Title of Invention

METHOD AND DEVICE FOR AUTOMATIC MONITORING OF REPETITIVE SEQUENCES OF AN INJECTION MOULDING MACHINE

Abstract The invention relates to a device and method for automatically monitoring repetitive operational sequences on injection molding machines and/or on handling units to injection molding machines, and for identifying abnormalities or disruptions with regard to the operational sequence. During normal production operation, the repetitive operational sequence of at least one partial process is provided, and a measurable quantity (1), which is influenced by the partial process, regarding the operational sequence monitoring is automatically recorded in short sampling intervals (2). The sampled values are compared with limit values (6) that were formed, by using an algorithm, from a multitude of previous sampled values of the same operational sequence. The sensitivity is automatically adapted to the process variations with regard to a decision to alert or terminate. This novel solution permits the carrying out of monitoring with the highest possible sensitivity, however, without a negative response, e.g. repetitive traveling motion, but also for operational sequences without moving parts such as the d.c. link voltage in converters or during controlled processes for monitoring a quantity that is not controlled.
Full Text Method and Device for Automatic Monitoring of Repetitive Sequences
of an Injection Moulding Machine
Technical Scope
The invention pertains to a method for monitoring repetitive sequences of an injection
moulding machine and/or handling devices for injection moulding machines and
identification of fluctuations or errors with respect to the sequence, and also a device for
monitoring and repetitive sequences of injection moulding machine and/or handling
devices of injection moulding machines and identification of fluctuations or errors with
respect to the sequence.
State-of-the-art Technology
A first characteristic feature of injection moulding machines lies therein that sequences
with several assembly groups, partly independent and partly working in a coordinated
manner, are involved with movable parts. In the centre of the injection moulding
machine is the casting mould. The mould lock is an independent assembly group, with
which the movable mould half is brought into a closed position for casting readiness, and
after the injection moulded parts have been cast and sufficiently cooled again, brought to
the open position. On the opposite side of the machine, the injection aggregrate is
arranged. The injection worm can be guided to the fixed mould half and after completion
of the injection sequence again guided away. It is a pre-requisite, that for the working
position a really large force is required:
In case of mould closing, the force must be so large that the melting pressure
required in the mould of say 1000 - 2000 bar may not open the mould halves, so
that fluid melt can exit between the mould halves.

2
In case of the aggregrate pressing, similarly it holds good, that the pressing force
has to be so large that between the injection nozzle and the corresponding mould
entry opening no gap is created and no fluid melt can exit.
In both cases it is almost an art to be able to select the pressing or closing forces
sufficiently large, but however not unnecessarily large. Therefore one endeavours
for an optimum pressing force within a highest run and a lowest run. Independent
of the optimum force, it has to be ensured that in automatic operation no machine
damages occur. The closing mechanism should not apply the mould closing force
"blindly", if for example entire injection moulded parts or plastic residues of the
previous injection cycle are still adhering to the mould half. In case of the
injection nozzle there should not be any solidified plastic melt between the mould
and the injection nozzle. In both cases a "blind" automatic movement sequence
of the mould closing or the drive up movement of the nozzle can result in an
erroneous situation or even cause damages to the concerned construction element.
At least in the case of the mould lock, nowadays a mould lock safety is a pre-
requisite. The mould lock safety is supposed to, in the case of any fluctuations in
movement sequence, e.g. not regular increases in mould lock forces, immediately
introduce an emergency stop for the mould movement.
DE 43 45 034 shows such a mould protection safety mechanism. As solution the
following are suggested:
a) slowing down of the closing movement to 20% or 10% in the region shortly
before contact of both mould halves;
b) measuring and storing the effective force requirement of all forces acting against
the movement of the mould half till contact, particularly frictional forces, in a first
closing sequence, so that
c) every following mould protection phase can be controlled and monitored on the
basis of the tolerance range laid down, and in case of excess forces to introduce a
stop immediately.

3
At the same time, one starts on the basis of an ideal force. Any fluctuation, generally an
increase in closing force, is detected as error and accordingly an emergency stop is
introduced.
In EP 0 25 987 it is suggested that the pressure- or force graph be monitored and
optimised during its stroke movement, e.g. the tool carrier, according to the already
described closing movement, especially with respect to the traverse part or the
corresponding rate values.
Both solutions have the disadvantage that the control of the drive movement and the drive
safety is conducted together, i.e. on the basis of the same agents. The drive safety is
dependent on the proper functioning of the control of the drive movement as well as its
sensor agents.
EP 0 296 183 too deals with a method for protection of moulds in injection moulding of
plastics, during closing. The mobile mould half is moved with a pre-given speed profile
against the fixed mould and the hydraulic closing pressure is restricted to a value that lies
above its standard value by a pre-determined adjustable amount. For this, it is suggested
that the standard value has been laid down by continuous measurement of the time
sequence of the hydraulic pressure, at least in the locking phase of the closing movement
of a previous closing sequence, and the pressure progression is monitored during its
closing sequence. The closing sequence is always interrupted if the hydraulic pressure in
the pressure progression lies above the measured standard value by a pre-determined
amount. As a special extension, it is further suggested that the pressure progression in
the immediately preceding closing sequence be selected as standard value. EP 0 096 183
suggests a constant tolerance range that makes sense only in a particular situation, mainly
in the locking phase of the closing movement of the mould tool. The preceding phases
cannot be protected in this way.
Another solution often use in the part is that a fixed value is selected for the mould
protection. The advantages of such solutions lie on the simplicity on the one hand and,

4
on the other hand, in the fact that no computer-/storage agents are required. The big
disadvantages however are that the drive safety is disproportionately insensitive and can
be applied only for the actual mould protection. In all solutions according to the state-of-
the-art technology an earlier identification of errors is hardly possible.
A second characteristic feature is the cyclic progression of individual sequences. At least
observed from outside, all movements, sequences and progressions repeat themselves
from cycle to cycle apparently without any changes. The higher the quality demand and
the automation degree, the more one can identify the variation in the progression in the
individual parameters. In the state-of-the-art technology one tries to at least counter a
part of the variation through more and more perfectionist regulation, or to reduce the
variation to an acceptable extent. Here two problem areas crop up, in the first case with
moved parts, in the second case without moved parts:
• In a regulated sequence, for example in injection, one parameter is always
regulated, e.g. the speed of movement of the injection axis. This results in an
injection pressure on the melt. The injection pressure is monitored with respect to
adherence to pre-given ranges.
• A second problem area lies in sequences without moved parts. An example for
this is the intermediate circuit voltage of inverters. Even here a particular range is
monitored.
It is interesting that in all mentioned characteristic features the critical parameters are
monitored. This can be a highest value or a lowest value. With the help of the
corresponding sensors or rated values of concerned components, corresponding alarm
signals are giving to the machine control mechanism in case of exceeding or falling
below, on the basis of which an immediate switching off or a stop action can be released.
However, the fact that at least in the case of serious fluctuations during which an
immediate switching off or immediate stop has to be introduced, it often takes place too
late. If the limits are set too close then this could lead to several unnecessary switch-offs.

5
The solutions according to the state-of-the-art technology do not allow any earlier
identification of discontinuity, fluctuations or errors in sequence.
It is now the task of this invention to look for a method as well as a device that can be
generally applied, and which with the maximum possible sensitiveness allows an
automatic monitoring safety of repetitive sequences, particularly the drive movement in
injection moulding machines, i.e. a high safety for an emergency stop, however no
unnecessary stop of the drive movements and early identification of errors.
Presentation of the Invention
The method as per the invention has the characteristic feature that in normal production
the repetitive sequence of at least a part process is pre-given and a measurable parameter
influenced by the part process is identified with respect to the sequence monitoring in
short scanning intervals and compared with limit values that are formed from several
earlier scanned values of the same sequence through an algorithm.
The device as per the invention has the characteristic feature that in normal production
the repetitive sequence or progression can be pre-given, and the device has storage-
/computer agents, particularly sensor agents, through which measurable parameters
influenced within a part process can be automatically identified with respect to the
sequence monitoring in short scanning intervals and can be compared to limit values that
have been formed from several earlier determined scanned values of the same sequence
through an algorithm.
It is very interesting that the new invention allows three new solution paths.
1) One can realise a drive path safety working with very high standard.
2) In individually regulated sequences, e.g. injection, the respective not regulated
parameters can be monitored according to the invention.

6
3) However, even in components without moved parts, e.g. the inverter voltage can
be monitored according to the invention.
As will be described below, the new invention not only monitors the moment of an error
but, works on the basis of earlier similar sequences and already identifies occurring
defects. One takes the preceding results and assesses at every moment in a predicted
manner any possible developing error, whereby the sensitivity with respect to an alarm or
interruption decision is automatically adapted to the process control for monitoring and
maximum possible sensitivity, however without erroneous activation.
It has been determined by the inventors that according to the first solution path the basic
requirement of monitoring of drive movement in the state-of-the-art technology has not
been adequately considered. The basic requirement lies in the special uniqueness of
injection moulding machine, in which almost all movements are determined by the
sequence/progression of the injection moulding process. In the movable axes of the
injection moulding machine one does not know the classical division of almost forced-
less drive up to the working position and thereafter a greater use of the working tool, e.g.
in a milling or boring machine. The drive up itself can already be afflicted with great
problems. The problems of any nature occur in the interior of the machine, partly even
completely within the tool.
The objective of automatic monitoring therefore has to be:
• Monitoring of individual or all sequence phases for movements forward and
backward;
• Monitoring and drive safety without erroneous activation;
• For normal production no purely empirical input of limiting values should be
used; rather during an automatic operation a self-optimisation should be possible
to match up to all possible normal operating conditions.
• The pre-requisite has to be created, that a maximum sensitivity be automatically
in the process dispersions, from shot to shot.

7
The new invention therefore foresees that repetitive drive movements are monitored by
movable components of the injection moulding machine and/or handling devices of
injection moulding machine with the help of momentary actual values of pressure/force
determined in scanned intervals at least in part sections of the drive movement. This
includes a constant optimisation of the drive safety. Division of the part section is
absolutely important, so that the drive safety in each situation can be adapted over the
entire drive part.
The invention allows a large number of advantageous extensions. For this, please refer to
the claims 4 to 17 and 19 to 25. Interval limiting values or group limiting values are
determined with the algorithm and laid as a basis of the following cycle. This has the
decisive advantage that - as in the state-of-the-art technology - it is not a single safety
range that is laid down for the entire drive movement, but from situation to situation a
momentary optimum limiting value is provided, whether it be for a start of the normal
and often almost force-less shifting movement or the great force build up at the end of the
movement, like e.g. during form locking or while pressing on the injection nozzle.
The group limiting values are statistically determined and obviously used as prescribed
values for forming an operation reference graph, whereby the determination takes place
at least over several cycles. This means that the limiting values are adapted in each
situation over the drive path, which ensures a maximum safety for example with respect
to possible accidents, whether it is pertaining to persons or machine parts.
A central concept of the new solution lies in the assessment of a group of scanned values.
This allows us to measure, analyse the pressure-Zforce values in scanned intervals of
milliseconds or fractions of seconds and then to immediately react according to the
situation. In this way fluctuations in the highest progression (peaks) as well as in the
lowest progression within a group of group limiting values can be optimally considered.
whereby each group includes several scanned intervals, e.g. ten or more. A further very
important point is that the individual scanned values are evaluated for monitoring the
drive movement as a function of the group limiting values. A scanned value is therefore

8
evaluated not with the individual preceding scanned value at the same point but as a
function of an entire group of earlier limiting values. There is further the possibility that
the number of scanned values per group and the location of the group along the drive
movement can be determined according to the situation.
Another advantageous extension concept is that fluctuations in the highest progression
(peaks) are determined within a group of successive scanned values and changes in the
group limiting values from cycle to cycle
a) are considered immediately for the following cycle in case of increase in the
scanned values; and
b) considered with delay in case of drop in the scanned values.
An increase in scanned value in peak run could announce a danger situation. One has to
react immediately. A drop would then hardly mean any danger, so that a delay would be
justified.
The situation is reversed in lowest run. Fluctuations in lowest run are determined within
a group of successive scanned values and changes in group limiting values from cycle to
cycle
a) are considered in a delayed manner for the following cycle in case of increase in
scanned values; and
b) considered immediately in case of drop in scanned values.
If one falls short of the required lowest run for the function, then generally the sequence
function is no longer guaranteed. Depending on the progression of the drive path it could
be advantageous to calculate comparative limiting values for one or more groups. The
individual limiting values for monitoring the scanned values are formed from the group
limiting values with a calculating function.
From several successive scanned values a group limiting value is determined and from all
determined group limiting values a reference graph is formed. The number of scanned

9
values per group and the position or the moment of the group in the sequence is laid
down according to the situation in a part process. It is however also possible, that
comparative limiting values are determined only in part regions of the drive movement.
The reference graph has the big advantage that a part of or the entire drive path is
monitored for all local situations, without continuously having to do calculations
unnecessarily. The group scanned values for monitoring the drive movement are
continuously evaluated as a function of the reference graph.
According to another very advantageous extension concept, during the first drive
movement while commissioning the machine a start reference graph with a start safety is
laid down, whereby for the following cycle the start reference graph is self-adapted stage-
wise and reduced to an operation comparison graph. During first commissioning it is
important that one does not immediately switch off on account of any small defect/error
due to safety measures. It is a pre-requisite that the erector is here also part of the
monitoring, with the highest degree of attention. For one or more scanning groups, on
the basis of the comparison limiting values a tolerance is laid down for alarm or
interruption decision for continuing the movement. A switch-off is decided statistically
or if required on the basis of experience, whereby the tolerance is preferably laid down in
such a way that for two or more instances of exceeding the start comparison limiting
value, a signal is given out to the machine control mechanism for interrupting the
movement. The new solution allows adaptation of the sensitivity with respect to an alarm
or interruption decision automatically to the process dispersions with the objective of
monitoring with the highest possible sensitivity, however without erroneous activation.
According to another advantageous solution path, it is suggested that an absolute drive
path safety be laid down with pre-given fixed limiting values for the scanned values,
whereby with the fixed limiting values also an upper limit can be laid down for the peak
run of the adaptive comparative values that can be laid down. In this way the client is
free to drive continuously with the absolute drive path safety corresponding to the
solution in the state-of-the-art technology. He has the advantage of, on the one hand,
having reliable employees and, on the other hand, a mechanical safety against damages to

10
the machine parts. There is also the possibility that the user for example works while
commissioning with absolute drive path safety and only later switches on an automatic
drive path safety, during which the movement over at least a part or the entire drive
movement is ensured without adjustments.
It has been found out by the inventors that the drive safety has to be organised
independent of or separately from the control and regulation instructions for the drive
movement. The momentary pressure-Aforce actual values have to be continuously
determined. By pressure-Vforce actual values one understands values that set in as
reaction to the movement, e.g. in the form of local carrier-, system- or bearing loads. It
can also be a hydraulic pressure or a motor torque. It is very important that dynamic
forces from acceleration as well as retardation and frictional forces, on the one hand, and
real defects, e.g. crack in a lubrication film or erroneous retention of whole or part of
injected parts, on the other hand, can be determined. All reactive forces acting against
the desired movement have to be determined and accordingly the real actual values of the
reaction forces against the drive movements will have to be determined. A disturbance
can also be in the form of lubricating film that it badly formed. If it has to do with
relatively small masses to be moved like ejectors, however with significant forces for the
corresponding function, then the running motor current consumption would be sufficient.
If it has to do with large masses to be moved, then the actual reactive forces from the
movement, whether it is on fixed or moved component, have to be determined.
A core concept of a new invention is that the drive safety is at the same time completely
active in the 'background' already in the first injection cycle, without the operator having
to adjust anything. In practice, a new tool is built in by the erector. He can then through
service points check and align not only the movement progression but also the
adaptability of the tool halves to one another. This generally takes place manually. It is
not at all significant, as to what extent automatic agents are playing a role in monitoring.
It is important that determination of the scanned values is involved for manual operation
or semi-automatic operation. Scanning of the resistance forces takes place in any case, so
that values above the drive path are stored already for the first injection cycle. The drive

11
safety should preferably be part of the machine control mechanism and can be released
only by trained personnel, e.g. for service jobs. For the regular working purpose the
drive safety is always active already in the erection phase and continuously determines
the relevant reactive pressures and/or forces.
According to a particularly advantageous extension, the path of the movable parts is
divided into several small individual situations and the drive movement is protected with
respect to the following individual situation. Thus the drive safety can work according to
the situation and allows protection of the movement through at least one part or the entire
drive movement without any adjustment. This gives the distinctive advantage that any
drive situation can be determined and monitored:
starting
constant movement
acceleration/retardation
acting of huge inertia forces of the concerned masses.
Another very interesting concept is that the pressure-/force values are measured and
analysed in the range of milliseconds or fractions of milliseconds. Measurement and
analysis require one or more high capacity computers because a decision is taken within
several milliseconds.
The new solution suggests that already during first commissioning of the moved parts the
pressure -/force values be measured and analysed, whereby the drive safety is designed in
an adaptable or learnable manner in such a way, that situations can be pre-given on the
basis of an input of earlier similar cases, or that errors during first test runs are detected
by the erector himself and prevented. Here one can see the advantages of the drive safety
according to situations. For example, the first very short starting peak can be handled
specially and on exceeding a pre-given fixed safety value, unnecessary stop can be
prevented. The same hold goods even for strong acceleration and retardation of large
masses. In the concrete extension it is suggested that the pressure-/force measured values
be measured and analysed during each drive movement, whereby for every section or

12
individual situation a reference point is calculated and from the result of the reference
point a reference graph is formed, especially by linear interpolation. The reference graph
is equipped for all special situations. A compulsive result is that the safety march for an
emergency stop at least for the most important situation is selected differently. It is also
important that not every short-term force or pressure in fact can already release an
emergency stop, but for example two or three impacts above the set limit would only
release a movement stop. The pressure-/force measured values are continuously analysed
during each drive movement and a possible error situation is determined according to an
algorithm, whereby the algorithm itself determines an error situation. The algorithm can
be a simple safety ratio number or a mathematical formula. The repetitive drive
movement is pre-given with the help of the machine control mechanism independent of
the drive safety. The drive safety determines error situations and on occurrence of an
error situation gives the machine control mechanism the corresponding signals for
introducing or preparing the necessary measures. The pressure-/force monitoring for the
drive safety takes place independent of the pressure- and/or force sensors that the
machine control mechanism uses for the drive movement. This holds good at least for
the mould lock as well as the aggregate movement. The drive safety is particularly used
for the movable part of the mould locking. As the greatest masses and forces of the
injection moulding machine come into play here, a reactive force or a reactive pressure is
determined, whether it be on the directly loaded part, the standstill part or the movable
part of the mould locking. This can be in the region of the fixed mould clamping plate of
the locking mechanism or the drive plate. The new solution is however suitable as drive
safety for all moved parts, e.g. for the aggregate shifting, the axial or rotating movement
of the plasticization or extruding worm and/or the movable parts, the ejector and/or the
core draws and/or the devices for injection moulded parts removal from the open mould
and/or for parts of a shot pots.
The injection moulding machine control mechanism has a storage agent for storing the
optimum sensitivity determined for a special injection job and instruction data for a
subsequent new injection job, in such a way that at the beginning of a first injection cycle
the drive safety is already laid in such a way that the injection job can be carried out

13
without pre-setting the drive safety, so that for each subsequent job with the same
injection moulding tool one can start with the optimum sensitivity determined
beforehand. The learning drive safety is a part of the basic software and can be used by
different axes with different measuring and regulating parameters. It can be the learning
drive safety also for injection pressure monitoring while injecting instead of the present
two-stage monitoring. Further application examples could be monitoring of the speed of
the injection axes during post-pressure and monitoring intermediate circuit voltage during
a cycle or a phase of the cycle.
According to the second solution path a respective not-regulated parameter, particularly a
speed, a pressure, a force, a path or a torque of a sequence or progression is monitored.
The big advantage is that the experience of the past is completely taken into account in
the monitoring concept and thus the requirement of best possible safety along with least
possible occurrence of error quota is ensured.
According to the third solution path the cyclic progression of a component without
moved parts is monitored. Depending on the component, it could be monitoring of the
current (A) or the voltage (V) of say the inverter.
Short Description of the Invention
The invention is described in details below on the basis of a few examples. The
following are shown:
Fig. 1 A schematic example for the progression of the scanned values and
limiting values formed with the help of an algorithm on the basis of earlier
scanned values;
Fig.2 formation of a reference graph from group limiting values;
Fig.3 group reference value formation as first step for forming the reference
graph according to the new invention;

14
Fig.4 formation of the reference graph according to the new invention as
theoretical graph;
Fig.5 a reference graph generated based on an interpolation;
Fig.6 the concretely measured actual values with respect to a prior determined
reference graph Rn... Rm;
Fig.7 optimisation of the sensitivity over a time slot of 10 m/s;
Fig.8 the step-wise cycle to cycle improvement of the reference graph;
Fig.9 a practical example for a measured value progression over the entire
movement of the closing tool, with an optimally adapted reference graph;
Fig. 10 an example for application of the new solution for injection monitoring;
Fig. 11 for post-pressure;
Fig. 12 an injection aggregate;
Fig. 13 and 14 an example for monitoring the inverter voltage;
Fig. 15 a simplified depiction of an injection moulding machine;
Fig. 16 a solution according to the state-of-the-art technology as per
EP 0 025 987;
Fig. 17 a solution according to the state-of-the-art technology as per
EP0 096 183.
Ways and Execution of the Invention
Fig. 1 and 2 show purely schematically the progression of the scanned values 1 that are
taken in scanned intervals 2. The scanned interval 2 can, as shown in the example, be in
milliseconds. It is however possible that the scanned interval 2 amounts to a greater time
range of lesser or more than 1 millisecond. A very interesting point is the combination of
several scanning intervals 2 into scanning groups 3. Theoretically for each scanned value
1 a limiting value 4 can be determined on the basis of the corresponding algorithm. The
new solution however goes in a different way, in that scanning groups 3 are formed as
one can see in fig. 2. The new invention suggests that group limiting values 5 be
determined on scanning groups 3. Only from the group limiting values 5 comparative
limiting values 6 and from the comparative limiting values 6 a reference graph 7 is

15
determined. As one can see in fig. 2, the scanning groups 3 do not necessarily have to
have the same number of scanning intervals. For example, in the start phase few scanned
value can be determined or many scanned values in shorter time. For a uniform drive
movement the number of scanning intervals can accordingly be the much greater. At the
end of the drive movement a situation can prevail like during the start. In the practical
example an operation reference graph 7 is set higher by a minimum value of say Sn > 0.5
KN than from a theoretical reference graph obtained. For the start a start reference graph
8 is laid down that lies higher than the operation reference graph 7 by a safety value "S'according to fig. 8, with a start correction value 9.
Fig. 3 shows the group limiting value formation as first step for formation of the
reference graph. For each scanning group respectively the highest scanned value is
selected for forming a group limiting value.
Fig. 4 shows a second step in the formation of the reference graph. If the desired value
parameters are changed (position, speed) then the reference graph has to be structured
afresh. In the first cycle the fine tool safety does not work, as information has to be
collected first. As during the first commissioning still no reference graph exists in the
sense of the invention, one works with an absolute drive path safety, e.g. according to the
solution in the state-of-the-art technology. As in the first cycle always irregularities and
run-ins occur, the information for the reference graph has to be collected over several
cycles, so that a statistically representative statement can be made. In order to prevent
erroneous activation of the tool safety during the phase where adequate information
(number of cycles) has not yet been collected, a start safety value "S" is added to the
reference graph, which ensures that fluctuations do not lead to an erroneous release of the
tool safety.
Fig. 5 shows of the interpolation of the intermediate value of the reference graph.
Fig. 6 shows the comparison of the reference graph with the measured pressure -
/force/actual values. Exceeding the reference graph is counted in sequence. Each

16
shortfall of the reference graph sets the counter back. If the counter for example reaches
three exceedings in succession, then the criterion "exceeding'" is considered to have
happened, so that immediately a corresponding signal is given out to the machine control
mechanism.
Fig. 7 shows a suggestion for optimising the sensitivity of the drive safety. Through run-
in effects the drive resistance can become lesser and regular. So that the reference graph
can also learn run-in effects, the algorithm shown in fig. 7 can be used. For each time
slot it is checked whether the following condition is always fulfilled:
Reference graph R 1ST signal-learning correction value L If this condition is fulfilled during the entire slot, then a counter belonging to the time slot
is increased by +1. When the counter reaches the learning duration P (number of cycles),
the reference graph in this time slot is corrected (subtraction) by 1/2 learning correction
value.
If the condition is not fulfilled in a cycle, then the counter is set back to 0. Meaningful
values for the tool safety are:
Learning duration T =- 100 cycles
Learning correction value = 0.1 kN
Fig. 8 shows structuring of the reference graph. While starting the tool closing there are
great dispersions, as the start position of the tool closing movement can disperse up to 0.2
mm. Especially after manual operation the start position cannot be reproduced.
Moreover, the drive resistances are higher after longer standstill periods, as a lubrication
film has to first get formed again. After standstill periods greater than stand time Tl not
only the learnt reference graph is used, which in this situation could be very sharp, but
one proceeds as in the case of the structuring of the reference graph, except that one starts
with value 1. So that while starting, the position uncertainties that could not be

17
statistically learnt (due to manual operation) do not lead to erroneous releases of the tool
safety, one generally does not release the tool safety during the first T
STARTVERZOGERUNG (start delay) during tool closing. T Startverz --= 50 ms.
Adaptive drive safety: if the machine is operated the first time with the new tool, then the
control mechanism knows very little about the force measuring signal to be expected
during the tool movement. This measured signal is complex and without additional
information can only be evaluated in a rudimentary manner. In this situation only a
simple limiting value can come in question. During the first movement only a simple
limiting value "basis drive safety" is effective.
First movement means: the first movement after changing the parameters that can
influence the force measuring signal:
• tool movement and tool position parameters
• eventually pressing on the aggregate, ejector, handling (indirect influence).
The actual tool safety in the sense of the state-of-the-art technology is active only over
the last one-third of the tool stroke, so that adequately high sensitivity can be attained.
One can think of defining an own limiting value for the erection movements as the
sensitivity could increase due to the low speeds. This basis drive safety is effective in all
operating conditions while closing the tool. If after parameter changes the machine is
operated the first time in such a way that the movement of the tool over the entire stroke
in automatic mode (not tilt operation) drives with production speed, then the force
progression can be analysed and a limiting graph can be calculated from it. This graph
contains the information about the force progression of the closing movement for the
given conditions. These movement cycles repeat themselves at random, but are mainly
identical, however differ, in that each graph contains statistical variations than can lead to
erroneous activation of monitoring.
An important task of the algorithm that makes formation of the reference curve and the
evaluation of the error condition possible is, on the one hand, to take into consideration

18
these dispersions and achieve a high sensitivity, however on the other hand, to rule out
erroneous releases. Additionally there are run-in effects, temperature-related forces and
influences that are dependent on the structure of the lubricating film. Till today, for this
the operator had to fall back empirically on the limit by always increasing the sensitivity
till he determined the erroneous activation. Thereafter the sensitivity is somewhat
reduced and it was hoped that now there would no longer be any erroneous activation,
which generally also functioned for an operating point. The observations hold good for
all movements.
User's point of view:
The user does not have to adjust or set anything. There is a configuration level where the
algorithm can be parametered in order to be in command of special cases.
Parameter: maximum force during closing movement
FXXX=50kN(l -99kN)
Default: 5% of the maximum closing force
Parameter: percentage of the stroke with active tool safety
YXXX-30%(0- 100%)
Default: 30% of the programmed stroke.
Sensitivity test of the drive safety:
As the entire optimisation takes place in a hidden manner and an operator does not have
to contribute anything to it, there could be the desire to quantify the result of
optimisation. For this a simple test is suitable: in the automatic cycle an error (defined
force signal) is superimposed on the actual force signal without allowing the algorithm to
learn this. By changing this error a release can be generated and hence it can be forecast,
at which sensitivity level the activation threshold for the actual production cycle lies.
The error can be set with respect to the amplitude and the position of the drive path
through parameters.

19
The operator gets three parameters: test switch on/off
Position of the drive path at which the error
effect gets released becomes the force
amplitude of the error.
Limits for adaptation:
On the one hand, adapting to the actual situation is a strength of the applied algorithm; on
the other hand, slow changes that the adaptation can follow should not lead to attaining
impermissible operating ranges. If for example, during a traverse draw the lubrication
gets worse and worse, with each cycle the friction becomes greater till ultimately the
traverse draw "eats up". In order to avoid this undesirable effect, limits are laid down
within which the algorithm can adapt itself. On exceeding this limit a corresponding
error case is generated and the error reaction released; information for detecting a drive
safety case and test of reaction of the test safety.
If the monitoring gets activated, a system alarm is generated and the following
parameters are given: position and speed at release point. Both these values permit an
assessment of the remaining brake path and can be evaluated with the optical result of a
jammed part. If the part is crushed very strongly, then the speed at the release point has
to be reduced. The speed at the release point is the only parameter at the disposal of the
operator in order to influence the result of a drive safety case. Ideally, the experiments
for optimisation of the reaction can be done, without parts getting jammed. This
possibility can be offered, in that the operator can release a drive safety occurrence in the
production without a part getting jammed.
On the basis of information of the system alarms it can be assessed whether a speed at the
release point is too high or has to be increased further so that an optimum cycle time can
be attained.

20
The operator gets three parameters: test switching on/off
Position of the drive path at which the error
is released becomes force amplitude of the
error (amplitude to be set at maximum).
Reaction to drive safety cases:
The reaction constitutes the quickest possible delay of the movement till standstill
without damaging the drive or shifting the machine. Depending on whether the
movement can be completely stopped or there is still a residual speed during crushing of
the part, there occurs a force peak that acts on the tool. In order to have, apart from the
result of the crushed part, a quantifiable magnitude for the occurrence, a system alarm is
given out with the parameters force of the force speed and the corresponding position.
Fig. 9 shows a practical example for a measured value progression over the entire
movement of the closing tool with an optimally adapted reference graph.
The invention is presented below on the basis of drive safeties of the closing movement
of the mould lock, particularly the tool safety. In principle, each linear axis can be
monitored in the manner described here, irrespective of whether the drives are hydraulic,
electrical or pneumatic. The force information is obtained from a force receiver, a
differential pressure measurement through cylinders, oil engines or pneumatic engine or
from the torque information of electrical drives. It is obvious that not all these variants
deliver the same quality of monitoring.
The objective of tool safety is, during a closing sequence not to allow any damaging
forces to act on the tool due to undesired occurrences (generally jammed plastic parts).
This pre-requires that such occurrences be identified and that one reacts fast enough. In
order to identify the occurrence, the force that acts on the movable mould plate is
measured on the toggle lever by expansion measuring strips. As jammed parts typically
allow the force to increase continuously till the tool movement get stopped, the release
sensitivity influences the stop sequence only to the extent that one reacts a few ms earlier

21
or later. It is very important that no erroneous releases occur because they lead to
bothersome production interruptions that cannot be tolerated. As the tool movement
continuously generates inertia forces due to acceleration and retardation that the sensor
measures, a way has to be found to filter out these forces. For this purpose, the actual
situation has to be first determined. During the first determination of the situation the
tool safety is not yet sensitive. According to the new solution always a minimum tool
safety has to be active. After determining the actual situation a sensitive tool safety can
become effective. The tool safety can be sensitive only if slow changes, e.g. friction or
thermal changes, can be filtered out. This however has the intrinsic danger that creeping
changes would never lead to activation of the tool safety. Therefore, a limitation of the
adaptation has to be foreseen, so that traverse draws with insufficient lubrication can lead
to activation before they eat up.
Movements from other axes (ejector, aggregate, handling etc.) can also have effects on
the signal offeree measurement through the inertia forces, which has to be taken into
account. After a tool safety occurrence is detected, a stop reaction is released. The
reaction speed of the drive determines the result: the faster one can apply the brakes the
lesser the crushing force gets built up, the kinetic energy is taken up by the drive instead
of by the crushed part and the sensitive tool.
For this two tasks have to be fulfilled:
• Reliable detection of the drive safety errors without erroneous release;
• Quick stopping of the drive in order to limit the crushing force.
In order to be able to detect the drive safety errors, the measured values from the force
signal have to be compared with reference values.
• In all operating conditions an error situation has to be identified by the force
progression.
• In the operating condition of production an optimum sensitivity to error situations
has to be achieved without manual intervention. Erroneous releases cannot be
tolerated.

22
• Creeping changes are filtered out up to an adjustable limit.
For each of the above mentioned three cases own reference values have to be determined.
Fig. 10 shows the situation during injection. During the injection sequence the speed of
the injection axis is regulated and it results in an injection pressure. This is obtained from
the mechanical (tool, worm geometry etc.) and the physical (temperature of the melt,
velocity etc.) surroundings. Here the injection pressure is used as the parameter to be
monitored and a learning graph is plotted. The method of plotting the comparative graph
is the same as for the tool safety, because the same software code is used. The distance
between the comparative graph and the starting phase is accordingly adjusted through a
configuration.
Fig. 11 shows the situation during post-pressure. During the post-pressure phase the
pressure is regulated and this results in the movement of the injection axis. The speed or
the path can be measured as the magnitude to be monitored and monitored with the help
of a comparative graph. Even here the comparative graph(s) is/are plotted over the speed
or the path with the same piece of software. Only the initial magnitude/parameter is not a
force but a speed or a position. In this example it could be meaningful to monitor not
only the upper limit but also the lower limit. For this one requires two comparative
graphs, one above and one below the actual value progression. Both the graphs are
created independent of one another with the same algorithm.
Fig. 12 schematically shows an injection aggregate.
Fig. 13 schematically shows the components for the intermediate circuit voltage. The
objective is to monitor the intermediate circuit voltage over a single cycle and from cycle
to cycle. The result can also be used for quality monitoring.
Fig. 14 shows monitoring of the intermediate circuit voltage over a cycle. The graphic
shows a possible progression of intermediate circuit voltage, over which a comparative

23
graph can be laid. The initial parameter for the drive safety is in this case a voltage.
Here too it would be meaningful to monitor the upper and lower limit.
Fig. 15 symbolically shows an injection moulding machine 30. Out of the several
individual components "X" only the following are indicated: drive or axis, mould
movement 31, ejector 32, movable mould or tool 33, worm 34, linear drive of the worm
(linear axis) 35 and rotation drive of the worm (rotation axis) 36, aggregate 37, protective
cover 38.
Fig. 16 shows fig. 2 of EP 0 025 987. In this solution according to the state-of-the-art
technology the drive safety is at the same time integrated into the control mechanism of
the drive path of the tool carrier. The vertical axis is marked "O", whereby P represents
the time intervals. The horizontal axis Q is the time axis. According to EP 0 025 987 the
progression of the curve R is adapted in an optimum manner to a pre-given value. In the
region marked T and U first upper values either pre-given by a computer or manually
determined are given, so that in the following injection cycle corrected graphs T or U set
in.
Fig. 17 shows fig. 1 of EP 0 096 193 and depicts a solution for tool safety frequently
applied in the state-of-the-art technology. Setting of the device requires conducting of a
closing sequence with empty tool with a minimum pressure progression po(t) and with
pre-given data of a fixed tolerance value and the tolerance range Pfs(t) obtained from it.
In this way any hindrance of closing can be immediately identified and the signal for
immediate ending of the closing sequence can be used.

24
Patent Claims
1. Method for control and monitoring of repetitive sequences in injection moulding
machines and/or handling devices for injection moulding machines and
identification of fluctuations or errors with respect to the sequence, whereby for
drive safety at least one measurable magnitude/parameter influenced by the
sequence of the part process is automatically determined with respect to the
sequence monitoring in short scanning intervals and with pre-give-able limiting
values, whereby during first commissioning a start reference graph with a
start safety can be laid down,
having the distinctive feature that
a) the limiting values are formed from scanning groups of earlier scanned
values of previous cycles of the same sequence through an algorithm; and
b) are used in a self-learning manner as prescribed data for formation of
comparative limiting values and an operating reference graph; and
c) the determined measurable parameter is compared with the operating
reference graph,
d) whereby during first drive movement a start reference curve is
dropped for the following cycles in a stage-wise self-adapting manner
and formed into an operating reference graph.
2. Method as per claim 1,
in which a maximum sensitivity is automatically adapted to the process
dispersions from shot to shot for monitoring and maximum possible sensitivity,
however without erroneous activation; and in scanning intervals the
momentary pressure- /force actual values of the drive movements are
monitored and the drive safety of each situation is adapted over the entire
drive path.

25
3. Method as per claim 2,
in which the sensitivity is adapted with respect to an alarm or interruption
decision automatically to the process dispersions.
4. Method as per one of the claims 1 to 3,
in which during the first sequence a start reference with a start safety is laid
down, whereby the start comparative value graph for the following cycles is
continuously reduced to an operating comparative graph.
5. Method as per one of the claims 1 to 4,
in which repetitive drive movements of movable components of the injection
moulding machine and/or of handling devices for injection moulding machines
are monitored with the help of momentary pressure- /force actual values
determined in scanning intervals, at least in part sections of the drive movements.
6. Method as per one of the claims 1 to 5,
in which with the help of the algorithm interval limiting values or group limiting
values are determined for the operating reference graph and laid down as basis for
the following cycles.
7. Method as per one of the claims 1 to 6,
in which fluctuations in the highest run (peaks) are determined within a group of
successive scanned values and changes in group limiting values from cycle to
cycle
a) are taken into account immediately for the following cycle in case of
increase in scanned values; and
b) taken into account delayed in case of decrease in scanned values.
8. Method as per one of the claims 1 to 7,
in which fluctuations in the lowest run (peaks) are determined within a group of
successive scanned values and changes in group limiting values from cycle to

26
cycle
a) are taken into account delayed for the following cycle in case of increase
in scanned values; and
b) taken into account immediately in case of decrease in scanned values.
9. Method as per one of the claims 1 to 8,
in which the individual limiting values for monitoring the scanned values are
formed from the group limiting values with a calculating function.
10. Method as per one of the claims 1 to 9,
in which from several successive scanned values a group limiting value is
calculated and from all calculated successive group limiting values with reference
graph is formed.
11. Method as per one of the claims 1 to 10,
in which the number of scanned values per group and the position or the point of
time of the group in the sequence is laid down according to the situation within a
part process.
12. Method as per one of the claims 1 to 11,
in which comparative limiting values are determined only in part regions of the
drive movements.
13. Method as per one of the claims 1 to 12,
in which for one or more groups of scanned values a tolerance for an alarm or
interruption decision is laid down on the basis of the limiting value for
continuation of the sequence.
14. Method as per claim 12 or 13,
in which the switching off is decided statistically, whereby the tolerance is
preferably is laid down in such a way that in case of exceeding the reference

27
graph two or more kinds a signal for movement interruption is given out to the
machine control mechanism.
15. Method as per one of the claims 1 to 14,
in which selectively an absolute drive safety with pre-given data of fixed limiting
values is laid down or evaluating the scanned values and with fixed limiting
values an upper limit is laid down for the peak run of the adaptively fixable
comparative values, or the automatic drive path safety is switched on, in
which the drive movement is protected free of adjustments.
16. Method as per one of the claims 1 to 15,
in which respectively a not-regulated parameter of a sequence or progression is
monitored, especially a speed (m/s), a pressure (kg/cm2), a force (N), a path or a
torque.
17. Method as per claim 16,
in which a component without moved parts monitors a voltage (V) or the current
(A).
18. Device for control and monitoring repetitive sequences of an injection moulding
machine and/or handling devices for injection moulding machines and
identification of fluctuation and errors with respect to the sequence, for drive
safety, whereby in normal production operation in repetitive sequence or
progression limiting values are pre-given and the device has storage-/computer
agents and sensor agents, with the help of which measurable parameters
influenced within a part process can be automatically determined with respect to
the sequence monitoring in short scanned intervals and can be compared with
limiting values that can be pre-given,
having the distinctive feature that
the storage- /computer agents are designed for forming limiting values from
scanned groups of previous scanned values of preceding cycles of the same

28
sequence through an algorithm; the storage- /computer agents are designed
to use the limiting values in a self-learning manner as pre-given data for
forming comparative limiting values and an operating reference graph, in
order to compare the determined measurable parameter with the operating
reference graph, whereby the storage- /computer agents are designed in such
a way that in case of commissioning of the machine the operating reference
graph can be reduced with respect to the limiting values that can be pre-
given.
19. Device as per claim 18,
in which it has storage- /computer agents and particularly sensor agents for
monitoring repetitive drive movements of movable parts of the injection moulding
machine and/or handling devices for the injection moulding machine, with the
help of which momentary pressure-/force actual values can be determined with
the help of scanning intervals at least in part sections of the drive movement,
whereby the drive movement can be actively pre-given through the machine
control mechanism.
20. Device as per one of the claims 18 or 19,
in which it has storage- /computer agents through which the limiting values can
be adaptively guided from drive movement to drive movement and which are
designed for automatic identification of errors and for giving out error message
signals to the machine control mechanism.
21. Device as per one of the claims 18 to 20,
in which the moved parts pertained to the mould lock and/or the mould height
adjustments and/or the movable parts, the injection aggregate and/or the worm
injection movement and/or the worm rotation and/or an injection piston and/or the
movable parts, the ejector and/or the core draws and/or the devices for injection
moulded parts removal from the open moulds.

29
22. Device as per one of the claims 18 to 21,
in which the sequence pertains to the injection, particularly the not-regulated
parameter as force (N) or speed (V).
23. Device as per one of the claims 18 to 22,
in which the movable parts pertain to the specific axes of a PET-machine for
producing pre-forms, particularly a removal robot and/or a hand over station
and/or a cooling station.
24. Device as per one of the claims 18 to 23,
in which the sequence or progression pertains to the voltage (V) and/or the
current (A).
25. Device as per one of the claims 18 to 24,
in which the injection moulding machine control mechanism has storage agents
for storing an optimum sensitivity determined for a specific injection moulding
job and data instructions for a subsequent new injection moulding job in such a
way that the drive safety is already laid while beginning a first injection cycle in
such a way that the injection moulding job can be carried out with complete drive
safety without pre-setting the drive safety.



The invention relates to a device and method for
automatically monitoring repetitive operational sequences on injection molding machines and/or on handling units to injection molding machines,
and for identifying abnormalities or disruptions with regard to the operational sequence. During normal production operation, the repetitive
operational sequence of at least one
partial process is provided, and a
measurable quantity (1), which is
influenced by the partial process,
regarding the operational sequence
monitoring is automatically recorded
in short sampling intervals (2). The
sampled values are compared with
limit values (6) that were formed, by
using an algorithm, from a multitude
of previous sampled values of the
same operational sequence. The
sensitivity is automatically adapted to
the process variations with regard to
a decision to alert or terminate. This
novel solution permits the carrying
out of monitoring with the highest
possible sensitivity, however, without
a negative response, e.g. repetitive
traveling motion, but also for operational sequences without moving parts such as the d.c. link voltage in converters or during
controlled processes for monitoring a quantity that is not controlled.

Documents:

04113-kolnp-2007-abstract.pdf

04113-kolnp-2007-claims.pdf

04113-kolnp-2007-correspondence others.pdf

04113-kolnp-2007-description complete.pdf

04113-kolnp-2007-drawings.pdf

04113-kolnp-2007-form 1.pdf

04113-kolnp-2007-form 2.pdf

04113-kolnp-2007-form 3.pdf

04113-kolnp-2007-form 5.pdf

04113-kolnp-2007-international exm report.pdf

04113-kolnp-2007-international publication.pdf

04113-kolnp-2007-priority document.pdf

4113-KOLNP-2007-(28-11-2013)-ABSTRACT.pdf

4113-KOLNP-2007-(28-11-2013)-CLAIMS.pdf

4113-KOLNP-2007-(28-11-2013)-CORRESPONDENCE.pdf

4113-KOLNP-2007-(28-11-2013)-DESCRIPTION (COMPLETE).pdf

4113-KOLNP-2007-(28-11-2013)-DRAWINGS.pdf

4113-KOLNP-2007-(28-11-2013)-FORM-1.pdf

4113-KOLNP-2007-(28-11-2013)-FORM-2.pdf

4113-KOLNP-2007-(28-11-2013)-FORM-3.pdf

4113-KOLNP-2007-(28-11-2013)-OTHERS.pdf

4113-KOLNP-2007-(28-11-2013)-PA.pdf

4113-KOLNP-2007-(28-11-2013)-PETITION UNDER RULE 137.pdf

4113-KOLNP-2007-CORRESPONDENCE OTHERS 1.1.pdf

4113-KOLNP-2007-CORRESPONDENCE OTHERS 1.2.pdf

4113-kolnp-2007-form 18.pdf

4113-KOLNP-2007-INTERNATIONAL SEARCH AUTHORITY REPORT.pdf

4113-KOLNP-2007-OTHERS.pdf

4113-KOLNP-2007-PCT REQUEST.pdf

4113-KOLNP-2007-TRANSLATED COPY OF PRIORITY DOCUMENT.pdf

abstract-04113-kolnp-2007.jpg


Patent Number 259845
Indian Patent Application Number 4113/KOLNP/2007
PG Journal Number 14/2014
Publication Date 04-Apr-2014
Grant Date 28-Mar-2014
Date of Filing 26-Oct-2007
Name of Patentee NETSTAL-MASCHINEN AG
Applicant Address INDUSTRIESTRASSE, CH-8752 NAFELS
Inventors:
# Inventor's Name Inventor's Address
1 WERFELL, FRIEDRICH NEUGADEN 157 CH-8752 SCHWANDI
2 FAH, MARKUS STEINENBRUGG 5, CH-8733 ESCHENBACH
PCT International Classification Number G05B 19/048
PCT International Application Number PCT/CH2006/000220
PCT International Filing date 2006-04-21
PCT Conventions:
# PCT Application Number Date of Convention Priority Country
1 751/05 2005-04-28 Switzerland