Spindle sounded like death (video)

[Ydna]

We were running some experiments the other day and I'd like to share the results and pose some questions. Bear with me whilst I explain everything. The tests were simply cutting some straight slots in aluminum, and watching the spindle torque during the cuts. Trying to find the breaking point of *something* be it the spindle torque, tool shatter, or whatever. We sortof found the limit at around 1:34 in the video below, where the machine makes a very noticeable "noooo more" growl.
YouTube - Diamondback aluminum rougher slot mill test
Someone was actually filming the torque gauge while I was filming the cutting, but the gauge didn't do a whole lot so I didn't even bother to include it in the youtube video. It was hanging out around 40% during the first workpiece, then hopped to around 60% during the second workpiece.

Anyway, during the 0.625" deep 150-IPM slot, the spindle sounded like it was dying. As a result I called out "that might be it" (the machine's limit) but here's the thing! the load gauge didn't do anything out of the ordinary. It still hovered around 60% even though it sounded like the spindle speed had dropped a good 15-25%, something that I can only assume would cause the torque meter to PING like no other. I can't say for certain though since I've never maxed out this particular machine. (or any machine for that matter, except for my SuperMini)

We had the Haas control's spindle load limit set to 115%, and it recorded a peak of 114% (lol) however I believe this was only at the beginning of the slot when the cutter broke into the wall.

So I pose this question to you people, what happened in this video? I have ideas but I do not know.
- Transmission to blame?
- Belt slipping? (what kind of belt is it, can it slip?)
- Axis ballscrew, vibration?
- Chip weldup in the flutes? I originally thought this was happenning but the appearance of the slots would beg to differ, both sides look perfectly consistent, not like you'd see with a weldup cutter. The cutter is just fine, I was hogging with it today actually.
- Tool flexure? I wouldn't think so being as it didn't shatter, and the floor finish is consistent too.
- Spindle feedback system (or the gauge) is screwed up?
- Something else??

This machine is older so it doesn't have the actual speed displayed, or any other axis torque besides the spindle...I wish it did. It's a 1997 VF-2 (brushed servos) with 10,000-RPM and gearbox option. The sticker says 15 horsepower although I thought I saw somewhere that the 10,000/gearbox option was 20? I probably wrong. Anyway despite its age the machine has around 3k hours on it being as it's only used by the college. It has the analog spindle gauge BTW.

THOUGHTS???

[rawen2]

Originally Posted by Ydna This machine is older so it doesn't have the actual speed displayed, or any other axis torque besides the spindle...I wish it did. It's a 1997 VF-2 (brushed servos) with 10,000-RPM and gearbox option.

I don't have any answers to your questions but I thought that Haas switched to brushless servos in 1996?

[Ydna]

Indeed they did switch. But we're told the machine is somehow a bastard in-between combination of parts. As you can imagine it makes having work done to it a tad more difficult.
(yeah the sticker even says brushless, but stickers lie!)

[Machineit]

Remind me not to buy a used machine from a school!!!

Mike

[Ydna]

ya because you get people like me performing "experiments" LOL

Nah truly we keep the machines in spectacular condition except for the fact that they're crashed every few months. We use dummy tools and plastic gauge blocks until everyone is to the point where they think before jogging. We have a 1993 VF1 that cuts like a champ...a slow champ...

[Donkey Hotey]

Originally Posted by Machineit Remind me not to buy a used machine from a school!!!

That's called Manufacturing Engineering. That's how cutter speeds & feeds are determined and how high speed machining was discovered. At Westec, Haas runs most of their demos fairly hard like this. They do it to demonstrate the spindle loads that the machine can sustain.

I don't have any suggestions as to the problem with that spindle dropping speed but, I'll be sharing that video when I get to work.

[Wizzard of H]

If what you have is a 1997 brush machine, the drive system to the spindle is a "variable frequency drive". It operates "open loop". That is, without feedback. It is not a vector drive and cannot react like the newest stuff. At the highest speeds, you cannot get high torque or high current that would be indicated by the load meter.

[rmarchjr]

I will bet that sound was the required HP need for the cut not being there.

Just guessing but the first few cuts created a good bit of heat in the spindle motor - reducing the HP output enough to slow the the spindle.

We go through a lot of 6061 - and usually run 200 - 300 fpm but only @ .25"DOC @ 5000rpm.

I am very surprised that the load meter didn't register much higher it should have.

[fastfrank]

Greetings All,

Ydna,

You have no worries on your spindle. That is just the way a DC (i.e. brushed servos) machine stalls. It's very spooky the first time a person hears it because it feels like it goes on forever. If you stall that machine 5 more times, it will act the same way every time. By then you will say, "OK that was normal."

I'm not sure were peak torque is on a 10000 rpm spindle but pay close attention by ear how long it takes to get to commanded spindle speed from a dead stop in 200 rpm increments. The torque falls off very fast just after peak. You will hear it really start to take quite a bit longer to go a little faster. When you get to that point, back up about 2 or 300 rpm. That will be the rpm where you want to let it eat! On a 15 hp 7500 rpm machine, it will have 3 times the ass at 6000 as it will at 7000 rpm.

You will be impressed with how much more grunt you will have for your test after messing around a little to find the sweat spot rpm.

A Vector drive is a completely different animal so none of this applies.


Hope this helps

[Ydna]

Ya I'd love to see a definitive torque curve for it. I've never actively looked for one outside of the online manuals, but those are obviously for a different breed of machine. (they don't even have the info anyway, since the machines are so different)

If a VFD was governing things then it makes perfect sense as to why the spindle load turned into a placebo. Well not a placebo but rather watered down reading. Controller still registered a jump in the load at the beginning of the cut, but gives me more appreciation for the newer machines. I think it might be time to recreate a similar slot mill test on different animal, which I can easily do...except the one I have is an SMM which will happily ping the load if a feather lands on it. Still a valid test though as far as I'm concerned.

Heat building is certainly a variable that could influence it, I didn't really think of that. Although I should mention there was about a 15 minute gap between cutting the first workpiece compared to cutting the second. The machine was just sitting there though, no coolant circulation or whathaveyou...

Anyway, I have no reservations against breaking tools and pinging the machine's load meter as long as I get some bona fide numbers on what happened. I mean we *did* set it to error out just above 100%, which we do for all these similar tests....except this is the first time we've ever taken a deep enough cut as to make the machine respond (audibly, anyway). For machines that have such decreased power compared to the heavy hitters, I stick by the statement that with the correct combo of software and tooling you can straight up ANNIHILATE some workpieces.

FWIW the tool manufacturer claims their 3/4" version of this rougher will withstand a 0.040 inch per tooth cut (0.120 IPR) at 18,000 RPM if the machine has the fantastic amount of beef to drive such a thing. Guy said they had a video of it though I haven't seen it (though I have no doubt the tool might not last long at those rates).

I can't say enough good things about these roughers, I mean they've basically restored my faith in using an aluminum rougher all together It's really changed how I program and that's even when running parts on the SMM. Sure the tool doesn't last "forever" but I can still run a ton of parts cutting full depth slots. Tools are cheap in comparison.....