Method and corrugator

Abstract

The invention relates to a method, wherein a plurality of paper webs (4, 6, 8) are guided in a feed direction (F) through a corrugator (2), wherein a measuring device (22) is integrated in the corrugator (2), having a light source (24) and an SWIR camera system (26), wherein the SWIR camera system (26) has a field of view (28) which extends transversely to the feed direction (F), wherein one of the webs (4, 6, 8) is guided through the field of view (28) and an optical absorption measurement is repeatedly carried out in a number of spectral bands (32) using the light source (24) and the SWIR camera system (26) and measurement data (34) is thus generated, wherein a humidity (42) of the web (4, 6, 8) is calculated from the measurement data (34) in a spatially resolved manner. The invention also specifies a corresponding corrugator (2).

Description

[0001] Description

[0002] Process and corrugator

[0003] The invention relates to methods, in particular for determining the moisture content of a web within a corrugated board plant, as well as to a corresponding corrugated board plant.

[0004] A corrugated board line generally has a wet end and a subsequent dry end. In the wet end, several paper webs are combined to form a corrugated board web. This regularly results in the formation of intermediate products, such as half-waves, which are made up of a flat and a corrugated paper web. The corrugated board web is then made up into individual corrugated board sheets (short: sheets) in the dry end. Along with its temperature, the moisture content of the paper is one of the most important parameters in the entire manufacturing process. Humidity significantly influences the strength, dimensional stability, formability, printability, penetration behavior and bonding properties of the paper webs and the corrugated board webs and sheets produced from them, as well as the intermediate products manufactured for these.In the operation of a corrugated board plant, considerable technical effort is regularly expended to precisely adjust moisture levels throughout the production process. Therefore, it is desirable to be able to measure moisture levels as cost-effectively, simply, and accurately as possible.

[0005] As part of the operation of a corrugated board plant, inline measurement of the absolute moisture content of the paper is particularly advantageous, for example, for controlling the production process within the corrugated board plant and for reducing the process energy required for this. In a corrugated board plant, very large quantities of paper are processed per unit of time across a wide width and at high web speeds in several different process steps. Therefore, real-time moisture measurement over as large an area as possible is desirable, especially in a corrugated board plant.

[0006] In principle, a variety of different methods are conceivable for measuring moisture content. Using a weighing method, the absolute weight of paper or corrugated board can be measured using samples. The samples can then be overdried in an oven under defined conditions to calculate the absolute moisture content from the weight difference. This measurement method is only destructive and is time-consuming. Measuring the electrical resistance or electrical conductivity of the material is also possible. However, the measured values ​​in paper exhibit a high degree of scatter, and contact with the paper with a measuring probe is necessary. Capacitive, radar, and microwave measurement methods can utilize the dielectric differences between paper, air, and water.For example, the influence of humidity on the measuring field of a capacitor or the influence of permittivity on the propagation of electromagnetic waves is determined. These measuring methods can be sensitive with regard to the distance of the sensor used from the paper web and must be calibrated separately for each paper quality. Furthermore, only point-by-point measurements can be carried out regularly. Ultrasonic measuring methods can utilize the influence of humidity on the phase velocity of the paper. As humidity increases, the mass of the paper increases and, at the same time, its stiffness decreases, which also causes the phase velocity to decrease. However, these measuring methods are only suitable to a limited extent due to the high orthotropy and inhomogeneity of paper. Thermal measuring methods can use the change in heat capacity to infer the moisture content based on the temperature curve during heat application.These measurement methods are energy-intensive and interfere with the process technology. Spectral measurement methods can utilize the absorption bands of free water in the material. Based on the spectral information obtained with such a measurement method, it is possible to determine the moisture content. However, corresponding sensors, such as spectrometers, only measure at specific points and are relatively expensive.

[0007] US 2017 160 192 A1 describes an optical multi-channel detector unit for measuring the moisture or dryness of paper or cardboard during their production or further processing. The multi-channel detector unit comprises several optical measuring channels made of optical fiber cables, which are configured to transmit the optical infrared radiation received from the paper. Each of a plurality of optical detectors is configured to temporarily receive infrared radiation from an optical measuring channel via an optical filter and convert an intensity of the optical radiation into an electrical signal. A movement mechanism causes the optical infrared radiation to be filtered by the optical filters in order to temporarily measure only in a specific absorption band.

[0008] EP 3 805 733 A1 describes a quality measurement method for a long sheet material. First, an infrared LED light source emitting light at a wavelength required for measuring paper quality is positioned on the opposite side of the long sheet material from an infrared camera. A quantity of transmitted light attenuated due to absorption by a measurement target contained in the long sheet material and a quantity of transmitted reference light with a similar wavelength not absorbed by the target are measured. Furthermore, the weight of the target is measured and determined based on a measurement equation obtained through prior calibration.Examples of measurement targets include cellulose, fillers and water, which are the main components of paper, as well as substances to be coated such as resin, binder and silicon, which are typically measured using an infrared absorption method.

[0009] Against this background, one object of the invention is to determine the moisture content of at least one of the webs processed in a corrugated board plant (i.e., paper web, corrugated board web, intermediate products) as simply, accurately, and cost-effectively as possible over the widest possible width. In particular, it is desirable to be able to determine a geometric distribution of the moisture content. Furthermore, real-time and inline moisture measurement should be possible.

[0010] The object is achieved according to the invention by a method having the features according to claim 1 and by a corrugated board plant having the features according to claim 15. Advantageous embodiments, further developments, and variants are the subject of the dependent claims. The statements in connection with the method also apply mutatis mutandis to the corrugated board plant, and vice versa. If steps of the method are described implicitly or explicitly below, advantageous embodiments for the corrugated board plant result from the fact that it is designed to perform one or more of these steps. For this purpose, the corrugated board plant has, in particular, a correspondingly designed control unit.

[0011] The method is used in particular for operating a corrugated board plant. As part of the operation, the method particularly includes a moisture content determination. In the method, a plurality of paper webs are generally guided through the corrugated board plant in one conveying direction. The webs can each be individual paper webs or webs composed of several such paper webs, specifically a corrugated board web produced by the corrugated board plant, or intermediate products therefor. Typically, several paper webs are each unwound from an unwinder of the corrugated board plant and then assembled in pairs in one or more single facers to form the corresponding number of half-waves. One or more half-waves are then assembled with another paper web (laminating web) in a double facer to form the corrugated board web. The double facer marks in particular one end of a wet end of the corrugated board plant.This is followed by a dry end of the corrugator, in which the continuous web of corrugated cardboard is cut into individual sheets.

[0012] A measuring device with a light source and a SWIR camera system is integrated into the corrugator. A SWIR camera system is a camera system designed for measurements in the near-infrared range (SWIR = "short wavelength infrared", also known as NIR). This spectral range contains several absorption bands of water, which are used in the moisture determination process.

[0013] The SWIR camera system has a field of view which extends transversely to the conveying direction, i.e. transversely to the webs, in particular in a transverse direction perpendicular to the conveying direction. Preferably, the field of view extends at least over the total width of at least one of the webs; typically, all webs are of a similar width. As part of the method, one of the webs is then guided through the field of view. In this process, an optical absorption measurement (in short: measurement) is repeatedly carried out in a number of spectral bands using the light source and the SWIR camera system, and in the process, i.e. during the absorption measurement, corresponding measurement data is also generated. “A number of” is generally understood to mean “one or more”.

[0014] The SWIR camera system has a pixel array, also known as a sensor array, with multiple pixels, onto which a point (measurement point) on the web is imaged at a given time. The measurement data then contains a measured value for each point on the web. The number of measured values ​​per measurement is defined by the number of pixels. In particular, each pixel is assigned a single spectral band (fixed or time-varying depending on the design), either by using a corresponding filter or by the light from the light source having a correspondingly restricted spectrum. Each pixel is an individual sensor and is made of InGaAs, for example. Since the web is moved in the conveying direction during operation, the absorption in one of the spectral bands along a longitudinal line (parallel to the conveying direction) on the web is measured with repeated measurements from each pixel.This makes it possible, in particular, to use a cost-effective and only one-dimensional sensor array.

[0015] Optical absorption measurement is preferably performed in the dry end of the corrugator and is then used primarily for final inspection of the corrugated board web after the wet end, optionally also for feedback control of the corrugator depending on the moisture content. The corrugated board web is preferably not yet finished, but absorption measurement on the finished corrugated board web, i.e., on the individual sheets, is also advantageous. Alternatively or additionally, absorption measurement directly behind an unwinder for one of the paper webs is also advantageous in order to characterize this web before further processing in the corrugator and then control it appropriately.

[0016] By combining the measurement data at different points in time, i.e., at different longitudinal positions of the web, a surface measurement is now realized. The moisture content of the web is then calculated with spatial resolution from the measurement data. This is done in particular by means of a corresponding evaluation unit of the corrugator. The evaluation unit is, for example, part of the measuring device. How precisely the moisture is calculated from the measurement data is initially of secondary importance. For example, in a separate training process under known conditions, a moisture value is permanently assigned to each measured value. More importantly, a SWIR camera system is used for the absorption measurement, the field of view of which is aligned transversely to the web, so that in conjunction with the conveying of the web in the conveying direction, a two-dimensional moisture profile of the web is determined, i.e.The humidity is calculated as a function of the position on the track, which then also determines a geometric distribution of the humidity.

[0017] The invention is essentially based on an imaging measurement of moisture in the short-wave infrared (SWIR) range. Water has pronounced absorption bands at 975 nm, 1200 nm, and 1470 nm. As the moisture content of the paper increases, the absorption of light with the corresponding wavelength increases, which changes the optical properties, i.e., reflection and transmission, of the web. In particular, these properties are also influenced by the type of paper and the configuration of the web (one or more layers, corrugated / non-corrugated, proportion of glue). Accordingly, a SWIR camera system is used here, so that the measurement is carried out in the SWIR range (approx. 800 nm to 3000 nm). Optionally, the measurement is also carried out at least partially in the VIS range (approx. 400 nm to approx. 800 nm). In particular, an individual spectral band has a bandwidth such that it contains a maximum of one single absorption band of water.Suitable bandwidths are in the range of 1 nm to 50 nm.

[0018] By imaging the reflection or transmission properties of the web in the SWIR range, the local distribution of moisture is measured. More precisely, the absorption of the light from the light source in the paper is measured, and the moisture content at the relevant location is calculated from this. If the measurement is carried out in several spectral bands, the moisture content can be determined not only qualitatively but also quantitatively. The more spectral bands used, the more accurate the measurement and the lower the effort required for initial training of an algorithm used for evaluation. The effort required to calibrate the measuring device for different paper types also decreases with the increasing number of spectral bands, as these paper types can now be conveniently differentiated based on their spectral fingerprint, similar to how water in paper is detected based on its spectral fingerprint.

[0019] In principle, a measurement in just a single spectral band is possible (monochromatic measurement). In this case, two measurements are expediently carried out at different points on the corrugator along the conveying direction, with an intermediate process step, so that a change in moisture due to the process step can then be determined by comparing the two measurements. However, this only provides a qualitative result. Therefore, in the present case, it is preferred that the optical absorption measurement is carried out in several spectral bands (multispectral measurement), and preferably in at least six different spectral bands. This then leads to a quantitative result, i.e. the moisture content of the web is determined absolutely. A measurement in six to eight spectral bands has proven optimal, but other numbers of spectral bands are also suitable.In at least one spectral band, a reference measurement is advantageously carried out outside the absorption spectrum of water.

[0020] Especially in a corrugated board plant, there is a need to determine the moisture content of different paper types as reliably as possible. This can generally be achieved by calibrating the measuring device accordingly for each paper type. As already indicated above, however, such grade-specific calibration is advantageously unnecessary if a sufficient number of spectral bands are used in the absorption measurement, namely the at least six spectral bands described above. In this case, several spectral bands typically lie away from water absorption bands, while several other spectral bands each contain an absorption band of water. Then, even with a calibration for only a few paper types, the moisture content of other paper types for which no calibration is available can be reliably determined.

[0021] The field of view of the SWIR camera system and its orientation relative to the web determine which part of the web is imaged onto the sensor array of the SWIR camera device for measurement. In this case, a longitudinal section, also referred to as a transverse stripe or transverse line, of the web is imaged so that several measuring points adjacent to one another on the web with respect to the conveying direction can be examined simultaneously. Since the web is continuously conveyed through the corrugator, the web is also guided past the SWIR camera system, so that the system scans the web line by line, so to speak, and then, through repeated measurements, the web is measured across its entire surface in two directions: both in the conveying direction and transversely to it.

[0022] Optionally, the SWIR camera system has a mirror arrangement or similar to appropriately align the field of view relative to the track and to redirect reflected / transmitted light from the track to the sensor array accordingly. However, without loss of generality, it is assumed below that the SWIR camera system is directed directly at the track, meaning that light coming from the track reaches the sensor array of the SWIR camera system directly without redirection.

[0023] In a suitable embodiment, the optical absorption measurement is carried out in transmission, i.e. light from the light source enters the web on one side and exits the web on the opposite side, before being guided to the SWIR camera system. Alternatively, a reflection absorption measurement is also suitable, in which light from the light source, which is reflected on one side of the web, is guided to the SWIR camera system. In contrast, transmission measurement is more advantageous because it measures the web across its entire thickness, which is not necessarily the case with reflection measurement. However, for very thick webs that cannot be sufficiently penetrated by light, reflection measurement is more advantageous.

[0024] The absorption measurement is preferably carried out on the continuous web, i.e., before the web is cut into individual sheets using appropriate cross-sections. However, the invention is also fundamentally applicable to the finished corrugated cardboard web, i.e., to the individual sheets.

[0025] The light source suitably emits light diffusely. Halogen lamps or SWIR LEDs are suitable as light sources. In a preferred embodiment, the light source and the SWIR camera system are designed and arranged such that, during the optical absorption measurement, light from the light source (on the way to the SWIR camera system) penetrates the web, namely with a penetration depth of at least 1 mm, preferably at least 2 mm. The penetration depth is in particular the depth at which the intensity of the incident light has decreased by 50%. The penetration depth depends in particular on the grammage of the paper and the intensity of the light source. Single-wall webs or papers up to approximately 380 g / m 2Fully translucent. Due to the penetration of light into the web, depth information about the web can be advantageously obtained, i.e., the absorption measurement does not merely examine the surface of the web, but the web as a three-dimensional object. The measuring device then expediently has a depth analysis unit, which, based on the measurement data, determines an internal condition or internal property of the web, i.e., depth information (for this purpose, a double-sided measurement is also advantageous, as described below). Aside from determining the moisture content, this enables a more precise examination of the interior of the web; accordingly, the internal condition or property is the course of a wave in the web, i.e., a corrugated layer, or a glue content. In this way, defect control is realized, with which, in particular, defects that are not visible from the outside can be detected, e.g.a faulty wave or faulty gluing.

[0026] In a suitable embodiment, depth information (internal nature or internal property of the web) is determined by carrying out several optical absorption measurements, namely with different intensities of the light from the light source. The multiple absorption measurements are expediently carried out at the same position on the web, i.e. at the same point when viewed on the surface. Depending on the intensity, however, the light penetrates the web to different depths, whereby the absorption is measured for different penetration depths. The greater the intensity, the deeper the light penetrates. The depth analysis unit is designed in particular such that it calculates the absorption measurements with different intensities with one another in order to calculate a measurement result for each absorption measurement at a specific penetration depth and thus at a specific depth position in the web.For example, a first absorption measurement is carried out at a first intensity and a second absorption measurement at a second intensity, whereby the second intensity is greater than the first intensity and the light therefore penetrates deeper into the web during the second absorption measurement. The light therefore has a greater penetration depth during the second absorption measurement than during the first absorption measurement, but naturally also has to overcome the first penetration depth. The first absorption measurement is then subtracted from the second absorption measurement (by appropriate calculation, not necessarily by simple subtraction), so that the overall absorption is determined, in particular with pinpoint accuracy, at several depth positions and is determined as a function of the depth position within the web.The depth position indicates how far a point within the web is from its surface (in particular, measured perpendicular to the surface). Preferably, the moisture content of the web is determined as a function of the depth position (a so-called "moisture profile" as depth information), for which the moisture content at a given depth position is then determined, in particular based on the absorption at this depth position. In this way, it is advantageously possible to determine whether the web is moister or less moist in its interior than on the surface. This is particularly well possible with a combination of corresponding absorption measurements on both sides of the web (bilateral measurement).

[0027] Regardless of the use of depth information, defect detection based on the measurement data from the absorption measurement is also generally advantageous. In a practical embodiment, a defect in the web is then detected based on the measurement data. For this purpose, the corrugator has, in particular, a corresponding defect detection unit, which is, for example, part of the measuring device. Defects that are detected include, for example, cracks, defects, faulty corrugation formation, and / or faulty glue application on individual glue lines. In response to a detected defect, the section of the web containing the defect is then, for example, rejected.

[0028] Suitably, the SWIR camera system for optical absorption measurement comprises a multispectral camera or a hyperspectral camera as the camera.

[0029] A multispectral camera is preferred because it is more cost-effective than a hyperspectral camera and less sensitive to ambient temperature fluctuations. However, measurements with a multispectral camera are typically less accurate. For example, a multispectral camera has a mosaic filter to measure in different spectral bands simultaneously. The mosaic filter is designed so that different spectral bands are mapped to different pixels. Instead of a mosaic filter, other technical implementations for a multispectral camera are also possible.

[0030] A hyperspectral camera, such as a push-broom camera, differs from a multispectral camera in that it splits the incoming light rather than filtering it. This makes it possible to record a larger number of spectral bands—more precisely, even a continuous spectrum. The sensitivity of a hyperspectral camera is also typically higher than that of a multispectral camera.

[0031] The wavelength-dependent reflection / transmission of the web also represents a fingerprint of its chemical composition. The range around 1450 nm is particularly sensitive to the amount of water contained in the paper; other spectral ranges outside the water absorption bands are then advantageously used analogously to determine the chemical composition of the paper (pulp content, grammage, recycled content, etc.). This can be implemented specifically with a hyperspectral camera, but for the application described here in a corrugated board plant for measuring different papers, it can also be achieved sufficiently well with a multispectral camera.

[0032] A particularly preferred embodiment is one in which the SWIR camera system has only a single line sensor for each spectral band. The measuring device is therefore particularly compact and cost-effective. To generate area information, use is made of the fact that the web is moved in the conveying direction and is then repeatedly scanned line by line by the line sensor. Using multiple spectral bands results in various options for arranging the respective pixels for each spectral band. For example, multiple line sensors are arranged one behind the other, resulting in a two-dimensional sensor array, but only a single line sensor is used for each spectral band. This makes it possible to achieve a particularly high resolution across the conveying direction, and every point on the web can essentially be examined in every spectral band.Alternatively, the sensor array comprises only a single line in which the line sensors for different spectral bands are combined, e.g., by arranging the pixels for different spectral bands alternately next to each other in a periodic sequence. This results in a particularly compact arrangement. If different light sources with different spectra are used to realize a measurement in different spectral ranges, a single line sensor is sufficient.

[0033] The absorption measurement is also influenced by the atmosphere between the light source and the web on the one hand, and between the web and the sensor array on the other. To reduce this influence, a corresponding calibration measurement is expediently carried out at regular intervals. In a suitable embodiment for this purpose, the measuring device has at least one reflector with which a calibration measurement for the SWIR camera system is repeatedly carried out. “Repeatedly” is understood here specifically to mean that the calibration measurement is carried out again after a certain time, e.g. after 15 minutes. The reflector is, for example, simply a white ceramic surface. The reflector is expediently arranged as close as possible to the web so that the path of the light from the light source to the sensor array during the calibration measurement corresponds as closely as possible to the path of the light during the actual absorption measurement.

[0034] Suitably, the reflector is permanently arranged in the field of view of the SWIR camera system, i.e. both during the calibration measurement and during the absorption measurement. The reflector extends only over part of the total width of the web or the field of view is dimensioned such that it projects laterally beyond the web so that this is not created by the reflector. For example, three reflectors are provided, one in the middle of the web and two at the two side edges of the web. Alternatively, a design is also suitable in which the reflector is movable and is retracted from outside the field of view for the calibration measurement and then moved out of the field of view again. An equivalent design is in which the field of view is moved appropriately and thus aligned with the reflector, e.g. by pivoting the sensor array or corresponding mirrors.

[0035] However, the permanent arrangement in the field of vision is simpler in terms of equipment.

[0036] The illumination profile of the light source is also relevant for the absorption measurement, especially if it is spatially inhomogeneous, but also if it changes over time, e.g., due to aging. Therefore, in a suitable embodiment, a calibration is performed to compensate for the illumination profile of the light source. In this calibration, a reference surface is measured across the entire field of view of the SWIR camera system. The reference surface is, for example, a reflector as described above. The reference surface is expediently a combination of a white reference and a black reference, each across the entire field of view, so that a white measurement and a black measurement are performed successively for the calibration.Analogous to the reflector described above, the reference surface is also expediently arranged as close as possible to the web so that the path of the light from the light source to the sensor array during calibration corresponds as closely as possible to the path of the light during the actual absorption measurement. The black reference is implemented, for example, by means of a shutter on the SWIR camera system (e.g. a cover on a lens of the camera system), which closes the SWIR camera system for the measurement in such a way that no light enters it. Calibration is typically carried out much less frequently than the calibration measurement described above; it is also conceivable that just a single calibration during assembly or when commissioning the corrugator. However, recurring calibration is more practical, e.g. every 24 hours or when changing the light source.

[0037] The calibration measurement or calibration with a reflector as described above is particularly suitable for absorption measurements in reflection. If the absorption measurement is performed in transmission, the reflector and reference surface are not required; the calibration measurement or calibration is conveniently carried out using a completely dry sheet as a white reference, ideally using a corresponding calibration pattern. Since the reference surface occupies the entire field of view during calibration, it is conveniently positioned outside the field of view and moved into the field of view for calibration. After calibration, the reference surface is then moved out of the field of view again. For this purpose, either the reference surface is movable or the field of view is moved accordingly.

[0038] Another advantageous embodiment is one in which the measuring device is designed such that the optical absorption measurement is carried out on both sides of the web (double-sided measurement), so that measurement data is generated for both sides. For example, a light source and a SWIR camera system for a respective optical absorption measurement are arranged on both sides of the web, as already described. Alternatively, a single light source and / or a single SWIR camera system are used for both sides, e.g. in alternating operation and with suitable deflection, or by combining a measurement in transmission on one side with a measurement in reflection on the other side. The measurement data now available for both sides enable a more detailed evaluation and thus increased accuracy, but can also optionally be advantageously used for defect detection.It is advisable to compare the measurement data from both sides of the web and then, based on such a comparison, a quality or property of the web is determined, e.g., depth information as described above. Furthermore, double-sided measurement also offers a cost advantage, because typically the camera (multispectral camera) is expensive and the light source is inexpensive, or conversely, the light source is expensive (multispectral illumination) and a monochromatic SWIR camera system is correspondingly inexpensive. By using the cheaper module on both sides of the web and the more expensive module only once, low expenditure is combined with the advantages of a double-sided system design. While reflection measurement is generally preferred for determining moisture, transmission measurement is used in particular to determine defects in the web, e.g.of the wave in it, or missing adhesives are particularly easily recognized. These two aspects are then combined cost-effectively. The evaluation, i.e. the processing of the measured data and the calculation of the moisture content from it, is basically possible in various ways. Especially when using many, e.g. six, spectral bands, it is advantageous to use methods from chemometrics for the evaluation. It is advisable to use a training process in which reference measurements are taken on web samples of defined moisture content to build up a regression model and then use it to calculate the moisture content. In a suitable embodiment, the moisture content is then calculated using a PLS regression model (PLS = "partial least squares"). Machine learning is expediently used for the training process, in which an algorithm for evaluation is used with known training data, e.g.Measurement data on different webs or papers with known moisture content is used.

[0039] The measurement data is appropriately processed first. The measurement data contains, in particular, a spectral curve for each measurement point, i.e., the measured absorption in the path as a function of wavelength in the various spectral bands. If not already automatically specified by the design of the SWIR camera system, the wavelength range to be used for evaluation is restricted to the spectral bands under investigation. The measurement data (i.e., the spectral curves) are then smoothed, thereby reducing any noise in the measurement data. To better highlight relevant signals, the first derivative of each spectral curve is optionally calculated. The measurement data is also normalized.This completes the processing and, based on the now processed measurement data, a regression model is created using the partial least squares method and, using a suitable algorithm, an evaluation is then carried out to calculate the humidity with spatial resolution.

[0040] A corrugated board system according to the invention is designed to carry out a method as described above. To carry out the method, the corrugated board system has a control unit into which one or more parts of the measuring device are integrated, in particular the aforementioned evaluation unit, the depth analysis unit, and / or the defect detection unit.

[0041] In summary, with the method and corrugator described here, moisture measurement with an accuracy of up to 0.1 mass percent is possible. A one-time training session is usually sufficient; after that, no further calibration for different paper types is necessary. The measurement described here is advantageously carried out non-contact, in real time, and inline in the corrugator. The number of measuring points is essentially unlimited and, transverse to the conveying direction, is only limited by the resolution of the SWIR camera system. In the conveying direction, however, the resolution is limited in particular by the web speed and the measuring time of the SWIR camera system. The measurement is insensitive to temperature fluctuations, ambient lighting (provided no halogen spotlights are used for ambient lighting), and web vibrations.In addition, the measuring device can also perform one or more functions of a standard image processing and monitoring system, especially when measuring in the VIS range, e.g., detecting defects such as cracks or defects or monitoring the corrugation of the corrugated layers. The measuring device can also advantageously monitor the glue application, even for individual glue lines within the web. The solution described here is also simple and cost-effective.

[0042] In the following, exemplary embodiments of the invention are explained in more detail with reference to a drawing. In each case, the following schematically show:

[0043] Fig. 1 a corrugated board plant,

[0044] Fig. 2 a method,

[0045] Fig. 3 a measuring device of the corrugator, Fig. 4 a sensor array of the measuring device,

[0046] Fig. 5 a variant of the sensor array,

[0047] Fig. 6 a representation of the spatially resolved moisture of a web,

[0048] Fig. 7 to 9 show processed measurement data of an absorption measurement.

[0049] Fig. 1 shows an exemplary corrugated board plant 2, and Fig. 2 shows a method which is used here to operate the corrugated board plant 2 in Fig. 1. In the method, a plurality of paper webs 4, 6, 8 are generally guided in a conveying direction F through the corrugated board plant 2. The webs 4, 6, 8 are individual paper webs 4, intermediate products 6 which are composed of a plurality of such paper webs 4, and a corrugated board web 8 which is produced by the corrugated board plant 2. In the exemplary embodiment shown, two paper webs 4 are each unwound from a dispenser 10 and assembled with a single facer 12 to form a half-wave 6, which is then assembled with another paper web 4 (laminating web) in a double facer 14 to form the corrugated board web 8. The double facer 14 marks one end of a wet end 16 of the corrugated board plant 2.This is followed by a dry end 18, in which the continuous web of corrugated cardboard 8 is finished, i.e. cut into individual sheets 20.

[0050] A measuring device 22 is integrated into the corrugator 2, comprising a light source 24 and a SWIR camera system 26. An exemplary embodiment of the measuring device 22 is shown in a perspective view in Fig. 3. The SWIR camera system 26 has a camera designed for measuring in the near-infrared range (SWIR = "short wavelength infrared", also NIR). The SWIR camera system 26 shown here has a multispectral camera or a hyperspectral camera as the camera for the optical absorption measurement.

[0051] The SWIR camera system 26 has a field of view 28 which extends transversely to the conveying direction F, i.e. transversely to the webs 4, 6, 8, in this case in a transverse direction Q perpendicular to the conveying direction F. In the exemplary embodiment shown, the field of view 28 extends over a total width 30 of the web 8. Within the scope of the method, one of the webs 4, 6, 8, in this case the corrugated cardboard web 8, is guided through the field of view 28. In a first step S1, an optical absorption measurement (in short: measurement) is carried out in a number of spectral bands 32 using the light source 24 and the SWIR camera system 26, and corresponding measurement data 34 is generated in the process. This happens repeatedly. The measurement data 34 are stored, in this case by way of example, in a control unit 36 ​​of the corrugator 2.

[0052] The SWIR camera system 26 has a pixel array, also referred to as a sensor array 38, with a plurality of pixels 40, onto which a point (measurement point) on the web 4, 6, 8 is imaged at a given time. The measurement data 34 then contains a measured value for each point on the web. Two examples of sensor arrays 38 are shown in Fig. 4 (24 pixels) and Fig. 5 (72 pixels). In this case, each pixel 40 is also assigned an individual spectral band 32 by using a corresponding filter. Since the web 4, 6, 8 is moved in the conveying direction F during operation, the absorption in one of the spectral bands 32 along a longitudinal line (parallel to the conveying direction F) on the web 4, 6, 8 is measured accordingly with repeated measurements of a respective pixel 40. The field of view 28 of the SWIR camera system 26 and its orientation relative to the path 4, 6, 8 determine which part of the path 4, 6, 8 is imaged onto the sensor array 38 for measurement.In the present case, a longitudinal section of the web 4, 6, 8 is imaged so that several measuring points lying next to one another on the web 4, 6, 8 with respect to the conveying direction F can be examined simultaneously. Since the web 4, 6, 8 is continuously conveyed through the corrugator 2, the web 4, 6, 8 is also guided past the SWIR camera system 26 so that the latter scans the web 4, 6, 8 line by line and then, through a repeated measurement, the web 4, 6, 8 is measured across its surface in two directions, namely both in the conveying direction F and transversely thereto in the transverse direction Q. By combining the measurement data 34 at different times, i.e. at different longitudinal positions of the web 4, 6, 8, a surface measurement is now realized. In a second step S2, a moisture content 42 of the web 4, 6, 8 is then calculated with spatial resolution from the measurement data 34. This is done by means of a corresponding evaluation unit 44.

[0053] 4 and 5, the respective SWIR camera system 26 has only a single line sensor 50 for a respective spectral band 32. In Figs. 4 and 5, the assignment of a respective pixel 40 to a spectral band 32 is indicated by its color; accordingly, three different spectral bands 32 are used as examples. To generate area information, use is made of the fact that the web 4, 6, 8 is moved in the conveying direction F and is then repeatedly scanned line by line sensor 50. When multiple spectral bands 32 are used, various possibilities for arranging the respective pixels 40 for the respective spectral band 32 arise, as is clear from Figs. 4 and 5. For example, as shown in Fig. 5, multiple line sensors 50 are arranged one behind the other, so that although a two-dimensional sensor array 38 is present, only a single line sensor 50 is ever used for a respective spectral band 32.Alternatively, as shown in Fig. 4, the sensor array 38 has only a single row in total, in which the line sensors 50 for different spectral bands 32 are combined by arranging the pixels 40 for different spectral bands 32 alternately next to one another in a periodic sequence.

[0054] The optical absorption measurement is carried out in the dry end 18 in the present case and is used for the final inspection of the unfinished corrugated board web 8, which is still present as a continuous web, after the wet end 16. The concept described here can, however, in principle also be applied at any other point in the corrugator 2. An absorption measurement on the individual sheets 20 or directly behind a unwinder 10 for one of the paper webs 4 is also possible. In Fig. 3, the absorption measurement is carried out in reflection, with light from the light source 24, which is reflected on one side of the web 4, 6, 8, being guided to the SWIR camera system 26 on the same side. Alternatively, the absorption measurement is carried out in transmission (not shown), i.e. light from the light source 24 enters the web 4, 6, 8 on one side and exits it again on the opposite side and is then guided to the SWIR camera system 26.In an embodiment not explicitly shown, the measuring device 22 is configured such that the optical absorption measurement is performed on both sides of the web 4, 6, 8, so that measurement data 34 are generated for each side. For example, starting from Fig. 3, a second light source 24 and a second SWIR camera system 26 are arranged on the other side of the web 4, 6, 8.

[0055] The light source 24 emits light diffusely in this case and is a halogen spotlight. Furthermore, the light source 24 and the SWIR camera system 26 are designed and arranged such that, during the optical absorption measurement, light from the light source 24 penetrates the web 4, 6, 8 on its way to the SWIR camera system 26, namely with a penetration depth of at least 1 mm. The measuring device 22 has a depth analysis unit 46 which, based on the measurement data 34, determines an internal condition or internal property of the web 4, 6, 8, i.e., depth information. Aside from determining the moisture content 42, this enables a more precise examination of the interior of the web 4, 6, 8 or a defect inspection with which defects that are not visible from the outside can be identified. Optionally—independently of the use of depth information—a defect in the web 4, 6, 8 is detected based on the measurement data 34.For this purpose, the corrugator 2 has a corresponding defect detection unit 48. Defects that are detected include, for example, tears, defects, faulty corrugation formation, and / or faulty glue application on individual glue lines. In response to a detected defect, the section of the web 4, 6, 8 containing the defect is then rejected.

[0056] The absorption measurement is also influenced by the atmosphere between the light source 24 and web 4, 6, 8 on the one hand, and between web 4, 6, 8 and sensor array 38 on the other. To reduce this influence, a corresponding calibration measurement is carried out at regular intervals. For this purpose, the measuring device 22 in Fig. 3 has at least one reflector 52, with which a calibration measurement for the SWIR camera system 26 is carried out repeatedly, e.g., every 15 minutes. In the present case, the reflector 52 is permanently arranged in the field of view 28 of the, i.e., both during the calibration measurement and during the absorption measurement. The reflector 52 extends only over part of the total width 30 of the web 4, 6, 8. In Fig. 3, three reflectors 52 are present as an example, one in the center of the web 4, 6, 8 and two at the two side edges of the web 4, 6, 8.

[0057] The illumination profile of the light source 24 is also relevant for the absorption measurement, especially if it is spatially inhomogeneous, but also if it changes over time. Therefore, in the present case, a calibration is carried out to compensate for the illumination profile, in which a reference surface 54 is measured across the entire field of view 28. The reference surface 54 is, for example, a combination of a white reference and a black reference, each across the entire field of view 28, so that a white measurement and a black measurement are carried out one after the other for the calibration. Calibration is typically carried out much less frequently than the calibration measurement described above. Since the reference surface 54 occupies the entire field of view 28 during calibration, it is arranged outside the field of view 28 in the present case and is moved into the field of view 28 for the calibration and moved out again after the calibration.For this purpose, either the reference surface 54 is movable or the field of view 28 is moved accordingly.

[0058] Fig. 6 shows a representation of the spatially resolved moisture content 42 of an exemplary web 4, 6, 8. The scale at the bottom edge shows the color coding for the moisture content in a value range from 0 to 26 percent by mass. The calculated moisture content 42 shown in Fig. 6. The evaluation, i.e. the processing of the measurement data 34 and the calculation of the moisture content 42 therefrom, is fundamentally possible in various ways. For example, with the help of training, in which reference measurements are taken on web samples of defined moisture content, a regression model is built and then used to calculate the moisture content 42. The moisture content 42 is then calculated using a PLS regression model (PLS = "partial least squares", method of least squares).

[0059] In this case, the measurement data 34 are first processed; the result of this is shown in Figs. 7, 8 and 9 for three papers with different moisture contents 42, with Fig. 7 showing the driest paper and Fig. 9 the wettest. In each case, a plurality of curves is shown, each of which displays a smoothed, simply derived and normalized spectral profile. The wavelength 56 is indicated on the horizontal axis (in the range from 1100 nm to 1600 nm), and finally, the absorption 58 of the light in the paper is indicated on the vertical axis. Each curve contains an absorption measurement for all spectral bands 32 at a single point on the paper, i.e. a spectral profile for a given measurement point. For the evaluation, the wavelength range to be used is restricted to the spectral bands 32 being examined.Based on such processed measurement data 34, a regression model is then created using the partial least squares method, and the evaluation is then carried out using a suitable algorithm, which calculates the humidity 42 with spatial resolution. The representation in Fig. 6 is the result of using eight spectral bands 32, i.e., not all spectral bands 32 shown in Figs. 7-9 were used.

[0060] List of reference symbols

[0061] 2 corrugated board lines

[0062] 4 webs (single paper web)

[0063] 6 track (intermediate, half-wave)

[0064] 8 web (corrugated board web)

[0065] 10 dispensers

[0066] 12 single facers

[0067] 14 Double Facers

[0068] 16 Wet End

[0069] 18 Dry End

[0070] 20 sheets

[0071] 22 Measuring device

[0072] 24 light source

[0073] 26 SWIR camera system

[0074] 28 field of view

[0075] 30 total width

[0076] 32 spectral bands

[0077] 34 measurement data

[0078] 36 Control unit

[0079] 38 sensor array

[0080] 40 pixels

[0081] 42 Humidity

[0082] 44 Evaluation unit

[0083] 46 Depth Analysis Unit

[0084] 48 Defect detection unit

[0085] 50 line sensor

[0086] 52 reflector

[0087] 54 Reference surface

[0088] 56 wavelength

[0089] 58 Absorption

[0090] F Conveying direction

[0091] Q Transverse direction S1 step (“optical absorption measurement and generation of measurement data”)

[0092] S2 Step (“Calculating the humidity”)

Claims

Claims 1. Method, a. wherein a plurality of paper webs (4, 6, 8) are guided through a corrugated board plant (2) in a conveying direction (F), b. wherein a measuring device (22) is integrated into the corrugated board plant (2), comprising a light source (24) and a SWIR camera system (26), c. wherein the SWIR camera system (26) has a field of view (28) which extends transversely to the conveying direction (F), d. wherein one of the webs (4, 6, 8) is guided through the field of view (28) and, in the process, an optical absorption measurement in a number of spectral bands (32) is repeatedly carried out using the light source (24) and the SWIR camera system (26), and measurement data (34) are generated in the process, e. wherein a moisture content (42) of the web (4, 6, 8) is calculated in a spatially resolved manner from the measurement data (34).

2. The method according to claim 1, wherein the light source (24) and the SWIR camera system (26) are designed and arranged such that during the optical absorption measurement, light from the light source (24) penetrates into the web (4, 6, 8), namely with a penetration depth of at least 1 mm, wherein the measuring device (22) has a depth analysis unit (46) which determines an internal condition or internal property of the web (4, 6, 8) based on the measurement data (34).

3. Method according to claim 1 or 2, wherein an internal condition or internal property of the web (4, 6, 8) is determined by carrying out a plurality of absorption measurements, namely with different intensities of the light from the light source (24).

4. The method according to any one of claims 1 to 3, wherein the SWIR camera system (26) for the optical absorption measurement comprises a multispectral camera.

5. The method according to any one of claims 1 to 3, wherein the SWIR camera system (26) for optical absorption measurement comprises a hyperspectral camera.

6. The method according to any one of claims 1 to 5, wherein the SWIR camera system (26) has only one line sensor (50) for a respective spectral range (32).

7. The method according to any one of claims 1 to 6, wherein the measuring device (22) has at least one reflector (52) with which a calibration measurement for the SWIR camera system (26) is repeatedly carried out.

8. The method according to claim 7, wherein the reflector (52) is permanently arranged in the field of view (28) of the SWIR camera system (26).

9. The method according to any one of claims 1 to 8, wherein, to compensate for an illumination profile of the light source (24), a calibration is carried out in which a reference surface (54) is measured over the entire field of view (28) of the SWIR camera system (26).

10. The method according to claim 9, wherein the reference surface (54) is arranged outside the field of view (28) and is moved into the field of view (28) for calibration.

11. Method according to one of claims 1 to 10, wherein the measuring device (22) is configured such that the optical absorption measurement is performed on both sides of the web (4, 6, 8), so that measurement data (34) are generated for each side.

12. The method according to one of claims 1 to 11, wherein the moisture content (42) is calculated using a PLS regression model.

13. The method according to one of claims 1 to 12, wherein a defect in the web (4, 6, 8) is detected based on the measurement data (34).

14. Method according to one of claims 1 to 13, wherein the optical absorption measurement is carried out in a dry end (18) of the corrugator (2).

15. Corrugated board plant (2) which is designed to carry out a method according to one of claims 1 to 14.

Images