Re: W.E.T.[CAD_CAM_EDM_DRO] Re: Z axis drive

Yeah, my fingers can't keep up with my head and I tend to proof read and
see what I want not what I wrote. "Without" would be correct.

One problem with large machines whose axes are heavy is that they
generally have high inertia's and long travels. We retrofit one of our
8" Gray Horizontals when I worked for a pressure vessel manufacturer.
This machine had 62 feet of horizontal or x axis travel, 30 feet of
vertical or y axis travel, 9 feet of spindle or z axis travel and 4 feet
of column or w axis travel. The machine was fitted with a GE MC2000
control and GE DC Servos, axis feedback was through Farrand Inductosyn
Scales.

The x-axis had a huge moving mass and we wanted to keep the traverse
rates a as hish as possible, our target was 200 ipm, originally as a
manual machine it was 110 ipm. While this sounds like a high rate for a
machine of this size, if you do the math, the travel is 62 feet or 744
inches. If we were working on the far end of the machine from the home
limits and needed to recheck home or the setup at the start of a shift,
traverse to home and back in itself is 8 or 9 minutes of just riding the
machine. As a result the axis was fitted with 2 DC Servo drives of about
15 HP each through gear boxes. To control axis backlash these boxes a
tied together with a torsional shaft which intentionally detimes the two
driving pinions forcing one to load the forward flank and the other to
load the reverse flank of the rack. Weight or mass is never a cure for
axis backlash or an excuse to accept it. Backlash usually results in low
axis stability at rest no matter what the machine weights.

In the setup of the control, an error limit is used in the calculation
of axis gain inside the control. The control looks at the positional
error reported by the scales and compares it to a preset "dead zone", If
the machine is within this dead zone tolerance, no correction is
attempted. If it is outside the "zone" however the machine uses the
positional error to apply gain correction. One form of the correction is
simply position based, some amount of additional control voltage is
applied to the drive input for each .0001" the machine is out of
position. Another gain is time based, after the machine moved outside
the dead zone, a control voltage is added to the drive which increases
exponentially with the time that the machine is outside of the "zone".
These gain voltage will be applied until the machine moves back into
position or the drive is maxed out or receiving a full 10VDC analog
input attemptin gto move the machine back into the dead zone

When the machine is commanded to move it starts to calculate a series of
theoretical positions (typically what is seen on the control display)
and compares this theoretical position to the actual position reported
by the machine scales or encoders (actual position which is rarely seen
on the axis display). The difference is used to calculate an analog
voltage sent to the drive to command it to move using the machine gains.
If the feedrate is not constrained with a time based acceleration rate,
the control calculates theoretical position as if the axis travels with
respect to time is at a constant rate with instantanous acceleration to
maximum conmmanded rate. As the machine starts to move, even the
lightest of axis has some inertia and cannot accelerate from 0 to
feedrate instantaneously. This difference of actual to theoretical
position is sometimes called the "following error" and most controls
have a setup value that controls the maximum error this is allowed. As a
machine starts to move, all will show some degree of following error
especially at traverse. The gain applied by the control attempts to
correct this error by applying extra analog voltage to the drive over
what is required to theoretically move the axis at programmed rate in an
attempt to reduce the "following error". If the machine cannot get
control of the following error and it continues to increase until it
equals or exceeds the preset "Error Limit" the control recognizes that
it is not within the accuracy of the path required and will never. The
machine immediately shuts down all axes with a "Following Error". This
is usually done by dropping the CNC "Drive Enable Output" to the servo
drives. Depending upon how the drives are setup they will either coast
to stop or brake to stop.

On heavy machines which have huge inertia, you cannot accelerate and
axis quickly even with equally huge servo motors because as the servo
motor becomes larger and more powerful it also has increased inertia's
so we never catch up. So there are three solutions.

1) Run the axis without regard to error limit and allow them to arrive
in position whenever they can but inhibit processing the next block
until all axes are in position. This was the common approach in the
"good old days" (typically anyone who talks affectionately about the
"good old days" wasn't actually there). While this works the actual
machine path may vary considerably from the programmed path. This can be
a problem if you have to thread your way between clamps or around
features of the part.

2) Run the axis at very low feed and traverse rates, with potential
machine accuracy controlling the rate through the Error Limit setup. On
small machines this may not be a big deal (20 ipm on a Sherline Mill,
patience factor is still contrllable), on large machine it is - consider
traveling 700" at 30 ipm.

3) Best way is to divide each movement of the axis into three sections.
The first section is a move where the actual feedrate is varied by the
control or accelerated from 0 to programed rate based upon time. A
second section where the axis is moved at the programed rate and last
section where the axis is varied or decelerated from programmed rate to
0 based upon time. For short moves the axis was never rise to programmed
feedrate because the length of the acc and dec ramps are longer than the
commanded move.

Now the reason for this ramble, on the original post we were talking
about counterbalances. The simplest is a weight that offsets the weight
of the head usually connected to the head by a chain or cable passing
over a pulley like an elevator in a large building. This allows the head
to be raised or lowered with very little energy since no work is being
done where work is defined in engineering terms as raising the mass of
the head to a higher energy state. In a small machine there is very
little mass (inertia) to start with so doubling the mass using a
counterweight to equalize the head weight has little if any real inertia
impact. Think about it as if you are being paid to do a job. If you are
being paid almost nothing to start with - doubling the pay won't make
any real difference, you are still being paid almost nothing.

Now consider a bigger machine, the head may have considerable inertia
from it's own weight and be difficult to deal with in terms of
acceleration (think of a drag race between a Corvette and an empty Semi
Trailer truck, both may have similar horsepower but very different
accelerations). Now if a counterweight is added the weight of the head
is offset and the energy to move the head is diminished but the total
moving mass (inertia) is considerably increased. (Now think of a drag
race between the Corvette and a fully loaded Semi Truck, on level ground
the rolling resistance does increase but at a fraction of the increase
due to higher inertia).

The whole idea of using a no-weight counterbalance is to obtain the
added force to offset the weight of the head and make it easier to move
without, increasing the inertia of material being moved.In a large
machine the mass of the piston rod in a hydraulic counterbalance is
negligible compared to the mass of the moving axis. For most hobby class
and small job shop machines the change in the inertia properties from
addign a counter weight is not likely important unless very high
feedrate are required. This is one reason that Linear Induction Motors
are so well suited to high speed machines, they add almost no inertia to
the drive system at all and apply large forces without the added inertia
added by large rotating armatures.

Now to the second question, very large machine tools have been around
from a very long time. At the same company as above we had a Betts Twin
Column Vertical Boring Mill. This machine had a 39 foot table and could
swing 44 feet. The table was mounted on tapered rollers running in an
outer race diameter of about 24 feet and a bronze center post bearing.
Table capacity was 200 Tons and rated at a maximum of 3 RPM's which is
scary fast on a machine this size. A loose tee nut left in the table
will put a hole through the wall if it comes out at speed.

This machine had originally been installed in the Brooklyn Naval Yard
and used to inspect and final bore large ship propellers. The kicker
here is that this machine was erected in 1903. It appeared that this
machine was assemble from a huge number of relatively small castings
that were machined individually and hand fitted into very large
assemblies. Every thing seemed to have been scraped for flatness,
straightness and alignment. I thing these large machines were literally
hand made from large hunks of metals by artists of extreme skill. Of
course once a large machine was produced, it could be used to reproduce
itself on a slightly larger scale with a higher degree of starting
accuracy.


gary



wthomas@... wrote:

>Gary and All:
> Hi, I have a question on one of you excellent comment.
>When you were talking about faulting the machine in acc or dec did
>you not mean "without faulting the machine"?
> Those Farrel machines a usually all BIG ONES are they not?
>In Michigan upper penn. they have a hoisting drum that is 40 feet in
>dia. and a large pieced together 40 foot pump engine flywheel. I
>wonder what machine they were turned on back in the 1800 w/o
>electricity or electronics? (If you know please reply directly to
>me)
> GOD'S BLESSINGS
> Bill
>
>
>Many machines even with mechanical counterweights use an electric
>brake on the Ball Screw that is tied to the Servo Drive "Drive OK"
>output. This brake is engaged in its default state to prevent the
>weight of the head from back driving the ball screw and free falling
>and requires power to be applied to disengage. This power is
>controlled by the Servo Drive "Drive OK" and sometimes by the CNC
>"Drive Enable" outputs in series. If either output goes low the brake
>voltage is dropped and the brake mechanically engages preventing free
>fall. Usually this will occur so quickly that the brake is set before
>the drive output has fully decayed and the head will nt drop at all.
>Similarly when the machine is restarted the drive is fully energized
>and in control of the axis before the brake is released.
>
>On large machines that I have retro-fitted where the head could be
>over 7,500#, we generally use a hydraulic counterbalance. One example
>of this was a 24 Foot Farrel Vertical Boring Mill. The z-axis ram was
>20" square, have a 50 HP Spindle inside of it and would extend out 9
>feet, it weighted 16,000#. This used a simple ram circuit with the
>cylinder mounted parallel to the ram and connected to the end pushing
>up on the ram. We apply a variable hydraulic pressure to the
>hydraulic
>cylinder ram to take a major part of the head weight and allow the
>cylinder to be back filled by the pump as the head rises maintaining
>constant load. When the head stops the cylinder ram maintains a
>constant load on the head. When the head is driven downward a check
>valve closes the supply line from the pump cylinder and the ball
>screw drives against the cylinder increasing the hydraulic pressure.
>A
>separate pressure relief valve in he circuit then opens to vent the
>cylinder back to the tank. Again there is usually a normally closed
>solenoid valve in this circuit behind the relief valve that is held
>open by the Drive OK/Drive Enable Signal which closes if the signal
>goes low maintaining counterbalance in the event of a power failure
>or
>machine health problem.
>
>The biggest reason that we use a counterbalance however is to balance
>the dynamic reactions of the machine in the up and down direction. On
>a heavy head without a counter balance it is not unusual for the
>servo
>drive to be firing upward all the time even when the machine head is
>going down or at rest. It is easy to understand that when the head is
>being raised that the drive has to fire upward, what is not so
>obvious
>is that as the head feeds downward the ball screw essentially offers
>no resistance to back driving and the head tends to free fall. The
>servo has to act like an engine brake on a truck to maintain
>control.
>You can see this if you monitor the current up and down. The
>counterbalance allows us to balance the drive current and performance
>independently for both up and down travel. The pump pressure is
>varied
>to assume a large portion of the head weight to keep the up currents
>and accelerations reasonable. The pressure relief valve in the return
>circuit is then adjusted separately to prevent head free fall and
>offer enough resistance to match the down current and accelerations
>to
>the up travel. This allows us to run a much tighter error limits and
>higher acc and dec ramps with faulting the machine with a following
>error.
>
>For lighter weight heads we have also used a hydraulic ram connect
>directly to a nitrogen accumulator where the nitrogen pre-charge
>maintains the load.
>
>gary
>
>skykotech wrote:
>
>
>
>>>All the commercial VMCs I've worked on use hydraulic or pneumatic
>>>counterbalencing on the Z axis, usually hydraulic and with the
>>>cylinder in a 2:1 compound arangment with chains like on a forklift.
>>>Reduces the inertia problem since the counterballance is some chain
>>>and a piston as opposed to a giant weight.
>>>
>>>Pete C.
>>>
>>>
>>>
>>>
>>>
>>Hmmm, my Shizuoka B-3V bedmill does not use a counterweight for the Z
>>axis, and the z axis alone probably weighs 1000lbs. I wonder why?
>>
>>I think it has some clutch arrangement to prevent the z axis from
>>falling when the servo is off.
>>
>>I wonder if adding a counterweight would be beneficial to my
>>mill...hmmm....
>>
>>
>>
>>
>>
>>Addresses:
>>FAQ: http://www.ktmarketing.com/faq.html
>>FILES: http://groups.yahoo.com/group/CAD_CAM_EDM_DRO/files/
>>Post Messages: CAD_CAM_EDM_DRO@yahoogroups.com
>>
>>Subscribe: CAD_CAM_EDM_DRO-subscribe@yahoogroups.com
>>Unsubscribe: CAD_CAM_EDM_DRO-unsubscribe@yahoogroups.com
>>List owner: CAD_CAM_EDM_DRO-owner@yahoogroups.com,
>>wanliker@..., timg@... Moderator: pentam@...
>>indigo_red@... davemucha@... [Moderators] URL to this
>>group: http://groups.yahoo.com/group/CAD_CAM_EDM_DRO
>>
>>OFF Topic POSTS: General Machining
>>If you wish to post on unlimited OT subjects goto:
>>aol://5863:126/rec.crafts.metalworking or go thru Google.com to reach
>>it if you have trouble. http://ww
>>
>>I think it has some clutch arrangement to prevent the z axis from
>>falling when the servo is off.
>>
>>I wonder if adding a counterweight would be beneficial to my
>>mill...hmmm....
>>
>>
>>
>>
>>
>>Addresses:
>>FAQ: http://www.ktmarketing.com/faq.html
>>FILES: http://groups.yahoo.com/group/CAD_CAM_EDM_DRO/files/
>>Post Messages: CAD_CAM_EDM_DRO@yahoogroups.com
>>
>>Subscribe: CAD_CAM_EDM_DRO-subscribe@yahoogroups.com
>>Unsubscribe: CAD_CAM_EDM_DRO-unsubscribe@yahoogroups.com
>>List owner: CAD_CAM_EDM_DRO-owner@yahoogroups.com,
>>wanliker@..., timg@... Moderator: pentam@...
>>indigo_red@... davemucha@... [Moderators] URL to this
>>group: http://groups.yahoo.com/group/CAD_CAM_EDM_DRO
>>
>>OFF Topic POSTS: General Machining
>>If you wish to post on unlimited OT subjects goto:
>>aol://5863:126/rec.crafts.metalworking or go thru Google.com to reach
>>it if you have trouble. http://www.metalworking.com/news_servers.html
>>
>>http://groups.yahoo.com/group/jobshophomeshop I consider this to be
>>a sister site to the CCED group, as many of the same members are
>>there, for OT subjects, that are not allowed on the CCED list.
>>
>>NOTICE: ALL POSTINGS TO THIS GROUP BECOME PUBLIC DOMAIN BY POSTING
>>THEM. DON'T POST IF YOU CAN NOT ACCEPT THIS.....NO
>>EXCEPTIONS........ bill List Mom List Owner
>>
>>
>>Yahoo! Groups Links
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>
>
>[Non-text portions of this message have been removed]
>
>
>
>Addresses:
>FAQ: http://www.ktmarketing.com/faq.html
>FILES: http://groups.yahoo.com/group/CAD_CAM_EDM_DRO/files/
>Post Messages: CAD_CAM_EDM_DRO@yahoogroups.com
>
>Subscribe: CAD_CAM_EDM_DRO-subscribe@yahoogroups.com
>Unsubscribe: CAD_CAM_EDM_DRO-unsubscribe@yahoogroups.com
>List owner: CAD_CAM_EDM_DRO-owner@yahoogroups.com,
>wanliker@...,
>timg@... Moderator: pentam@...
>indigo_red@... davemucha@... [Moderators] URL to this
>group: http://groups.yahoo.com/group/CAD_CAM_EDM_DRO
>
>OFF Topic POSTS: General Machining
>If you wish to post on unlimited OT subjects goto:
>aol://5863:126/rec.crafts.metalworking or go thru Google.com to reach
>it if you have trouble. http://www.metalworking.com/news_servers.html
>
>http://groups.yahoo.com/group/jobshophomeshop I consider this to be
>a sister site to the CCED group, as many of the same members are
>there, for OT subjects, that are not allowed on the CCED list.
>
>NOTICE: ALL POSTINGS TO THIS GROUP BECOME PUBLIC DOMAIN BY POSTING
>THEM. DON'T POST IF YOU CAN NOT ACCEPT THIS.....NO
>EXCEPTIONS........
>bill List Mom List Owner
>
>
>Yahoo! Groups Links
>
>
>
>
>
>
>
>

[Non-text portions of this message have been removed]