WO2021058529A1

WO2021058529A1 - Automated manufacturing process, and manufacturing system for bending glass panes using an integrated digital image

Info

Publication number: WO2021058529A1
Application number: PCT/EP2020/076512
Authority: WO
Inventors: Muhammed ALTIN; Achim ZEICHNER; Arthur PALMANTIER; Thomas Bobek
Original assignee: Saint-Gobain Glass France
Priority date: 2019-09-27
Filing date: 2020-09-23
Publication date: 2021-04-01
Also published as: CN112955835A

Abstract

The invention relates to an automated manufacturing process for bending panes, wherein a pane can be processed using movable system parts, and the movable system parts can be controlled by a stored-programmable controller on the basis of parameter values which can be entered manually. The stored-programmable controller can output control signals to actuators of the movable system parts and receive sensor signals from sensors in order to detect actual states of the actuators. Manually entered parameter values for controlling the movable system parts are transmitted both to the stored-programmable controller in order to control the actuators as well as to a digital image of the automated manufacturing process. The digital image ascertains simulated target sensor data on the basis of the parameter values, and the digital image receives actual sensor data and ascertains a deviation of the actual sensor data from the target sensor data for each sensor. In the event that the deviation of the actual sensor data from the target sensor data for a sensor exceeds a specifiable value, information is output on a monitor and/or the manufacturing process is stopped.

Description

Automated manufacturing process and manufacturing system for bending glass panes with an integrated digital image

The invention lies in the technical field of manufacturing glass panes and relates to an automated manufacturing process for bending glass panes with an integrated digital image of the manufacturing process. The invention also extends to an automated manufacturing system for performing the automated manufacturing process according to the invention for bending glass panes.

In the manufacture of glass panes for automobiles, flat glasses are cut to size, pre-processed and then subjected to a glass bending process at high temperatures in the range from 500 ° C to 750 ° C in order to form the curved geometry typical of automobiles. The safety of the occupants is of central importance, especially in the field of glazing of passenger cars. Since untreated glass represents a considerable risk of injury in the event of breakage, toughened safety glass is used as a windbreak, stain or side window, among other things. Tempered safety glass is produced from normal glass through a thermal toughening process, which consists of heating and subsequent rapid cooling. The internal stresses introduced in this way increase the breaking strength. At the same time, they ensure that the glass breaks into small pieces with blunt edges. Bending and thermal toughening of the panes usually takes place in a combined manufacturing process in which the heating of the panes to the bending temperature is used for thermal toughening.

In WO 2004/087590 and WO 2006072721 a method is described in which the pane is first pre-bent on a bending frame by gravity, followed by press bending by means of an upper or lower bending form. In EP 255422 and US 5906668 the bending of a disk by suction against an upper bending form is described. EP 1550639 A1, US 2009/084138 A1 and EP 2233444 A1 each have a device in which a press frame can be transported between bending stations on a carriage that is slidably mounted on a stationary support.

DE 102005043022 A1 shows a method for controlling and / or monitoring a movement of a free body in an industrial machine, the movement of the free body being simulated. In the industrial series production of glass panes, automated production systems are used in which movement control of movable system parts takes place by means of actuators, for example electric motors, and sensors. For example, a servo motor, which consists of an electric motor and a sensor, in combination with a motor controller, can be used to move to fixed positions. The sensors (encoders), for example rotary encoders, record the actual states of the actuators and encode them into digital signals.

Actuators usually have individual controls (e.g. motor control). The motion controls of actuators are typically controlled by at least one higher-level, programmable logic controller (PLC). This contains the control logic for the entire production process and brings together all process data at a central point. The PLC coordinates the manufacturing process by transmitting setpoints at the right time to the subordinate motion control systems and monitoring the process sequence via feedback from sensor values in the process. The PLC is therefore the central control instance for the automated manufacturing process.

What is essential here is that a human operator can influence the automated process sequence via a human-machine interface by entering specific manipulated variables (parameter values) to control the manufacturing process. The manufacturing process is parameterized for this purpose. Changing the values of process parameters does not change the programming of the PLC.

The operator has an important task here, since it is usually necessary to change parameter values in the automated production process when process conditions have changed. For example, actuators are to be controlled differently when a tool change has taken place or generally another, for example time-optimized method with reduced cycle times is to be carried out. This requires well-trained operators and is challenging, especially since modern systems for automated glass bending are becoming more and more complex due to additional functionalities.

Maintaining machine safety is particularly important when changing process parameters, and collisions between system parts must be avoided under all circumstances. Collisions can damage parts of the plant and possibly result in longer downtimes in the production plant. However, it is due to the steadily increasing The degree of complexity of automated production systems can sometimes be difficult to set the parameters in such a way that machine safety is always ensured.

To make matters worse, the bending of glass panes takes place in a hot environment with limited spatial accessibility, so that visual monitoring of the manufacturing process is difficult and not at all possible from certain viewing positions or viewing angles. It is sometimes difficult or even impossible for the operator to recognize whether parts of the system are coming dangerously close.

Energy savings and short cycle times are also important for the practice of series production. The operator's task here is to reduce the distances traveled by moving system parts by entering optimized parameter values, to increase their speeds and accelerations, if necessary, and to ensure that a pane of glass to be processed is accessed in rapid succession in order to generally compress the production process over time to reach. However, this increases the risk of collisions between movable system parts.

In practice it has been shown that inaccuracies or errors in the movement of movable system parts can occur, with an actual position of a system part differing from its target position. In the worst case, this can lead to a collision of parts of the system.

In contrast, the object of the present invention is to provide an improved automated manufacturing process and an automated manufacturing system for bending panes with which these disadvantages can be avoided. In particular, errors in the positioning of system parts are to be recognized and collisions between system parts are to be reliably and safely avoided. Such errors should be able to be corrected automatically. The operator should also be able to observe the process flow from different, even inaccessible, angles. In addition, it should be possible to optimize parameter values with regard to a selectable process property, preferably the cycle time for processing wafers. These and other objects are achieved according to the proposal of the invention by an automatized manufacturing process for bending panes and an automated manufacturing plant for performing the method according to the independent claims. Advantageous refinements of the invention emerge from the subclaims.

According to the invention, an automated manufacturing process for bending panes is shown in which a pane can be machined using movable system parts, the movable system parts being controllable by a programmable logic controller (PLC) on the basis of manually enterable parameter values. The programmable logic controller can output control signals to actuators of the movable system parts and receive sensor signals from sensors for detecting actual states of the actuators.

According to the present invention, in the automated manufacturing process, manually entered parameter values for controlling the movable system parts are transmitted by an operator to the programmable logic controller for controlling the actuators as well as to a digital image of the automated manufacturing process. The programmable control device advantageously transmits the parameter values to the digital image. The programmable logic controller sends corresponding control signals to the actuators based on the parameter values and regulates the automated production process, with actual sensor data being received. At the same time, a simulation of the automated production process is carried out in the digital image on the basis of the manually entered parameter values. The real manufacturing process and the manufacturing process simulated by the digital image are carried out at the same time. The digital image determines, in particular, simulated target sensor data on the basis of the parameter values. In the context of the present invention, the term "target sensor data" denotes target data generated by the digital image on the basis of the entered parameter values, which sensor data should correspond to the real sensors. It goes without saying that this is not data received from real sensors, but rather simulated data.

What is essential here is that the digital image receives actual sensor data from the sensors of the real production process and determines a discrepancy between the received actual sensor data of the real production process and the (simulated) target sensor data for each sensor the actual sensor data from the target sensor data for a sensor exceeds a predeterminable value, information is output on a monitor and / or the production process is stopped. In the context of the present description of the invention, a “movable system part” is understood to mean a system part that is held on a forced movement path by an actuator, in contrast to a free body such as a disk. In contrast to a movable system part, a free body cannot be kept on a forced movement path by an actuator alone. Accordingly, a disk is not a movable part of the system for the purposes of the invention.

The present invention is based on the knowledge that an integrated digital image (also referred to as a “digital shadow”) of the automated manufacturing process for bending panes can be used usefully in the real manufacturing process. In particular, the aforementioned problems with the automated bending process can be avoided as a result, which is shown in detail below. The digital image simulates the real production process in a program implementation (software) that is implemented in logic modules that are already available or that are additionally provided for this purpose.

The digital image of the automated manufacturing process requires automatic, computer-aided processing of the process data, in particular the control and sensor data. The use of a programmable logic controller (PLC) is required so that this data is available in digitized form and can also be influenced. The automated production process is controlled by an operator by manually entering specific control variables (parameter values).

The digital image includes a three-dimensional (3D) simulation of the kinematics of the system parts of the production system that can be moved (by a respective actuator), in particular of the complete production system, and optionally a visualization of the simulated kinematics of the movable system parts on at least one monitor. A higher-level control of the kinematics simulation is carried out by a PLC, which can be implemented in hardware or software (emulation). Furthermore, a communication and data device is provided which enables bidirectional communication via a preferably standardized machine-to-machine (M2M) communication, so that there is access to current process data at a sufficient update rate. The communication and data device can preferably also store process data. As an interface to the real process, the communication and data device is directly connected to the PLC of the real manufacturing process. The kinematics simulation draws the required process data from the communication and data facility and is also connected to the PLC of the digital image. The digital image depicts the behavior and properties of the automated manufacturing process in a level of detail that is appropriate for its intended use of supporting the human operator in the operation of the process. The aim is to gain knowledge about the past, present and / or future of the automated manufacturing process, with the present description of the invention being essentially based on a currently running manufacturing process. This gain in knowledge is used to support the operator in his task as the person responsible for the process and the highest decision-making authority.

A key component of the digital image is the kinematic simulation of the automated manufacturing process. First of all, a geometric model of the production plant, in which the production process to be mapped takes place, is built within the kinematic simulation model. The level of detail is at the discretion of the modeler and must be adapted to the objectives of the digital image and the specific manufacturing process. On the one hand, even detailing down to individual screws can be useful if, for example, these limit the movement of other system parts or are otherwise relevant for the course of the manufacturing process. On the other hand, larger, complex assemblies can also be replaced by placeholder geometries or left out completely in other scenarios. A digital 3D geometry model is preferably created with a computer-aided design (CAD) system, which is usually constructed as an edge, surface or volume model. After the geometric model of the production plant, the second element of the simulation is the kinematic model. The kinematic model is linked to the initially immobile geometric model and thus allows all degrees of freedom of movement of the real manufacturing process to be reproduced. The influence of the real actuators on the manufacturing process is advantageously simulated in a simplified manner by means of kinematic constraints. For example, a table is considered that is connected to an electric motor via a spindle and can thus be moved linearly in a uniaxial manner in a guide. The electric motor is controlled by a motor controller, which receives setpoints for the axis position of the table, the speed and the acceleration as parameters. In the simulation model, the 3D geometry model of the table is advantageously assigned a uniaxial, linear degree of freedom so that its position can be influenced directly by a variable. This simplification strongly abstracts from the original mechanical behavior, but still allows the sequence of the manufacturing process to be accurately reproduced. Another component of the kinematic simulation is the sensor system. In the real production process, the controller (PLC) needs as Input values the sensor data of the process so that it can respond to setpoint values for the actuators. Accordingly, these must be reproduced by the simulation model. With the 3D kinematics simulation, the process sequence is mapped by the movement behavior of the production plant influenced by the actuators. The primary state variables are the position variables of the moving system parts. The inclusion of the glass pane in the process flow is optional.

Because the PLC control program can be parameterized externally, any parameter configuration and its effects on the process sequence can be simulated. Relevant information for the process operator is first of all whether the control program is running correctly with the selected parameter values. Related questions can be whether all positions can be reached with the given speeds, whether the further conveyance of the glass pane is successful or whether collisions occur. In addition, the effects on process variables such as the cycle time are relevant. A visualization of the simulated process flow by means of a 3D animation on at least one monitor is particularly advantageous, which enables more effective information transfer of the flow of a complex manufacturing process with new parameterization to people than merely text-based information.

The PLC controls the process and movement sequence in the automated manufacturing process. The sensors are connected to the inputs of the PLC and the actuators to the outputs. The use of the digital image for the representation of process states requires a simulation of process states that is independent of the real process. Therefore, the 3D kinematics simulation must be controlled separately, analogous to the functionality of the PLC. The most important goal is the transferability of the simulation results to the control of the real process. Ideally, the same parameter configuration causes identical motion sequences, regardless of whether the PLC used in the process, another PLC in hardware or an image (emulation) of the PLC is used for the kinematics simulation. The control within the simulation also allows a complete abstraction of the sensors within the simulation model, since the control program can directly access all simulated process states and process variables. For example, a PLC emulation is linked to the 3D kinematics simulation. Since the real control program is used in the PLC emulation, all of the sensors of the automated production process required for motion control must process can be implemented in the simulation model and the sensor signals are available in a PLC-compatible form. In addition, the sensors and actuators of the 3D kinematic simulation must be connected to the virtual inputs and outputs of the PLC emulation.

The digital image can therefore particularly advantageously generate information about a future process status or status of the manufacturing system with the aid of the kinematics simulation. The kinematic simulation enables the production process to be represented independently of the current process data of an actual process that is currently running. This enables the digital image to make statements about the effects of parameter changes on any process variables. Building on this, a suitable optimization strategy can be used to optimize process parameters with regard to defined criteria. In particular, collisions between parts of the system can be avoided.

It is also possible that the digital image uses the process data of the actual manufacturing process in operation and thus continuously puts itself in the same state as the real process. Based on this data, the digital image can visualize the current status of the production plant through a display on at least one monitor (e.g. 3D animation). This data stream can also be used for process analysis purposes. New process-related variables can be generated from the existing data through aggregation, reduction or calculation and these can be visualized accordingly. The digital image can also carry out safety checks by monitoring safety-relevant process variables and comparing them with previously defined rules.

It is also conceivable that the processing of previous process data will enable the digital image to analyze and display the causes of errors.

Results of the digital image are transmitted to the operator via a visual display on at least one monitor and, if necessary, text-based. Text-based information can also be displayed, for example, by the MMS.

The communication and data device defines which data is processed and stored in which form and transmits this data to the 3D kinematics simulation as required. The process data required for the real process is read out through a connection to the PLC, which processes it centrally. The main task of the communi- Communication and data facility is therefore the transmission of historical and current process data to the simulation in a compatible data format and a sufficient update rate.

The operator's interaction with the digital image can basically take place on three different levels. First, the digital image can perform a purely informational function, i.e. it provides the person with information that they have to interpret independently and, if necessary, implement in a process intervention. The information can either be requested manually by the operator or it can also be provided automatically by the digital image. In the method according to the invention, there is optionally a visual representation of the process sequence based on previously entered parameter values on at least one monitor.

The second stage is the submission of suggestions for action by the digital image, which can either be accepted or rejected by the user. The suggestions can either be triggered manually by humans or automatically.

The third stage describes the fully automatic action through the digital image, which puts the operator in a passive role. This monitors the automatically executed actions and can react to them at his own discretion. However, the operator is the highest decision-making authority and manually decides which information is actually fed back into the real process.

As already stated, it is provided according to the invention that the real production process and the production process simulated by the digital image run at the same time, where the digital image receives the actual sensor data from the sensors of the real production process, preferably in real time. A permanent comparison of the actual sensor data with the target sensor data takes place for each sensor, with the corresponding deviations being determined. If the deviation of the actual sensor data from the target sensor data for a sensor exceeds a predeterminable or predefined value, information is output on a monitor and / or the production process is stopped. This makes it possible in an advantageous manner that possible errors in the movement or positioning of movable parts of the system can be detected, in particular to reliably and safely avoid collisions between parts of the system. According to an advantageous embodiment of the automated manufacturing process for bending panes, a simulated sequence of movements of the movable system parts based on the parameter values is displayed on at least one monitor by means of the digital image. This is particularly advantageous because the view from the outside is usually very limited due to the special features of the hot bending zone in the real production plant and a visual inspection of plant parts is often not possible. The simulated movement sequence of the movable system parts is particularly advantageously shown in an enlarged representation and / or different viewing angles on the at least one monitor, which enables the operator to make a precise visual assessment of the process sequence.

According to a further advantageous embodiment of the automated manufacturing process for bending panes, the simulated sequence of movements of the at least one movable system part is displayed on the at least one monitor with a time delay. This enables the operator to study the simulated process sequence very precisely, in slow motion, as it were.

According to a further advantageous embodiment of the automated manufacturing process for bending panes, in addition to the visual representation of the sequence of movements of movable parts of the system, further information, which can be useful for the operator, is also displayed on the at least one monitor. At least one path-time diagram (cyclogram) of at least one movable system part is preferably displayed on the at least one monitor. This makes it easier for the operator to analyze the process flow.

According to a further, particularly advantageous embodiment of the automated manufacturing process for bending panes, the parameter values are changed by the digital image of the automated manufacturing process in such a way that the determined deviation of the actual sensor data from the target sensor data for one or more sensors is the respective falls below the specified value. This measure enables the risk of a possible collision to be reduced.

The parameter values entered by an operator are typically suboptimal with regard to a selectable process property, such as the cycle time when machining wafers. According to a further advantageous embodiment of the automated manufacturing process for bending panes, the digital image of the automated manufacturing process takes place using an optimization algorithm based on the actual sensor data and / or the simulated data Target sensor data an automatic optimization of the parameter values in relation to a selectable process property, preferably the cycle time.

According to an advantageous embodiment of the automated manufacturing process for bending panes, the actual sensor data received are stored, which enables errors to be analyzed subsequently.

The automated manufacturing process for bending panes, which can be found in the digital image, preferably comprises the following (e.g. successive) steps:

Providing a sheet heated to bending temperature in a bending zone, for example directly below a mold. Fixing the disc on a contact surface (of a tool) of the mold. The pane is advantageously fixed by blowing a gaseous fluid on it. Alternatively and preferably in addition, the disk is fixed by suction on the contact surface of the mold. Positioning a frame within the bending zone, for example directly below the form, while the pane is attached to the form, and placing the pane on the frame. The frame is used to transport the pane on top of it, whereby the pane can be bent by gravity.

Optionally, a press frame is provided within the bending zone, with the pane being pressed between the mold and the press frame. Optionally, the pane can be placed on the press frame.

Optionally, a prestressing frame is provided in the bending zone, the pane being transported on the prestressing frame to a cooling device for thermal prestressing of the pane. During transport on the prestressing frame, the pane can be bent in the inner area of the pane by gravity.

For example, a press frame is first provided in the bending zone, followed by pressing the pane between the mold and press frame and then providing a prestressing frame in the bending zone, and placing the wafer on the prestressing frame.

It goes without saying that the disk can be fixed to several shapes one after the other in order to produce complex geometries. The at least one shape is preferably only moved up and down in a translatory manner in the vertical direction. The at least one frame is preferably moved back and forth translationally in only one horizontal direction.

The invention also extends to an automated production system for bending panes, which is suitably set up for carrying out the method according to the invention. The production plant comprises movable plant parts for processing a pane, which can be controlled by a programmable control device on the basis of manually inputted parameter values. The programmable logic controller can output control signals to actuators of the movable system parts and receive sensor signals from sensors for detecting actual states of the actuators. The production system has a digital image of the automated production process and optionally at least one monitor for displaying content relating to the process flow. The production system is programmed in such a way that manually entered parameter values for controlling the movement of movable system parts are transmitted both to the programmable logic controller for controlling the actuators and to a digital image of the automated production process, the digital image being simulated target sensor data based on the Determines parameter values, receives the digital image of actual sensor data and determines a deviation of the actual sensor data from the target sensor data for each sensor, with a predeterminable one in the event that the deviation of the actual sensor data from the target sensor data for a sensor If the value exceeds the information output on a monitor and / or the production process is stopped.

In the context of the present description of the invention, the term “pane” generally refers to a pane of glass, for example a soda-lime glass.

The automated manufacturing system for bending panes advantageously comprises several structurally and functionally delimitable zones. An essential component is a bending zone for bending hot panes, which is advantageously equipped with a heating device for heating panes. In particular, for this purpose the bending zone can be brought to a temperature that enables the plastic deformation of panes and is, for example, in the range from 500.degree. C. to 750.degree. The bending zone is preferably designed as a heatable chamber that is closed or closable towards the external environment. For bending panes, the bending zone comprises at least one shape that can be equipped with a tool for fixing a pane, as well as at least one frame (eg an annular frame) on which the pane can be placed. Typically, the pane rests on the frame only with the pane edge. The tool has a contact surface for contacting the disc. The contact surface is designed to be suitable for a desired bending of the disk. The frame is used to store the disc and, if necessary, to press the edge area of the disc with a mold. In the form of a press frame, the frame has a press surface that is complementary to the contact surface of a tool of a mold. The frame is advantageously designed to be suitable for surface prebending by gravity in the inner region of the pane, with the inner region of the pane being able to sag downward by gravity. For this purpose, the frame can be open, that is, it can be provided with a central opening, but it can also have a full surface configuration, as long as the inner region of the pane can sag. An open design is preferred with regard to simpler processing of wafers.

In one embodiment, the bending zone has at least one mold and a press frame assigned to the at least one mold, the mold and press frame being displaceable in the vertical direction relative to one another so that the pane can be pressed in the edge area between the mold and press frame. Preferably, the shape can only be moved in a translatory manner (one-dimensional or uniaxial) in the vertical direction. The press frame can preferably only be moved in a translatory manner in a horizontal plane. This enables easy control of the mold and press frame. For example, the bending zone has only a single shape and associated press frames. For more complex pane geometries, the bending zone can also have two or more shapes and at least one associated press frame, for example, with the pane being bent in several stages.

The at least one mold preferably has a means for fixing a pane on its contact surface, for example a pneumatic suction device for sucking in a gaseous fluid, in particular air, through which the pane can be drawn against the contact surface by means of negative pressure. The contact surface can be provided for this purpose, for example, with at least one suction hole, advantageously with a plurality of suction holes, for example evenly distributed over the contact surface, at each of which a negative pressure can be applied for a suction effect on the contact surface. The suction device generates a typically upwardly directed flow of a gaseous fluid, in particular air, which is sufficient to hold the disc on the contact surface. This makes it possible in particular to place a frame for receiving the pane fixed on the contact surface below the pane. Alternatively or in addition, the means for fixing a disc on the contact surface comprises a pneumatic blowing device for generating a gaseous fluid flow, in particular an air flow, which is designed so that a disc is blown from below by the gaseous fluid flow, thereby raised and against the Contact surface of the form can be pressed. The fixing of a disk on the contact surface of a mold is not necessarily associated with a bending process, but can lead to a bending of the disk.

The automated production system advantageously has a preheating zone with a heating device for heating panes to the bending temperature, as well as a transport mechanism, in particular of the roller bed type, for transporting panes from the preheating zone to the bending zone, in particular to a removal position (e.g. directly) below a mold . The roller bed is advantageously designed so that individual slices can be transported to the removal position one after the other. The removal position can in particular correspond to an end section of the roller bed.

The automated production system also advantageously has a thermal prestressing zone with a cooling device for thermal prestressing of a pane, whereby a prestressing frame (e.g. prestressing ring) can be provided for transporting a pane from the bending zone into the prestressing zone. The thermal pre-stressing (tempering) creates a temperature difference between a surface zone and a core zone of the pane in order to increase the breaking strength of the pane. The pretensioning of the pane is advantageously generated by means of a device for blowing a gaseous fluid, preferably air, onto the pane. Preferably, the two surfaces of a disc are acted upon simultaneously with a stream of cooling air.

For example, the production system has at least one mold, a press frame (e.g. press ring) and a prestressing frame (e.g. prestressing ring), wherein the mold can be lowered and raised in the vertical direction by a reciprocating translational movement, and both the press frame and the prestressing frame each have a reciprocal translational movement can be offset in a horizontal direction, in particular in a position directly below the at least one shape. Thus, a disc can be picked up by the mold. men, pressed together with the press frame and then placed on the prestressing frame. The press and pretensioning frames are advantageously moved one after the other to a position directly below the mold.

The various embodiments of the invention can be implemented individually or in any combination. In particular, the features mentioned above and to be explained below can be used not only in the specified combinations, but also in other combinations or alone, without leaving the scope of the present invention.

The invention is explained in more detail below on the basis of exemplary embodiments, reference being made to the accompanying figures. It shows in a simplified, not to scale representation:

Fig. 1 is a schematic representation of an exemplary automated manufacturing

Disc bending process;

FIG. 2 shows a schematic illustration of a production system for bending panes in a plan view for the production process of FIG. 1; FIG.

3 shows a schematic illustration of the process control by the PLC;

FIG. 4 shows a diagram to illustrate the information flows in the manufacturing plant of FIG. 2;

FIG. 5 shows a diagram to illustrate the information flows in the manufacturing plant of FIG. 2 with an integrated digital image;

6 shows a diagram to illustrate the information flows in a first application of the digital image in the production plant of FIG. 5;

FIG. 7 shows a diagram to illustrate the information flows in a second application of the digital image in the production plant of FIG. 5; 8 shows a diagram to illustrate the information flows in a third application of the digital image in the production plant of FIG.

First of all, FIGS. 1 and 2 are considered. FIG. 1 uses a schematic representation to illustrate an exemplary automated production process for bending panes in automobile glazing. In the manufacturing process, flat, two-dimensional glasses are processed that have previously been cut and pre-processed. The resulting product is what is known as single-pane safety glass with a geometry that can be freely designed within the framework of certain boundary conditions. To do this, a disk is machined in two steps in a production facility. First of all, the disc is bent to give shape by pressing under the action of a strand and then pretensioned by means of controlled cooling. FIG. 2 uses a schematic representation to illustrate an exemplary production system for the automated production process from FIG. 1 in a top view from above. In the schematic representation of FIG. 1, the manufacturing process runs from left to right.

In this case, a pane 2 is first heated over a heating section, since it is not possible to reshape the glass when it is cold. The pane 2 is heated in a preheating zone 12 by radiant heat 3, which is supplied from above and below a roller bed 4 on which the pane 1 is supported for its transport. The disk 1 is fed to a bending zone 5 on the roller bed 4. Within the bending zone 5, the pane 1 is blown from below with hot air 6 and picked up by a vertically movable mold 7. The mold 7 is to be provided with a suction device for the disc 1 in order to generate a negative pressure on its surface. The surface of the mold 7 is specially designed to achieve a desired geometry of the disc 2 to be produced. By adapting the hot glass to the surface of the mold 7, a deformation of the pane 1 is already achieved. Now a horizontally movable, hot press ring 8 is moved under the mold 7 as a counterpart to the mold 7. In contrast to the mold 7, the press ring 8 does not depict the complete geometry of the disk 1, but only offers a contact surface for the edge of the disk 1. The mold 7 is then lowered and the disc 1 is pressed between the mold 7 and the press ring 8 to give it a shape. The disk 1 remains after the pressing process with the aid of a negative pressure generated on the surface of the mold 7 on the mold 7 until the press ring 8 is retracted and a horizontally movable, cold pre-tensioning ring 9, which was previously located in a pre-tensioning zone 10 next to the bending furnace 5, whose position has taken. The vacuum is now released and the disk 1 is placed on the pretensioning ring 9. The disk 1 is transported on the prestressing ring 9 from the bending furnace 5 into the prestressing zone 10 and is prestressed with a stream of cold air 11 and cooled. After pretensioning, the process is complete and the pane 1 can be removed. In Figure 2, the linear movements of the three central elements form 7, press ring 8 and pretensioning ring 9 are illustrated schematically by means of arrows.

In the manufacturing plant 1, the discs 1 are automatically fed in, and finished discs are automatically removed and transferred to downstream manufacturing steps. The sequence of the production process within the production system 1 is completely automated, with the mold 7, press ring 8 and pretensioning ring 9 each being able to be moved uniaxially by actuators (e.g. servomotors). The course of movement of the mold 7, press ring 8 and pretensioning ring 9 controlled by actuators is decisive for the transport and the resulting geometry of the disk 1. In addition to the actuators for moving these central elements of the production system 1, the additional actuators are used to influence the process in a targeted manner. For example, the supply of cold air and cold air is controlled by flaps that are moved by actuators, and the separation of various furnace areas by movable doors. The process is determined by the control of the axes of the mold 7, press ring 8 and preload ring 9 of the production plant 1, since their movements are mutually dependent and they operate in the same work area. For example, the mold 7 has to lower itself into the press ring 8 in order to be able to carry out the pressing step. Small deviations in the position or the chronological sequence of movements can result in undesired collisions, which results in an expensive downtime in the production process, but also in severe damage to the tools and the production system due to the high speeds and forces of the servomotors 1 can cause yourself. A control of these various axes is important for the successful execution of the production process, but their control is based on the movements of the central elements of the production plant 1.

The automated manufacturing process for bending disks illustrated with reference to FIGS. 1 and 2 comprises a single mold and a pressing and pretensioning ring. This is only to be understood as an example, it being understood that, in principle, several molds can be used in order to manufacture very complex disk geometries. Thermal pre-tensioning of the pane is also optional.

The process is controlled by a central PLC, which is connected to all the sensors in the production plant 1 and, on this basis, determines the setpoint specifications for the various axes to be controlled. This is illustrated schematically in FIG. Accordingly, the PLC issues motion controls for the respective actuators on the basis of received sensor data in front. Subordinate motor controls take over the control of the actuators based on the setpoints of the PLC. In addition, the PLC controls the non-kinematic process influences, for example furnace temperatures and flow pressures. The operator can access the PLC through the MMS and control the process sequence, with specific parameters being entered in the MMS for this purpose.

FIG. 4 uses a diagram to illustrate the various information flows in the automated manufacturing process for bending glass panes, as is carried out, for example, in the manufacturing plant 1 of FIG. The role of the human operator is to monitor and parameterize the manufacturing process. The MMS is available to the operator for this, via which the manufacturing process can be started or stopped, and parameters for controlling the manufacturing process can be entered. An exemplary input mask for the MMS is shown, in which a parameter value (here e.g. 200) for a "pre-position B" and a further parameter value for a "pressing position 1" (here e.g. 250) can be entered manually. The parameters are transferred to the PLC, which controls the glass bending process accordingly, whereby the PLC uses sensor data for this purpose. Various process information that is provided by the PLC can be brought to the attention of the operator via the MMS, which is not shown in detail here. The actual process sequence recorded by cameras is displayed on monitors, which provides the operator with supporting process information. As a rule, however, the operator has no direct visual insight into the production process, as the axis movements mostly take place within the closed bending zone. In addition, visibility through cameras is limited, as the environment is high-temperature and special cameras with comparatively poor resolution and viewing angles have to be used. Due to the structure of the system, at least four views are generally required in order to fully reproduce the process flow. The operator monitors the production in particular with regard to system malfunctions, such as the loss of a pane within the system, which can be caused by an incorrect flow of hot air or a break in the negative pressure.

The process parameterization is important for the correct flow of the manufacturing process. Especially after the mold has been retooled with a new tool, the parameters must be adapted to the changed process and the new tool geometry. The programming of the PLC specifies the existing movement positions of the axes and their basic sequence structure. The programming is only changed for profound process changes take place, for example when a completely new movement step is introduced. The concrete axis position values of a specific movement step as well as the associated speed and acceleration are the subject of parameterization by the operator. There are preset parameter values for each tool type, but these may have to be adjusted manually to the condition of the wheel or to the prevailing conditions. The operator enters all parameter values manually into the MMS and overwrites the existing parameterization of the PLC after pressing the start button. The process is then carried out with the new parameters.

According to the present invention, the automated manufacturing system for bending panes is provided with a digital image of the automated manufacturing process, which supports the operator. The existing IT infrastructure is used to seamlessly incorporate the information flow of the digital image. The control of the digital image is therefore integrated into the MMS. As shown at the beginning, the digital image is a program-technically implemented kinematic simulation of the glass bending process.

This is illustrated with the aid of the schematic representation of FIG. 5, in which the various information flows during the glass manufacturing process are shown, including the digital image, in a representation analogous to FIG. In order to avoid unnecessary repetitions, only the differences from FIG. 4 relating to the digital image are explained and otherwise reference is made to the above statements.

The digital image connected directly to the PLC comprises several components. The central component is a 3D kinematics simulation of the glass production process, which is controlled by a programmable (emulated) PLC, for example. (Simulated) sensor data can be sent to the emulated PLC. Connected to the 3D kinematics simulation is a combination of a communication server and a database, through which communication with the PLC for the real production process takes place. Here the communication server is designed, for example, in the form of an OPC UA server. OCT UA is a bidirectional machine-to-machine communication protocol that was developed for industrial automation. The details are known to the person skilled in the art and are not relevant for an understanding of the invention, so that they do not have to be discussed in more detail here. Specifically, the use of OPC UA enables the transfer of current process data between the PLC of the real manufacturing process and the 3D kinematics simulation. The database saves the process data stream entering the OPC UA server. The 3D kinematics simulation can be carried out using the OPC UA server access current and previous process data. The emulated PLC executes the original PLC control program for the real glass bending process. A visualization of the 3D kinematics simulation is shown to the operator on the monitors in addition to the camera images.

The 3-D kinematic simulation comprises a geometric model of the production plant in which the production process to be mapped takes place, in conjunction with a kinematic simulation model. The 3D kinematics simulation depicts the process sequence of the production plant influenced by the actuators, with the primary state variables being the position variables of the movable plant parts.

Various applications of the integrated digital image of the glass bending process of the production plant 1 are described below.

A first application is illustrated with the aid of FIG. 6, in which the various information flows in the glass production process are shown, including the digital image, in a representation analogous to FIG. This essentially involves a simulation of the automated manufacturing process using current sensor data, which runs parallel to the real manufacturing process.

The flow of information from the glass bending process in operation to the digital image is one-sided and automatic. Sensor data are transmitted to the 3D kinematics simulation via the PLC and the OPC UA server and simultaneously archived in the database. Archiving in the database is only optional. On the basis of this sensor data stream, the simulation of the glass bending process is carried out automatically without any action by the operator, as long as no other control commands are triggered by the operator. The mapping of the current process status continuously puts the simulation model in the status of the real process, so that in particular the movements of the axes of the main components of the production system can be shown as 3D animation on the monitors. The problem of poor visibility into the process due to the high temperature environment can thus be solved. In addition, information that may already be displayed text-based by the MMS is clearly visualized in the 3D animation. Temperature measurements can be displayed as a text annotation in the correct area of the bending zone and blowing with hot and cold air can be visualized through animations and provided with additional context, such as current pressure and temperature. The compression of the information and the visualization in one 3D animation gives the operator a better and faster overview of the process.

This function can be further supported with the method of process analysis. Instead of permanently defining the selection of information in the visualization statically, the most important information can be displayed dynamically in accordance with the current process status. For this purpose, the actual process data are assigned to the movement steps of the axes and then prioritized depending on the deviations from the setpoints or average values. In addition, throughput statistics can be calculated as new values from the process data. While the MMS displays a static cycle time based on the duration of the programmed movement sequence in the PLC, the average throughput, which is influenced by delays in the upstream production steps, can be calculated in the 3D kinematics simulation environment.

The use of the current process data stream (actual sensor data) also enables safety checks. Since the simulation model depicts the geometric state of the system and, in contrast to the PLC, has data on the dimensions of the tools currently installed, the digital image can be used to determine the distances between the axes while the process is running. Distances that fall below the limit values can be highlighted for the operator in the 3D animation. In cases of particularly serious deviations, the simulation can also intervene in the process independently and send a command to stop the system to the PLC via the OPC UA server.

The same applies to the monitoring of the negative pressure that holds the disc on the mold after pressing. If the 3D kinematics simulation detects a break in the negative pressure on the basis of the data from the pressure sensors, the exact system status can be displayed and frozen in the 3D animation when the fault occurs, so that the furnace operator has a better basis for decision-making than the limited view through the cameras for the need process intervention can be presented.

A second application is illustrated with the aid of FIG. 7, which again shows the various information flows during the glass production process, including the digital image, in a representation analogous to FIG. This is essentially about an error analysis after the occurrence of a fault in the real glass bending process. In contrast to the application of FIG. 6, the error analysis is triggered manually by the operator if necessary. After the command has been triggered via the MMS, this is forwarded to the 3D kinematics simulation. The simulation requests previous sensor data from the OPC UA server. These sensor data are read from the database and serve as a new data source for the 3D kinematics simulation. The simulation model then runs through the past process states and visualizes them in the same way as current process data is used.

Two exemplary scenarios are particularly relevant for this application in production. Firstly, system errors such as unplanned downtimes, glass loss within the furnace or collisions can be investigated. Using the time stamp of the error, the faulty cycle can be identified and visualized for the operator. The comparison of the process data of the faulty manufacturing cycle with earlier, fault-free cycles with the same parameterization and tooling from the database enables unusual values or values deviating from the target value to be highlighted. The digital image helps to analyze the cause of errors by identifying and visualizing potential sources of error. An initial assessment of the extent of the defect and the possible need for repairs can be carried out directly on the system by the responsible personnel without additional aids. The second usage scenario of the failure analysis concerns defective glass panes. Each manufactured pane has a unique identification number that is stored in the database together with the previous process data. Each pane can therefore be assigned to the data image of its individual production cycle. If the measuring station downstream of the glass bending process detects an increasing number of deviations from the target geometry, the operator can use the pane number to visualize the production of this pane again using the digital image. Analogous to the investigation of system errors, the 3D animation of the production cycle of the defective panes contains additional information from the comparison with previous cycles of the same configuration. This supports the operator in deciding whether and how a parameter adjustment has to be made or whether incorrect system behavior is the cause of the geometry deviations.

A third application is illustrated with reference to FIG. 8, in which the various information flows in the glass manufacturing process are shown in a representation analogous to FIG. 5, including the digital image. This essentially involves checking new parameters, in particular to avoid collisions between parts of the system, and then automatically optimizing them based on this. The test of new parameters for the operation of the manufacturing process is an important support function of the digital image for the operator. The frequency of parameter changes in connection with the serious consequences of a single incorrectly entered parameter results in a simple and reliable method for risk-free advance review of changes. To use this method, the operator preferably enters the new parameter values in the normal input mask of the MMS and, for example, presses a newly added button to trigger the simulation instead of directly parameterizing the PLC of the real production process with the new values. The parameters are transferred via the PLC and the OPC UA server to the 3D kinematics simulation, which the emulated PLC re-parameterizes. For example, the 3D kinematics simulation stops the ongoing processing of current process data and is therefore independent of the ongoing manufacturing process. The emulated PLC executes the control program and controls the actuators of the simulation model using the sensor values it has read. The simulation environment simulates the process sequence with the original control program using the operator's new parameterization proposal and visualizes the process sequence on the monitors. At the same time, the simulation calculates the resulting cycle time of the new parameter configuration and whether collisions have occurred. The 3D animation enables the operator to visually and qualitatively validate the new process sequence, including any collisions that may occur. The simulation results with regard to the cycle time and any system components involved in collisions are, for example, also transmitted to the MMS as text and displayed. At the end of a check cycle, the final decision on the application of the new parameterization lies with the operator, who can confirm or reject it with the push of a button.

An automatic optimization of parameters, which is preferably carried out in addition, runs largely analogously to the test of a new parameterization. For this purpose, the operator enters a set of parameters into the MMS as the basis for the optimization process, which is processed from the digital image. Instead of one simulation run, several runs are now carried out which, in order to save time, preferably do without the 3D animation. Between the runs, the parameters are adjusted step by step using an optimization algorithm, for example to minimize cycle times. The parameter configuration with the shortest cycle time, which however still runs properly and without collisions, is visualized and the associated parameters are shown in the input mask of the MMS. The operator must finally confirm the parameterization suggestion before it is applied to the real process. The parameters are preferably automatically optimized with a view to reducing the deviations of the actual sensor data from the target sensor data, as a result of which the risk of collisions can be avoided.

From the above it follows that the invention provides a novel automated manufacturing process for glass bending and a manufacturing system with a digital image of the automated manufacturing process, in which a collision of system parts due to errors can be reliably and safely avoided. In addition, process parameters can be optimized with regard to avoiding collisions and the cycle time.

List of reference symbols

I production line 2 disc

3 heating radiation

4 roller bed

5 bending zone

6 hot air 7 mold

8 press ring

9 preload ring

10 preload zone

I I cold air flow 12 preheating zone

Claims

1 . Automated manufacturing process for bending panes, in which a pane can be machined by means of movable system parts, the movable system parts being controllable on the basis of manually enterable parameter values by a programmable logic controller, the programmable logic controller outputting control signals to actuators of the moveable system parts and sensor signals from sensors can be received to detect actual states of the actuators, manually entered parameter values for controlling the movable system parts are transmitted both to the programmable logic controller for controlling the actuators and to a digital image of the automated production process, the digital image being simulated target sensor data determined on the basis of the parameter values, the digital image receives actual sensor data and a deviation of the actual sensor data from the target sensor data is determined for each sensor, in the event that the Ab The deviation of the actual sensor data from the target sensor data for a sensor exceeds a predefinable value, information is output on a monitor and / or the manufacturing process is stopped.

2. Automated manufacturing process for bending panes according to claim 1, in which the parameter values are changed by the digital image of the automated manufacturing process in such a way that the deviation of the actual sensor data from the target sensor data for a sensor falls below the predeterminable value.

3. Automated manufacturing process for bending panes according to claim 1 or 2, in which the parameter values through the digital image of the automated manufacturing process using an optimization algorithm based on the actual sensor data and / or the simulated target sensor data in relation to a selectable process property be optimized.

4. Automated manufacturing process for bending disks according to claim 3, in which there is an optimization of the parameter values in relation to a cycle time in the disk machining.

5. Automated manufacturing process for bending panes according to one of Claims 1 to 4, in which a simulated sequence of movements of the movable system parts based on the parameter values is displayed on at least one monitor.

6. Automated manufacturing process for bending discs according to claim 5, in wel chem the simulated sequence of movements of the movable system parts in an enlarged Dar position and / or different angles is shown on the at least one monitor.

7. Automated manufacturing process for bending panes according to claim 5 or 6, in which the simulated sequence of movements of the movable system parts is displayed on the at least one monitor in an externally inaccessible viewing angle.

8. Automated manufacturing process for bending panes according to one of claims 5 to 7, in which the simulated sequence of movements of the movable system parts is displayed with a time delay on the at least one monitor.

9. Automated manufacturing process for bending panes according to one of claims 5 to 8, in which at least one item of information relating to the process sequence that goes beyond the display of the sequence of movements of the movable system parts is displayed on the at least one monitor.

10. Automated manufacturing process for bending disks according to claim 9, in which at least one path-time diagram of a movable system part is displayed on the at least one monitor.

11. Automated manufacturing process for bending panes according to one of Claims 1 to 10, in which the actual sensor data received are stored.

Storage of the actual sensor data

12. Automated manufacturing plant for bending disks to carry out the method according to one of claims 1 to 11, which has movable plant parts for processing a disc, the movable plant parts being controllable by a programmable logic controller based on manually enterable parameter values, the memory programmable control device can output control signals to actuators of the movable system parts and receive sensor signals from sensors for detecting the actual states of the actuators, which has a digital image of the automated production process and which is programmed in such a way that manually entered parameter values are used to control the movable system parts both to the programmable logic controller to control the actuators as well as to a digital image of the automated production process, the digital image determines simulated target sensor data on the basis of the parameter values, the digital image receives actual sensor data and a deviation of the actual sensor data from the target sensor data determined for each sensor, in the event that the deviation of the actual sensor data from the target sensor data for a sensor exceeds a predeterminable value, information is output on a monitor and / or the production process is stopped.