Serious crash - I need some help to understand why


(Luke) #1

Hi Guys

I’m looking for some help to find out why I just had a serious crash on my S3. I was milling aluminium when almost at the end the spindle took the work piece and threw it around my enclosure. I was milling 16mm aluminium sheet, bolted down to an aluminium table with 4 M6 screws… Not pretty. I’ve wrecked a brand new £15 end mil, a £15 sheet of aluminium the V wheels. The rest needs a proper inspection.

http://a360.co/2AyTFFJ

thats the file - I was being incredibly conservative. It handled the holes like a champ, it went wrong on the far back tab.


(mikep) #2

That’s…substantial… material to be tearing off the table.

Lubrication? If you start to get galling in the end of the end mill (how big?) it can fuse/weld with the material and really rip things up. Did the screws start to back out due to vibration? I assume you get the “screech” - this high frequency vibration can back things out pretty easily. Did it tear the whole sheet off, or just the part? I can’t see how you tabbed this…if it wasn’t tabbed, I wouldn’t be too surprised that it would get picked up and thrown round.


(Curious in Portland) #3

Were you just finishing up milling the perimeter?

Did the tabs fail, or did the hold downs give way?

I am looking at your model and I see two CAM setups with 4 operations each. Which operation failed?


(Luke) #4

It was all lubed up with copious amounts of wd40.

The tabs stayed in tact, the whole lot was ripped out. I’d not seen anything like it.

It was the first cam operation the 2nd one was a previous one. However I realise I was playing last night - after and it now might be out of date.

It happened near a tab and I wonder if I needed more of them. I had 4 tabs - triagualr and I wonder if it picked it up a fraction but then got carried away and caught. The mill end has a 20mm doc, this was only 16 and there was no melting of Ali, all chips.


(mikep) #5

Wait a sec… you ripped a 16mm plate, screwed to an aluminum table, clean off? Not just the cut part, the whole plate? And this didn’t seriously damage the table (pulled threads out of a hole, etc)?


(Luke) #6

Correct, the sheet was on 20mm ply and my hand over my kill switch. The brackets and screws took a beating and were damaged too


(Evan Day) #7

I’m not sure what the cause is, but if you have pics of the destruction, would like to see that just out of morbid curiosity.


(Mark Bellon) #8

Sorry to hear about your crash. Welcome to the world of CNC! :wink:

A CNC crash is almost inevitable. Even those with extensive experience have one every now and then. Crashes generally happen for one of several reasons:

  1. Feeds and speeds (Not applicable here because you were close to the end and things were fine)

    The most common result in these cases is a broken machine tool but in severe cases spindles can be damaged and stock/fixtures can go flying.

  2. Lubrication and cooling. (Not applicable here because you were close to the end and things were fine)

    The most common results in these cases are bad surface finish but in severe cases spindles can be damaged. Because milling generates heat somewhat unusual effects are exposed including end mills with stock welded to them (very easy with plastics but this can also happen with metals (i.e. 6061) and “work hardening” (where the stock becomes harder and the F&S speeds is now wrong.

    One of the most grave mistakes newbies make is misunderstanding lubrication and cooling. In reality, except for highly unusual circumstances, lubrication and cooling is not a major factor - SWARF REMOVAL IS. Getting those “chips” out of the work area so they are not remachined is highly important.

    Those CNC videos where the stock is practically immersed in coolant/lubricant is often the root cause for the mistaken thinking. The “flood” is mostly to remove swarf. Most commercial lubricants are highly diluted, 1 part in 20 is not uncommon. The fluid one sees is mostly water.

    Please don’t mistake what I’m saying lubrication is necessary in many cases but the effect is often small, necessary for the best possible finish and highest possible F&S. Cooling is necessary to protect the tool, its coating, and prevent certain materials from moving into problematic material issues (e.g. work hardening).

  3. Fixture failure. (This is what I think most likely happen in your case)

    CNC machines generate a lot of vibrations and place large forces on the stock and fixtures. Bolts loosen, fixtures “pop off”, and adhesives “give”. Things go flying.

Once a fixture “gives” flying stock and fixtures never cease to amaze me as to how fast they can move and how hard they can impact things. I’ve seen end mills embedded in polycarbonate, windows cracked by flying stock, and steel (and aluminum) fixtures bent.

While the X/Y/Z drive doesn’t have that much power, the torque of a non-trivial spindle can be converted into linear motion with amazing force. An 600W-800W spindle is ~ 1HP. How much can a horse pull? A GREAT DEAL! Large spindles, more force.

You do not want to be around a 10 HP 30K RPM spindle when a fixture lets go (in a “professional” CNC machine). Tens of thousands of dollars of damage.

mark


(Luke) #9

This isn’t my first crash, but it is the first since I moved to a 2.2kw spindle :joy:

I think in this case the tabs on the part I was milling failed to a level, in turn the bit caught on the spindle, this then led to the spindle putting too much force on the fixtures which in turn broke free the work piece from the base. In turn this made a right old mess between me observing and killing it off - thankfully I have a big ass kill switch these days.

Thankfully the pltboard sheet saved my bed from serious damage, but it chewed right through it.

Here are some pics for those who enjoy my pain :rofl:


(Griff Carpenter) #10

Thanks for posting, I much prefer learning from others mistakes rather then my own. :stuck_out_tongue_winking_eye:
How will you avoid this in future? More tabs or?


(Luke) #11

Ha ha, snap!

I’m taking a different approach. The part it 16mm deep so instead I’m going to mill it from 20mm flat bar in a low profile vice with a 10mm cutter. Then flip it over, shave off 4mm excess, and then mill the holes with a 1/4 bit and remove the tabs completely


(Mark Bellon) #12

The details make perfect sense… fixture failure. Sigh, it happens to all of us…

The 2.2 KW spindle puts a lot more torque on the stock than a 600-800W spindle. Your tabs not being sufficient explanation makes double sense with the new spindle. Your work flow expectation wasn’t calibrated to the new reality. You need bigger/better tabs.

Ever wonder why machining vises are so expensive? $2K is not uncommon. Now you know. Massive overkill prevents this kind of thing. :joy:

I glad to hear the damage wasn’t so bad. As I said earlier, I’ve seen things go REALLY BAD with costs about US $10K. Tool destroyed, spindle shot, vise bent, stock wasted, machine out out of square…

mark

P.S.

Just a reminder to everyone. Do not leave a CNC machine running unattended. They are a fire hazard. I know someone who burned down his detached garage leaving his CNC machine running when he walked away. Alway stay without ear and eye shot.


(Davion Hill) #13

Hi - I had a similar problem, which I think was caused by too much depth per pass (0.015”). It handled it fine for a few parts, but I think it cracked a v-wheel and I started getting wobbly cuts like what you’ve shown. The walls were chewed up, and it resonated itself into explosive fury. My part was well fixed but the machine started losing rigidity.

See video - crash happens around 8:20. I found later that readjusting all v-wheels on the z-axis carriage, replacing the x-direction belt, and double checking all assembly screws on the z-carriage put me back in business.