms. It's the 'other' process for thread rolling but also adaptable to knurling. A farm-out process to be sure, probably larger numbers than you envisage too. Just added 'for the record'.

If you are using two cutters then mount them on opposite sides for a balanced cut. Kinda like a balanced boring bar.

Ed.

Heat Treat D2 for Knurling Wheels

lawmate said:

Hi

I have a job that I need to make some custom pattern knurling wheels.

They are for form knurling some 8mm 420 stainless rods up to 130mm along the length. I have tried a few methods so far without good results.

I am going to make a large heavy duty scissor type knurler to hold the knurling wheels. The wheels will be ~80mm diameter and 130mm wide (full width). The style of the knurled pattern means that I am unable to make a small wheel and traverse. It has to knurl the full width at once. This means I need a relatively heavy duty tool. The wheels will be supported at both ends by arms made from 100m x 20mm steel bars.

My question is regarding the knurling wheels themselves. I am planning on using D2 as it has been recommended for this, but I am unsure about what hardness to aim for. I have a pretty accurate kiln but I haven't found any info on what hardness to give the best strength/toughness combination.

Any suggestions appreciated

Thanks

Click to expand...

D2 62-64Rc

Roger

swarfless said:

I think this job of yours would be a no-fuss routine thing for a FLAT knurling machine but if the knurling is decorative rather than purely functional that 420 stainless might demand more than usual fine tuning. In case you're unfamiliar, two parallel platens, one fixed, one moving with workpiece in between .. rather like rolling a plasticine sausage between your palms. It's the 'other' process for thread rolling but also adaptable to knurling. A farm-out process to be sure, probably larger numbers than you envisage too. Just added 'for the record'.

Click to expand...

That sounds like the answer. I am not that familiar with the method but a local screw machine shop makes motor shafts that are turned on a screw machine, then put through a flat knurl to make a straight knurl for a pulley pressed on. The operator picks the part from the screw machine and feeds it through the knurls. Zing! and the knurl is done. No worries about flexing and very fast. I can direct you to the shop if you want to discuss it with them.

Bill

Have you considered traversing a straight knurling cuter then single pointing a thread whose pitch was equal to that of the knurl?

Just thinking that the form roller does not have to form the whole length in one go:

If you imagine a pair of full length rollers then cut a helix along them . Provided the two rollers are synchonized , there is only a relatively small area forming the surface at any one time.

Your rollers are much larger than the rod so the rod could do several turns as the helical form worked its way down the length of it.