DE19834797C2 - Method and device for condition-dependent process control in the processing of plastics - Google Patents

Abstract

The aim of the invention is to increase production reliability and quality in the processing of thermoplastic or thermosetting plastics by pressing or injection molding. DOLLAR A The task is solved by continuously monitoring the processing process with known measuring principles, such as the recording of dielectric or acoustic parameters with the aid of sensors attached to the tool and a computer program with expert knowledge anchored therein, and intervening in the processing machine in a controlling manner. On the one hand, each individual part can be monitored in production, in the event of deviations from target values, the part is immediately classified as a scrap and the machine can be stopped. On the other hand, minor deviations can be interfered with in the control, e.g. B. the tool can be opened when defined material properties are reached. DOLLAR A The area of application concerns the entire range of plastics processing by means of pressing and injection molding.

Description

Thermoplastic and thermosetting plastics are usually processed on injection molding machines. Pressing is also used for thermosets. Thermoplastics are processed by molding in the molten state. For this purpose, they are injected as a melt into the mold, the temperature of which is below the solidification temperature. With thermosets, cold or preheated (presses) or melted materials (injection molding) are brought into the mold. The temperature of the tool is so high that the chemical crosslinking reaction starts and the material hardens. The manufacturers of plastic parts are facing increasing quality requirements from their customers. If the consistent introduction of ISO 9000 ff certification has already led to stricter framework conditions with regard to quality management, the requirements of the QS 9000 quality standard initiated by the American auto industry, which requires the individual proof of the quality of each part delivered, heralds a further tightening of the quality requirements , The product liability law and the enormous costs associated with the loss of image of recall campaigns therefore force zero-defect production.

To what extent is zero defect production in the manufacture of plastic parts according to the state of technology already implemented?

Plastic processing machines (e.g. injection molding machines, presses, etc.) have a high level of automation today. As a rule, computers are integrated into the machines, which carry out the entire process control and process monitoring. Characteristic process parameters such as pressure (injection pressure, mold cavity pressure, etc.), temperature (temperature in the mold, temperature in the plasticizing unit, etc.) and paths (e.g. mold closing path, plunge edge movement, etc.) are constantly checked and regulated kept constant. As an option, the parameters can also be recorded statistically and saved for documentation. Examples are known that a computer is integrated into every injection molding machine in production and these are all networked with one another with the central production server / R. Doh, "On the pulse of the machine", Netrunner issue 4 ( 1995 ) /.

Once you have the connection between Process parameters and quality manufactured, it is sufficient from now on to the process parameters control and keep constant. Detected deviations in the process parameters can then react adequately and trigger an alarm, e.g. B. if the Tool heating, unusual pressure curve, jammed parts etc.

In practice, however, this is not enough for quality assurance. In any case, there must be a at least random quality control of the produced part is connected, since compliance with the process parameters alone does not offer a sufficient guarantee of quality. The reason for this are inevitable deviations in the source material, which one can have a serious impact on quality. Plastics are complex materials, which are usually manufactured discontinuously. So deviations from Avoid batch to batch. Here is z. B. on the different Molar mass distribution from batch to batch referenced. Coming with thermosets Add inaccuracies in the ratio of resin and hardener. The plastics also react partly sensitive to storage conditions. So the absorption of air humidity has one  sometimes serious influence on processing behavior. Works with thermoplastics Water as a plasticizer and changes the softening temperature and melt viscosity. Water is a component of thermosets in the polycondensation molding compounds the reaction chemistry. Higher water shares accelerate the reaction and lead to Over-curing. In addition, many thermosets are sensitive to storage because of the chemical reaction also runs at low temperatures, which makes it a superimposed layer Pre-networking is coming.

Especially in manufacturing, there are often great uncertainties as to how long this will take Material may no longer be used. This question arises in small processing companies because the delivery containers often contain quantities that exceed current requirements. To all To be able to take these influences into account would require a comprehensive one Incoming goods inspection and a complicated, based on many years of experience Experience to adjust the machine to the fluctuations in the Adapt the source material. The effort for this is often considered to be unacceptably high, or as viewed practically not enforceable.

If there are also material fluctuations within a container, you currently have no chance to stabilize the production and manufacture in accordance with quality. Stay in here usually only the rejection of the entire delivery, with all problems, up to Proof to the supplier that such fluctuations really exist.

The patent literature contains a variety of suggestions for solving the described Problems. For example, in DE 44 34 654 a method for influencing is carried out cyclically Processes described in particular for plastics processing machines. By test Manufacture of products and specification of characteristic properties Process key figures determined and machine settings derived from them. In the DE 195 14 535 z. B. a causal relationship between cavity pressure and Realized product properties and derived forecast data. EP 0 566 738 has one Method for determining the injection conditions to the content, in which Countermeasures when errors occur in an expert system are included and the machine setting can be modified automatically. A method for Finding conditions for injection molding using an expert system is also described in US 5,350,547. Another knowledge-based control method an injection molding machine is patented in US 5,539,650. Here is the from Operators programmed the processes from an expert system for plausibility checked, stored machine settings can be accessed.

The solution according to the invention addresses the problem differently than it does Meaningful or critical points in the tool and possibly also in the plasticizing unit during injection molding, with state-specific material parameters dielectric and / or ultrasonic sensors continuously recorded and with the help of a Expert systems can be evaluated by a computer. The one from such sensor data derivable material parameters are e.g. B. the memory module, the loss module and the Viscosity. Known sensors are suitable as sensors for these material parameters Electrode arrangements for measuring conductivity, dielectric constant and dielectric loss or sound transmission and reception sensors, especially in Ultrasonic range, for measuring the speed and attenuation of sound. The physical relationships between the named parameters and the sought Material properties are well known from the literature.  

The use of dielectric measurement methods is known e.g. B. described in DE 38 22 716 and in the following literature:
SD Senturia, NF Sheppard jr. "Dielectric analysis of thermoset cure", Advances in polymer science Vol. 80, Springer Verlag 1985 , pp. 1-49
W. Michaeli, M. Stöger, "Process Monitoring in Thermoset Processing Using Dielectrometry", Kunststoffberater, No. 1, 1993, pp. 38-43
W. Michaeli, G. Burkhardt, M. Stöger: "Quality Assurance in Thermoset Processing", Plastics No. 11, Vol. 81, pp. 1014-1017, 1991

For the ultrasonic method, the patent DE 37 29 031 and the following quotations are given:
KG Häusler; P. Hauptmann; E. Schröder; R. Neat; P. Wagnitz, "For Network Characterization Using Ultrasound", Plastics and Rubber No. 9, Vol. 30, pp. 485-488, 1983
I. Alig; M. Fedtke; KG Häusler; W. dancers; S. Wartewig, "Modified Epoxies as Studied by Ultrasonic Methods", Colloid and Polymer Science, Vol. 81, pp. 54-58, 1988

In the method according to the invention, the material characteristic values in the process of one Computer system constantly queried, the values processed in characteristic curve fields, on the monitor output online and with the programmed one or in dialog with the operator predetermined target characteristics compared. In the event of deviations, the measuring computer programmed reactions triggered. These can be discarded as part of the committee, the Stop production with alarm triggering up to targeted changes in Lead process parameters.

A particularly favorable solution is the opening, which is controlled by the material condition reached the tool, d. H. the realization of an intelligent machine that always has the highest parts Quality produces. In the case of thermoplastics, this would mean if, due to the cooling of the Plastic well-defined values of the memory and / or loss module and / or the Viscosity are reached to demold. With thermosets, this would correspond to the achievement of one Preselected degree of networking, which in turn is reflected in the specified parameters reflects.

It is also advantageous that manufacturing-related errors such as insufficient cavity filling, insufficient compression and air pockets can be recognized immediately.

The method and the device are to be illustrated using a few examples:

Embodiment 1

One device is shown in Figure 1. The part of the tool 1 in which the cavities 2 are located (ejector side) is shown. Two cavities are shown by way of example, in which measuring sensors 3 for dielectric or ultrasound are embedded flush. A particular advantage of ultrasonic sensors is that the sensors do not necessarily have to be in contact with the molding compound, but can also "see through" the wall of the tool. The sensors are deliberately arranged at the end of the cavity because here, in addition to the actual material parameters, spraying errors such as insufficient compression, air entrapment and underfilling can be detected. The sensors are connected to a measuring interface card 4 in a computer 5 via measuring cables 6 . The computer is connected to the processing machine via a control line 7 and an interface connector 8 . A separate measuring computer is shown in the picture. This can of course also be integrated into the processing machine, or the existing machine computer itself can be used.

Embodiment 2a

The method in cooperation with the device should now be based on the example of thermoplastic Injection molding for a 4-cavity mold are described in more detail.

Four measuring sections with paired commercially available ultrasonic sensors with a resonance frequency of 4 MHz are integrated in the tool in a sonic arrangement. For this purpose, identical sensors are arranged flush in the wall on the nozzle side and the ejector side. The measuring sections are at the extreme end of the respective cavity. There is also a sensor section near the nozzle on the injection channel. All ten sensors of the five measuring sections are connected to the connector strip of the ultrasonic measuring card 4 in Figure 1 in the computer 5 via coaxial lines 6 .

The ultrasound measurement card 4 realizes that in multiplex operation, every 5 measurement sections receive a transmit pulse at 1/20 second intervals and the received pulse is recorded with regard to transit time and signal amplitude. The measurement is triggered by triggering (e.g. from the machine) simultaneously with the Close tool command. In real-time operation, the speed of the sound and the damping are then calculated from the sensor signals, and the storage and loss modulus in each measurement section are calculated based on the known relationships. The change in the speed of sound and damping in the process is output on the monitor. After the injection molding process, the values are stored fully automatically in the computer and transmitted to a server via a data network (LAN). The curves are also evaluated online. Based on the programmed expert system, predefined tolerances can be monitored and an alarm can be triggered if they are exceeded. This allows the defective part to be discarded and production to be stopped. With the help of trend analyzes, a slow drift away from the stable area can be recognized and intervened in a regular manner. In particular, a condition-based control is possible, ie when a predetermined value of the speed of sound and / or damping is reached, which in turn is a measure of the cooling state, the machine can receive the signal for opening the tool. Adequate reactions are therefore reacted to fluctuations in the material, the injection temperature and the mold temperature. As a result, parts are manufactured with consistently high quality.

The following material parameters can be derived from the 5 measuring sections by measuring the material properties of sound speed and damping:

• - Measuring section on the injection channel: molding compound temperature, viscosity - conclusions on the Temperature in the plasticizing part of the processing machine, heating by friction on the Nozzle, fluctuations in the composition of the molding compound and the change in state of the Plastic.
• - Measuring section at the end of each cavity: flow time from the injection channel into the cavity, from it again viscosity, filling time of the cavity, inclusion of air, incomplete filling of the Cavity, insufficient compaction, degree of solidification of the material, determination of the Time of opening of the tool (due to the state of the plastic reached). The comparison of the 4 cavities also shows inhomogeneities, how different Flow resistance, uneven temperature distribution.

The measured values of all 4 cavities are compared in the computer with learned setpoints. When the most critical cavity for the quality has reached the set of parameters specified for sufficient dimensional stability, the machine receives the control line 7 and the command to open the tool according to the condition.

In this way, the machine can automatically manufacture in a stable area. It is special important for start-up processes, where it often takes a long time for the dynamic to change  Temperature balance in the tool. Are outside the setpoints Measured values for critical quantities, such as material temperature and viscosity in the injection channel, Flow time, filling level, cooling time, uniformity of the 4 cavities detected, an alarm is triggered immediately triggered and the process stopped, the part discarded as scrap. The drift of relevant parameters, such as B. the melt temperature is tracked and derived from it Control commands (e.g. for a moderate change in tool temperature) are sent to the machine transmitted via the control line.

Embodiment 3 thermoset

In thermoset processing, besides the filling of the tool, the achievement of one is defined curing condition is crucial for the quality of the molded part. Both under and overcuring must also be avoided. In particular, the variation of Material thickness in the molded part is often a major problem, since thick material can be used for both Undercuring due to the low heat conduction and the later start of the Reaction as well as over-hardening up to thermal destruction due to the inner Overheating can come from the strongly exothermic reaction.

In a thermoset tool for a molded part with uneven thickness distribution (Switch housing) become several at significant points of different thickness dielectric sensors are inserted flush into the surface so that they are in contact with the Stand molding compound. The control of the sensors with an AC voltage is in the Multiplex operation made by a dielectric measurement card in the computer, the Sensor current measured separately on the card according to real and imaginary current. From that become including the material thickness, the dielectric constant and the dielectric loss calculated. The characteristic values are output on the monitor, which is how the display is structured is that the tool is shown stylized and the measured values at the location of the sensor alphanumeric or as a graphic symbol, e.g. B. Bar, pointer display including display of Tolerance limits are output.

The programmed expert system can differ at different points Evaluate the thickness at the most critical point when the desired hardening condition is reached (usually with the greatest thickness) open the tool. If the deviations are too large The state of hardening at points of different thickness can regulate according to the expert knowledge be intervened by z. B. the hardening time is adjusted accordingly.

Claims (7)

1. A method for the quality-controlled manufacture and processing of plastics using acoustic and / or dielectric sensors, associated measuring systems for dielectric and acoustic measurements and computing technology, characterized in that quality-determining material parameters are calculated from the measured acoustic and / or dielectric parameters and their Changes in the processing process are output online on the monitor, stored in the computer, evaluated with the help of an expert system with regard to the quality of the molded part and used for machine control. 2. Claim according to claim 1, characterized in that when optimal properties of the plastic are reached, a control command to the Processing machine is triggered, which causes the process to be terminated quickly Eject the part leads. 3. Claim according to claim 1, characterized in that continuously the measured material parameters with learned and saved target curves be compared and triggered in the event of deviations and the part as a committee is classified. 4. Claim according to claim 3, characterized in that carried out a fault analysis based on programmed expert knowledge and the Operator possible causes and remedial measures to eliminate the error be proposed. 5. Claim according to claim 3, characterized in that based on programmed expert knowledge, an error analysis is carried out and the Processing machine automatically controlled in the stable manufacturing area is returned. 6. Claim according to claim 5, characterized in that the drift of the material parameters continuously recorded from cycle to cycle, from there Trend analyzes are determined and early intervention is carried out to ensure quality to keep constant. 7. Claim according to claim 1, characterized in that processing errors based on their typical Measured values are determined and the parts are classified as rejects and information is given to the operator Causes of errors are given.