How close do you rapid?

[wheelieking71]

Yes, if you side crash on a mill you're most likely fine, you just shear the end mill off at the holder. But if you crash in Z, chances are you're replacing the spindle if it was full rapid.

Lathes are far more robust. We had a small Mazak, the guy crashed it every god damn day for months, the machine never came out of alignment. We have Mori's and Okumas that have been crashed a few times each, some minor in-house alignment and we're off and running. We did have one HUGE crash on a smaller Okuma lathe though that cost us quite a bit of money.

The brand of machine crashed matters in your lathe vs. mill crash opinion. BIG-TIME
It seems most of these guys are talking haas (not me, I get it!).
You crash a 40-taper haas mill? you are out $10K + 1-day down max.
You crash a haas lathe? Scrap it! They will never get it back in alignment!
If they do? You will puke when you see the labor bill! And, it will be far more than 1-day down-time!

Agreed, you get in to more well-built iron this gap closes. And, mill spindle costs skyrocket.

 

[Billy_C]

I don't think it's a Haas vs everyone else comparison. There are other linear guide lathes out there by better brands. It's the same problem. Linear guides just don't stand a chance against big parts coming un-chucked at high speeds. A little bump here and there on either a mill or lathe is usually always recoverable. I also think there is a place for linear guide lathes. Just don't crash the hell out of em.

 

[mhajicek]

In general, an average lathe will have a lot more energy involved in a crash than an average mill. I heard one half a shop away through a few walls that shook the whole building; felt like a bomb went off.

 

[paisenhi8]

I worked at I believe we would rapid in to .05 on Z on most jobs. Usually not any closer. Some of the CNC machines supposedly had enough inertia to overshoot when they came to a stop
hellodear.in

teatv.ltd

 

[Vancbiker]

……..Some of the CNC machines supposedly had enough inertia to overshoot when they came to a stop

No matter how large, any decent CNC will not overshoot the commanded position when decelerating from a rapid move unless there are some mechanical, electronic, or tuning problems.

 

[dandrummerman21]

No matter how large, any decent CNC will not overshoot the commanded position when decelerating from a rapid move unless there are some mechanical, electronic, or tuning problems.

We had a horizontal with a 0mc control on it years ago, that had the big color screen option, and lots of extra I/O. It shouldn't have been a 0 control IMO, later versions of this machine seem to have 16i controls.

Anyway, I would say it had "electronic or tuning" problems, but I don't know how to quantify it. But if you fed down in Z at a fast feedrate (100ipm or so was "fast" for said macine, it probably had 393ipm rapids), SOMETIMES it would overshoot by as much as .020 or .030". It was not repeatable on a specific move or part, but always at faster feedrates in Z. It could have very well done it in other axis directions as well. I am confident it was not mechanical (screw or thrust bearings)

The suspected culprit according to me, is the fact that the feedback of position on the screen was very slow and choppy. you know how if you handwheel or jog the axis it looks like it updates 100 times per second or so? I would guess this machine updated the position on the page maybe 5 times per second.

It doesn't matter much now, that machine is long gone and scrapped. But why would the display be so slow? Does the display on the screen depend on PMC cycle time, etc?

 

[mhajicek]

We had a horizontal with a 0mc control on it years ago, that had the big color screen option, and lots of extra I/O. It shouldn't have been a 0 control IMO, later versions of this machine seem to have 16i controls.

Anyway, I would say it had "electronic or tuning" problems, but I don't know how to quantify it. But if you fed down in Z at a fast feedrate (100ipm or so was "fast" for said macine, it probably had 393ipm rapids), SOMETIMES it would overshoot by as much as .020 or .030". It was not repeatable on a specific move or part, but always at faster feedrates in Z. It could have very well done it in other axis directions as well. I am confident it was not mechanical (screw or thrust bearings)

The suspected culprit according to me, is the fact that the feedback of position on the screen was very slow and choppy. you know how if you handwheel or jog the axis it looks like it updates 100 times per second or so? I would guess this machine updated the position on the page maybe 5 times per second.

It doesn't matter much now, that machine is long gone and scrapped. But why would the display be so slow? Does the display on the screen depend on PMC cycle time, etc?

Counterweight?

 

[gustafson]

what frequency do your old machine servos whine at?

Do the math.

1000hz
400ipm
.006 before it knows where it even is.

Most machines had conservative enough accel/decel curves to not crash

 

[Vancbiker]

what frequency do your old machine servos whine at?

Do the math.

1000hz
400ipm
.006 before it knows where it even is.

Most machines had conservative enough accel/decel curves to not crash

How on earth do you equate frequency of servo whine to the ability of a CNC control to accurately track axis position????

The position control loop of the modernish CNC operates at mHz speeds.

 

[Freedommachine]

Rapid... you mean full send?
Real men never lift!

M06 T10;
M03 S12000;
M08;
G43 H10 Z -0.25;
G00...