U.S. patent number 3,806,012 [Application Number 05/295,969] was granted by the patent office on 1974-04-23 for method and apparatus for maintaining registration in a multistation printing press.
This patent grant is currently assigned to J. Bobst & Fils S.A.. Invention is credited to Roger-Henri Roch.
United States Patent |
3,806,012 |
Roch |
April 23, 1974 |
METHOD AND APPARATUS FOR MAINTAINING REGISTRATION IN A MULTISTATION
PRINTING PRESS
Abstract
In a web-fed printing press having a plurality of stations, each
for printing one of a plurality of colors, registration is
maintained by controlling the tension or elongation of the web
between successive stations, in response to signals derived from a
plurality of sensors, such signals being representative of
registration error and the measured tension of the web.
Inventors: |
Roch; Roger-Henri (Ecublens,
CH) |
Assignee: |
J. Bobst & Fils S.A.
(Prilly, CH)
|
Family
ID: |
4403529 |
Appl.
No.: |
05/295,969 |
Filed: |
October 10, 1972 |
Foreign Application Priority Data
|
|
|
|
|
Oct 8, 1971 [CH] |
|
|
14749/71 |
|
Current U.S.
Class: |
226/2; 226/28;
101/484 |
Current CPC
Class: |
B65H
23/1882 (20130101); B65H 2557/2644 (20130101) |
Current International
Class: |
B65H
23/188 (20060101); B65h 023/18 () |
Field of
Search: |
;226/28-31,113,44
;101/228,DIG.21,248 |
References Cited
[Referenced By]
U.S. Patent Documents
Primary Examiner: Schacher; Richard A.
Attorney, Agent or Firm: Hill, Sherman, Meroni, Gross &
Simpson
Claims
I claim as my invention:
1. A method for maintaining registration in a multistation web-fed
printing press comprising the steps of determining the registration
error at a location downstream from the second station of two
successive printing stations, calculating a desired value for the
tension of the web extending between said two stations as a
function of said registration error, and controlling the tension of
the web along the entire length between said two stations in
accordance with said calculated value, so that the tension of the
web between said two stations is equal to said desired calculated
value.
2. The method according to claim 1, wherein said calculated value
is derived in response to said registration error and to the rate
of change of said registration error, relative to the longitudinal
dimension of said web.
3. Apparatus for maintaining registration in a multistation web-fed
printing press comprising misregistration sensor means disposed at
a location downstream from the second station of two successive
printing stations for producing a signal representative of the
registration error, computer means connected with said
misregistration sensor means and responsive thereto for producing
an output signal representative of a calculated value for the
desired tension of the web extending between said two stations, and
actuator means connected with said computer means and responsive
thereto for controlling the tension of the web between said two
stations, so that the tension is equal to said desired tension.
4. Apparatus according to claim 3, wherein said computer means
comprises means for deriving said output signal in response to said
registration error and to the rate of change of said error,
relative to the longitudinal dimension of said web.
5. Apparatus according to claim 3, wherein said computer means
comprises means for deriving said output signal as the sum of
signals representative of three parameters, one of said parameters
being proportional to said registration error, a second of said
parameters being proportional to the rate of change in said
registration error, relative to the longitudinal dimension of said
web, and a third of said parameters being proportional to the sum
of a plurality of said registration errors determined for equally
spaced locations along the longitudinal dimension of said web.
6. Apparatus according to claim 5, including means for establishing
predetermined limits on the value of said third parameter.
7. Apparatus according to claim 3, including means for limiting
said output signal in accordance with predetermined limits on the
maximum and minimum values for the tension of said web.
8. Apparatus according to claim 3, including manually operable
means for producing a control signal representative of the nominal
tension of said web, and means connecting said computer means with
said manually operable means, whereby the output signal produced by
said computer means is responsive to said control signal.
9. A method for maintaining registration in a multi-station web-fed
printing press comprising the steps of determining the registration
error at a location downstream from the second station of two
successive printing stations, calculating a value for the tension
of the web extending between said two stations as a function of
said registration error, controlling the tension of the web between
said two stations in accordance with said calculated value,
measuring the actual tension of the web between said two stations,
comparing the measured tension with said calculated value, and
changing the tension of the web between said stations so that said
measured tension equals said calculated value.
10. Apparatus for maintaining registration in a multi-station
web-fed printing press comprising misregistration sensor means
disposed at a location downstream from the second station of two
successive printing stations for producing a signal representative
of the registration error, computer means connected with said
misregistration sensor means and responsive thereto for producing
an output signal representative of a calculated value for the
tension of the web extending between said two stations, actuator
means connected with said computer means and responsive thereto for
controlling the tension of the web between said two stations,
tension responsive sensor means for producing a signal
representative of the tension of said web between said two
stations, and switch means for selectively connecting said computer
means with said tension responsive sensor.
11. Apparatus for maintaining registration in a multi-station
web-fed printing press comprising misregistration sensor means
disposed at a location downstream from the second station of two
successive printing stations for producing a signal representative
of the registration error, computer means connected with said
misregistration sensor means and responsive thereto for producing
an output signal representative of a calculated value for the
tension of the web extending between said two stations, actuator
means connected with said computer means and responsive thereto for
controlling the tension of the web between said two stations,
tension responsive sensor means for producing a signal
representative of the actual tension of said web between said two
stations, and means for controlling said actuator means in response
to the difference between said output signal and the signal
produced by said tension responsive sensor means.
Description
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a method and apparatus for
maintaining register in a web-fed printing press having multiple
stations.
2. Prior Art
Multistation printing presses are commonly used for color printing,
and it is necessary to provide some means of adjusting the position
of the image printed by one station relative to that printed by
another station, in order to correct for and eliminate
misregistration of the various colors. A number of systems for
making such adjustments have been developed in the prior art. In
one system, a compensating roller is placed between two successive
stations, and means is provided for moving the position of the
compensating roller, in response to a detected misregistration, to
effect a change in the length of the web extending between the two
stations. The misregistration is corrected when necessary by
providing a greater or lesser length of web between the stations,
with the result that the image printed by the second station is
shifted longitudinally relative to the image printed by the first
station.
In other known systems, the angular displacement of the printing
cylinder at one or the other of two successive printing stations is
changed to effect a desired correction. When this is to be
performed automatically, a sensor is employed to detect the
magnitude and direction of the change which is needed.
While the systems of the prior art have been effective in their
purpose of correcting misregistration, the operation has been
relatively slow, since only the magnitude and direction of
misregistration have been taken into account in the derivation of a
control function which restores registration to normal, and the
rate at which misregistration is being corrected has been ignored.
Because of the length of web between the two stations, and the
length of web between the second station and the misregistration
sensor, a time lag exists between the making of a corrective
adjustment, and the sensor's recognition of the adjustment. In
order to avoid instability, the response time of the corrective
system is extremely slow, and as the amount of misregistration
decreases, the rate at which it decreases changes exponentially so
that correct registration is approached asymptotically, and reached
only a relatively long time after the misregistration is initially
detected.
Therefore a method and apparatus for effecting a more rapid
correction of misregistration, is extremely desirable.
SUMMARY OF THE INVENTION
It is a principal object of the present invention to provide a
method and apparatus for effecting a rapid correction of
misregistration in a multistation web-fed press.
Another object of the present invention is to provide means for
effecting the most rapid correction of misregistration possible,
while avoiding breakage of the web by maintaining the web tension
under a critical limit.
A further object of the present invention is to provide a method
and apparatus for automatically maintaining the tension of the web
at a value which insures proper registration.
Another object of the present invention is to provide a method and
apparatus for employing the rate of correction as a factor in
deriving a control function which restores registration to
normal.
These and other objects and advantages of the present invention
will become manifest upon a consideration of the following
description on the accompanying drawings.
In one embodiment of the present invention there is provided, in a
multistation web-fed press, tension responsive sensor means
disposed between two successive stations for producing a signal in
response to the tension of the web, misregistration sensor means
disposed downstream of the second station for producing a signal
responsive to the magnitude and direction of a registration error,
and means for controlling the tension of the web between the two
stations in response to a signal derived from said tension
responsive sensor means and from said misregistration sensor
means.
The present invention makes use of the fact that the error in
registration which is measured downstream from the second station
is a result not only of the relative orientation of the printing
cylinders at the two stations, but also of the tension of the web
extending between the two stations. Accordingly, suitable control
of the tension of the web between the stations permits an optimal
correction of the registration error. The appropriate tension for
the web is calculated from the registration error detected
following the second station, and when the tension of the web is
held at the appropriate value, the registration error is rapidly
brought to zero. The appropriate tension for the web is calculated
by a computer which develops a signal in response to the existing
tension on the web and the magnitude and direction of the existing
misregistration error.
BRIEF DESCRIPTION OF THE DRAWINGS
Reference will now be made to the accompanying drawings in
which:
FIG. 1 is a diagrammatic illustration of a press having two
separate printing stations;
FIG. 2 is a functional block diagram of apparatus constructed in
accordance with the invention for correcting the tension of the web
between two stations of the press;
FIG. 3 is a functional block diagram of a computer employed in
connection with the invention;
FIG. 4 is a schematic diagram of an illustrative embodiment of the
present invention, employed with a press having two printing
stations;
FIG. 5 is a functional block diagram of an alternative embodiment
of the present invention;
FIGS. 6 and 7 both illustrate apparatus for imposing a
predetermined tension on the web;
FIG. 8 is a functional block diagram of apparatus for deriving the
tension of the web from the rate of change in registration
error;
FIG. 9 is a diagramatic illustration of apparatus for generating a
signal in response to web tension; and
FIGS. 10, 11, and 12 are a series of graphs demonstrating the
correction of an initial registration error e.sub.0.
DESCRIPTION OF THE PREFERRED EMBODIMENTS
Referring now to FIG. 1 of the drawings, y represents the upward
displacement, from a nominal position O, of an idler roller 20
which roller supports the web 21 between two successive stations 22
and 23 of a multistation press. The actual tension of the web
between the two stations is T, which is greater than the normal web
tension T.sub.O because of the displacement of the roller 20. The
value of T.sub.O of course depends upon the particular press, and
also on the physical characteristics of the web 21
As illustrated in the graphs 24, 25 and 26 of FIG. 1, which
respectively relate the registration error e, the displacement y
and the web tension T to the length x of the web 21 passing through
the two stations 22 and 23, the registration error is initially
e.sub.o, the displacement is initially zero, and the tension of the
web is initially T.sub.O. The most rapid correction possible is
illustrated by the solid line curves in the graphs 24, 25 and 26
(FIGS. 10-12). The tension of the web is raised quickly to a
relatively high value T by raising the position of the idler roller
20 by an amount y. As a result, the web is elongated by a length
equal to 2y and the registration error is decreased uniformly at a
rate proportional to the elongation of the web. The linear decrease
in the registration error e is illustrated in the graph 24 of FIG.
10, and the increase in the tension of the web from T.sub.O to T
and the maintenance of the tension T is illustrated in the graph 26
of FIG. 12. The value T is ideally just below the breaking point of
the web, to maximize the rate at which correct registration is
approached.
As the web travels through the press, the idler roller 20 gradually
changes its position so as to maintain the tension of the web at
the desired value T. The displacement y of the idler roller is
increased uniformly as illustrated in the graph 25 of FIG. 11.
When the registration error e has been reduced to O, the tension of
the web is immediately returned to the normal tension T.sub.O by
returning the idler roller to a position which gives that tension.
Thereafter the tension on the web is maintained at its rated value
T.sub.O, and registration is maintained.
If the registration error should subsequently increase to some
value, the process is repeated as described above. The dashed lines
included in the graphs 24, 25 and 26 illustrate, for comparative
purposes, the effect of moving the idler roller 20 immediately to
its final position to make a correction in registration error, as
shown in the graph 25. From the graph 24, it is apparent that the
correction is asymptotic and therefore takes a considerable time to
accomplish, as compared with the more rapid linear correction
function of the ideal case, illustrated in solid lines in the
graphs 24, 25 and 26. The present invention, in controlling the
tension of the web, approaches the ideal case much more closely
than the systems of the prior art.
Referring now to FIG. 2, a computer 1 is provided in the present
invention for the purpose of converting data, sensed from the
condition of the web 21, into a signal representative of the
computed tension T.sub.c, which it is the calculated optimum
tension of the web between the successive stations, in order to
bring about the corrective action described above. The signal
generated by the computer 1 is a control signal and is connected to
the input of a regulator 2. The regulator 2 receives a second input
a signal T.sub.m, which is proportional to the measured tension of
the web between the two successive stages of the press. The
regulator 2 functions to produce at its output a signal in response
to the difference between T.sub.m and T.sub.c and furnishes such
signal to an actuator 3, which is the motive means which moves the
position of the idler roller 20. The actuator 3 may conveniently be
a solenoid, or a servo motor, both of which are well known. The
movement of the idler roller 20 is in the direction which increases
the existing tension of the web 21, if the quantity T.sub.m is
initially less than quantity T.sub.c, in order to bring T.sub.m
into identity with T.sub.c. Similarly, the web tension is reduced,
by moving the idler roller downwardly (as seen in FIG. 1) if the
quantity T.sub.m is greater than T.sub.c.
The computer 1 is illustrated in more detail in FIG. 3, which is a
functional block diagram showing the various elements of the
computer 1. The input of the computer 1 is a signal representative
of the registration error e.sub.i, and this is used to calculate
three separate parameters which are added together in a summing
amplifier 8. The first parameter, which is derived in the block 4,
is the error e.sub.i multiplied by a constant, K.sub.p. The second
parameter, derived in the block 5, is the sum of the errors,
beginning at some arbitrary point, multiplied by a different
constant K.sub.I. The parameter produced by the block 5 is derived
by adding together the registration errors developed at successive
sensings of the registration error, which sensings occur in regular
succession as registration marks pass under a sensing head or
misregistration sensor, as well known in the art. The third
parameter, derived in the block 6, is equal to the difference
between two successive registration errors, multiplied by a
constant K.sub.D.
The parameter derived in the block 5 is limited in the block 7 to
establish a maximum for the effect that the sum parameter may have
on the subsequent calculations.
All three parameters are summed in the summing amplifier 8 and then
passed through a limiter 9 which limits the maximum signal which
can be passed to the actuator 3, in order to establish upper and
lower limits on the tension of the web 21.
One additional input to the summing amplifier 8 is representative
of the integer "1", so that when the misregistration error is O (as
well as the parameters derived in the blocks 5 and 6), the output
of the summing amplifier is unity.
The output of the summing amplifier 8 corresponds to the ratio of
the tension T.sub.c which is needed to correct misregistration and
the nominal tension T.sub.O. This output is furnished to one input
of a multiplier 10, which receives as another input a signal
representative of the nominal web tension T.sub.O which is derived
by means of a device 11. The device 11 is preferably manually
adjustable, so that a correct nominal tension may be selected by
the operator for any particular web material which may be used. The
product generated by the multiplier 10 is the required signal
T.sub.c.
The function of the limiter 9 is to limit the maximum amount by
which the web tension may be changed from nominal, in order to
avoid breaking the web, and to maintain sufficient tension for
proper operation of the printing press.
Referring now to FIG. 4, the misregistration sensor 12 is shown at
its location downstream from the second station 23. It develops the
error signal e.sub.i. The sensor 12 is connected with the computer
1, which has been described in connection with FIG. 3.
The sensor 13 produces a signal in response the tension of the web
between the two stages 22 and 23 of the press, and develops the
T.sub.m signal which is furnished to the regulator 2. The computer
1 and the regulator 2, as well as the actuator 3, are connected as
described in connection with FIG. 2, and function to control the
position of the idler roller 20.
Although only two stations of the multistation press are
illustrated in FIG. 4, it is evident that a separate idler roller
may be associated with each two successive stations of the press,
and the equipment illustrated in FIG. 4 is then repeated for each
pair of stations, so that all registration errors are avoided.
Referring now to FIG. 5, there is shown, in functional block
diagram form an alternative embodiment of the present invention.
The construction and operation of the apparatus of FIG. 5 is the
same as that which has been discussed above in connection with FIG.
3 except that the integer "1" input to the summing amplifier 8 is
omitted, and the block 5, which derives the sum parameter, is
provided with an additional input which is connected from the
output of a differential amplifier 14. The amplifier 14 performs
the same function in the apparatus of FIG. 5 as the regulator 2
performs in the apparatus of FIG. 2, namely producing a signal
representative of the difference between two input signals. The
function of the separate input to the block 5 in the apparatus of
FIG. 5 is to initially set the value of the quantity represented by
the output from the block 5 equal to the output of the differential
amplifier 14. This is accomplished by initially operating the
single pole double throw switch 27 to connect the output of the
differential amplifier 14 to the line 28, via the terminal Ma. This
position of the switch 27 is referred to as the "manual" position,
and it functions to cause the block 5 to produce an output
corresponding to the quantity represented by the signal T.sub.m
produced by the sensor 13. This is so because the initial output of
the summing amplifier 8, during start up, is zero, with the result
that the output of the differential amplifier 14 is equal in
magnitude to T.sub.m. With the output of the block 5 thus set, the
output of the differential amplifier 14 drops to zero. The switch
27 is then moved to its other position, in which the output of the
amplifier 14 is connected to the actuator or motor 3. This position
of the switch 27 is the "automatic" position, because the operation
of the apparatus automatically controls the web tension. With the
apparatus of FIG. 5, the nominal tension T.sub.O is not required
for proper operation, because the calculated tension T.sub.c is
computed directly by the summing amplifier 8, rather than the
tension ratio which is computed in the apparatus of FIG. 3. It is
necessary, however, to modify in this event the values of the
constants K'.sub.P, K'.sub.D and K'.sub.I, which are employed in
deriving the outputs of the blocks 4, 5 and 6 of FIG. 5.
By the use of the apparatus of FIG. 5, the initial value of the
calculated tension T.sub.c is equal to the actual tension T.sub.m
at that time, and so sudden changes in the tension of the web are
avoided.
In a typical case for the computer of FIG. 3, the nominal tension
of the web is 50 Kg, and the tension controlling apparatus
functions to increase or decrease the actual tension 50 percent of
nominal, to a value between 25 Kg. and 75 Kg. The constants for the
computer of FIG. 3, and the limit values for the limiters 7 and 9
are:
K.sub.P 80 K.sub.I 20 K.sub.D 40 limiter 7 -0.5 to + 0.5 limiter 9
+0.5 to + 1.5 To 50 Kg.
For the same case, when the apparatus of FIG. 5 is employed,
K'.sub.P 40 K'.sub.I 10 K'.sub.D 20 limiter 7 25 to 75 Kg. limiter
9 25 to 75 Kg.
Preferably the computers of FIGS. 3 and 5 are analog in nature, and
update their outputs approximately 5 times per second, the rate at
which registration marks printed by the press pass under the
misregistration sensor 12 when such marks are printed one per
meter, and the web is moving at a velocity of about 300 meters per
minute. An analog representation of the value of the registration
error is placed in storage at the time of passage of each
registration mark, and the necessary parameters are produced
therefrom. Except for the sum, which is accumulated to continuously
represent the parameter produced by the block 5, it is necessary to
store only the last previous registration error. The computation
may obviously be carried out by digital means, instead of by analog
means, if desired.
In certain applications, and particularly in the case of webs
formed of material which is easily stretched such as webs of
synthetic material, it is difficult to measure the tension of the
web directly by means of the sensor 13. In this case it is
preferable to measure the tension inferentially, by measuring
auxiliary parameters which depend on the tension, or to impose a
calculated tension on the web by means of an open loop system, in
which case the tension responsive sensor 13 is not required.
When an inferential measurement of the tension of the web is to be
made, it is preferable to employ an on-line computer in connection
with the press, which calculates a value for the computed tension
T.sub.c on the basis of an imperically derived relationship and
controls the position of the idler roller 20 accordingly.
When the tension is to be maintained at the calculated value by
means of an open loop system, an apparatus which may be called a
"tension source" is employed. Although such a tension source may be
formed in a variety of ways, one preferred construction of such a
tension source comprises a pair of rollers 29 and 30 (FIG. 6) for
causing the web to form a loop 31 therebetween, with a roller 32
forced toward the end of the loop by a piston in a pneumatic
cylinder 33. The web tension is then proportional to the air
pressure within the cylinder 33, which is in turn controlled by the
output of the regulator 2 or the differential amplifier 14.
An alternative tension source is shown in FIG. 7, where the web 21
is caused to pass around a roller 34 having a shaft 35. The shaft
35 is supported by a lever arm 36, which in turn is supported at a
fulcrum 37. A movable weight 38 is supported on the arm 36 on the
opposite side of the fulcrum 37 from the roller 35, and serves to
counterbalance the latter. The position of the weight 38 is movable
along the length of the arm by means of a motor 39, which responds
to the output of the differential amplifier 14, which receives one
input from a sensor 40 in response to the position of the weight
38. The tension imposed on the web 21 is proportional to the
distance of the weight from a balance point on the arm 36 so the
output of the sensor 40 is proportional to the tension imposed on
the web 21.
One means for determining the web tension inferentially from a
change in the registration error is illustrated in FIG. 8. In FIG.
8, two successive stations of a printing press are illustrated,
with two sensors 41 and 42 disposed at positions following each
station, respectively. A third sensor 43 senses the vertical
position of the idler roller 20, which is controlled by the motor
3, in the same manner which has been described.
Inputs from the three sensors 41, 42 and 43 are connected to a
computer block 44, which calculates a function for the measured
tension from the three input signals. The sensor 43 is also
connected to the computer 1, which has been described in connection
with FIG. 2, and the regulator or differential amplifier 2
energizes the motor 3 in accordance with the difference between the
measured tension, represented by the output of the block 44, and
the calculated tension represented by the output from the computer
1.
The block 44 derives a value which is proportional to the actual
tension of the web 21 by making use of the fact that the tension of
the web is directly proportional to its elongation. The elongation
of the web is calculated by calculating the rate of change of the
misregistration error (by subtracting the current value of the
error from the previous value, as in the block 6 of FIG. 2) at the
location of each of the sensors 41 and 42 in blocks 45 and 46,
respectively, and subtracting one from the other in the block 47,
to give a signal representative of the net increase (or decrease)
in the quantity of the web which has entered the area between the
two stations 22 and 23 (or departed from that area) within the
period between two successive registration marks. This quantity is
accumulated, by the block 48, so that the output of the block 48 is
proportional to the total change, from some initial value, of the
quantity of the web present between the two stations 22 and 23.
This quantity is subtracted, in the block 50, from the output of
the block 49, which is representative of twice the change, during
the preceding period between two registration marks, in the height
of the roller 20. The output of the block 50 therefore gives the
change in elongation of the web between the stations 22 and 23
during the period between two registration marks. This quantity is
added (or subtracted if the quantity is negative) in the block 51
to the total of the changes in elongation to get the total
elongation which is proportional to web tension.
Referring now to FIG. 9, there is illustrated a diagramatic
illustration of apparatus for deriving a signal in proportion to
the tension of the web 21, which may be substituted for the sensor
13 of FIG. 4. In FIG. 9, a roller 52 is provided to support a loop
of the web 21, formed between rollers 53 and 54. The roller is
rotatable about a shaft 55 and the shaft 55 is supported by a
journal 56, which in turn is supported by a leaf spring 57 secured
to the frame 58 of the printing press. The spring 57 is secured to
the frame 58 at one end, and the journal 56 at its other end. An
arm 59 extends horizontally from the journal 56 and the vertical
position of the arm 59 moves as the spring 57 is flexed in response
to changes in the tension of the web 21. An increase in tension
moves the arm 59 downwardly as shown in FIG. 9, and a decrease in
tension allows the spring 57 to move the arm 59 upwardly. A
conventional induction type sensor 60, employing, for example, a
differential transformer with a movable slug disposed within the
transformer coils, derives an electrical signal as a function of
the movement of the arm 59. Although the apparatus of FIG. 9 has
been described as if FIG. 9 were an elevation view, it will be
understood that the apparatus of FIG. 9 functions equally well in
any attitude.
From the above description, it is evident that the apparatus of the
present invention accomplishes the objective of permitting a rapid
elimination of a registration error by controlling the tension of
the web in accordance with a function derived from the existing web
tension and the existing registration error, and for maintaining
the web tension at a level which maintains proper registration.
Since the signals derived by blocks 5 and 6 of the computers of
FIGS. 3 and 5 are related to the longitudinal dimension of the web
21, rather than to time, the operation of the apparatus is
independent of the velocity of the web 21.
* * * * *