Where did this bolt come from?

[rbkhs]

Hello

We are running an Auto Route 8 Pro +

I have a student working on a wooden shield and the rough cut was going to take 30 hours to create dome of disks with a straight bit and then about 10 hours to smooth everything out with a ball nose.  About 20 hours into the flawless rough cut (and about 15 minutes after we had last checked the machine) something happened.  When we came back in the room the cutter had clearly gotten off course, plowed into the work, shifted the whole thing over on the platform and was now cutting in the air about 14 inches off target.

If you can see the photos the bit clearly plowed into one side of the dome and then must have jammed into the other side when it did a cross move.  I am assuming that that deep plunge must have made the tool pretty angry and is probably why the whole workpiece shifted.  By the way, the piece is mounted to a piece of plywood and the plywood was held down with 8 clamps.



I found this 1.5" bolt on the floor but I can't figure out where it came from.  I am wondering if it came loose and somehow caused the machine to spin off target? 

I shut down the machine, rebooted, and homed all 3 axes and everything seemed to work fine.

[dadealeus]

Great looking project!

However, a cut time of 30 hours for a roughing path is *very* extreme for a workpiece of this size. Your roughing stepover is a lot smaller than it could be, resulting in excessive cut times.

For example, I generated a sample shield shape 24" in diameter, 5" in height in Fusion 360 and ran a pocketing pass with 1/4" cutter at a 1/8" depth of cut, assuming I would cut the shape from a solid block of 24"x24"x5" wood - and it comes in at nearly 1/10th of your cut time. If you swapped to a 1/2" cutter with a 1/4" depth of cut, you could likely complete that roughing path in less than 30 minutes.

Reduced cut times are going to reduce the wear on your machine as well as the likelihood of a mishap during the cut stemming from machine or clamping failures.

----

As to this particular issue, can you elaborate on your toolpath generation? It looks like you used Vectric Aspire to generate a 3D Roughing path using a raster pattern, but I'm not sure.

It's tough to tell which happened first:

1) The workpiece came loose, shifted, and then jammed up the cutter causing the machine to skip steps - compounding the failure.

-or-

2) The machine skipped steps, driving the cutter too deeply into the wood and overcoming your clamping force, shifting the workpiece and compounding the failure.

Personally, I don't believe the screw is related to this particular failure as anywhere it would have fallen out of that would probably cause skipped steps (like from one of the roller bearing flanges) would have 3 more just like it still in place. You should definitely check your Y ball-screw though to make sure it didn't come loose from there.

Honestly, it looks like this may have been as simple of an issue as inadequate clamping. What types of clamps were you using and how were they tightened? I've found that (especially on long cuts) hand tightening the axiom clamps was not enough as the combination of constant vibrations and shifts in the workpiece geometry (as material was removed) would often cause the workpiece to come loose prematurely and ruin the cut.

In fact, I made this to address this exact issue:

axiomprecision.proboards.com/thread/1085/axiom-holding-clamp-wrench

There are a couple different styles made by different users who all had this issue in that thread.

While I'm not sure that's what happened in your instance, I believe it was the more likely cause of failure than skipped steps due to a single missing bolt. However, if that wasn't the first bolt to fall out, that could definitely be the cause and you may not see the malfunction by simply homing the machine.

I'd say run another quick test cut that puts some lateral forces on the cutter in both the X and Y directions. If those cuts work fine, it was almost assuredly inadequate workpiece holding that caused the failure.

[dadealeus]

May 10, 2022 16:55:10 GMT -5 rbkhs said:

I have a student working on a wooden shield and the rough cut was going to take 30 hours to create dome of disks with a straight bit

It should also be noted that straight cutters are not your friend.

I used them exclusively when I first started using CNC mills because I had seen some videos on YouTube comparing cut quality of downcut, upcut, and straight cutters and (at first glance) it appeared like a straight cutter was the best of all worlds (largely eliminating tearout on both the top and bottom of the workpiece).

However, they MASSIVELY reduce your cutting capabilities compared to upcut/downcut/compression tools due to how they plow through the material.

Think of whittling a piece wood using a knife. If you were to place a block of wood down and try to cut it across its entire thickness like a piece of cheese, you're likely not going to get anywhere because you are attempting to cut through too much material at once. However, if you were to turn the block to the edge and cut away thin slivers starting from a corner, the wood cuts like warm butter.

A straight cutter works the same way as cutting the wood like a block of cheese - with every revolution of a cutting blade, it is attempting to cut the entire stock from the surface of the wood to the bottom of the depth of cut simultaneously. This basically results in a machine-gun like chatter every time the flute hits the wood, scraping out a chunk of wood and then ceasing cutting until the next flute hits the wood.

With an upcut or downcut, the cutting blades overlap one another vertically. This allows a smaller portion of cut to take place at a very constant rate - vastly reducing that "machine-gun chatter" in the straight cutter and allowing you to push much faster through the material.

Here is a perfect, slow motion example of this:



Observe how much more violent the straight cutter is than all the others - shaking the wood to pieces as it takes large, infrequent chunks out of the wood.

Switch to an appropriate upcut/downcut or compression tool and you're likely going to double or triple your material removal rate. It sucks having to manually change out cutters, but the tradeoffs are absolutely worth it.

[rbkhs]

Thanks for the tips.

I should clarify that we were using a 1/4" downcut bit with a .125 cutting depth and held the project down with 8 of the axiom clamps, 2 in each corner of a plywood square that was sticking out from under the dome. All were tightened with a strap wrench to avoid...well...this. We were using VCarve Pro to cut a Z level roughing pattern.

Last night I checked the toolpaths again and realized that our feed rate was somehow set too slow (we had to reinstall the software recently and I guess I missed updating this one) so when I upped it to 4 ips this morning it cut the roughing time down to 12 hours. Still not as good as your 3 hours, but much better.

That seems to be the only bolt that has magically appeared. It is not out of the Y ball-plate. That was my first guess since the cutter was SO far of track in the Y direction after this failure that I thought the Y axis may have been messed up. But that's not where this bolt came from. I am going to open up the gantry covers to see if it came out of the X.

We are going to come up with a better clamping plan and also try to figure out how to cut out the damaged segments and replace them. Should be interesting.

[aluomala]

Check underneath your machine for the cable management system (drag chain, etc). I found a bolt when I unboxed my AR4, and couldn't figure out where it came from. Eventually I realized that it had been part of the cable management system and had dropped out during shipping, and noticed things "drooping" under the table BEFORE anything got damaged. As I recall, the bolt in question looks like the bolts used for that purpose, as they use the slots in the bottom of the spoilboard (the underneath surface) the same way that is used for the rotary axis mounts, or the clamps that we use to hold things down.

I mentioned to Axiom about the concept of using LocTite for those types of things but apparently they'd rather people discover that things come loose via the shipping, or through normal use of the machine.... It sucks that one has to check ALL of the bolts on a machine before using it, but that's the way it goes, I guess. On a related note, check the bolts holding down the spoilboard, as they come loose and can cause issues with Z-height (don't ask how I figured this one out.... granted, they are much easier to check than bolts underneath the machine, so it should be part of routine maintenance.

Allan

[dadealeus]

May 11, 2022 8:22:42 GMT -5 rbkhs said:

I should clarify that we were using a 1/4" downcut bit with a .125 cutting depth and held the project down with 8 of the axiom clamps, 2 in each corner of a plywood square that was sticking out from under the dome. All were tightened with a strap wrench to avoid...well...this. We were using VCarve Pro to cut a Z level roughing pattern.

Sure thing.

Assuming that's a 2-flute downcutter and you're open to testing, give these settings a shot for on a piece of waste from the same species of woods you're working with. I think they'll work quite a bit better for you and should drop your cut times even more. Just make sure your stickout on the tool is no greater than about 1".

If your machine is in good condition (clean, lubricated ballscrews and a sharp tool):

Feed: 175 IPM

Plunge Feed: 85 IPM

RPM: 9000

Depth of Cut: .25"

Stepover: .2"

If your machine has been punished a bit, those settings will likely skip some steps; try these:

Feed: 122 IPM
Plunge Feed: 60 IPM
RPM: 6000
Depth of Cut: .25"
Stepover: .2"

Should knock quite a chunk off that 12 hours.

[dadealeus]

If you're curious about Fusion 360, here's an 8 min video from start to finish of a similar shield shape that I whipped up:

[savannahdan]

As to the bolt that you found. I had one of the small bolts in the top of the spindle clamp fall out and saw it before my dust collection system sucked it in with the saw dust from the job I was working on. I think the bolt I found was a bit smaller than yours. I've had cuts like that that I solved with grounding of the system.