Several times mention has been made of dancing machines, and the importance
of firm anchoring, level-ness, etc.
Suppose a machine -- cnc or manual -- was placed on a very light but very
rigid base, which was then suspended in the air by 4 or so guy wires from
So the machine would be level and un-twisted because of this light rigid
base, but it would be free to swing.
How would that affect its performance, accuracy, etc.?
I would hazard to say that there would be no ill effects at all, cuz of
Heh, if there were springs in the guy wires, sure would help with building
But, I was wrong once before, and it could happen again.....
Would probably help with small vibrations, but without some sort of
damping there might be a chance of an escalating harmonic in the
system. Still probably wouldn't affect operation any more than being
bolted down...until the springs or cables finally failed.
Air-supported tables like these are the standard in optics labs, and
we put our wafer probes and microscopes on them at Unitrode:
The large leg in each corner contains a rolling diaphram air cylinder
controlled by a position regulator that does a pretty good job of
keeping the table level, much better than springs would. They work
very well until you need to torque larger fasteners, then they bob and
rock like a diving float.
Einstein was good at "thought experiments". You're not. And Newton
is gonna kick yer ass in five or six different dimensions if you ever
try this for real.
First, there's that stuff about how a body at rest tends to remain
at rest unless acted upon by drugs or alchohol or something. That does
NOT refer to the whole body, if said body is made of parts that can move
with respect to each other. What it refers to is the center of the
body's inertial mass. So when you hit the rapid button and make the
machine's table move to the left, everything else about the machine is
going to move a bit to the right, so the center of mass stays where it
was. And if you move the saddle away from you, the rest of the machine
will come toward you. Etc. And that's gonna be a problem.
If the guy wires were all taught and balanced and everything before
you moved the table, they're not going to be that way after you move it.
The platform and machine body are now in a different place, having
moved to compensate for the change in mass distribution when you jogged
X. So the tension on the wires is now unequal. And, gravity being what
it is, the whole machine and base will start another bit of movement,
swinging pendulum style toward a new location where the tension in the
wires is equalized again. And that swing won't go away till friction
eats it, which can take a very long time. And every new axis motion
will add to the swing. And add to it. And ADD to it. And so on.
If you attempt to move Z the wires will either loosen a bit (Z-
motion) or tighten a bit (Z+), for the same Newtonian reasons. And,
since wires (or anything else, if you stress it enough) are elastic,
they'll just naturally want to "rebound" a bit when the motion stops.
So in addition to swinging back and forth from XY motions, your poor
dizzy Fadal will now be bouncing up and down, platform and all. And
don't even THINK about adding springs!
Worse, none of the moving parts of your Fadal is likely to have its
own center of mass on a line that goes through the center of mass of the
whole machine. That means the forces described above will form
"couples", which will cause rotation. When the table moves left or
right, the machine won't just pendulum swing in the other direction.
It'll also try to twist itself around it's center of mass. Viewed from
the operator's position, it'll twist clockwise if you move left, and ccw
if you move right, or vise versa, depending on whether the X mass is
located above or below the main mass center. Same problem when you move
Y or Z (top of machine twists toward you or away from you). This causes
even more distress for the poor guy wires, which will now be out of
balance in god knows how many ways, and will need to hang on while the
whole machine does some MORE pendulum/gravity swinging to equalize
everything all over again.
And before you know it, you and your poor machine will be swinging
back and forth like a 10,000 pound tether ball, twisting in at least
three axes simultaneously, and screaming like a kid on one of those
insane multi-dimensional amusement park rides that's specifically
engineered to make people lose their lunches from both ends at once.
And then, since you asked, there's the small problem of vibration.
And here's the relevant question: If you were a machine tool, and if
you had a whole bunch of nasty little compression waves running all over
the inside of your body, and if you really REALLY wanted those waves to
go away, becaue they hurt, and itch, and make your eyes water, and more,
what would you most like to grab hold of for stability?
A. Some long, skinny, flexible, stretchy pieces of cable?
B. A big-ass chunk of concrete, conveniently attached to the entire
earth, with all its zillion trillion gazzillion tons of rock and iron
If you guessed A, you'd better own a good helmet and a generous
supply of barf bags.
Nuff said. Have fun!
Did you read what HalfNutz said, in one of my recent rants? That I could
I think we are at least neck and neck, eh? :)
Well, I didn't doubt for a minute that pendulum-like motion and rotation
would accrue, but would it affect the machining *results* of a part?
And, regarding the motion, suppose the machine were constrained at the
coners by horizontal springs, and in the vertical by springs. Couldn't this
actually *relieve* internal stresses in a machine, with no sacrifice of part
I wonder how this would affect the tuning of the servo system?
The more important question, is how much coolant are you going to dumping out of
coolant tank due to wave action.
"Additionally as a security officer, I carry a gun to protect
government officials but my life isn't worth protecting at home
in their eyes." Dick Anthony Heller
Why not float it on a raft in a pool of coolant?
The first time you walk up next to it and punch in a rapid traverse
it's going to run you down like a rhinoceros.
Does AFV have a category for that?
Nice one, KG
I think the real "utility" to this Q, beyond any impact on accuracy (which I
don't think there will be) is that a floating design might be more
"structure/building friendly" say, for machines mounted on a second floor.
Or, in garage-type ditties, more neighbor friendly ito of transmitted
vibration to the next house, etc.
Not a biggie, just never saw the big payoff to bolting shit down like an
earthquake was coming.
A proper floating design would have no tendency to walk, either.
Okay. Put your safety glasses on. Stand on the platform in front of your
machine. Adjust wires, springs, reductape. Go to MDI. Type in 10000 M3.
Press cycle start. Type in M5. Press cycle start. HANG OOOOOOONNNNN!!!!!!!
I once worked at a dealership under the main air compressor which was hung
from the (metal) roof trusses by steel rods and compression spring brackets.
We called it "Skylab", and just ignored it after the first few days of
working there. It was there for 10 years before I started with no problems
and might still be there now, 30 years later.
All the money I ever wasted was spent trying to impress somebody who was
OTOH, this is how new things get invented. A crazy idea, everyone laughs
and points out the flaws, he thinks, "OK, it's gonna move, how can I
constrain it" and the next thing you know you end up here -
Basically the same idea, with some of the more egregious bugs worked
I've had to deal with this issue my entire working life Dan.
High speed robotics is right at the core of the injection molding industry
today. It is if you want to compete anyway, and you have the same issues to
deal with in what EA is mulling over. Tool changers causing part
imperfections has been a real issue in the milling industry right along.
In the mid 90's I bought a big beam robot out of the BK of a Texas molding
company to use on a 630 ton Toshiba hydraulic press.
All I wanted was the metalwork and we stripped it completely when it
arrived. I replaced the air/oil drive with a fast pitch ball screw set that
Beaver Precision built for me and hooked it up to a pair of Fanuc's high
speed alpha series (3600 RPM) servo's.
The thing was lightning fast and worked pretty good until the vacuum chuck
was holding the twelve pound part we wanted to remove from the machine. The
vacuum wouldn't hang on to the part when the robot retracted out of the
machine. I should have stopped and though about that for a minute but what I
did instead was install a spring activated mechanical clamping system we
cobbled together. The chuck would move in and when it grabbed the part, six
fingers engaged to clamp the part from behind. I figured the part wouldn't
fall of then, and it didn't.
When the robot accelerated the part upwards I ended up with the entire mess
sitting in what was left of the discharge shute at the bottom of the machine
and a big old pile in my britches. You see, I hadn't bothered to set up the
acceleration ramping time constants on the drives, only the decel so the
servo just came on at 2000 RPM. The screws were six pitch. OOPSIE!
Anyway, I'm just poking fun at PV/EA/whatever.
The M3/M4 thing at ten percent below max RPM is something I actually do when
commissioning a machine however. Busted the 35 Kw geared head on a Formosan
machine once. Took about twenty minutes. Smoke everywhere.
Sales guy had said you could reverse the spindle all day long at 3000 RPM
when I asked him about rigid tapping.
I guess they have really short days in Formosa......
That would also explain why all of these guys think they are well hung.
Imagine what ten inches is if an entire day is 35 minutes long.
Talking about lift off: wonder if NASA needs any machinists in space?
Chips just float away, pump in out side atmosphere for coolent?
What ever happened to"sENOIR ManuaL" swinging the door every shot?<g>
We had Wittman robots on the Engels - Sweeet! If there was one bad
cavity they prgrammed it to drop that one in the scrap pile, while
others were placed individually in trays. Cool shit. I did alot of EOA
tooling for them whilst getting my wings on setup & processing. Those
side robots are really freakin cool. Ever see the Engel training
video, done here loacally at Winsler? Now thats a 1st class facility.
They'd have black tie events on the molding floor.
Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here.
All logos and trade names are the property of their respective owners.