Problems with CNC cut knurl – sub-contract order

[Nick anonParticipant @nickanon93441]

Hello, I thought I would join the forum to see if anyone could shed some light on an issue that I'm having with CNC knurling. Basically I'm setting out making various specialist tools as a fledgling business. One of the parts is a standard 1 inch in diameter knurled 316L thumbwheel with a centre hole and opposing allen screws to hold it firmly on a keyed shaft. I can manage knurling brass on my lathe, but stainless is a bit of a strain, so I sub-contracted the job to a UK cnc company.

When the wheels were delivered, they were all, to varying degrees, suffering from a random drunken waviness to the lines of knurls, as if the tool had wobbled or wandered from side to side, up to 1mm. The MD had assured me that he could produce a good 14 tpi diamond knurl without damage or waviness and they are a well respected firm. When I emailed to tell him about the issue, he pretty much played it down and said that the knurl was cut, not rolled and that the speed and feeds must have made the pattern irregular. You can only really notice the wavering when you put the wheels on a shaft and spin them, but it is really frustrating that a professional company can't deliver the basics, to a standard that I can achieve on a home lathe. I'll use the first batch, but it's never good to use sub-standard parts.

My question is, would anyone have an explanation for why this has happened, so that I can feed this back to the boss of this firm in order to correct the issue, if I go ahead with a repeat order.

Many thanks in advance, Nick.

[Nick anonParticipant @nickanon93441]

[jason udallParticipant @jasonudall57142]

To be frank if your supplier doesn’t already know what went wrong then nothing you are going to tell him will help.
316 as you know can work harden rapidly
.so knurling is a challenge. ..
Hence you farmed these out.
Suggestions. .
Drill / tap cross holes after knurle
..interrupted cuts can set off “chatter”…

Allow turning to size if you were using stock size ( again work hardened skin on stock)
.
Be clear on the “standard” acceptable for finished part….”he ” might not have the same criteria as you ..reach agreement…
.

..

[Nick anonParticipant @nickanon93441]

Many thanks for the reply Jason. I've just checked the specs and they actually used 303, not 316L.

The chatter due to the interrupted cut, explains why the deviation is worse around the holes. On some of my previous home workshop components, I've knurled and then used a 3mm endmill to create a flat, so that you can then centre drill for the cross holes. I'll take your advice and try to explain what I need and that I'll only accept the order if it meets the brief.

I've actually got 200 unusable versions of the same components from a previous order with another online machining company, who actually subbed it out to China. They were worse than the English based company, wavy, dented and damaged diamonds on every wheel. It's a really hairy exercise trying to face off a work hardened knurl, so that you can re-do it with a finer pitch, to try and save the components. I never new that knurling stainless was this challenging for professional outfits. You do see very crisp well formed knurls on cheap stainless items such as fishing bank sticks, so it is possible. I'm going to have a hunt around to see if a third company can deliver what is needed.

Thanks again, Nick.

[David Clark 13Participant @davidclark13]

Hi Jason

Try Verwood Precision in Verwood Dorset.
No connection other than being an unhappy ex employee.
Their work was good (I was employed there).

They may of course no longer exist.

regards David

[Ian S CParticipant @iansc]

Would it be possible to drill the cross hole after the knurling is done. The way I visualize it being done is to knurl the uncut bar, then part off the thumb wheel, and drill it, sounds like a capstan lathe job.

Ian S C

[BazyleParticipant @bazyle]

Could the material have been slipping in the chuck? I assume if it is CNC cut they programmed a multistart thread left and right handed onto a bar that was parted off to make wheels, or would they make and knurl each one individually? 1mm deviation in the spin test implies quite a big error that must show up in different sized diamonds and implies they would be utterly useless at producing a threaded component.
Do CNC software suites include a knurling utility or is it manually programmed? Perhaps the software is defective and nobody has ever looked closely before.

Maybe just change to a straight cnurl? (that's a CNC knurl -gettit &nbsp I prefer straight knurls as they stay cleaner or are easier to keep clean.

Edited By Bazyle on 23/09/2014 13:06:27

[Nick anonParticipant @nickanon93441]

I assumed that when he said it was a cut knurl, he meant they were made using a 5 axis cnc machine, milling the knurls in relief with a single point tool. Flex of that single point tool would account for the deviation of the rows of diamonds.

Bazyle, my first thought was that high stress had caused the actual work to flex/wobble. I've seen this on my lathe when knurling with a cheap Chinese tool from Chronos or RDG, ithe tool arms wobble from side to side.

I might consider a machined pattern like squares if that gets round the probably.

thanks to everyone for the suggestions, will update when I get it sorted.

Cheers, Nick.

[Clive HartlandParticipant @clivehartland94829]

If the S/steel is work hardening then consider cutting a plain portion in the middle of the job which will reduce the load on the knurl, this will also give you a clear area to drill and tap for the securing screws. To me it would look better!

Clive

[Anonymous]

I find it difficult to believe that the knurls would be cut groove by groove, time equals expensive. When they said 'cut' may be they meant something like these tools:

**LINK**

which cut knurls in one pass, rather than forming them.

Regards,

Andrew

[jim’Participant @jim11037]

Posted by Andrew Johnston on 23/09/2014 15:08:06:

I find it difficult to believe that the knurls would be cut groove by groove, time equals expensive. When they said 'cut' may be they meant something like these tools:

**LINK**

which cut knurls in one pass, rather than forming them.

Regards,

Andrew

That's what i use, only give trouble if they are not set correctly.I'd also drill hole after knurling

[Tony Pratt 1Participant @tonypratt1]

Hi Nick,

You will have to somehow produce two 'good' samples which you are happy with, one for you and one for the vendor to work to so there is no quibbling as to what is required, it's very difficult to quantify a good knurl on paper.

No need to specify the method of manufacture as it's down to the vendor as to how they make the product as long as it is to sample.

Also did you pay up front for these parts/

Tony

[Nick anonParticipant @nickanon93441]

thanks Tony, I had thought about doing this, but I think it is the way to go. To be honest, I'm shocked that the firm in Hertford were happy to supply the wheels in that condition. I'm sure that a wavy knurl would not be acceptable on a component as part of a new machine or tool, so for the MD to underplay the issue seems unprofessional.

Can I ask, is it normal practice to define a quality spec and then reject the parts if they do not meet this standard? It just seems a difficult situation to get in to, because they are then going to be out of pocket. I guess if they take the job, then they should be confident that the part can be made properly.

I did pay upfront after a pro-forma invoice, so couldn't withhold payment. I don't think that any engineering company would commence production without payment.

I've emailed around for some more quotes and emailed the MD to see if he'll be willing to make a sample to show that he can deliver the original spec that I requested.

all the best, Nick.

[Tony Pratt 1Participant @tonypratt1]

Yes it is normal to supply sizes, tolerances & finish on an engineering drawing, then everyone knows what is required and the vendor would quote against the drawing.

If you require tight tolerances and good finish you would expect to pay more, also as I said defining a knurl in words isn't perhaps so easy and is open to interpretation as in this case.

Tony

[jason udallParticipant @jasonudall57142]

Just had a look through a large box of “spare”..screw gauges…
Bare with me..
These are knurled for purely practical reasons…….
These appear “perfectly”matched across 50 different go and no go and even one taper gauge..the no go use a undercut bit to leave two knurled bands..

In any case these were made in hardened tool steel and a couple in carbide( I suspect the outter is steel not carbide but..)..
It is possible to do this work ..I haven’t in 316.. but it is do able…..
Btw 303 is very much more forgiving than 316.. ..
And is it normal to reject non conforming parts…
Too right it is.
It isn’t even your place to have to sort the conforming/ non conforming parts.. Common practice would be to bounce the lot and make the supplier sort them..
I used to work in a company running with reject rated in the ppm..
And by god didn’t we fight our corner before accepting any reject..
But we fought to the drawing….and order…

[Tony Pratt 1Participant @tonypratt1]

This is not meant as a criticism but if an engineering drawing is not dimensioned correctly you can have problems, either paying too much money if you tie every feature up to plus/minus nothing etc. or in this case as the quality was not defined exactly Nick has unacceptable parts.

You would hope the guy making the parts would ask questions but obviously his standards are not of the best!

Tony

[Nick anonParticipant @nickanon93441]

Point taken, it's something I've had to pick up as I went, so probably as much my fault. My big issue is this, I have never in all my years, seen a finished product , that had a wobbly knurl. That being the case, I thought it a relatively sound assumption to make, that a professional engineer would take as given ,the need for a knurl to be straight and true.

I've just had another reply, after passing on the key suggestions that everyone's made. He got back to me and said the the knurl was formed first and then drilled and tapped. He also stated that the small wobble was caused by the pitch not matching the circumference and as a result it tried to climb on to the next row???? I had specifically mentioned in a pre-order email that the diameter could be varied to match the chosen pitch. To be honest though, I've never seen deviation side to side, caused by a non-tracking pitch. I don't think this was a pitch mis-match, I would guess that it was a forced deep cut that deviated the rows.

In addition to the info, he hasn't exactly been friendly or "customer focused", so I'll cut my losses, I'll be moving on to another company or just buying a cnc mill and making them myself. Who would think it could be so hard to get a standard everyday part made, without it coming out unlike any thumbwheel I've ever seen.

many thanks guys for all your help, much appreciated.

cheers, Nick.

[Keith LongParticipant @keithlong89920]

Posted by Nick anon on 24/09/2014 18:05:45:

Who would think it could be so hard to get a standard everyday part made, without it coming out unlike any thumbwheel I've ever seen.

Taking of "standard everyday parts" have you had a look here or here to see if either of these companies already make what you want in their "standard parts" catalogue – they may be able to do an adaptation of one of their standard items for sensible costs as well. I've not dealt with either of these companies direct but I did deal with "Components4machinery" which was part of the first comapny and found them very good to deal with, fast delivery at very sensible prices even for 1 and 2 offs.

Keith

[Roy MParticipant @roym]

I designed a knurling tool for use on cnc lathe, the concept is simple and well tried. The company I work for had the need for 3 different diameters of knurled nut/ bolts. I made three bespoke tools and the set ups were simple and the results consistent , with no strain on the machine. If you are interested I can post some details and pictures.

Roy M

[JonParticipant @jon]

Posted by Nick anon on 24/09/2014 18:05:45 the knurl was formed first and then drilled and tapped

He also stated that the small wobble was caused by the pitch not matching the circumference and as a result it tried to climb on to the next row

Its the right way to do it and also the quickest having done many over the years.

Its also very plausible having encountered the same mismatch almost daily. They don't care they are just cnc repetition with lower standards pushing the boundaries, else would have to start job again. I would put money the cut knurl is the tool in link above. All that's needed is an alteration to the plunge in depth which should have been sorted in test or setup stage.

Material wont make much difference though brass and manganese the easiest followed by aluminiums, steels just need more pressure or beefier machine. G303 no problem, 304 not too bad, G316 certainly S11 will take some effort.

[Author]

Posts