I've gone on a few rants recently about the pervasive idea constantly repeated in the machining world that negative rake inserts are horsepower hungry monsters that need to be buried up to the clamping screw to work, while positive rake inserts are the one true path for the hobby machine...

I'm not sure really how this persists, and it's a phenomenon that happens right across all aspects of life. As children, we don't take being told anything well, we have to test things. 'Don't touch the stove, its hot...' So, invariably, what do you do? 'Wet paint.' But is it, really? I guess at some point in our lives, whilst lying on our back on the cold wet floor with a sore bum and hands that smell like burnt paint, we tend just start accepting the things we're told as a given?

Those who've read those rants will know I'm a fan of a Tungaloy CNMG120404-11 insert (the -11 chipbreaker code being the important part here), and have made such outrageous claims as it being a better insert for light cuts than the almighty CCMT, to the point that this negative rake insert would be just fine running on the average Myford/South Bend/whatever.

The point of my heresy being, that positive/negative designation is completely irrelevant given modern molded inserts, a 'negative' insert can present a positive cutting edge (and most indeed do).

To help illustrate this point, here's a side by side of aforementioned CNMG insert, installed in it's holder, up against a Tungaloy CCMT with a -PS chipbreaker, installed in it's (LH) holder:

IMG_20240407_112350_1400x1050.jpg

I think we can agree that 'negative rake' insert certainly doesn't look any less positive at the cutting edge than said CCMT, and may even be a bit more positive (hard to compare the angle inside the CCMT chipbreaker groove.

And just to labour the point a bit more, here's a lineup of inserts:

IMG_20240407_110637_1400x1050.jpg

ALL of these go into a CNMG negative rake holder, but only one of them is actually presenting a negative cutting edge. Clockwise from bottom left, my favourite CNMG -11, some fairly standard looking Kennametal CNMG, a Kennametal CNMP, and finally a CNMA - the only true negative insert in there, due to its flat top.

Well, so what, the crowd asks. So, I did a series of experiments, in order to as I believe the young kids are saying now 'bring the receipts' and proved myself both right and wrong at the same time.

1020 750rpm 0-25mm.jpg

1020 1150rpm 0-25mm.jpg

This is a chunk of 1020, started off at 30mm, and was about 28-29mm at this point. The 4 bands have each been turned with a different insert, taking off 0.25mm from the diameter (0.125mm insert engagement), and a feed of 0.00270"/rev. The only difference between these two photos is that the upper was done at 750rpm, the lower at 1150. The same inserts were used in the same locations. Fairly unremarkable, not what you might call a fantastic finish, merely serviceable.

Keep in mind though - this IS 1020, and taking 0.25mm off the diameter. Hindsight at this point informed me I could have made a better choice of steel, so I grabbed a 32mm diameter chunk of what behaves fairly similar to 4140.

4140 speed.jpg

Now this first shot looks like another insert test. It isn't. All of these were done with the same insert. The right hand band is the 0.25mm test for this insert at 1150rpm, the grotty looking one is the same insert at about 0.25mm when cleaning up to set a consistent start diameter at 750rpm, and the two to the left are the same depth of cut, but I stopped in the groove and turned the speed up to 1150rpm. Feed remains at 0.00270"/rev for all tests.

4140 0-25mm 1.jpg

This is the actual insert test. The same 4 inserts, in the same locations, 1150rpm, 0.25mm off the diameter.

Have a good look at these, and see if you think you can guess what insert did each band, going by conventional wisdom. I'll put the candidates up lower down, and I think one of them will blow some minds.

Before that though, this was a good result, so lets go back to that bit of 1020 and try something else. Click into 1800 rpm, and a quick test cut. Yikes. Holy gear noise Batman. Find earmuffs, ponder how long it takes tinnitus to manifest, and do the actual test:

1020 1800rpm 0-25mm 1.jpg

Again, 0.25mm off the diameter, at 1800rpm, same inserts in same locations. Once again, a decent spread of results, some of them starting to approach the realm of decent.

So, now we come to the big reveal - what inserts was I using?

lineup.jpg

By deliberate choice, at the unsupported outboard end (to give it any disadvantage), the far right band was done with my pet, the NS530 CNMG -11. The next band in, done with perhaps the most common CCMT found on eBay, the 'VP15tf' Mitsubishi knockoff. Third band in, Tungaloys NS530 grade with -PS chipbreaker in a CCMT insert. Here's where the mind blowing part comes in - that left hand band was done with a true negative rake insert, an SNMA tool.

Not only did it not completely suck, it arguably outperformed the VP15TF CCMT in some tests - at 0.125mm depth of engagement. Certainly there was not much to pick between them. My pet insert took line honours here, and followed up by its brother in the NS530 CCMT. I suspect the NS530 grade has a bit to do with this. I seem to have good luck with cermets, before I moved to Tungaloys NS530, I was using an NX2525 Mitsubishi cermet that worked reasonably well.

So having established, somewhat to my surprise, that even a true negative insert could give pretty comparable results to a bargain basement CCMT, I decided to see who else I could off, and changed the inserts.

1020 1800 rpm 0-25mm 2.jpg

Still 1800rpm, 0.25mm off the diameter on 1020. Outboard end is still my pet CNMG, for a reference. The next band in, a CNMP insert. Performed exactly how I thought it would - its great for removing lots of material with low cutting force, but crap for finish at low engagement. A common theme I've noticed for 08 nose radius tools with that simplistic high positive cutting face. Third band in, more evidence for that theory, a CCMT with an 08 nose radius. No surprise for me again.

The last band..... Well, turns out the 'who else' is me, in this case. I was rummaging around wondering what insert to use for the last band, and thought of the WNMG tool I bought very very cheaply on eBay with a 10 pack of inserts, tried briefly a couple of times and decided the inserts were crap, and shoved it up on a shelf. I expected a horrible mess - what I got was a slightly better surface finish than my pet tool.

So, I called it good, and wandered off to do something useful.

Only to find myself back here doing this:

1020 cnmg 1800rpm 0-1mm.jpg

Third band from the left, trying to redeem my favourite, with a 0.1mm off the diameter cut at 1800rpm. (The band to the left of that with the colour change was me playing with the NS530 grade CCMT and rpm)

(That outboard end was me trying to get the VP15TF to hang with the negative rake boys at light depths of cut. It couldn't. No way, no how.)

Unfortunately, the WNMG matched the CNMG. And then surpassed it.

1020 wnmg 1800rpm 0-02mm.jpg

Now read this line carefully, this is NOT a typo. That second band from the left is a WNMG, 1800rpm, taking 0.02mm off the diameter. Confirmed by micrometer, and I even swapped the insert over for a new one out of the box and tried again, just to make sure there wasn't a chip in the cutting edge that created a nice keen cutting edge (I've seen trashed inserts before deliver beautiful polished results). At this depth of cut, the clearance between the trailing body of the insert and the work is a problem, as soon as any bit of swarf gets stuck in here it machines its own clearance and ruins the finish.

This is where the slightly pointier nature of the CNMG helps, and so does the design of the chip breaker. It was easier to get a clean run with my CNMG at the same 0.02mm without the swarf damage. The other salient point is that winding the WNMG tool back up the bed without withdrawing the cross slide took noticeable shavings off, indicating tool or work deflection of some sort. Whereas my favourite CNMG made the tiniest dust on the return trip. Not a problem on the diameter we're talking, but for thin stuff, the CNMG would obviously hold less of a taper.

Just to round this out, the CNMG at 0.05mm off the diameter:

1020 cnmg 1800rpm 0-05mm.jpg

The camera picks up quite a lined pattern, probably backing the feed down would help here - but the reality is, I don't take this small a depth of cut. I usually do two final passes at between 0.25mm and 0.5mm off the diameter, and measure to make sure I'm taking what I'm asking. We're right in the 'Crap, I mis measured and now need to fix my mistake' zone at this sort of depth of cut, otherwise known as probably really the territory of a cylindrical grinder.

The point really here is that even some extremely cheap 'negative rake' inserts are capable of taking TINY cuts. And when I say cheap - there is NO label on the box of WNMGs, even to identify them as a WNMG. I have no idea what nose radius they're supposed to be (although they look REALLY small), what grade they're pretending to be, nothing.

As I said in a another thread - inserts ain't inserts. Positive/negative is completely irrelevant when it comes to cutting performance, as I hope I've demonstrated here. Any hobby machine could run any of these inserts at the depths of cut we're talking here.

Of far more importance is nose radius, insert material (cermet, coated etc), chipbreaker layout. And of course the actual material you're turning - I find most low carbon grades of steel a pain in the bum regardless of the insert loaded, much prefer 1045, 4140 etc, as they finish up much nicer at a lower rpm. I'm quite often using that CNMG at 750 rpm and getting good finishes, it's often very tolerant of a wide range of feeds and speeds. Yes, I could run that WNMG at 1800rpm for the chewing gum stuff, as I've now discovered, but I don't think I'm brave enough to spin any of my jawed chucks at that speed, so that only works for stuff that fits in the collet chuck...

Hopefully this has inspired someone else to question the status quo, and ask themselves WHY will such and such not work, WHY is it so, and try experiments for themselves - not just in regard to inserts, either.
Attached Images




Read the full thread at metalworkforums.com...