Cutting method and cutting device for scoring components made of glass or ceramic, as well as method for splitting components made of glass or ceramic.

Description

[0001] The present invention relates to a cutting method and a cutting device for superficially scoring preferably plate-shaped components made of glass or ceramic, preferably glass panes, as well as a method for dividing preferably plate-shaped components made of glass or ceramic, preferably glass panes, e.g. float glass panes or laminated glass panes, in particular laminated safety glass panes, into individual component blanks, preferably glass pane blanks.

[0002] Flat glass refers to any glass in the form of glass panes, regardless of the manufacturing process used.

[0003] Laminated glass panes are a laminate of at least two individual glass sheets, each bonded together by an adhesive interlayer made of plastic. In laminated safety glass, the interlayer connecting the individual glass sheets is usually made of an elastic, tear-resistant high-polymer film, so that if the panes break, the fragments adhere to the film. This reduces the risk of cuts or punctures. The laminated safety glass also retains some residual load-bearing capacity after breakage.

[0004] Flat glass panes can consist of a single glass plate (single glass panes) or of several glass plates glued together (laminated glass panes).

[0005] The division of individual glass panes is generally achieved by cutting or scoring the glass pane to be divided with a cutting wheel, either on the surface or on one of its two glass pane surfaces. The glass pane is then broken along or at the resulting cutting line, scoring line, or scoring track. Breaking occurs by bending the glass pane, e.g., by bulging the glass pane on both sides of the scoring line. The terms "cutting" and "scoring" are used synonymously for surface scoring in the field and also in the context of the present application.

[0006] When splitting laminated glass panes, both glass panes are first scored on the surface, preferably synchronously, and then the two glass panes are successively broken along the scored line. This creates a scoring effect on both surfaces of the laminated glass pane. The film sandwiched between the two glass panes is then severed. This can be done mechanically using a knife and / or thermally. The film can also be separated beforehand using a laser.

[0007] DE 101 57 833 B4 discloses a device for dividing a laminated glass pane into individual laminated glass pane blanks. The device comprises a support table for the laminated glass pane, a cutting device with a scoring, breaking, and separating tool for dividing, and a positioning bridge arranged above the support table for moving back and forth to the cutting device for displacing the laminated glass pane on the support table. The positioning bridge has a gripping device with grippers for gripping the laminated glass pane, which is designed such that, with respect to the counter-rotating movement of the positioning bridge, the laminated glass pane can be gripped both in its rear edge region facing away from the cutting device and in its front edge region facing the cutting device.

[0008] The goal when dividing glass panes is to achieve good and as consistent a quality as possible at the broken edges, especially for good edge strength. The broken edges should be as smooth and even as possible, as well as perpendicular to the surface of the glass pane, for a better visual effect, greater manufacturing precision, and / or to achieve the final dimensions with less grinding during subsequent grinding. Especially in construction, the application limits of flat glass are significantly restricted by the edge strength. If the broken edges are of insufficient quality, depending on the application, extensive rework may be necessary to improve the fit and / or edge strength.

[0009] Scoring creates deep, radial, and lateral cracks in the glass pane. The easy breakage of the glass pane is due to the intended strength-reducing effect of the deep crack. However, lateral and radial cracks should be avoided or reduced as much as possible, as they reduce edge strength. In particular, such microcracks in the fracture edge can continue to negatively impact edge strength even after subsequent grinding.

[0010] Depending on the glass's application, the quality of the broken edge can be assessed with varying degrees of importance based on the deviation from a perfectly vertical and straight edge, the presence of disruptive cracks, or the measured mechanical strength of the edge. Edge strength is determined in particular as described in the article "Cut edge of annealed float glass: crack system and possibilities to increase the edge strength by adjusting the cutting process," Springer Nature Switzerland AG 2019, published by TU Darmstadt. The quality of the broken edge can also be assessed using DIN 1249-11-2017-05.

[0011] Factors influencing poor edge quality include suboptimal machine technology, incorrect cutting tool selection, incorrect process parameters and / or the state of wear and / or damage to the cutting wheels.

[0012] For consistently good fracture edge quality, the correct cutting pressure is important, among other things. If the cutting pressure is too high, for example, this leads to an increased formation of unwanted lateral and radial cracks.

[0013] An optimal and precise angle of the cutting edge of the cutting wheel in contact with the glass surface is also crucial for good breakage quality. Furthermore, the cutting wheel must be rotationally symmetrical.

[0014] The cutting wheel is also subject to continuous wear in the contact area with the glass surface, which can be counteracted to a certain extent by adjusting process parameters. For example, the cutting edge of the cutting wheel becomes blunt over time. To maintain good break edge quality, the cutting pressure must therefore be increased over time. This is initiated by the operator of a glass cutting system if, for example, they notice that the break edge quality is deteriorating based on the visual appearance of the break edge.

[0015] Compared to these rather slow changes, sudden, local damage to the contact area with the glass pane surface is particularly critical, as this is often not immediately recognized and can lead to poorer fracture edge quality.

[0016] For this reason, the scoring process, and especially the quality of the break edges, must be continuously monitored and controlled. It is well known that experienced operators of a glass cutting system can sometimes assess the quality of the scoring process based on the visual appearance of the scoring line and / or the break edge during the scoring process and, if necessary, initiate appropriate countermeasures.

[0017] US Pat. No. 3,880,028 A discloses a glass cutting device with monitoring of the broken edge quality. The glass cutting device comprises a holding device for holding a glass pane in a scoring position, a scoring tool mounted to contact the glass pane in the scoring position, a device for generating a relative movement between the glass and the scoring tool to score the glass, and a sound recording device for sound waves generated during scoring. The sound recording device generates an electrical signal as a function of the intensity of the sound waves.

[0018] JP H07 257935 A discloses a method for cutting a glass plate, in which a cutting wheel pressed against the glass plate is moved to create a scoring line on the surface of the glass plate. The method determines whether the scoring line is good or bad by measuring the vibrations of the cutting wheel or the cutting noise during scoring.

[0019] The object of the present invention is to provide a cutting method and a cutting device for cutting or scoring, preferably plate-shaped, components made of glass or ceramic, preferably glass panes, which are intended to ensure good breaking edge quality.

[0020] Another task is to provide a process for splitting glass panes, which should also ensure good breaking edge quality.

[0021] The objects are achieved by a cutting method according to claim 1, a cutting device according to claim 11 and a method according to claim 9. Advantageous developments of the invention are characterized in the respective subsequent subclaims.

[0022] The invention is explained in more detail below using a drawing as an example. The drawings show: Figure 1: Highly simplified and schematic plan view of a cutting device according to the invention Figure 2: Highly simplified and schematic side view of a cutting head with a cutting wheel Figure 3: A cutting-side view of a cutting wheel Figure 4: A frequency spectrum of a longitudinal section, cut with a microstructured cutting wheel Figures 5a-c: Frequency spectra of longitudinal sections, cut with conventional cutting wheels with different cutting pressure Figures 6a+b: Frequency spectra of a cross section and a longitudinal section, cut with the same conventional cutting wheel Figures 7a-c: Frequency spectra of conventional cutting wheels with different degrees of wear Figures 8a+b: The frequency spectrum of the conventional cutting wheel with medium degree of wear Figure 7b and a deflection in the time signal Figures 9a-c: Time-based evaluation of the time signals of the cutting wheels according to the Figures 7a-cFigure 10: A more detailed side view of a cutting head with a cutting wheel Figure 11: A schematic representation of a sound pressure sensor with a horn and with the solid angle range of the sound that is captured by the sound pressure sensor

[0023] The cutting device 1 according to the invention ( Fig. 1) is used for cutting glass panes 2 and preferably also for dividing glass panes 2, e.g. single glass panes or laminated glass panes, into individual glass pane blanks. For this purpose, the cutting device 1 has a support table 3 for receiving the glass pane 2 to be cut and a cutting bridge 4 extending in a horizontal y-direction with a cutting head 5 for scoring a glass pane surface 2a of the glass pane 2 and for introducing scoring lines or scoring tracks 6 into the glass pane surface 2a. The cutting bridge 4 spans the support table 3 on the top and / or bottom. In addition, the cutting bridge 4 can be stationary or movable back and forth over the glass pane 2 in a horizontal x-direction perpendicular to the y-direction. In this case, the cutting device 1 has corresponding drive means for moving the cutting bridge 4.

[0024] The cutting head 5 is mounted on the cutting bridge 4 in a manner known per se, so that it can be moved back and forth in the y-direction. The cutting bridge 4 has corresponding drive means for moving the cutting head 5.

[0025] In addition, the cutting head 5 also has, in a manner known per se, a cutting head housing 7, a cutting wheel 8, and a cutting wheel carrier 9. The cutting wheel carrier 9 is mounted on the cutting head housing 7 so that it can be moved back and forth in a vertical z-direction perpendicular to the glass pane surface 2a. For this purpose, the cutting head 5 has corresponding drive means. In addition, the cutting wheel carrier 9 is preferably mounted on the cutting head housing 7 so that it can rotate about a carrier rotation axis 9a parallel to the z-direction. The cutting wheel carrier 9 can be freely rotatable or driven about the carrier rotation axis 9a. The cutting wheel 8 is mounted on the cutting wheel carrier 9 so that it can rotate freely about a horizontal wheel rotation axis 8a parallel to the glass pane surface 2a.

[0026] In addition, the cutting device 1 preferably has at least one positioning bridge 10 for moving the glass pane 2 in the x-direction on the support table 3. The positioning bridge 10 also spans the support table 3 on the top and / or bottom and has, for example, clamping grippers 11 for gripping the glass pane 2 in the region of a glass pane edge 2b and / or suction grippers for gripping the glass pane 2 at the glass pane surface 2a. In addition, the positioning bridge 10 can be moved back and forth in the x-direction. For this purpose, the cutting device 1 has corresponding drive means. Of course, the cutting device 1 can also have other means for moving and / or gripping the glass pane 2.

[0027] The cutting device 1 can also have means for rotating the glass pane 2 about an axis of rotation parallel to the z-direction. The rotation can be achieved, for example, by means of a suction gripper.

[0028] The cutting wheel 8 ( Fig. 3) has a circumferential, V-shaped cutting edge 12 with a cutting angle α. The cutting edge 12 can be ground in the circumferential direction (conventional cutting wheels) or toothed (microstructured cutting wheels). These toothed or microstructured cutting wheels 8 (not shown) thus have a tooth structure on their cutting edge 12.

[0029] In principle, any desired cutting contours, in particular arcuate cutting contours, can be created with the cutting device 1 by combining the movement of the cutting head 5 along the cutting bridge 4 in the y-direction, the movement of the glass pane 2 in the x-direction and / or a movement of the cutting bridge 4 in the x-direction and / or a rotation of the cutting wheel carrier 9 about the carrier's rotation axis in a conventional manner. However, mainly scoring lines 6 are created that are parallel to the x-direction or parallel to the y-direction. A scoring line 6 parallel to the y-direction is created by moving the cutting head 5 in the y-direction along the cutting bridge 4 (=cross-section). A scoring line 6 parallel to the x-direction is created by moving the cutting bridge 4 and / or the glass pane 2 in the x-direction (=longitudinal cut). During scoring, the cutting wheel 8 rolls along the glass pane surface 2a in a manner known per se and scores it.cuts these.

[0030] The glass pane 2 can also be rotated about an axis of rotation parallel to the z-direction during the scoring process in order to produce arcuate scoring lines.

[0031] In addition, the cutting device 1 preferably has means for breaking the glass pane 2 after scoring, preferably in a conventional manner. For example, the cutting device 1 has a breaking roller, which is also mounted on the cutting head 5. However, it can also be provided, for example, with a breaking bar with hold-down devices. In this case, the cutting device 1 serves not only for cutting but also for dividing glass panes 2.

[0032] If the cutting device 1 is a laminated glass cutting device, it also has a separating means for separating the laminated glass film. For example, this is a mechanical separating means, e.g., a cutting blade. Alternatively or additionally, the separation can also be performed thermally. It can also be performed using a laser. The separation can be performed, for example, after the breaking process or even before the breaking process using the laser. The laminated glass film can also be weakened using a laser before the breaking process.

[0033] In addition, the cutting bridge 4 in a laminated glass cutting device is designed as a double cutting bridge. The double cutting bridge has an upper and a lower cutting bridge section. The upper cutting bridge section spans the support table 3 on the upper side, while the lower cutting bridge section spans the support table 3 on the lower side. A cutting head 5 is mounted on both the upper and lower cutting bridge sections. The two cutting heads 5 are mounted on the respective cutting bridge section, in particular so as to be movable synchronously in the y-direction. In addition, the support table 3 is divided or slotted in a conventional manner in the area of ​​the double cutting bridge, or has a gap or slot extending in the y-direction. The gap serves to accommodate the cutting, breaking, and separating means.

[0034] Within the scope of the invention, it has now surprisingly been discovered that it is possible to monitor and improve the fracture edge quality if the scribing process is monitored and analyzed using acoustic emission analysis. This is preferably done online or in real time.

[0035] Acoustic emission testing (AETT) takes advantage of the fact that sudden changes in the structure of a material, triggered by chemical or thermal processes or under mechanical stress, lead to the emission of sound. Recording and analyzing these acoustic emission signals can then be used to characterize, for example, the manufacturing quality or the damage status of material composites, components, and parts. Sensors, such as piezoelectric sensors, are used to detect the acoustic emission signals.

[0036] These can convert the transient deflections in the material as the wave passes through it into a voltage signal, which can be digitized and further evaluated using data analysis software.

[0037] However, acoustic emission analysis also has its limitations. For example, the presence of interference sources (e.g., machine noise, electromagnetic noise, or crack edge friction) can be problematic.

[0038] In the present case, for example, there are numerous mechanical joints in the interface chain—cutting wheel 8—wheel axle 8a—cutting wheel carrier 9—wheel carrier mount—cutting head 5—which dampen high-frequency vibrations in particular. The very small component size also hinders the installation of sensors in the immediate vicinity of the cutting wheel 8. Furthermore, the scoring process generates high relevant frequency ranges. The rotational frequency of a cutting wheel 8 at a cutting speed of 1 m / s and a wheel diameter of 5.6 mm is, for example, 57 Hz. And the tooth meshing frequency of a microstructured cutting wheel with a tooth pitch of 50 µm and a cutting speed of 3 m / s is, for example, 60 kHz.

[0039] Within the scope of the invention, it was nevertheless discovered that it is possible to monitor the scoring process or the quality of the scoring process, and to detect and characterize certain disruptive phenomena, by recording and analyzing the acoustic emission signals of a scoring process. In particular, it was discovered that some cutting parameters, e.g., the cutting pressure and / or the wear condition of the cutting wheel 8 and / or the type of wear of the cutting wheel 8 and / or the cutting speed, can be monitored, adjusted, controlled, and / or regulated using acoustic emission analysis.

[0040] For this purpose, the acoustic waves (= sound signals or sound emission signals) generated during the scratching process are measured using at least one AE sensor or

[0041] Sound pressure sensor or microphone 13 ( Fig. 2), in particular a piezoelectric sensor, is detected, converted into electrical analog time signals, digitized, and recorded. The sound pressure sensor 13 has a circuit board 18 in a conventional manner.

[0042] For digitization, the sound pressure sensor 13, in particular the circuit board 18, preferably has a digitization device, e.g., an analog-to-digital converter. Early digitization can reduce or prevent interference during signal transmission.

[0043] Preferably, the sound pressure sensor 13 is connected, in particular via cables 19, to another signal converter 20. The signal is recorded in an evaluation device 21. After the signal is recorded, a signal analysis is performed, preferably online, also by means of the evaluation device 21. The evaluation device 21 is preferably part of a control device 22 of the cutting device 1.

[0044] The signal analysis can be carried out in the time domain and / or in the frequency domain (frequency analysis).

[0045] In particular, during the signal analysis and evaluation, at least one actual cutting wheel signal is generated and compared with at least one previously determined or defined target characteristic of the cutting wheel signal. This can be done automatically using software of the evaluation device 21 and / or by an operator. The cutting wheel signal with the target characteristic and the corresponding actual cutting wheel signal can each be an electrical, analog, or digital signal generated based on the measured acoustic emission signals. For example, it can be an analog or digital time signal or a frequency spectrum.

[0046] The target characteristics of the cutting wheel signal are determined, for example, based on empirical values. For example, tables already exist in the field that specify specific cutting wheel types for various glass panes to be cut, along with the associated cutting parameters, such as cutting speed and / or cutting pressure, which should be used / set to achieve good break edge quality. Using these specified settings, scoring processes can then be performed, and the corresponding cutting wheel signal with the target characteristics can be recorded. If the break edge quality is not sufficient despite the specifications, the cutting parameters must of course be adjusted accordingly to determine the target characteristics of the cutting wheel signal. Thus, characteristics of the recorded signals are correlated with the assessment of the edge quality.

[0047] Cutting wheel types differ, for example, in their grinding type and / or cutting angle and / or diameter, which will be discussed in more detail below. The different cutting wheel types each exhibit a characteristic cutting wheel signal. For example, the cutting wheel signals of microstructured and non-structured cutting wheels differ. However, even different non-microstructured cutting wheels exhibit slightly different signal shapes a priori.

[0048] Through signal analysis and evaluation, features are extracted and conclusions are drawn based on collected data and experience. Artificial intelligence methods, particularly machine learning, can also be used. For example, causal relationships and patterns can be recognized / derived based on larger amounts of data. Predictions and conclusions can be calculated using a prediction algorithm.

[0049] For example, conclusions can be drawn from collected data regarding the state and / or type of wear in correlation with the determined edge quality. Limit values ​​can also be defined for the target characteristics of the cutting wheel signal and predictions can be made regarding the service life of the cutting wheel. In particular, it can be determined that the cutting wheel must be replaced once it reaches a certain degree of wear, which correlates with a specific characteristic of the cutting wheel signal. It can also be monitored whether the cutting wheel is subject to a normal degree of wear. The target characteristics can therefore also depend on the degree and / or type of wear.

[0050] If the signal analysis is based on frequency analysis, the recorded digital time signal is transformed into the frequency domain, preferably using a Fourier transformation. The resulting actual frequency spectra are then further analyzed and evaluated, preferably compared with target characteristics. This can be done automatically using software in the evaluation device and / or by an operator. For example, as described above, a cutting wheel has a target characteristic of the frequency spectrum, which is then compared with the measured actual frequency spectrum. In particular, individual frequency ranges of the target characteristic and the actual frequency spectrum are compared with each other.

[0051] As is well known, continuous, aperiodic signals are decomposed into a continuous frequency spectrum using Fourier transformation. Thus, a digital signal can be decomposed into its frequency components and subsequently analyzed.

[0052] As already explained, monitoring or acoustic emission analysis is preferably carried out online.

[0053] According to the invention, at least, and possibly exclusively, acoustic emission signals in the inaudible ultrasonic range (16 kHz to 1.6 GHz) are recorded and / or analyzed. Within the scope of the invention, it was determined that a more precise evaluation and monitoring of the scratching process is possible in this frequency range.

[0054] Furthermore, the sound pressure sensor 13 is arranged near the cutting wheel 8 and attached to the cutting head 5. This is shown schematically in Figure 2The sound pressure sensor 13 thus serves to detect near-emission sound signals or the nearby sound emission signals.

[0055] The sound pressure sensor 13 is also preferably directed toward the cutting wheel 8 (directional microphone). For this purpose, the sound pressure sensor 13 has a horn 23. This significantly reduces the detection of background noise and improves the measurement results. In particular, ambient noise, generated, for example, by the motion axes and drives, is suppressed, and broadband cutting noise may also be amplified.

[0056] In the context of the invention, it was surprisingly discovered that, in particular due to the sound horn 23, glass panes 2 with different thicknesses can be cut without the position of the sound pressure sensor 13 relative to the support table 3 having to be changed in the z-direction. This is because, with glass panes 2 of different thicknesses, the glass pane surface 2a is necessarily displaced in the z-direction relative to the sound pressure sensor 13 (see Figure 10 ). This means that the sound source (the point at which the cutting wheel 8 rests against the glass surface 2a) also shifts relative to the sound pressure sensor 13. The direction of the shift is in Figure 10 and 11marked as a double arrow. As a result, the sound source is no longer positioned exactly centrally or centered in relation to the sound horn 23, and the travel path of the individual sound waves is no longer uniform. For example, the glass thickness of the glass panes 2 to be cut is between 3 and 19 mm. This corresponds approximately to a displacement of 5.6 mm. Nevertheless, it was determined within the scope of the invention that the signal quality is still sufficiently good, since the travel paths change significantly less even with a displacement of 5.6 mm. Half the sound wavelength that could lead to local cancellation of the sound is 1.9 mm at 90 kHz, for example.

[0057] Furthermore, the sound pressure sensor 13 preferably has a protective membrane 25 arranged in front of or at the open end of the sound horn 23. In particular, the protective membrane 25 covers the sound horn 23 at its open end facing the cutting wheel 8.

[0058] The protective membrane 25 protects the sound horn 23 from the penetration of solid and liquid dirt particles, e.g. from the penetration of cutting oil. They are therefore also particularly suitable as splash protection. The protective membrane 25 is therefore, on the one hand, sound-permeable but, on the other hand, is designed in such a way that the solid and liquid dirt particles cannot penetrate through. The protective membrane 25 is preferably made of a fabric. The fabric is also preferably made of metal, more preferably stainless steel. In addition, the fabric preferably has a mesh size (= distance between the fabric threads) of 40 to 500 µm, preferably 55 to 100 µm. The protective membrane 25 should be as thin as possible in order to dampen the short sound waves as little as possible. The fabric threads preferably have a thread thickness (diameter) of 30 to 50 µm.

[0059] Alternatively, the mesh can also be made of plastic. However, since metal has lower damping properties than plastic, metal, especially stainless steel, is preferred.

[0060] Furthermore, the sound funnel 23 preferably has at least one ventilation opening 26 (schematically shown in Figure 11 shown), which is also intended to prevent contamination of the sound pressure sensor 13. The ventilation opening 26 extends continuously from the outside to the inside through a funnel wall of the sound horn 23. An airflow is created through the ventilation opening 26, which ensures the continuous cleaning and exchange of air in the sound horn 23. As a result, dirt particles are continuously removed from the interior of the sound horn 23. It was found that despite the airflow, particularly due to the high frequencies of the signals, the signal quality is not impaired.

[0061] To suppress interference signals, in an advantageous development of the invention, at least one additional sound pressure sensor 14 can be provided, which is arranged further away from the cutting wheel 8 than the first sound pressure sensor 13. The additional sound pressure sensor 14 thus serves for the additional detection of distant sound emission signals. It detects a larger proportion of ambient noise or interference noise. For example, the interference noise is continuous or short-term background noise, e.g., in the hall in which the cutting device 1 according to the invention is arranged. Preferably, the additional sound pressure sensor 14 is also arranged on the cutting head 5.

[0062] Preferably, the additional sound pressure sensor 14 is also designed analogously to the first sound pressure sensor 13 and thus also has a circuit board 24, preferably with a digitizing device. However, it is preferably not designed as a directional microphone.

[0063] Preferably, the further sound pressure sensor 14 is also connected, in particular by means of cables 19, to the further signal converter 20 and via this to the evaluation device 21.

[0064] The length of the cables 19 is preferably ≤ 0.5 m each.

[0065] Preferably, the first and the further sound pressure sensor 13;14 as well as the signal converter 20 and the evaluation device 21 and the control device 22 are arranged in a housing 27 attached to the cutting head 5 ( Fig. 10 ).

[0066] Using the evaluation unit 21, a signal analysis of the sound emission signals from the two sound pressure sensors 13; 14 is then preferably performed, and the results are calculated. This means that the background noise is removed from the near-field sound emission signals using background noise suppression. The result is a characteristic cutting wheel signal calculated based on at least two sound pressure sensors 13; 14. The cutting wheel signal, in turn, has a particularly high correlation with good or poor cutting behavior.

[0067] To reduce computational effort, it is also advantageous to determine whether the near-range sound emission signals are evaluated or not based solely on the far-range sound emission signals. If a threshold value for the background noise detected by analyzing the far-range sound emission signals is exceeded, the near-range sound emission signals are excluded from evaluation to prevent the background noise from being evaluated, which could otherwise lead to misinterpretation. In this case, the far-range sound emission signals essentially serve as control signals.

[0068] For further optimization and improved monitoring accuracy, it is also possible to not record and / or evaluate the acoustic emission signals during the entire cutting process. For example, only the acoustic emission signals that occur when noise sources are far away from the cutting wheel 8 are detected and / or evaluated. This is the case, for example, with longitudinal cuts when the linear guides and drives are far away.

[0069] The advantages of the invention are explained in more detail below using exemplary embodiments: In the exemplary embodiments, the sound emission signals from each scratching process were detected, recorded, and evaluated using two sound pressure sensors (near-field sound pressure sensor and far-field sound pressure sensor). Both longitudinal and cross-sectional cuts were performed. The longitudinal cuts were performed by moving the cutting bridge in the x-direction. The cross-sectional cuts were performed by moving the cutting head along the cutting bridge in the y-direction. A Rapidline cutting device from the applicant was used.

[0070] Identical microphones were used to record the near-field sound emission signals and the far-field sound emission signals. The signals were digitized directly on the circuit board on which the microphones were located. Both signals were combined in a converter box attached to the cutting head and transmitted in real time to the evaluation device via a cable. The near-field sound pressure sensor was located at a distance of approximately 60 mm diagonally in front of the cutting wheel and 40 mm above the surface of the glass pane. A turned metal part approximately 30 mm long and with a diameter of approximately 20 mm was positioned in front of the near-field sound pressure sensor. This part was machined on the inside and aligned with the contact area between the cutting wheel and the surface of the glass pane so that the near-field sound signal generated in the contact area between the cutting wheel and the glass pane hit the near-field sound pressure sensor as in-phase as possible, thus achieving a directivity or directional effect.Amplification occurred. The remote sound pressure sensor was positioned approximately 200 mm from the contact area between the cutting wheel and the glass pane surface and above the cutting head, so that it primarily recorded ambient noise. The remote sound pressure sensor was not equipped with a directional pickup pattern.

[0071] During signal analysis, the two sensor signals were not combined to suppress background noise. Rather, the near-field sound emission signals were not evaluated if background noise exceeded a certain level.

[0072] The transformation of the digital time signals into the frequency domain in the Figures 4 to 7The frequency spectra shown were each analyzed using online FFT (Fast Fourier Transform). A window with a width of 2 ms was used, which was shifted in steps of 1 ms. The Hanning window was used as the window function. This resulted in a window width of 375 samples and a time resolution of 1 ms with an overlap of 50%. The frequency resolution was 501.3 Hz. The signal was also recorded at a sampling rate of 187.5 kHz. This is known to allow frequencies up to half this frequency, i.e., 93.75 kHz, to be resolved (https: / / de.wikipedia.org / wiki / Nyquist-Shannon-Abtasttheorem).

[0073] In Figure 4A frequency spectrum of a longitudinal cut made with a microstructured cutting wheel is shown. The acceleration of the cutting head was 4 m / s 2< , the constant cutting speed was 100 m / min, and deceleration was 6 m / s 2< with a tooth pitch on the cutting wheel of 50 µm.

[0074] Clearly visible in Figure 4 is envelope 15, which initially rises during acceleration, then runs in a straight line at constant cutting speed, and falls again during deceleration. It can also be seen that envelope 15 has a wavy shape. One period of the waves corresponds to one revolution of the cutting wheel. Harmonics 16 are shown above envelope 15.

[0075] From the waviness, the average meshing frequency in the constant cutting speed range was determined to be 32.8 kHz. This corresponds to a constant cutting speed of 98.4 m / minute. The target meshing frequency was 33.3 kHz (calculated from the target constant cutting speed / tooth pitch). The slip was thus 1.6%.

[0076] From the waviness, it was also calculated that the meshing frequency fluctuates (32.3 kHz - 33.3 kHz). This means that the cutting wheel appears to stall at certain angular positions once per revolution, i.e., at approximately 94.9 Hz. The waviness thus indicates an asymmetry of the cutting wheel.

[0077] In the Figures 5a-cFrequency spectra of a longitudinal section are shown. Cutting was performed using a non-microstructured, conventional cutting wheel with a wheel diameter of 4.1 mm and a cutting angle of 145°. In addition, a single sheet of 4 mm glass thickness was cut. The cutting parameter, cutting pressure, was varied ( Figure 5a 0.3 bar, Figure 5b 0.7 bar, Figure 5c 1,1 bear).

[0078] The increase in cutting pressure is clearly evident in a more pronounced high-frequency range. In particular, an increase in cutting pressure leads to a significant increase in broadband cutting noise across the entire frequency range up to 50 kHz.

[0079] In Figure 6a and 6b are the frequency spectra of a cross section ( Figure 6a ) and a longitudinal section ( Fig. 6b ), each cut with the same non-microstructured cutting wheel.

[0080] It can be seen that the frequency spectrum of the longitudinal section exhibits less interference. Strong broadband noise can be detected in three frequency ranges in the frequency spectrum of the cross-section. This is due to the fact that interference sources, especially linear guides, are further away in the longitudinal section.

[0081] Furthermore, conventional cutting wheels with different wear levels were tested in further examples. The corresponding frequency spectra are shown in the Figures 7a-c shown. Figure 7a shows the frequency spectrum of a new cutting wheel. Figure 7b shows the frequency spectrum of a cutting wheel with medium wear and Figure 7c shows the frequency spectrum of a worn cutting wheel with a notch.

[0082] It can be seen that with increasing wear, the "events" or outliers become more frequent. These are recorded in the frequency spectra of the Figure 7b and7c by recurring spikes or outliers 17 in high frequency ranges. In Figure 7c the spikes or outliers 17 are significantly increased and more frequent and, above all, more regular than in Figure 7b . The cutting wheel made of Figure 7c You can also read the time interval between two deflections 17. This corresponds approximately to the circumference of the cutting wheel, 17.6 mm. The cutting wheel obviously has a dominant defect that causes a deflection 17 once per revolution.

[0083] Wear can therefore not only be detected in principle, but different degrees and types of wear can also be distinguished.

[0084] The cutting wheels of the Figure 7b and 7c would have the cutting wheel made of Figure 7b can continue to be used in real operation and cutting wheels made of Figure 7cshould have been replaced earlier to ensure good fracture edge quality.

[0085] In Figure 8a The frequency spectrum of the conventional cutting wheel with a medium degree of wear is shown again. The cutting wheel had a diameter of 5.6 mm and a cutting angle of 152°. Figure 8b shows a deflection of 17 in the time signal. Figure 8b It becomes clear that the deflections are a real signal and not an artifact. It is a very broadband signal with a duration of approximately 0.2 ms.

[0086] Figure 8bmakes it clear that the sampling frequency could ideally be even higher, because each peak is resolved by only one sample. For signal analysis, as described above, 375 samples weighted by the Hanning window were subjected to a Fourier transformation every 1 ms, i.e., offset by 187.5 samples. This achieves a time resolution of 1 ms and a frequency resolution of 501.3 Hz in the frequency spectrum. With fewer samples in the Hanning window, the time resolution would be better at the expense of the frequency resolution.

[0087] In Figure 4For example, the stalling of the cutting wheel would be better resolved in terms of time, but less so in the frequency domain. At a typical cutting speed of 100 meters per minute and a cutting wheel diameter of 5.6 mm, the time for one revolution is 0.0105 seconds. If 375 sample values ​​are subjected to frequency analysis at one-millisecond intervals, this results in an average of 10.5 frequency analyses per revolution of the cutting wheel.

[0088] Figure 8bHowever, this also highlights the limitations of signal analysis across the frequency range. The event shown here has a temporal length of approximately 0.3 milliseconds, which, at a sampling frequency of 187.5 kHz, represents 56 samples. With a Hanning window width of 375 samples, the duration of the "event" represents only a certain proportion. Time periods without an "event" therefore have greater weighting. The "event" becomes clearer in terms of time resolution due to a narrower Hanning window width, which, however, comes at the expense of frequency resolution.

[0089] Figures 9a-c show time signals of the cutting wheels according to the Figures 7a-cFor the analysis, the absolute value of the difference between the consecutive audio time signals was averaged using a moving average of 50 values. The result is a one-dimensional curve that is easy to evaluate and can also be easily analyzed using algorithms. The differences in the degree of wear are also clearly visible in this representation. In particular, the peaks or outliers 17 are clearly visible.

[0090] Figures 9a-c demonstrate the advantages of suitable signal analysis in the time domain. These include, in addition to the likely lower computing power required (depending on the method chosen), better time resolution. This cancels out the very short-term "events" as in Figure 8b The project, which was carried out within the Figures 9a-c The very simple algorithm used requires little computing power.

[0091] The embodiments clearly show that the cutting method according to the invention and the cutting device according to the invention enable machine operators to monitor and specifically adjust important parameters of the cutting device and / or the cutting method, such as cutting speed and / or cutting pressure, and to detect the degree and / or type of wear of the cutting wheel or machine errors.

[0092] This is because a clear influence of the cutting parameters on the time signal and / or frequency spectrum can be seen, which is similar across different cutting wheels. A change in the cutting parameters can be seen in the time signal and / or frequency spectrum.

[0093] The analysis of signals in the frequency domain (frequency analysis) is therefore an alternative to evaluation in the time domain. Depending on the application, signals are evaluated either in the time or frequency domain. Evaluation in the frequency domain is advantageous, for example, with conventional cutting wheels, since the frequency spectrum of unworn cutting wheels is quite constant over time. Furthermore, an increase in cutting pressure causes higher frequency components to appear, even with a new cutting wheel. With increasing wear, more "events" occur that cause a short-term change in the frequency spectrum, particularly in the sense that higher frequency components increase. However, these are even more clearly visible in the time domain.

[0094] Furthermore, for microstructured cutting wheels, frequency analysis can be used to determine the theoretical meshing frequency based on the known tooth pitch and cutting speed. Since the actual frequency is somewhat lower, this indicates slippage. If a slight ripple in the frequency response is noticeable once per rotation of the cutting wheel at a constant cutting speed, this indicates a slight stalling of the cutting wheel at certain angular positions.

[0095] Finally, as part of the signal analysis and evaluation, based on the physical understanding of the cutting processes and a priori knowledge, for example, about the cutting wheel used and the specific process parameters (diameter, microstructure, and if so, how, cutting speed, etc.), a characterization or classification of the recorded signals is performed to determine whether the cutting parameters are correctly set, whether there are component defects, and / or whether the cutting wheel is worn. For example, 1 / revolution effects are detected, which indicate component defects or a defect at a specific point on the cutting edges.

[0096] Monitoring the scribing process is particularly advantageous for thin glass / display glass, but also for conventional glass thicknesses for demanding automotive applications and large-format glass facades. In the display glass sector, microstructured cutting wheels are used exclusively, except for smartphones (laser beam cutting).

[0097] It is also within the scope of the invention that the breaking and, in the case of a laminated glass pane, the separation of the laminated glass film, do not take place in the cutting device, but in separate devices. Furthermore, the breaking and separation can also be performed manually.

[0098] In addition, the glass pane can be divided in such a way that - depending on the cutting pattern - the entire glass pane is first cut or scored and then broken and, if necessary, separated. Or a scoring line is first created (in the case of laminated glass panes, on both surfaces of the glass pane), and immediately afterwards the glass pane is broken along the scoring line and, if necessary, separated. The separated glass panes are then scored and divided again. This is arbitrary. The glass pane can also be rotated on the support table to create the next scoring line and / or separated glass panes can be returned and then cut again. This is described, for example, in DE 10 2018 217 660.4 for laminated glass panes.

[0099] Furthermore, as already explained, other components made of glass or ceramic can also be scored using the cutting method and / or the cutting device according to the invention. And plate-shaped components can, in principle, be divided into individual component blanks in a similar way to the process described for glass panes.

[0100] Furthermore, it is within the scope of the invention to use a bowl-shaped sound reflector with a parabolic mirror or a sound tube instead of the sound horn 23 to focus the sound onto the microphone point of the sound pressure sensor 13. However, the sound horn 23 is preferred because it ensures better signal quality, especially with different glass thicknesses.

Claims

1. Cutting method for the superficial scoring of a, preferably plate-shaped, component made of glass or ceramic, preferably a glass sheet (2), wherein a scoring line (6) is introduced by means of a cutting wheel (8) into at least one component surface, preferably into a glass sheet surface (2a) of the glass sheet (2), characterized in that the scoring process is monitored by means of acoustic emission analysis, wherein in the context of the acoustic emission analysis, a signal analysis and evaluation of digitized acoustic emission signals detected during the scoring process is carried out, the signal analysis and evaluation being carried out in the time domain and / or in the frequency domain.

2. Cutting method according to claim 1, characterized in that for determining the degree of wear of the cutting wheel (8) and / or the type of wear of the cutting wheel (8), the acoustic emission signals are monitored in relation to peaks (17) which represent a brief change in the frequency spectrum or the time signal such that higher-frequency frequency components increase.

3. Cutting method according to any of the preceding claims, characterized by a) sensorial detecting acoustic emission signals from the cutting wheel (8) during the scoring process, b) converting and digitizing the acoustic emission signals into at least one electrical time signal, preferably a voltage signal, and recording the time signal, c) analyzing and evaluating the time signal.

4. Cutting method according to any of the preceding claims, characterized in that during analyzing and evaluating the signal, at least one actual cutting wheel signal is compared, preferably automatically, with at least one nominal characteristic of the cutting wheel signal, wherein preferably, a) the degree of wear of the cutting wheel (8) and / or the type of wear of the cutting wheel (8) are monitored and the degree of wear of the cutting wheel (8) and / or the type of wear of the cutting wheel (8) are determined on the basis of peaks (17) in high frequency domains in the actual cutting wheel signal, the actual cutting wheel signal being an electrical, digital time signal generated on the basis of the measured acoustic emission signals or a frequency spectrum, and / or b) the cutting wheel signal having the target characteristic and the corresponding actual cutting wheel signal are each an electrical digital time signal or a frequency spectrum generated on the basis of the measured acoustic emission signals.

5. Cutting method according to any of the preceding claims, characterized in that a) the time signal is transformed into at least one frequency spectrum during the signal analysis and the transformation preferably is performed by means of Fourier transform, preferably Fast Fourier Transform, and / or b) cutting parameters determined by means of the acoustic emission analysis, preferably the cutting pressure and / or the degree of wear of the cutting wheel (8) and / or the type of wear of the cutting wheel (8) and / or the tooth meshing frequency of a microstructured cutting wheel and / or the cutting speed are monitored, and / or c) at least acoustic emission signals in the inaudible ultrasonic range from 16 kHz to 1.6 GHz are recorded and / or analyzed, and / or d) the acoustic emission signals are detected when longitudinal cuts are made, which are generated by movement of a cutting bridge (4) and / or the glass sheet (2) and without movement of a cutting head (5) along the cutting bridge (4).

6. Cutting method according to any one of the preceding claims, characterized in that the acoustic emission signals of the cutting wheel (8) are detected and analyzed during the scoring process, wherein preferably at least one first sound pressure sensor (13) arranged in the vicinity of the cutting wheel (8) is used for detecting near acoustic emission signals of the cutting wheel (8), wherein preferably a) a sound pressure sensor (13) directed towards the cutting wheel (8) is used with means for focusing the sound, preferably with a sound funnel (23) or a sound reflector with parabolic mirror, and / or b) the sound pressure sensor (13) is protected against contamination, in particular by cutting oil, preferably by means of splash protection agents.

7. Cutting method according to claim 6, characterized in that at least one further sound pressure sensor (14) is used for detecting remote acoustic emission signals, which is arranged further away from the cutting wheel (8) than the first sound pressure sensor (13), wherein preferably a) the near acoustic emission signals and the remote acoustic emission signals are set off against each other in such a way that noise is calculated out of the near acoustic emission signals by means of noise suppression, or b) based on the level of the noise detected by means of the remote acoustic emission signals, it is determined whether the near acoustic emission signals are evaluated.

8. Cutting method according to any one of the preceding claims, characterized in that a) methods of artificial intelligence, in particular of machine learning, are used for the signal analysis and evaluation, and / or b) the acoustic emission analysis is performed online or in real time.

9. Method for dividing of, preferably plate-shaped, components made of glass or ceramics, preferably glass sheets (2), wherein the component is scored at least on one of its two component surfaces and then broken along the scoring line (6), characterized in that the scoring is performed according to the cutting method according to any one of the preceding claims.

10. Method according to claim 9, characterized in that a) by means of monitoring the scoring process, the breaking edge quality, in particular the breaking edge strength, of breaking edges of the divided components is monitored, and / or b) the target characteristic of the cutting wheel signal is determined by setting the cutting parameters as a function of a desired breaking edge quality, in particular a breaking edge strength, and / or c) the glass sheet (2a) is a laminated glass sheet consisting of two glass sheets joined together by a laminated glass film, wherein the laminated glass sheet is scored on both glass sheet surfaces (2a), the two glass sheets are successively broken along the scoring line (6), and the laminated glass film is separated.

11. Cutting device (1) for introducing scoring lines (6) into at least one component surface of a plate-shaped component made of glass or ceramic, preferably into a glass sheet surface (2a) of a glass sheet (2), in particular cutting device (1) for carrying out the cutting method according to one of the claims 1 to 8 and / or the method according to one of the claims 9 to10, comprising: a) a support table (3) for receiving the component to be cut, b) a cutting bridge (4) spanning the support table (3), c) at least one cutting head (5) having a cutting wheel (8) mounted to be movable back and forth along the cutting bridge (4) in a y-direction and, d) drive means for moving the cutting head (5) along the cutting bridge (4), characterized in that the cutting device (1) has means for monitoring the scoring process by means of acoustic emission analysis, a digitizing device for digitizing the acoustic emission signals, and an evaluation device (21) for signal analysis and evaluation.

12. Cutting device (1) according to claim 11, characterized in that the cutting head (5) comprises at least a first sound pressure sensor (13) arranged in the vicinity of the cutting wheel (8) for detecting near acoustic emission signals of the cutting wheel (8), wherein preferably the cutting device (1) has means for protecting the first sound pressure sensor (13) from contamination, in particular splash protection means for protecting the first sound pressure sensor (13) from contamination by cutting oil.

13. Cutting device (1) according to claim 12, characterized in that the first sound pressure sensor (13) is directed towards the cutting wheel (8) and preferably has means for focusing the sound, preferably a sound funnel (23), the first sound pressure sensor (13) preferably having means for preventing dirt from entering the sound funnel (23), wherein preferably a) the sound pressure sensor (13) has a sound-permeable protective membrane (25) which covers the open end of the sound funnel (23) facing the cutting wheel (8) and serves to prevent dirt from entering the sound funnel (23), wherein preferably the protective membrane (25) consists of a fabric which is preferably made of metal, preferably of stainless steel, or of plastic, and / or which preferably has a mesh size of 40 to 500 µm, particularly preferably 55 to 100 µm, and / or which has fabric threads with a thread thickness of 30 to 50 µm, and / or b) the sound funnel (23) has at least one ventilation opening (26) which extends continuously from the outside to the inside through a funnel wall of the sound funnel (23).

14. Cutting device (1) according to one of the claims 12 to 13, characterized in that the cutting device (1) has at least one further sound pressure sensor (14) for detecting remote acoustic emission signals, which is arranged further away from the cutting wheel (8) than the first sound pressure sensor (13), the further sound pressure sensor (14) preferably being attached to the cutting head (5).

15. Cutting device (1) according to one of the claims 11 to 14, characterized in that a) the cutting bridge (4) is stationary or can be moved back and forth over the component in a horizontal x-direction perpendicular to the y-direction, and / or b) the cutting device (1) has means for breaking the component along the scoring line (6), and / or c) the cutting device (1) serves for cutting laminated glass sheets and the cutting bridge (4) is designed as a double cutting bridge which has an upper and a lower cutting bridge part, the upper cutting bridge part spanning the support table (3) on the upper side and the lower cutting bridge part spanning the support table (3) on the lower side, and a cutting head (5) being mounted both on the upper and on the lower cutting bridge part, wherein preferably the cutting device (1) has separating means for separating a laminated glass film, and / or d) means for monitoring the scoring process by means of acoustic emission analysis are configured to monitor the acoustic emission signals in relation to peaks (17) in high frequency domains for determining the degree of wear of the cutting wheel (8) and / or the type of wear of the cutting wheel (8).