DE102009001682B4 - Method for controlling an injection molding process - Google Patents

Abstract

A method of controlling an injection molding process, comprising the steps of: (a) taking images of an injection molded part taken from an injection mold during cooling and detecting the temperature while taking an image, (b) determining time and temperature dependent shrinkage curves from the steps in (step (c) extrapolating the dimensions of the injection molded part in the final state from the time and temperature-dependent shrinkage curves, (d) adjusting the injection molding parameters so that the dimensions of the injection molded parts produced by the injection molding process are within a predetermined tolerance range after cooling.

Description

State of the art

The invention relates to a method for controlling an injection molding process.

Injection molding processes produce molded parts from plastics. For this purpose, the plastic is first melted and then injected under pressure into a tool, which is a negative image of the molded part to be produced. In the mold, the molding is cooled down until the plastic has solidified. Then the molding is removed from the injection mold.

Due to the high coefficient of thermal expansion of plastics dimensional accuracy of plastic parts, which are produced by injection molding, difficult to handle. For example, upon cooling of the molded article produced after processing, severe shrinkage occurs. This is to be perceived more or less clearly depending on the geometry of the injection-molded part. If moldings are produced by injection molding of semi-crystalline plastics, so occurs during the cooling in addition to a phase transformation, the crystallization, which is also associated with a volume shrinkage. This makes the prediction of component dimensions even more difficult. The component dimensions are also influenced by parameters of the plastics processing, for example the amount of the reprint or the injection speed. These parameters continue to influence the component distortion. This is due to the fact that the internal structure of the plastics, the morphology, depends on the processing conditions.

A rudimentary prediction of the distortion and dimensional accuracy of injection molded components is possible, but only qualitative results are provided which can not be used to pre-correct the injection mold or to adjust the injection molding parameters.

A process control to optimize the dimensional accuracy is possible, but requires a complex training in which parts that are manufactured with different injection molding parameters are measured. By later evaluation of the measurement results, the injection molding parameters and the waveforms of the tool sensors forecasts for the influence of processing on the dimensional stability and the distortion of the components can be created. Based on these forecasts, component dimensions are predicted. However, a check is only possible by means of a re-measurement of the components.

A method in which mechanical properties and processing properties of an injection molded part are determined is, for. B. off DE 101 19 853 A1 known. To predict the properties with regard to the production of a plastic molded part is a hybrid model in which one or more neural networks and rigorous models are interconnected. The rigorous models are used to represent partial models that can be described using mathematical formulas, while the neural submodel maps processes whose connection is only in the form of data and can not be rigorously modeled. This makes it possible to make a prognosis with regard to the mechanical, thermal and rheological processing properties and with regard to the process time of a plastic molding. With the described method, however, it is not possible to make any statements about the dimensional accuracy of components produced by injection molding.

A method for evaluating the optical properties of a component is out US 2008/0 013 821 A1 known. The component may, for example, also be a plastic part produced by injection molding. To assess the optical properties of a recording of a component is created and from the recording properties are determined. The properties are stored in a vector. The optical properties are corrected with the process parameters, so that from given process parameters it is possible to conclude optical properties of a component produced with the corresponding process. Statements about the shrinkage of a component are not made.

The WO 2005/084913 A2 discloses a method for characterizing melt properties of thermoplastic powders by means of a digital camera. In this process, successive images of a rotational cast body are recorded during the process and various temperatures recorded during the recording in order to set the production parameters depending on them.

The WO 2008/009147 A1 discloses a method in which are taken in an injection molding process by means of a thermal imaging camera temperature-dependent images of the injection molded body at certain times to the basis of the surface temperature of the injection molded body at the time of leaving a manufacturing or processing device, or even within a certain period of time after leaving the device, the Adapt manufacturing or processing method. In this case, depending on these time-dependent temperatures, the parameters of the injection molding process are chosen such that the demolding takes place at the time at which the temperature of the Injection molded part is sufficiently below the melting temperature of the material of which the body consists.

The DE 10261498 A1 discloses a method for controlling the manufacture of injection molded parts in an injection molding tool of an injection molding machine, wherein the temperature of the tool is controlled.

The US 2011/0101555 A1 discloses a method in which, depending on the casting temperature and the melting temperature of the injection molded part, the shrinkage behavior is analyzed and the injection molding process is controlled.

The DE 19518804 A1 discloses a method for monitoring an injection molding process.

Object of the present invention is to produce dimensionally accurate injection molded components.

This object is solved by the features of claim 1.

Disclosure of the invention

• (a) Aufnahme von Bildern eines aus einem Spritzgießwerkzeug entnommenen Spritzgussteils während der Abkühlung und Erfassen der Temperatur während der Aufnahme eines Bildes,
• (b) Bestimmen von zeit- und temperaturabhängigen Schwindungskurven aus den in Schritt (a) aufgenommenen Bildern,
• (c) Extrapolieren der Maße des Spritzgussteils im Endzustand aus den zeit- und temperaturabhängigen Schwindungskurven,
• (d) Anpassen der Spritzgießparameter, damit die Maße der durch den Spritzgießprozess hergestellten Spritzgussteile nach Abkühlung in einem vorgegebenen Toleranzbereich liegen.

An inventive method for controlling an injection molding process comprises the following steps:

• (a) taking pictures of an injection molded part taken out of an injection mold during cooling and detecting the temperature during the taking of an image,
• (b) determining time and temperature dependent shrinkage curves from the images taken in step (a),
• (c) extrapolating the dimensions of the final molded part from the time and temperature dependent shrinkage curves,
• (d) adjusting the injection molding parameters so that the dimensions of the injection molded parts produced by the injection molding process are within a predetermined tolerance range after cooling.

By taking pictures during the cooling of an injection-molded part, the shrinkage of the injection-molded part can be determined by comparing the individual images. The data collected in this way can be used to adjust the injection molding parameters so that dimensionally stable components can be manufactured. Since the dimensions of the injection molded part produced also depend on the injection molding parameters, a suitable adaptation of the injection molding parameters can influence the shape of the injection molded part after cooling, so that it is possible to produce dimensionally stable components by means of suitable injection molding parameters.

Because the dimensions of the injection-molded part in the final state can be extrapolated from the images and the time-dependent and temperature-dependent shrinkage curves determined therefrom, it is also possible to check the dimensions of an injection molded part produced and the delay of the injection molded part in the production process, so that control parameters and control strategies of the injection molding process during operation can be adjusted.

To pick up the images of the injection molded part taken from the injection mold in step (a), parts are randomly discharged with a suitable handling system. These parts are then taken with a camera measuring system images. This allows an exact evaluation of the component dimension during operation.

In order to determine the distortion and the shrinkage of the component, the recording of the images in step (a) takes place continuously in one embodiment of the invention. Alternatively, it is also possible for the images to be recorded in step (a) at predetermined intervals. If the pictures are taken at predetermined intervals, the intervals between taking two pictures may be the same length, but it is also possible to choose different time intervals depending on the cooling rate. For example, at the beginning it is possible to take pictures at a smaller time interval and to increase the distance of the pictures with increasing cooling. Compared to the continuous recording of images, taking pictures at predetermined intervals has the advantage of requiring less data storage space. The disadvantage compared to a continuous recording, however, is that the course must be interpolated between two images.

By taking pictures of the component, measurement data of one or more significant dimensions of the component can be determined. In this case, on the one hand, it is possible for only one image of the component to be recorded in each case; alternatively, it is also possible to record a plurality of images from respectively different directions onto the component. The recording of several images from different directions allows a three-dimensional detection of the component whereas when taking pictures with only one camera from the same direction only a two-dimensional recording of the component is possible. However, if the significant dimensions of the component are all in the same direction, such a two-dimensional image of the component is sufficient.

In addition to the images of the component, the temperature of the component is simultaneously recorded. The detection of the temperature can also be carried out continuously or at predetermined intervals. If the temperature is recorded at predetermined time intervals, the temperature is measured simultaneously with the acquisition of an image. The detection of the temperature is preferably carried out without contact, in particular by means of an infrared radiation pyrometer. In addition to the temperature detection by means of an infrared radiation pyrometer, however, it is alternatively also possible to record the component with an infrared camera, for example, so that the recorded image delivers both measured values for the component dimensions and for the component temperature.

From the combination of the measurement data of the at least one significant dimension of the component and the temperature that the component has at the time of recording, the time- and temperature-dependent shrinkage of the component dimensions can be determined. Using suitable algorithms, the final state of the component dimensions can be extrapolated from these shrinkage curves. The suitable algorithms result, for example, from the course of the measured shrinkage curves.

The advantage of the extrapolation of the component dimensions in the final state is that adaptation of the injection molding parameters during operation is possible. The determination of the dimensions in the final state is generally not possible during operation, since the change in shape of the component is not completed even after days. However, the extrapolation of the final dimensions allows an estimation of the component dimensions already after a comparatively short time, so that not only after several days an adaptation of the injection molding parameters takes place, but already after completion of the measurements and extrapolation.

The adaptation of the injection molding parameters as a function of the extrapolated component dimensions in order to ensure compliance with tolerance limits of the component dimension is achieved by means of a suitable injection molding control which is known to the person skilled in the art. By means of such an injection molding control, for example, injection pressures, speeds and melting temperatures are set.

For this purpose, the injection molding control can be trained on the one hand with extreme parameter sets in boundary tests or, alternatively, on suitable algorithms, in order to record the dependence of the dimensions of a specific component on the processing or the injection molding parameters.

In order to avoid that an outlier for the determination of the injection molding parameters is measured, it is preferable to take pictures of several injection-molded parts and in each case also to record the temperature during the recording of the image. The injection-molded parts are preferably produced in each case with the same injection molding parameters. In this way, the shrinkage process can be followed for several parts, so that the data can be averaged and from this averaged data the injection molding parameters can be adjusted.

In order to achieve comparable values, it is furthermore preferred, when taking pictures of a plurality of identical injection-molded parts, to record the images in each case at essentially the same temperature of the injection-molded part. If the images of the individual injection-molded parts are recorded at different temperatures, it is necessary to use suitable algorithms to calculate measured curves from the determined data sets by suitable algorithms, for example by interpolating between two measured values, and to compare the curves thus obtained and suitable mean values to build.

In order to enable a process control that allows dimensionally stable components to be produced, it is advantageous if, for the adaptation of the injection molding parameters, respectively adjusting dimensions of the injection-molded part are detected at given parameter sets. For this purpose, it is possible, for example, to set parameters with which the component is injected, to produce some components with these parameters and then to acquire measured data for the components produced in this way. The acquisition of the measured data takes place, for example, by evaluating images which are taken by the cooling component. In this way, the shrinkage or a delay can be determined. On the basis of the data thus determined, a parameter set can be created which can be used to control the injection molding process. From this parameter set, it is then possible to determine, for example, how to influence certain measured deviations in order to achieve an improvement of the injection-molded part to be produced.

In order to produce suitable injection molding parameter sets with a manageable expenditure, it is advantageous to determine necessary parameter sets from known parameter sets in order to adapt the injection molding parameters. In this way it can be determined for each parameter set, for example, how a component produced with it will probably look like. The determination of the adaptation of the injection molding parameters necessary parameter sets from known parameter sets can be done for example by suitable algorithms or functions or a suitable control system. Suitable algorithms, functions and / or rule sets are known to the person skilled in the art. In particular, neural networks are suitable for determining the parameter sets necessary for adaptation of the injection molding parameters from known parameter sets. A check can then be made, for example, by random checks. These random checks can also be used to improve control by feeding the measured data as additional training sets to the injection molding control algorithm.

From the parameter sets thus trained, it is possible to determine, for example, what influence the change of a single injection molding parameter, for example the change of the pressure, the temperature, the injection speed or the like, has on the component. From this it is then possible to react to dimensional deviations of the component and the process can be adapted accordingly.

Brief description of the drawings

Embodiments of the invention are illustrated in the figures and are explained in more detail in the following description.

Show it:

1 a schematic representation of the principle of distortion correction with the aid of image processing,

2 an exemplary temperature-dependent shrinkage curve for a component dimension,

3 Cooling curves depending on the processing parameters.

Embodiments of the invention

In 1 is schematically illustrated by a flow chart, the principle of distortion correction using the image processing.

In a first step 1 In an injection molding machine, an injection molded part is produced with a predetermined parameter set. The predetermined parameter set includes, for example, injection speed, closing pressure of the tool, injection pressure, temperature of the melt, pressure in the tool, etc. Further suitable parameters are familiar to those skilled in the art.

Samples are taken from the injection-molded parts produced in this way. The samples will be like the arrow 3 represented a second step 5 fed. At the second step 5 the samples are measured. For this purpose, images of the injection-molded part are recorded continuously or at predetermined intervals. The dimensions of the component are determined from the pictures. By recording at intervals or the continuous recording, the change in the component dimensions can be determined during the cooling of the injection molded part. The change in the component dimensions results, for example, from the shrinkage due to the cooling and, for semicrystalline polymers, for example, also from phase transformation, ie crystallization. In addition to taking pictures, the temperature of the component is also detected at the time of taking a picture at the same time. The temperature detection is preferably carried out without contact in order not to influence by applied sensors measurement results during the dimensional change of the component. For temperature measurement, preferably infrared radiation pyrometers are used. Alternatively, however, it is also possible, for example, to detect both the temperature and the component dimensions by taking a picture by using a suitable infrared camera. The advantage of temperature detection with an infrared camera is, in particular, that a temperature distribution and different temperatures on the component can be measured.

From the data so collected, d. H. Time-dependent and temperature-dependent shrinkage curves are created for the measured temperature values and the dimensional changes. The curves are generated by suitable mathematical methods, which are familiar to the person skilled in the art. Commonly used methods are, for example, extrapolation methods.

The thus determined shrinkage curves are, as with arrow 7 shown in a third step 9 fed to a process control.

By means of the process control, for example, final dimensions of the component are extrapolated from the shrinkage curves. From the thus extrapolated dimensions of the final state of the component can then be adjusted if necessary, the injection molding parameters. It is possible to either change individual parameters or to change several parameters. Which parameters can be changed to what extent can be determined, for example, by suitable parameter sets, which are calculated, for example, using neural networks. With the so on the process control in the third step 9 determined values, the injection molding process in the first step 1 modified. This is with an arrow 11 shown. After changing the injection molding parameters, injection molded parts are sampled and measured again to check the influence of the modified injection molding parameters on the components. By using a suitable process control, which has sufficient parameter sets, from which the influence of the change of individual parameters can be taken on the components, the process can be easily adapted without producing a large number of rejects due to incorrect process parameters getting produced. In addition, the injection molding process can be intervened quickly, since the dimensions are extrapolated from the shrinkage curves and so not the entire cooling process has to be awaited until the final dimensions of the component are reached, which in some cases can only be achieved after several days.

In 2 exemplified a change in a component dimension as a function of the temperature. Here is on the x-axis 21 the temperature T and on the y-axis 23 the component dimension I applied. As in the curve 25 can be seen, the component dimension I changes with the temperature T, wherein the component dimension I with decreasing temperature T also decreases. As can be seen from the diagram, the component expansion increases more with increasing temperature. This means that when cooled at high temperatures, a greater decrease in component dimensions can be observed than at a lower temperature.

In 3 Cooling curves and maximum temperature are shown depending on the processing parameters and the time. On the x-axis 31 here is the time t and on the y-axis 33 the temperature T applied. As from the individual cooling curves 35a . 35b . 35c As can be seen, the temperature T decreases with increasing time t, the temperature running asymptotically towards a lower limit. It can also be seen that the temperature initially decreases rapidly and then always slower. The cooling curves shown here each have a different maximum temperature. The first cooling curve 35a starts with the lowest maximum temperature, the second cooling curve 35b already has a higher than the first maximum temperature maximum temperature and the third cooling curve 35c starts with the highest maximum temperature. The maximum temperature in each case indicates the beginning of the measurements of the injection-molded part and is thus below the injection temperature of the plastic in the injection mold. In general, the maximum temperature corresponds to the removal temperature of the injection molded part from the injection mold. Since the largest change in the injection-molded part takes place immediately after demoulding, the removal of the injection-molded part from the injection-molding tool should take place as quickly as possible.

In addition to the melt temperature, however, the maximum temperature may also depend on other parameters, for example the injection speed or the tool temperature.

The cooling curves 35a . 35b . 35c are not only dependent on the injection molding parameters, but also, for example, on the plastic used and the component dimensions. For this reason, different cooling curves are obtained for each polymer and each component. For example, the cooling curve is dependent on the ambient temperature and the temperature of the cooling medium in which the component is cooled.

Also in 2 shown temperature-dependent change in the component dimensions is in addition to the injection molding parameters depending on the polymer used and the component to be produced. For this reason, it is necessary here for each plastic as well as for different components to determine new curves.

From the into the 2 and 3 Then, parameter sets can be calculated on the basis of which the injection molding parameters can then be set with a suitable process control.

Claims (10)

Method for controlling an injection molding process, comprising the following steps: (a) taking pictures of an injection molded part taken out of an injection mold during cooling and detecting the temperature during the taking of an image, (b) determining time and temperature dependent shrinkage curves from the images taken in step (a), (c) extrapolating the dimensions of the final molded part from the time and temperature dependent shrinkage curves, (d) adjusting the injection molding parameters so that the dimensions of the injection molded parts produced by the injection molding process are within a predetermined tolerance range after cooling. A method according to claim 1, characterized in that the recording of the images in step (a) takes place continuously as a film. A method according to claim 1, characterized in that the recording of the images in step (a) takes place at predetermined intervals. Method according to one of claims 1 to 3, characterized in that images are taken of several identical injection molded parts. Method according to one of claims 1 to 4, characterized in that when taking pictures of several identical injection molded parts, the images are taken in each case at the same temperature of the injection molded part. Method according to one of claims 1 to 5, characterized in that the temperature of the injection-molded part is detected without contact. A method according to claim 6, characterized in that the non-contact detection of the temperature of the injection-molded part takes place with a radiation pyrometer. Method according to one of claims 1 to 7, characterized in that in each case adjusting dimensions of the injection molded part are detected at predetermined parameter sets for the adjustment of the injection molding parameters. Method according to one of claims 1 to 8, characterized in that for adjusting the Spritzgießparameter necessary parameter sets are determined from known parameter sets. Method according to claim 9, characterized in that the parameter sets necessary for adaptation of the injection molding parameters are determined from known parameter sets with the aid of neural networks.

Images