Unexpected slotting at 300% DOC and 100ipm...oops

[Julien]

So I made an “interesting” error today. I had a lot of deep (15mm / 0.6") pockets to cut in bamboo, and of course I went for adaptive clearing for roughing, and I thought “well, I’m sure I can pull off using a single 15mm depth pass if I keep my WOC small enough, to optimize cutting time”.

6mm 2-flute, 24000RPM, 100ipm, 0.03" optimal load (WOC).
Sure enough, cutting the pockets went just fine.

But…somehow I overlooked the fact that my adaptive clearing toolpath boundary was such that the same toolpath would also cut the outside perimeter of my piece, and I also happened to have declared a stock dimension smaller than it actually was. Which resulted in…Fusion generating a toolpath that would then plunge my endmill all the way down to 15mm depth, and then starting to contour cut at that depth. Except in reality I was not outside the stock at that point, so I ended up SLOTTING at 15mm DOC and 100ipm

I mean that plunge right there and the following straight cuts:

Now for the interesting part: it ALMOST worked. Don’t ask me why I did not hit the E-stop, curiosity got the better of me. The machine managed to cut almost all around the perimeter, leaving packed up chips chips in its trail that the dust collection was not able to remove, before the X stepper temporarily gave up, lost steps for a while, but then the cut continued almost normally. So I was apparently not far from the maximum acceptable force that machine can take.

I wanted to see how far this unintended cut pushed things in theory, so I fired up @gmack’s spreadsheet and got this:

Let’s ignore the power stats, as I did not fill the “router” section with the right values for my spindle.

Unless I made a mistake in there, this tells me I had 33lbf of machine force during that cut, while the Shapeoko limit is (experimentally) around 18lbf. And while I did lose steps, it only occured for a short while, I would expect twice the max cutting force to bring the machine to a halt, almost ?

@gmack (and others), your opinion please?

By the way, the Shapeoko is a TANK. No visible harm after that mishap, it’s now happily cutting a corrected version of this.

[Vince.Fab (Chippy McChipload)]

A little input from the Millalyzer. Not sure how hard bamboo is but here is what that cut in white oak would look like. The 10 for K-factor and your slot width input error is probably throwing things off a tad. The dynamic sim also predicted chatter.

From what ive seen, the machine can tolerate higher loads spikes than the 18lb max, its those prolonged high loads (especially at higher ipm) or sudden violent spikes that get ya.

Updated* - Birch

[Julien]

Ha, quite interesting thanks. I mistakenly picked the “10.00” K-factor from the “NYCCNC K-factors” tab in the worksheet but those seem to be default values, I should have actually looked at the “measured K-factors” tab, but it does have a 9.45 there for baltic birch for example, which janka factor (1200?) is not too far from bamboo (~1400). If bamboo’s K-factor was actually around 18-20, that would better match what I see. Does anyone know or have measured that ?

[gmack (Gerald Mackelburg)]

@Julien

Unless I made a mistake in there, this tells me I had 33lbf of machine force during that cut, while the Shapeoko limit is (experimentally) around 18lbf. And while I did lose steps, it only occured for a short while, I would expect twice the max cutting force to bring the machine to a halt, almost ?

Your DOC and WOC were a little off. Too bad you weren’t monitoring current or power - it should have been measurable. It appears that bamboo’s K-factor is about 17 cu-in/min/HP since the X-stepper was slipping (maybe at a hard spot in the bamboo?). 2020-11-11 Julien’s Bamboo Mishap.zip (21.9 KB)

HSMAdvisor uses 20 cu-in/min/HP for Acrylic (the closest currently available) but I used it for an (in?)sanity check and unit converter anyway.

2020-11-12 Update OOPS, I used 13mm rather that 15mm for DOC, so I went back and corrected the images.
I had to download your “Machine” from HSMAdvisor’s cloud since I did this on a different computer. It needed some fixing - maybe that’s why most of the “Machines” there look FUBAR?!

[gmack (Gerald Mackelburg)]

@Julien

If bamboo’s K-factor was actually around 18-20, that would better match what I see. Does anyone know or have measured that ?

@spargeltarzan seems to use 23 cu-in/min/HP for White Oak and 30.2 cu-in/min/HP for Paper Birch. I don’t know where those values came from.
@Vince.Fab The multitude of glue layers in Baltic Birch (plywood) likely makes it much harder to cut that regular birch. I recently posted a picture of how it chewed up the edge of my Amana 1/2" HSS “Aluminum Cutting” endmill.
Maybe someone with a Shapeoko PRO and a 800 W spindle (or Modkita) and/or Vince could measure and share some K-Factors.

[spargeltarzan (caveat calculator)]

The data for wood in millalyzer is from:

[20] Green, D.W., Winandy, J.E., Kretschmann, D.E. “Mechanical Properties of Wood”, in Wood handbook — Wood as an engineering material , Gen. Tech. Rep. FPL–GTR–113, Department of Agriculture, Forest Products Laboratory, Madison, WI. 1999.

(But it’s not K-factors, as mentioned before.)

The thing with wood is the variability - cutting through the fibres obviously requires a lot more force than cutting parallel (you can often even hear that), and batches of the same species can easily differ 30% in strength, more with different humidity, how it was stored and so on.

[spargeltarzan (caveat calculator)]

Regarding the force limits:

Since you were slotting, the force component that could load the stepper would be the one in the direction of feed - so if that force is consistently above the available thrust, it will loose steps. If, on the other hand, it is only a little higher for a very short time, the dynamics work to your advantage: At 2.5 m/min, you have a little linear momentum that a small/short impulse may not reduce much. Only if you exceed the motor torque long enough (couple of milliseconds or so) or violently enough will you slow down enough to loose a step.

Finally, the 18 lbf limit is based on measurements on one particular machine. The theoretical limit without losses and friction (2 pi T/pitch) is about 140 N (32 lbf), and friction is notoriously inconsistent: Your machine could just have a little less friction/losses: Machine well taken care of, bonus points for the maintenance guy!

[gmack (Gerald Mackelburg)]

@spargeltarzan

Since you were slotting, the force component that could load the stepper would be the one in the direction of feed - so if that force is consistently above the available thrust, it will loose steps.

Millalyzer’s effective K-Factor is 18 cu-in/min/HP for Black Locust. That’s pretty close to my corrected prediction of 17 cu-in/min/HP. Using that, Millalyzer predicts 18 lbf peak transverse and 10.2 lbf peak feed.
My guess is that the (single) X-Axis stepper motor slipped when the machine was feeding in the Y direction. Is that right @Julien?

[Vince.Fab (Chippy McChipload)]

@Julien would you know the tools helix angle off hand? I’m using 30 degree and gmack is using 25 degress, looks like it can make a big difference.

@gmack just setup the Mod-Kita again last night. No bamboo in sight though.

[gmack (Gerald Mackelburg)]

Look at how smooth 32 degrees should be at that DOC!

[Julien]

@gmack

My guess is that the (single) X-Axis stepper motor slipped when the machine was feeding in the Y direction. Is that right @Julien?

A closer looks at the aftermath indicates that it lost steps mostly on X, probably during a move along Y indeed, the vertical left part of the cut was package with chips and I had to use a screwdriver to get the out, also the stock was oriented such that cutting along Y meant cutting across the “grain”, while X moves where along the grain.

@Vince.Fab

@Julien would you know the tools helix angle off hand? I’m using 30 degree and gmack is using 25 degress, looks like it can make a big difference.

I don’t know for sure, but here’s a pic

[gmack (Gerald Mackelburg)]

@Julien

X moves where along the grain.

So cutting power and forces were probably higher when feeding in the X direction, but not enough to cause the two Y-Axis steppers to slip.

@Julien

I don’t know for sure, but here’s a pic

@spargeltarzan’s formula allows it’s calculation from measurements.

[LiamN (Liam Newcombe)]

If you’re right in the middle of X you get approx double Y stepper strength, this decays to about 1.1 times Y stepper strength at min and max X travel.