AT520658B1 - PROCESS MONITORING IN THE MANUFACTURING OF INJECTION MOLDED PARTS - Google Patents

Abstract

A measuring system for monitoring a casting process is described below. According to one exemplary embodiment, the measuring system has an ultrasonic transducer (8) which is designed to couple ultrasonic pulses into a casting tool (7). The measuring system also has a sensor circuit (22) coupled to the ultrasonic transducer (8), which is designed to control the ultrasonic transducer (8) with a control signal that has a sequence of pulses in order to generate the ultrasonic pulses and to process a received signal (u (t)) provided by the ultrasonic transducer (8) (or a further ultrasonic transducer) (8), which represents ultrasonic pulses reflected in the casting tool. The received signal (u (t)) has a sequence of pulses which corresponds to the control signal and which each comprise a main pulse and post-oscillation. The processing of the received signal includes - for each pulse of the received signal (u (t)) - the generation of a first measured value (UH (t)), which represents the amplitude of the main pulse, and a second measured value (UN (t)) , which represents the amplitude of the ringing.

Description

description

PROCESS MONITORING IN THE MANUFACTURING OF INJECTION MOLDED PARTS

TECHNICAL AREA

[0001] The present description relates to the monitoring of an injection molding process by means of an ultrasonic sensor.

BACKGROUND

Injection molding is one of the most important manufacturing processes in plastics processing. Increasingly higher quality requirements for the manufactured components with regard to mechanical and / or optical properties, the minimization of rejects (keyword "zero defect production") as well as the independent optimization (i.e. the shortening of production times or the automated compensation of changeable external influences) play an increasing role a role to meet the required quality requirements.

A so-called molding unit basically consists of an injection molding machine and the injection molding tool. While the injection molding tool has at least one negative mold - the so-called cavity - of the molded part to be produced, the injection molding machine processes the mostly granular plastic base material into a flowable melt and injects this material under high pressure into the tool, where it cools down (or reacts) in the cavity becomes). In many cases, the finished molded part can be detached (demolded) from the tool. The injection molding tool and the injection molding machine are often a purely structural unit without any connection by means of sensors.

If monitoring of the process by means of sensors cannot be dispensed with - be it for reasons of the necessary / prescribed quality assurance or because sensor data are required to control the injection molding process - so-called internal mold pressure sensors and molded part temperature sensors have become established on the market.

For an accurate measurement, however, it is usually necessary that the sensor head touches the molded plastic part, for which a hole into the cavity is necessary. In order to guarantee error-free measurement and tightness of the cavity at high pressures, strict tolerance requirements are placed on the bores in the tool, which results in corresponding costs. The resulting imprint of the sensor head visible on the finished molded part is a reason for exclusion for many applications in which a measurement would be useful - such as optical components (e.g. lenses, car headlight housings, etc.).

In addition, the sensor data from internal mold pressure sensors are usually used for process monitoring, which in the best case do not have to correlate with the product properties of the molded part produced. The molded part temperature sensors used as standard are generally not suitable for quantitatively determining the cooling rate (cooling rate) that is important for the product properties. The high pressures in injection molding require a massive coating of the actual sensor element (e.g. thermocouples), which results in long response times in the range of seconds. The sensor head is also in contact with the (temperature-controlled) tool, which can also falsify the result of the temperature measurement.

A promising possibility for the simultaneous monitoring of process and product properties during injection molding consists in the use of pulsed ultrasonic sensor systems. In a large number of publications (e.g. DE 197 37 276 C2, DE 198 34 797 C2, EP 2 657 801 A2, DE 20 2012 008 359 U1, US 5,951,163 A, WO 03/089214 A2) the measurement of the speed of sound is direct or indirect - and the sound attenuation in a reflection arrangement (ie ultrasonic transmitter = ultrasonic receiver) or a transmission arrangement. The direct coupling of viscoelastic or thermodynamic properties with sound damping or sound velocity theoretically enables a large number

Conclusions about the process and material condition during injection molding. The strong attenuation of the ultrasound in plastic melts can be seen as problematic; The attenuation can make the determination of acoustic parameters difficult or impossible, for example in the case of larger wall thicknesses as well as the presence of fillers (e.g. glass fibers, talc, etc.) in the plastic melt. For this reason, this type of ultrasonic measurement in injection molding has not found any relevant use in industrial practice.

Publication DE 2716833 A1 describes a system for monitoring one of e.g. hydrometallurgical processes by means of ultrasound technology, whereby an ultrasound pulse is fed into a container and that part of the signal that represents multiple reflections of the ultrasound pulse is evaluated in the received signal.

[0009] The publication US Pat. No. 7,017,412 B2 describes how to circumvent the above-mentioned problems of pulsed ultrasound systems by means of a continuously operating ultrasound system. However, the results are always strongly dependent on the local conditions (e.g. distance to the cavity, cavity thickness, coupling efficiency, individual sensor characteristics, etc.), and a resonance frequency must be determined on the system for each installation situation and mold temperature, which makes such a system less efficient industrial practice makes. The publication US 5,951,163 reports qualitatively for the first time about the possibility of using simple reflection measurements for flow front detection, with a quantitative evaluation method and suggestions for calibrating the system to take into account sensor aging effects and a decrease in coupling efficiency, which are, however, indispensable for permanent measurements in industrial practice is.

The inventors have set themselves the task of improving the signal processing / evaluation of pulsed ultrasound reflection measurements in injection molding tools, so that they can be used for a larger number of plastics (with or without fillers) and molded part wall thicknesses for process management and product monitoring Furthermore, sensor aging effects and changes in the coupling efficiency of the sensor to the tool can be compensated for by suitable signal normalization.

SUMMARY

The above-mentioned object is achieved by the measuring system according to claim 1 or the method according to claim 6. Various exemplary embodiments and further developments are the subject of the dependent claims.

A measuring system for monitoring a casting process is described below. According to an exemplary embodiment, the measuring system has an ultrasonic transducer which is designed to couple ultrasonic pulses into a casting tool. The measuring system also has a sensor circuit which is coupled to the ultrasonic transducer and which is designed to control the ultrasonic transducer with a control signal that has a sequence of pulses in order to generate the ultrasonic pulses and to control a signal from the ultrasonic transducer (or a further ultrasonic transducer) to process the received signal provided, which represents ultrasonic pulses reflected in the casting tool. The received signal has a sequence of pulses corresponding to the control signal, each of which comprises a main pulse and post-oscillation. The processing of the received signal includes - for each pulse of the received signal - the generation of a first measured value which represents the amplitude of the main pulse and a second measured value which represents the amplitude of the ringing.

A further exemplary embodiment relates to a method which has the following: coupling a sequence of ultrasonic pulses into a casting tool; receiving a corresponding sequence of reflected ultrasonic pulses and generating a received signal which has a corresponding sequence of pulses, each of which has a main pulse and a ringing. A first measured value and a second measured value are generated for each pulse of the received signal, each representing the amplitude of the main pulse and the amplitude of the ringing.

BRIEF DESCRIPTION OF THE DRAWINGS

In the following, exemplary embodiments are explained in more detail with reference to figures. The illustrations are not necessarily true to scale and the exemplary embodiments are not limited to the aspects shown. Rather, emphasis is placed on illustrating the principles on which the exemplary embodiments are based. In the pictures shows:

FIG. 1 contains timing diagrams to illustrate ultrasonic pulse signals and their envelope curve.

FIG. 2A is a time diagram to illustrate an ultrasonic pulse received by an ultrasonic receiver, which was reflected on the wall of a cavity of an injection molding tool.

FIG. 2B illustrates the occurrence of interferences due to multiple reflections in the wall layer of the cooling molded plastic part.

[0018] FIG. 3 is a block diagram to illustrate an example of a circuit for signal processing and signal evaluation of a reflected ultrasonic pulse.

FIG. 4 contains time diagrams to illustrate the resulting measurement signals in the course of an injection molding process.

FIG. 5 illustrates the measurable influence of the temperature of the injection molding tool on the growth of the edge layer of the cooling molded plastic part.

DETAILED DESCRIPTION

Before going into more detail on the figures, some general aspects of the exemplary embodiments described here will first be explained. According to the exemplary embodiments described here, short ultrasonic pulses are repeatedly generated by means of an ultrasonic transducer and emitted in the direction of the cavity of the injection molding tool. The ultrasonic transducer can be arranged in the injection molding tool or on its surface. The injection mold is usually made of metal, e.g. Steel. Before the plastic melt is injected, there is air in the cavity. Due to the very high density differences between metal and air, the sound energy is almost completely reflected at the metal / air boundary layer (i.e. on the inner wall of the air-filled cavity). The reflected signal can be received by the same ultrasonic transducer and converted into an electrical signal. This signal can now be further processed and evaluated by a signal processing unit. According to the examples described here e.g. a measured value is determined that represents the amplitude of the reflected signal (a so-called pulse or burst). For an air-filled cavity (before the injection of the plastic melt), this measured value can be used as a reference value; further measurements during an injection molding process are normalized (scaled) with this reference value. In this way, aging effects of the ultrasonic transducer and the coupling media, temperature differences in the tool steel and other possible effects can be eliminated, which enables the comparison of measured values over long periods of time.

During the injection molding process, plastic melt is injected into the cavity of the tool under pressure, whereby the plastic successively displaces the air from the tool and fills the cavity. As soon as the plastic melt reaches the position at which the ultrasonic pulse is reflected during this process, a reduction in the above-mentioned measurement value, which represents the amplitude of the reflected signal, can be observed, as part of the sound energy now passes through the metal / plastic interface ( ie through the inner wall of the cavity, which is now covered with plastic, into the plastic melt. In practice, around 1 - 5% of the coupled sound energy can be transmitted into the plastic melt.

The above-mentioned normalization to the measured amplitude of the reflected signal in the case of an air-filled cavity essentially corresponds to the calculation of the reflection coefficient

R on the inner wall of the cavity, which indicates what percentage of the coupled acoustic energy is reflected at an interface. Since this reflection coefficient R is practically 1 (100%) for a metal / air boundary layer, it can be determined very easily by expressing the measured values for the amplitudes of the reflected signal as a multiple of the reference value mentioned, which is the amplitude of the reflected signal at air-filled cavity. A standardized measured value of 1 therefore corresponds to a reflection factor of 100% and indicates an air-filled cavity. A normalized measured value less than 1 (e.g. between 0.95 and 1) indicates a plastic-filled cavity. With the help of the reflection coefficient R and its change, conclusions can be drawn about the density of the plastic during the injection molding process. Ultrasonic transducers can be arranged at several points on an injection molding tool, which enables measurements at different positions in the cavity. In such applications the ultrasonic measurements can e.g. can be used to detect the flow front of the plastic melt.

In addition to the detection of the flow front of the plastic melt in the cavity, suitable signal processing can also be used to determine the point in time when the cavity is completely filled with plastic melt. If the cavity is completely filled with plastic melt, an increase in pressure in the plastic melt leads to an increase in density and thus to a further reduction in the reflection coefficient. The time at which the cavity is completely filled is an important process parameter in injection molding. At this point in time, the speed-controlled injection of the plastic melt is usually switched to what is known as "holding pressure". Usually a (e.g. constant) pressure is exerted on the plastic melt in the cavity in order to reduce a volume contraction due to the cooling of the plastic. The ultrasonic measurements can be used to control the switching time. Too early switching can lead to incompletely filled cavities and too late switching to overfilling and thus to burr formation on the finished molded part at the mold parting line or, in the worst case, to overmolding of the mold, which causes the mold halves to open at the parting line during the process; this can lead to severe damage to the tool.

With the aid of the concept described here, the optimum switching point in time can be detected with the aid of the ultrasonic measuring system, and e.g. be signaled by sending a trigger signal to the injection molding machine. According to one exemplary embodiment, the aforementioned switchover time can be detected by detecting that the normalized measured value for the signal amplitude of the reflected signal falls below a threshold value.

The measured values determined by the measuring system for the amplitude of the reflected signals (ultrasonic pulses) can also be used to assess the homogeneity of the plastic melt with regard to undesired gas inclusions or unmelted plastic granulate. These two errors are primarily caused by moist plastic base material (granulate) or by defects in the plasticization of the base material and lead to serious optical and / or mechanical defects in the plastic molded part produced. Gas inclusions increase the reflection coefficient (due to a decrease in density compared to the plastic melt) and lead to an increase in the amplitude in the reflected signal. Conversely, unmelted granulate leads to a reduction in the reflection coefficient (because of the increase in density compared to the plastic melt).

Another detectable variable is the point in time when the molded part disappears from the cavity wall. From a certain point in time, the increasing cooling of the molded part does not allow any further transport of plastic melt into the cooling molded part due to the applied holding pressure. As a result, the volume contraction caused by the cooling can no longer be compensated for and the molded part shrinks and becomes detached from the inner wall of the cavity. This point in time is important because the air layer that forms between the cavity wall and the molded part drastically worsens the cooling effect on the molded part and a further stay in the cavity may unnecessarily increase the cycle time. The point in time of disappearing can be detected by a sudden increase in the measured value representing the amplitude of the reflected signal.

Additional information that can be determined with the aid of the concept described here is the thickness and the growth rate of the cooling edge layer of the molded part in the cavity. Inside the plastic, near the cavity wall, there is a further reflection of the ultrasonic signal at the boundary layer between the solidified plastic and the plastic melt. This further reflection can constructively or deconstructively superimpose a part of the main reflection (which takes place on the cavity wall), whereby interferences can be observed when evaluating the measured values for the signal amplitudes. These interferences provide information about the growth rate of the already solidified surface layer of the molded part in the cavity. A change in the growth rate in the molded part can indicate a malfunction in the tool cooling (see also FIG. 5). With the concept of monitoring a casting process described here, deviations from a desired course during the growth of the surface layer of the solidifying melt can be detected. Such deviations can, as described, indicate inhomogeneities (e.g. bubbles, incompletely melted plastic granulate, etc.) in the melt or incorrect cooling of the casting tool.

The timing diagrams in FIG. 1 serve to illustrate an ultrasonic pulse and an example of the measurement of the signal amplitude. Diagram (a) in FIG. 1 shows an ultrasonic pulse, more precisely the signal course (ie the instantaneous value of the signal as a function of time) of an electrical signal A (t) which is converted into an acoustic signal by an ultrasonic transducer and emitted or that is received by the ultrasonic transducer and generated by converting the received (reflected) acoustic signal. A pulsed ultrasonic signal signal usually consists of a sequence of pulses. Due to the limited bandwidth of real ultrasonic transducers, a pulse usually has the signal course of a burst signal (short: burst), i.e. a pulse comprises a few oscillations with a rapidly increasing and then decreasing amplitude, the frequency of the oscillation e.g. corresponds to the resonance frequency of the transducer. In Fig. 1, the maximum amplitude (approximately in the middle of a pulse) is denoted by Ap.

The mentioned reflection coefficient R can be defined as the quotient Ap / At of the amplitude type of an emitted ultrasonic pulse and the amplitude Ar of the corresponding reflected ultrasonic pulse (neglecting the attenuation of the sound wave). Taking into account the attenuation, R is equal to (Ar / Art) -e? - * (a> 0), where L is the distance between the ultrasonic transducer 8 and the reflective cavity wall and a represents an attenuation coefficient (see also Fig. 2B) . Since, as shown in FIG. 1, the amplitude of an ultrasonic pulse A (t) is not constant, the maximum amplitude A »can also be used instead of the amplitude to determine the reflection coefficient. One possibility of determining the maximum amplitude Ap of a burst signal by measurement is to first rectify the signal and then subject it to low-pass filtering. The rectified signal | A (t) | is shown in diagram (b) of FIG. 1 and the filtered rectified signal in diagram (c) of FIG. 1. This filtered, rectified signal also represents the envelope of the ultrasonic pulse and is here denoted by env {] A ( t) |}. The amplitude of the envelope corresponds to the maximum amplitude Ap of the ultrasonic pulse A (t), i.e. Ap = max {env {| A (t) |}. The amplitude of the envelope can therefore be used as a measured value for the amplitude of the ultrasonic pulse A (t) (and consequently also for determining the reflection coefficient).

Alternatively, the hatched area under the envelope curve env {| A (t) |} can also be used as a measured value for the amplitude. Since the pulse duration te »is essentially constant and given by the properties (including the bandwidth) of the transducer, the area is approximately proportional to the maximum amplitude Ap. This area can also be interpreted as a measured value for the signal energy (to be more precise, for the square root of the signal energy) of the ultrasonic pulse A (t). A measured value for the amplitude of the ultrasonic pulse A (t) can accordingly be determined by integrating the envelope signal env {| A (t) |} over the pulse duration tr. Instead of rectification (diagram (b) from FIG. 1), the signal A (t) could alternatively also be squared. In this case, the measured value obtained would be proportional to the amplitude square Ap * of the envelope (i.e. the signal energy), which is nevertheless a suitable one

The measured value for the amplitude of the ultrasonic pulse A (t) is.

2A shows a more realistic waveform of an electrical signal u (t) generated by an ultrasonic transducer (e.g. an electrical voltage), which represents the waveform of a received ultrasonic pulse that was previously reflected on the cavity wall of an injection molding tool. As can be seen in FIG. 2A, the signal u (t) (reflected ultrasonic pulse) has two sections which can be distinguished from one another, a first section, which is referred to as the "main pulse" and a second section, which is called " Nachschwingen "(ringing) is called. The main pulse has a pulse length t-, e.g. the time period ty can comprise a few (e.g. the first three) half-waves of the signal u (t). The ringing has its cause in the finite bandwidth of the ultrasonic transducer and the duration of the ringing is denoted by tv. In the area of this post-oscillation phase, interference occurs in the case of multiple reflections in the solidifying edge layer of the molded part (see also FIG. 2B). The end of the second section cannot be clearly seen from the signal curve, as the reverberation slowly subsides. However, one can define, for example, the end of the second section as the point in time at which the amplitude of the oscillation has fallen below a certain threshold value (e.g. 10% of the maximum amplitude). For example, the duration tn can be a certain number e.g. ten) of half-waves after the main pulse. The times ty and ty are approximately constant values for a specific measurement setup (pulse generator, ultrasonic transducer, geometry of the tool and cavity, etc.) and are therefore known system parameters.

During the injection molding process, ultrasonic pulses are continuously generated, reflected on the cavity wall (main pulse) and reflected back to the transducer. Some process parameters in injection molding can be recognized by the change in the amplitude of the main pulse. Other interesting parameters can be derived from the change in the amplitude of the ringing. For example, the following process parameters can be detected by evaluating the main pulse: time of flow front detection, time of complete filling of the cavity with plastic melt, checking / detection of the time of the beginning of the holding pressure phase, duration and end of the holding pressure phase, time of the molded part disappearing from the cavity wall, Assessment of the homogeneity and density of the melt. The evaluation of the post-oscillation enables the monitoring of the surface layer of the cooling plastic melt (surface layer growth rate or surface layer thickness).

[0034] FIG. 2B illustrates the origin of the aforementioned interferences in the post-oscillation phase of an ultrasonic pulse. 2B shows a sketch of the ultrasonic transducer 8, a tool 7 with a cavity 9 which is filled with plastic melt P. When the plastic cools / solidifies, a growing edge layer P ‘starting from the wall 9‘ of the cavity 9 forms in the molded part. That is, the thickness d of the surface layer increases during the cooling time. As sketched in Fig. 2B, part of the ultrasonic power is reflected on the wall 9 ‘. The non-reflected part penetrates into the boundary layer P ‘and is reflected back at the other end of the boundary layer (at the solid / liquid transition). The part of the sound reflected in the edge layer has a longer path (by twice the thickness d) than the sound reflected on the cavity wall 9 ', which corresponds to a transit time difference of At of 2d / c (co denotes the speed of sound in the Boundary layer P °). The transit time difference At is therefore proportional to the thickness d of the boundary layer P ‘and increases during cooling (since the boundary layer P‘ becomes thicker). In contrast to this, the running time 2L / cs in the steel wall 71 of the tool 7 remains constant (Cs denotes the speed of sound in the wall of the tool, L the wall thickness). In practice, the transit time difference At is shorter than the duration of an ultrasonic pulse, so that the pulse received by the transducer 8 is a superposition of the two reflections (on the cavity wall 9 'and in the edge layer P' °), this superposition being essentially in of the post-oscillation phase occurs, which is shown in FIG. 2B as the interference region | is designated. The change in the interferences and their meaning will be discussed later with reference to Figs.

Fig. 3 is a block diagram and shows an example of a possible implementation of a

Sensor circuit for the above-described ultrasonic measurements in an injection molding tool. It goes without saying that there are many other options for implementing the same or an equivalent function in terms of circuitry. In view of the above and the following explanations, a person skilled in the art will be able to easily implement the illustrated or other suitable implementations.

According to the example shown in Fig. 3, an ultrasonic transducer 8 is arranged in an injection molding tool 7 or outside on its surface. Also shown in FIG. 3 is an ultrasonic pulse which runs from the transducer 8 through the injection molding tool 7 to a cavity 9 arranged in the tool 7, is (at least) reflected on the wall 9 'of the cavity 9, runs back to the transducer 8 and from this is received again. During the injection molding process, plastic melt P is pressed into the tool. Fig. 3 shows the flow front of the plastic melt P in the cavity 9 and the flow direction.

The ultrasonic transducer 8 serves both as a transmitter and as a receiver. The sensor circuit 22 coupled to the transducer 8 can consequently be viewed as an ultrasonic transceiver (transceiver). The sensor circuit 22 thus has a transmission path and a reception path. The transmission path comprises an oscillator 1 which generates a signal s »p (t) with pulses of a defined pulse duration and an adjustable pulse repetition frequency PWF. For example, the oscillator 1 can be an astable multivibrator (relaxation oscillator). Other signal generation options can also be used. The pulse repetition frequency PWF can be set using a signal received at input 2. The oscillator 1 can be activated or deactivated via a trigger signal son (t), which can come from the control of the injection molding machine. There are many ways to implement this activation or deactivation. In the example shown, the trigger signal son (t) received by the injection molding machine at input 4 indicates via a certain logic level (e.g. high level) that the oscillator should be active. For this purpose, the oscillator signal sp (t) and the trigger signal son (t) are fed to an AND gate 3, which is transparent to the oscillator signal sp (t) as long as the trigger signal son (t) has a high level. If the level of the trigger signal son (t) changes to a low level, the oscillator signal sp (t) is blanked by the AND gate 3 and the signal level at the output of the AND gate 3 is approximately zero (low level). The trigger signal son (t) can e.g. a signal can be used which indicates whether the tool 7 of the injection molding device is closed.

The output signal sp ‘(t) of the AND gate 3 (corresponds to the oscillator signal sp (t) if the oscillator is active) is fed to a trigger input of a high-voltage pulse generator 5. The high-voltage pulse generator 5 is designed to generate a high-voltage pulse (i.e. a burst signal) of a defined pulse duration (e.g. a few microseconds) in response to a pulse received at the trigger input. As long as the oscillator 1 is active (i.e. the AND gate 3 does not blank the oscillator signal sp (t)), the high-voltage pulse generator 5 generates high-voltage pulses with the pulse repetition frequency PWF.

The sequence of high-voltage pulses generated by the high-voltage pulse generator 5 is fed via a transmitting / receiving splitter 6 to the ultrasonic transducer 8, which is designed to transmit corresponding acoustic ultrasonic pulses, i.e. to emit into the tool 7. The ultrasonic transducer 8 converts the electrical (burst) signal into an acoustic signal. The transceiver switch 6 is designed to prevent high-voltage signal components from being coupled into the receive path. Such transceiver switches are known per se and are therefore not explained in more detail here. The acoustic ultrasonic pulse emitted by the transducer 8 runs through the material (e.g. steel) of the tool 7 to the cavity 9, is reflected there on the cavity wall 9 and runs back to the ultrasonic transducer 8, which converts the reflected acoustic ultrasonic pulse into an electrical one Signal converted. The received electrical signal is fed to an analog filter via the transceiver switch 6, for example a bandpass filter 10, in order to suppress undesired frequency components, and is amplified (see amplifier 11). In the example shown, the band-pass filtered and amplified signal is denoted by u (t); an exemplary waveform of the signal u (t) was e.g. already shown in Fig. 2A. In this regard, reference is also made to the above

Referred to explanations. The received (filtered and amplified) signal u (t) also contains a sequence of pulses with the pulse repetition frequency PWF, only one pulse being shown in FIGS. 2A and 2B.

As explained above, there are various ways of determining a measured value that represents the amplitude of a pulse in the received signal u (t). In the present example, the area under the envelope curve of the pulse is determined by means of integration, with the main pulse and the post-oscillation (see FIG. 2A) being evaluated separately. As mentioned, the main pulse is the first received echo of the emitted ultrasonic pulse which, due to its relatively high amplitude, has no measurable interference phenomena with the (very weak) signal from the boundary layer of solidified plastic / melt from the interior of the molded part.

To determine the envelope curve, the received signal u (t) is first rectified (see FIG. 3, rectifier 12) and then low-pass filtered (see FIG. 3, low-pass 13). The effect of the rectification and low-pass filtering was explained above with reference to FIG. 1 and is therefore not repeated here. The resulting envelope signal env {lu (t) |} is fed to a first integrator 14 and a second integrator 15, which are regularly reset at the beginning of a pulse. For example, the oscillator signal sp (t) (alternatively also sp ‘(t)) can be fed to a reset input of the integrators 14, 15. The time windows of the integration are specified by the pulse generators 16 and 17, which can each contain, for example, a monostable multivibrator and a delay element. The pulse generators 16 and 17 are designed to generate a pulse with a defined length and a defined delay (relative to the oscillator signal) in response to the receipt of a pulse from the oscillator signal sp z (z). These pulses indicate the integration time window and can e.g. Enable inputs of the integrators 14, 15 are supplied.

The length and the time position (relative to the oscillator signal) of the pulses generated by the pulse generators 16 and 17 define the integration time window for the integrators 14 and 15. For the first integrator 14, the length of the time window is ty and for the second integrator 15 is the length of the time window tn. These time windows are also shown in FIG. 2A and correspond to the length of the main pulse and the ringing. The length and the temporal position of the time windows t4 and tn depend on the transit time of the acoustic signal and on the frequency and the bandwidth of the ultrasonic transducer 8 and are therefore known system parameters. At the end of the integration time windows ty and tv, the output signal of the associated integrator 14 or 15 is sampled by means of the sample & hold circuits 18 and 19, respectively. The resulting discrete-time output signals of the sample & hold circuits 18 and 19 are denoted by un [k] and un [k], where k is a time index. The sampling rate corresponds to the pulse repetition frequency PWF. The time-discrete output signals uH [k] or un [k] can be output at the outputs 20 or 21 and e.g. digitized and processed further by means of a processor (not shown). Alternatively, analog (quasi-continuous) signal processing is also possible.

According to one embodiment, the sensor circuit 22 is designed as an assembly, with a connection for the ultrasonic transducer 8, an input 4 for the trigger signal son (t) from the injection molding machine, an input 2 for the selection of the pulse repetition frequency and two outputs 20 and 21 for the output signals ufn [k] and un [k], which each represent the area under the envelope curve of the main pulse and the post-oscillation. As explained at the beginning with reference to FIG. 1, these areas can be used as a measured value for the (maximum) amplitude of the main pulse or of the post-oscillation.

The timing diagrams from Fig. 4 show an example of a typical signal course of the (discrete-time, quasi-continuous at high pulse repetition frequency) output signals ufn [k] and un [k], which represent the course of the (maximum) amplitude of the main pulses and the associated post-oscillation phases. As mentioned above, normalization to 100% ensures that aging effects and varying coupling efficiency of the ultrasonic transducer do not play a significant role. The signals shown in Fig. 4 are UHn [k] and Un ([k]

therefore based on the measured values for an empty (air-filled) cavity (e.g. UH (tk) = un [K] / uH [0] and Un (tk) = un [k] / un [0]). The time k = 0 (corresponds to to = 0s) represents the start of a measurement with a closed cavity still filled with air before the injection of plastic melt and the time tx corresponds to the multiple of the reciprocal value of the pulse repetition frequency (tk = k-PWF "").

Since there is air in the cavity at the time to, practically 100% of the sound energy is reflected on the cavity wall 9 ‘(cf. FIG. 3). The normalized signal UH (t) can therefore also be interpreted as the reflection coefficient R. If the flow front of the plastic melt reaches the "field of view" of the ultrasonic transducer 8 during the injection molding process, the reflection coefficient R and thus also the (maximum) amplitudes of the reflected main pulses (as well as the amplitudes of the associated ringing) decrease significantly. The flow front of the plastic melt P. (see Fig. 3) can e.g. can be detected in that the signal Upn (t) is monitored during the injection molding process and that point in time t-- is detected at which the standardized signal Uk (t) falls below a predeterminable threshold value Sfr. The threshold value Ser can e.g. are in the range of 98-99%. In this way, rapid and precise detection of the flow front is possible. Additionally or alternatively, the signal Un (t) can be used for the flow front detection. Changes in the measured value Upy (t) indicate changes in the reflection coefficient R and thus indirectly also to density fluctuations in the melt, e.g. can be caused by the aforementioned inhomogeneities. For process monitoring, e.g. An ultrasonic transducer can be placed on the casting tool (e.g. shortly after the injection point) in such a way that the entire injected melt has to pass the measuring field ("field of view") of the sensor. Short, sudden changes in the measured value Uu (t) (i.e. the reflection coefficient) can indicate inhomogeneities.

As soon as the cavity 9 is completely filled with plastic melt P (see. Fig. 3), there is a significant compression of the plastic material in the cavity, and consequently the reflection coefficient R and thus also the (maximum) amplitudes of the reflected main pulses (as well as the amplitudes of the associated ringing) continue to decrease significantly. The associated point in time tu of the complete filling of the cavity 9 can also be detected by means of a threshold value Sy which is smaller than the threshold value Ser (e.g. Se is in the range of 95-97%). The point in time tu is accordingly detected in that the normalized signal Up (t) falls below the threshold value Su. This point in time tu can be used to initiate the holding pressure phase, in which from a constant injection speed of the plastic melt to a e.g. constant pressure is switched. In this holding pressure phase, the molded part begins to cool down and a thicker surface layer is formed in which the plastic melt has solidified. Additionally or alternatively, the signal Un (t) can be used for the detection of the switchover time tu.

When the holding pressure is applied, the growth of the edge layer begins in the molded part, starting from the surface of the molded part and into the interior. Because of this (growing) edge layer P ‘, there is a reflection which is superimposed with the reflection on the cavity wall 9‘ (see FIG. 2B); this superposition (interference) is particularly noticeable in the (maximum) amplitude of the ringing - that is, in the signal Un (t). This superposition (interference) results in a kind of beat signal, which can be seen in the signal UNn (t) after the time tu.

The effect of the post pressure on the molded part can be seen from the signal Ux (t), which represents the amplitudes of the main pulses. Shortly after the switchover point tu, the injection molding machine regulates the pressure to the (usually constant) set holding pressure. The time at which this pressure level is reached can be seen in the signal Ux (t); the amplitude increases again slightly. The associated point in time is denoted by tn, starı in FIG. The duration of the holding pressure phase is usually set by the user on the injection molding machine. In fact, it can only work as long as plastic melt can be fed into the molded part through the liquid core of the molded part. The actual point in time In, at which the holding pressure no longer acts can be read from the course of the signal UH (t), since the pressure drop leads to a change in density and thus a sudden change in the

maximum amplitudes of the main pulses leads. The actual holding pressure duration acting on the molded part results from the difference tn = tn, end - IN, start-

In the signal Un (t), which represents the maximum amplitudes of the post-oscillation phases, the above-mentioned interference can be seen, which manifests itself in the signal Un \ (t) as an oscillation that occurs at the beginning of the post-pressure phase (approx tn, sar) begins. As mentioned, this oscillation is a result of the overlaying reflection on the cavity wall 9 ‘with the reflection on the growing edge layer P‘ of the molded part (see FIG. 2B). In this phase, the times tmax, 1, tmaxz of the maxima (additionally or alternatively also the minima) of the oscillation can be detected in the signal Uy (t) (e.g. by means of digital signal processing).

The times tmax, 1, tmaxz or the period of the oscillation Atyax can be monitored (and possibly stored and documented) during the injection molding process in order to monitor the constancy of the growth of the surface layer over several production cycles. The period of the oscillation Atmax can serve as a measured value for the growth rate of the surface layer. The actual growth speed can be calculated approximately from this measured value, taking into account the speed of sound in the plastic and the frequency of the ultrasound.

The cooling of the molded part results in a volume contraction, which in turn causes the molded part to shrink and detach itself from the inner wall of the cavity (“shrinking”). The point in time at which the molded part is detached from the inner wall of the mold can be clearly seen in both signals UpH (t) and Ur (t). Due to the resulting gap between the cavity wall 9 ° (see FIG. 3) and the molded part, the reflection factor rises again to approximately 100% and the signals rise again to their initial values.

The exemplary signal curves of the standardized output signals Um (t) and Un (t) make it clear that the concepts described here can be used both for monitoring and for controlling an injection molding process. In the examples described here, an ultrasonic transducer is used that works as both a transmitter and a receiver. Instead, two separate transducers can be used as transmitter and receiver. In this case, the send / receive switch is not required. Finally, it should be noted that the concepts described here are not only used for monitoring and controlling a (plastic) injection molding process (injection molding), but analogously also for monitoring and controlling a (metal) die casting process (die casting) or other casting processes can be.

Claims (10)

Claims 1. A method for monitoring a casting process, characterized by the following steps: coupling a sequence of ultrasonic pulses into a casting tool (7); Receiving a sequence of reflected ultrasonic pulses and generating a received signal (u (t)) mit having a corresponding sequence of pulses each having a main pulse and a ringing; Generating - for each pulse of the received signal (u (t)) - a first measured value (Uw (t)) which represents the amplitude of the main pulse; and generating - for each pulse of the received signal (u (t)) - a second measured value (UNn (t)) which represents the amplitude of the ringing, based on the change in the first and / or second measured values (UyH (t), Un (t)) one or more process parameters can be determined during a casting process. 2, The method according to claim 1, characterized in that the first measured values (Up (t)) and the second measured values (Un (t)) are normalized in such a way that they assume a predefined value when the casting tool (7) is empty a reflection coefficient of 100 percent on the inner wall of the empty casting tool can be assigned, with a later change in the standardized measured values indicating a corresponding change in the reflection coefficient. 3. The method according to claim 1 or 2, characterized in that at least one of the following is determined as the process parameter: the time (t--) of the detection of the flow front of the melt injected into the casting tool (7), the time (tu) of the complete Filling a cavity of the casting tool (7) with melt, point in time and duration of a holding pressure phase, growth rate of the surface layer of the melt in the casting tool, point in time when the molded part disappears from the wall of the cavity. 4. The method according to claim 1 or 2, characterized in that based on the change in the second measured values during a casting process in which the melt is introduced into a cavity of the tool (7), undesired deviations in the surface layer growth in the melt from a target profile are detected while the melt cools and solidifies. 5. The method according to any one of claims 1 to 4, characterized in that based on the first measured value (UH (t)) and its change, density fluctuations of the melt injected into the casting tool (7) are detected, in particular to add to inhomogeneities in the melt detect. 6. A measuring system for monitoring a casting process, comprising: an ultrasonic transducer (8) which is designed to couple ultrasonic pulses into a casting tool (7); a sensor circuit (22) coupled to the ultrasound transducer (8), which is designed to control the ultrasound transducer (8) with a control signal that has a sequence of pulses and one of the ultrasound to generate the ultrasound pulses -Transducer (8) or a further ultrasonic transducer provided received signal (u (t)), which represents the ultrasonic pulses reflected in the casting tool (7), to process, the received signal (u (t)) having a sequence corresponding to the control signal of pulses each having a main pulse and a post-oscillation, characterized in that the sensor circuit (22) for processing the received signal (u (t)) is designed - for each pulse of the received signal (u (t)) - one to generate a first measured value (U} (t)), which represents the amplitude of the main pulse, and a second measured value (Uyn (t)), which represents the amplitude of the ringing. 7. The measuring system according to claim 6, characterized in that the first measured value (Uy (t)) is based on the maximum amplitude of the main pulse and the second measured value (Un (t)) is based on the maximum amplitude of the ringing. 8. The measuring system according to claim 6, characterized in that the first measured value (U + H (t)) in each case on the area under the envelope curve of the main pulse and the second measured value (Un (t)) in each case on the area under the envelope curve of the Post-swing based. 9. The measuring system according to one of claims 8, characterized in that the sensor circuit (22) is designed to generate the first measured value (Ux (t)), the envelope of the respective pulse of the received signal (u (t)) in one to integrate the first time window (tn) and wherein, for generating the second measured value (Uyn (t)), the sensor circuit (22) is designed to generate the envelope of the respective pulse of the received signal (u (t)) in a second time window (tn) to integrate. 10. The measuring system according to claim 9, characterized in that the first time window (t4) begins a defined first delay time after the corresponding pulse in the control signal, and wherein the second time window (tn) begins a defined second delay time after the corresponding pulse in the control signal. In addition 3 sheets of drawings