DETERMINING THE STATUS FOR STRAND-GUIDING ROLLERS BY MEANS OF VIBRATION EVALUATION

Abstract

A strand-guiding section arranged downstream of a continuous casting mould of a continuous casting line. The strand-guiding section has at least one strand-guiding roller supporting a metal strand cast with the continuous casting mould, is mounted in roller bearings in the strand-guiding section, and has at least one sensor device, which has at least one vibration sensor acoustically coupled to the strand-guiding roller or its roller bearings to detect the vibrations occurring with the rotation of the strand-guiding roller in the roller bearings. The strand-guiding roller is cooled internally by a cooling water. The vibration sensor is acoustically coupled to the strand-guiding roller and/or to the roller bearings via the cooling water. The evaluation device determines the status of the at least one roller bearing by evaluating the vibration data.

Claims

1. A strand-guiding section of a continuous casting line, wherein the strand-guiding section is arranged downstream of a continuous casting mold of the continuous casting line, wherein the strand-guiding section has at least one strand-guiding roller for supporting a metal strand that is cast with the continuous casting mold, wherein the strand-guiding roller is mounted in roller bearings in the strand-guiding section, wherein the strand-guiding section has at least one sensor device, which in turn has at least one vibration sensor which is acoustically coupled to the strand-guiding roller or the roller bearings thereof, by means of which the vibrations occurring when the strand-guiding roller is rotated in the roller bearings can be detected, wherein the sensor device has an interface via which vibration data (SD) derived from the detected vibrations can be transmitted to an evaluation device, wherein the strand-guiding roller is cooled internally by cooling water supplied to the strand-guiding roller and the vibration sensor is acoustically coupled to the strand-guiding roller and/or to the roller bearings via the cooling water.

2. The strand-guiding section as claimed in claim 1, wherein for supplying and/or discharging the cooling water, the strand-guiding roller has a rotary feed-through on at least one end, via which supply lines for supplying and/or discharging the cooling water are connected to the strand-guiding roller, and in that the vibration sensor is arranged on the rotary feed-through.

3. The strand-guiding section according to claim 1, wherein the interface to the evaluation device is designed as a wireless interface, in particular as a radio interface, and the sensor device has a separate, internal energy source.

4. The strand-guiding section as claimed in claim 1, wherein the sensor device is designed in such a way that it transfers from a passive state (OFF) to an activated state (ON) due to an internally monitored timeout or due to an activation command (A) received via the interface, and transfers from an activated state (ON) into the passive state (OFF) due to an internally monitored timeout or a passivation command (P) received via the interface, and the vibration sensor detects the vibrations that occur when the strand-guiding roller is rotated only when the sensor device is in the activated state (ON).

5. The strand-guiding section as claimed in claim 1, wherein the vibration sensor detects the vibrations occurring when the strand-guiding roller is rotated with a sampling rate that is at least in the two-digit kHz range.

6. A determination system for the status (Z) of at least one roller bearing of a strand-guiding roller of a strand-guiding section of a continuous casting line arranged downstream of a continuous casting mold of the continuous casting line, wherein the strand-guiding roller can be cooled internally by supplied cooling water, wherein the determination system has at least one sensor device, which in turn has at least one vibration sensor which can be acoustically coupled to the strand-guiding roller or the roller bearings thereof via the cooling water, wherein the vibration sensor is designed in such a way that it can detect vibrations propagating in the cooling water, which occur when the strand-guiding roller is rotated in the roller bearings, by direct contact of the vibration sensor with the cooling water, wherein the sensor device has an interface via which the sensor device transmits vibration data (SD) derived from the detected vibrations to an evaluation device of the determination system, and wherein the evaluation device, by evaluating the vibration data (SD) transmitted to it, determines the status (Z) of the at least one roller bearing.

7. The determination system as claimed in claim 6, wherein the evaluation device is configured to transmit a message (M) to an operator and/or to a monitoring system at least whenever the evaluation device determines a critical state of the roller bearing during the evaluation.

8. The determination system as claimed in claim 6, wherein the evaluation device is configured to store the vibration data (SD) and/or the determined status (Z) of the at least one roller bearing in a memory in the form of a history.

9. The determination system as claimed in claim 6, wherein the evaluation device has an input for operating data and the evaluation device is configured to take into account operating data of the strand-guiding roller, of the strand-guiding section and/or of the continuous casting line in addition to the vibration data (SD), when determining the status (Z) of the at least one roller bearing.

10. The determination system as claimed in claim 7, wherein the evaluation device is configured to perform a two-dimensional spectral decomposition (FZ) with the vibration data (SD) transmitted to it, one of the two dimensions of the spectral decomposition (FZ) is the time (t) and one is the frequency (f), and the evaluation device evaluates the spectral decomposition (FZ) to determine the status (Z) of the at least one roller bearing.

11. The determination system as claimed in claim 10, wherein the evaluation device is configured to determine the maximum amplitude (MAX) for each time point of the transmitted vibration data (SD), the evaluation device compares the respective maximum amplitudes (MAX) with at least one limit (LIM1, LIM2) and the evaluation device determines the status (Z) of the at least one roller bearing depending on the comparisons.

12. The determination system as claimed in claim 11, wherein the at least one limit (LIM1, LIM2) is time-variable, in particular determined by the evaluation device as a function of the operating state (BZ) of the continuous casting line.

13. The determination system as claimed in claim 10, wherein the evaluation device is configured to determine the energy content (E) of the spectral decomposition (FZ) for a plurality of time points and in a number of predetermined frequency bands (FB) of the spectral decomposition (FZ), and the evaluation device determines the status (Z) of the at least one roller bearing on the basis of the determined energy content values (E).

14. The determination system as claimed in claim 13, wherein the evaluation device is configured, with or without prior filtering of the energy content values (E) within the respective frequency band (FB), to calculate a mean value (E1) of the energy content values (E) and a maximum value (E2) of the energy content values (E) for each frequency band (FB) and to determine the status (Z) of the at least one roller bearing on the basis of the calculated mean values (E1) and the calculated maximum values (E2) of the frequency bands (FB).

15. The determination system as claimed in claim 13, wherein the evaluation device is configured, for each of a plurality of frequency bands (FB), with or without prior filtering of the energy content values (E) within the respective frequency band (FB), a) to calculate a mean value (E1) of the energy content values (E) and a maximum value (E2) of the energy content values (E), b) to determine a start time point (ta) at which the energy content (E) first exceeds an intermediate value (E3) between the mean value (E1) and the maximum value (E2), c) to determine, starting from the determined start time point (ta), an end time point (te) at which the energy content (E) falls below the mean value (E1) again for the first time, d) to determine a resulting time point (tr), between the determined start time point (ta) and the determined end time point (te), at which the energy content (E) is a maximum, e) starting from the last determined end time point (te), to repeat steps b) to d) until all the resulting time points (tr) within the respective frequency band (FB) are determined, f) to then determine, on the basis of the determined resulting time points (tr), a respective period (TR) with which the energy content (E) varies within the respective frequency band (FB), and g) to determine the status (Z) of the at least one roller bearing on the basis of the periods (TR) determined for the frequency bands (FB).

16. The determination system as claimed in claim 13, wherein the evaluation device is configured to compare a frequency analysis of the determined energy content values (E) within the frequency bands (FB) with a rotation speed of the strand-guiding roller and to determine the status (Z) of the at least one roller bearing based on the comparison.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0074] The properties, features and advantages of the present invention described above and the manner in which these are achieved will become clearer and more comprehensible in conjunction with the following description of the exemplary embodiments, which are explained in more detail in connection with the drawings. In the drawings, in schematic representation:

[0075] FIG. 1 shows a continuous casting line,

[0076] FIG. 2 shows a strand-guiding section,

[0077] FIG. 3 shows a strand-guiding roller, its mounting and supply lines,

[0078] FIG. 4 shows a detail of FIG. 3 in an enlarged view,

[0079] FIG. 5 shows a block diagram with a sensor device and an evaluation device,

[0080] FIG. 6 shows a timing diagram,

[0081] FIG. 7 shows a timing diagram,

[0082] FIG. 8 shows a frequency-time component diagram,

[0083] FIG. 9 shows a flowchart,

[0084] FIG. 10 shows a flowchart,

[0085] FIG. 11 shows a flowchart,

[0086] FIG. 12 shows a flowchart,

[0087] FIG. 13 shows a temporal characteristic of an energy content,

[0088] FIG. 14 shows a flowchart,

[0089] FIG. 15 shows a flowchart,

[0090] FIG. 16 shows a flowchart,

[0091] FIG. 17 shows a Fourier-transformed signal, and

[0092] FIG. 18 shows another Fourier-transformed signal.

DESCRIPTION OF EMBODIMENTS

[0093] According to FIG. 1, a continuous casting line 1 has a continuous casting mold 2. Liquid metal 4, such as liquid steel, is poured into the continuous casting mold 2—for example via a schematically indicated immersion tube 3. The continuous casting mold 2 determines the format of the metal strand 5. The format can be a slab format, a billet format, or another format.

[0094] The liquid metal 4 solidifies on the walls of the continuous casting mold 2 and is extracted from the continuous casting mold 2 as a metal strand 5, which is initially partially solidified and later fully solidified. The metal strand is extracted with an extraction speed v which is generally within the range of a few meters per minute (m/min), for example between 3 m/min and 10 m/min.

[0095] The metal strand 5 is supported behind the continuous casting mold 2 in the extraction direction by means of a plurality of strand-guiding rollers 6. In FIG. 1, only a few of the strand-guiding rollers 6 are shown. The strand-guiding rollers 6 in their entirety form a strand guide downstream of the continuous casting mold 2. It is possible for the metal strand 5 to be rolled directly out of the casting melt behind the strand guide. Alternatively, the metal strand 5 can be cut to length, for example by means of a schematically indicated gas cutting machine 7.

[0096] The strand-guiding rollers 6 are arranged in strand-guiding sections 8 according to the illustration in FIG. 2. The strand-guiding sections 8 are therefore arranged downstream of the continuous casting mold 2, as are the strand-guiding rollers 6 themselves. FIG. 2 shows a typical design of a strand-guiding section 8. In this design, the strand-guiding section 8 is designed as a roller segment. The roller segment has a first and a second group G1, G2 of strand-guiding rollers 6, as can be easily seen from FIG. 2. The two groups G1, G2 of strand-guiding rollers 6 each have a plurality of strand-guiding rollers 6. Furthermore, the strand-guiding rollers 6 of the two groups G1, G2 are adjacent to opposite sides of the cast metal strand 5, where they roll off. In general, the distance between the strand-guiding rollers 6 of the first group G1 and the strand-guiding rollers 6 of the second group G2 can also be adjusted via actuating devices 9. For example, the actuating devices 9 can be designed as hydraulic cylinder units.

[0097] However, regardless of the design of the strand-guiding sections 8 as roller segments, at least one strand-guiding roller 6 is mounted in each strand-guiding section 8, as illustrated in FIG. 3. The strand-guiding rollers 6 are mounted via roller bearings 10. The roller bearings 10 are usually designed as anti-friction bearings.

[0098] In many cases the strand-guiding roller 6 is cooled internally by means of cooling water 11, as illustrated in FIG. 3. In this case, the strand-guiding roller 6 can have a rotary feed-through 12 on at least one end. For example, as illustrated in FIG. 3, a rotary feed-through 12 can be arranged at both ends of the strand-guiding roller 6. The rotary feed-throughs 12 are used to supply the cooling water 11 to the strand-guiding roller 6 and/or to discharge the cooling water 11 from the strand-guiding roller 6. Connections to the rotary feed-throughs 12 are made using rigid or flexible supply lines 13, via which the cooling water 11 is supplied to the strand-guiding roller 6 or discharged from the strand-guiding roller 6.

[0099] According to the enlarged illustration in FIG. 4, the strand-guiding section 8 has a sensor device 14. The sensor device 14 in turn has a vibration sensor 15, according to the block diagram of FIG. 5. The vibration sensor 15—for example a sound sensor, in particular an ultrasound sensor—is acoustically coupled to the strand-guiding roller 6 or its roller bearings 10. As a result, the vibration sensor 15 is able to detect the vibrations that occur when the strand-guiding roller 6 is rotated in the roller bearings 10. The vibrations can be caused, for example, by impurities on the roller bearings 10. The vibration sensor 15 outputs a sensor signal SS corresponding to the detected vibrations. The sensor signal SS is typically still analog.

[0100] The vibration sensor 15 detects the vibrations that occur during rotation of the strand-guiding roller 6 with a sampling rate that is at least in the two-digit kHz range, i.e. 10 kHz or more. According to the illustration in FIG. 6, a sampling time T of a maximum of 100 μs is present between two directly consecutive detections of the respective signal level of the vibrations and the corresponding output of the sensor signal SS. For example, the sampling time T can be 10 μs (equivalent to a sampling rate of 100 kHz) or 20 μs (equivalent to a sampling rate of 50 kHz) or 50 μs (equivalent to a sampling rate of 20 kHz).

[0101] As part of the embodiment according to FIG. 3, in which the strand-guiding roller 6 is internally cooled, according to the illustration in FIG. 3 the vibration sensor 15 is acoustically coupled to the strand-guiding roller 6 and/or to the roller bearings 10 via the cooling water 11. The vibration sensor 15 (more precisely: a vibration-sensitive surface of the vibration sensor 15) is thus in direct contact with the cooling water 11, as also indicated schematically in FIG. 5. The cooling water 11 thus acts as the transmission medium for the vibrations. In particular, the vibration sensor 15 can be designed as a hydrophone for good signal acquisition.

[0102] The arrangement of the vibration sensor 15 can be selected as required. If (at least) one rotary feed-through 12 is present, the vibration sensor 15 is preferably arranged on the rotary feed-through 12 (or one of the rotary feed-throughs 12). However, it can also be arranged at another location. If the signal is not acquired via the cooling water 11, the vibration sensor 15 can also be arranged at any other location where signal acquisition is possible, including in particular in the immediate vicinity of the roller bearing 10 to be monitored.

[0103] The sensor device 14 has additional components, as illustrated in FIG. 5. The most important additional component is an interface 16, via which the sensor device 14 transmits vibration data SD to an evaluation device 17. The vibration data SD is usually digital data that is produced by digitizing the acquired sensor signals SS, for example in an analog-to-digital converter 18. In any case, however, the vibration data SD is data derived from the detected vibrations or the sensor signals SS, the information content of which corresponds to the sensor signals SS. The interface 16 is preferably designed as a wireless interface, in particular as a radio interface.

[0104] As an additional component, the sensor device 14 also preferably comprises at least one separate, internal energy source 19. In this case, the sensor device 14 does not require any electrical cables to supply power. A control device 20 is also present, which controls at least the interface 16 and also the other components of the sensor device 14, if present.

[0105] It is possible that the sensor device 14 transmits the sensor signal SS—possibly after digitization—to the evaluation device 17 as is, i.e. as vibration data SD. However, the sensor device 14 preferably has additional components. For example, as illustrated in FIG. 5, the sensor device 14 can have a filter 21 in which filtering takes place. The filter 21 can be arranged before or after the analog-to-digital converter 18 in the direction of the signal flow as required. The filtering can be high-pass filtering, low-pass filtering, level limiting, etc., as required. In addition, the sensor device 14 can have an amplifier 22. The amplifier 22 amplifies the analog sensor signal SS—preferably before any filtering, but in any case before the analog-to-digital conversion.

[0106] According to the illustration in FIG. 7, the sensor device 14 is only temporarily active. For example, the sensor device 14 can transfer from a passive state OFF to an activated state ON whenever it receives an activation command A via the interface 16—see FIG. 5. Alternatively, it is possible that the sensor device 14—see also FIG. 5—has an internal timer 23 and transfers to the activated state ON whenever the timer 23 signals the expiry of a specific time period T′. In this case, the transition to the activated state ON is based on an internally monitored timeout, so that the sensor device 14 activates itself.

[0107] The sensor device 14 maintains the activated state for a certain time period T″. The time period T″ is usually in the single-digit minute range, for example, between 3 minutes and 8 minutes. The sensor device 14 then transfers from the activated state ON back to the passive state OFF. For example, the timer 23 can be used to monitor the expiry of the time period T″. In this case, the transition to the passive state OFF is thus based on an internally monitored timeout, so that the sensor device 14 deactivates itself. As an alternative possibility, the sensor device 14 transfers to the passive state OFF whenever it receives a passivation command P via the interface 16. In this case, the device which transmits the passivation command P to the sensor device 14 must ensure that the time period T″ is observed. This device can be, for example, the evaluation device 17.

[0108] The time period T″ during which the sensor device 14 is activated is usually considerably shorter than the time period T′ between two directly consecutive activations. For example, the time period T′ can be 1 hour or more, or several hours and even several days.

[0109] It is also clear from FIG. 7 that the vibration sensor 15 detects the vibrations occurring when the strand-guiding roller 6 is rotated only when the ON state of the sensor device 14 is activated. If an immediate transmission to the evaluation device takes place, the same applies to the transmission of the vibration data SD derived from the sensor signals SS to the evaluation device 17 via the interface 16.

[0110] The evaluation device 17 evaluates the vibration data SD which the sensor device 14 transmits to it. As part of this evaluation, the evaluation device 17 determines the status Z of the at least one roller bearing 10 of the strand-guiding roller 6. Possible procedures for determining the status Z will be explained in detail later. However, it should be mentioned immediately that when determining the status Z, the evaluation device 17 can take into account operating data of the strand-guiding roller 6, of the strand-guiding section 8 and/or the continuous casting line 1 as a whole, in addition to the vibration data SD. This will become apparent from the explanations of the possible forms of evaluation to be given below.

[0111] As illustrated in FIG. 5, the evaluation device 17 transmits a message M to an operator 24 and/or a monitoring system 25 at least whenever it determines a critical status of the roller bearing 10 during the evaluation. It is therefore possible that the evaluation device 17 also outputs a message M to the operator 24 and/or the monitoring system 25 if the determined status Z of the roller bearing 10 is not critical. However, in this case the transmission of the message M can also be omitted. If, on the other hand, the determined status Z is critical, the evaluation device 17 always outputs the message M.

[0112] In addition, the evaluation device 17 can store the vibration data SD in a storage device 26, in accordance with FIG. 5. The data is stored in the form of a history. Newly stored vibration data SD are thus added to the already stored vibration data SD, but do not replace it. Furthermore, time information is associated with the stored data, in particular time information for the acquisition of the vibration data. Additional data may be added, for example, about the type of evaluation or the time at which the evaluation is carried out. As an alternative or in addition to the vibration data SD, the evaluation device 17 can also store the determined state Z and, if applicable, other data in the storage device 26 in the same way. An example of other data that can be meaningfully stored is an acknowledgment by the operator 24, with which the operator 24 or the monitoring system 25 acknowledged a reported critical status.

[0113] As part of the evaluation of the determined vibration data SD, the evaluation device 17 first determines a two-dimensional spectral decomposition FZ of the vibration data SD. It thus determines a two-dimensional field in which, according to the illustration in FIG. 8, the two dimensions of the spectral decomposition FZ—i.e. the input variables of the spectral decomposition FZ—are the time t and the frequency f, and in which the output variable indicates the proportion of the vibration data SD which is present at the respective frequency f at the respective time t.

[0114] On the time axis, the spectral decomposition FZ extends from a start time t1 to an end time t2. The start time t1 and the end time t2 are the times during which the sensor signals SS are detected by the vibration sensor 14 during a given activation. The extension on the frequency axis can be chosen as required.

[0115] The proportion determined as a function of time t and frequency f can be a complex value or—equivalent thereto—a pair of real values—or a single real value.

[0116] In order to determine the spectral decomposition FZ, the evaluation device 17 can perform a short-time Fourier transformation of the vibration data SD in a step S1, for example, in accordance with the illustration in FIG. 9. The short-time Fourier transform first yields the complex components (or, equivalent to this, the division into sine and cosine functions for equal frequencies). In addition, the evaluation device 17 can then determine real components on the basis of the complex components in step S2. In particular, it can determine the magnitude or the square of the magnitude of the respective complex component. In this case, the spectral decomposition FZ directly specifies the energy content E for the respective time point t and the respective frequency f. In step S3, the evaluation device 17 then evaluates the spectral decomposition FZ. This evaluation is used to determine the status Z of the at least one roller bearing 10. Means for implementing step S3 of FIG. 9 are explained below in conjunction with the other figures.

[0117] Step S2 is optional and therefore only shown with a dashed line in FIG. 9. As a rule, however, step S2 is included. It is also assumed throughout the following text that the spectral decomposition FZ is performed according to steps S1 and S2, so that the resulting spectral decomposition FZ already delivers the square of the magnitude of the short-time Fourier transformation.

[0118] In a possible configuration of step S3 of FIG. 9, the evaluation device 17 can determine the maximum MAX of the spectral decomposition FZ in step S11, as illustrated in FIG. 10, for example. In this case, in step S12 the evaluation device 17 compares the determined maximum MAX with a lower limit LIM1. If the maximum MAX is below the lower limit LIM1, in step S13 the evaluation device 17 sets the status Z to the value “good”, symbolized by “+” in FIG. 10. Optionally, the evaluation device 17 can output a corresponding message M in a subsequent step S14. On the other hand, if the maximum MAX is above the lower limit LIM1, the evaluation device 17 proceeds to step S15. In step S15, the evaluation device 17 compares the determined maximum MAX with an upper limit LIM2. The upper limit LIM2 is, self-evidently, larger than the lower limit LIM1. If the maximum MAX is below the upper limit LIM2, the evaluation device 17 sets the status Z to the value “fair” in step S16 (symbolized by “0” in FIG. 10) and outputs a corresponding message M in step S17. If the maximum MAX is above the upper limit LIM2, the evaluation device 17 sets the status Z to the value “poor” in step S18 (symbolized by “-” in FIG. 10) and outputs a corresponding message M in step S19.

[0119] The lower limit LIM1 and the upper limit LIM2 are suitably chosen. For example, they can be determined and specified based on experiments. It is also possible that the lower limit LIM1 and/or the upper limit LIM2 are time-variable. For example, as illustrated in FIG. 11, the evaluation device 17 can receive an operating state BZ of the continuous casting line 1 in step S21 before carrying out steps S11 to S19, and determine the lower limit LIM1 and/or the upper limit LIM2 depending on the operating state BZ of the continuous casting line 1 in step S22.

[0120] The example above, in which a lower and an upper limit LIM1, LIM2 are defined, is purely exemplary. Only one limit can also be specified. More than two limits can also be specified.

[0121] In a further possible configuration of step S3 of FIG. 9, the evaluation device 17 can determine an energy content E of the spectral decomposition FZ in step S31, as illustrated in FIG. 12. In the simplest case, step S31 is trivial because the determination of the energy content E was already the object of step S1. Otherwise, it must be carried out explicitly. If, for example, the evaluation device 17 determines the (complex) component in step S1, in step S31 the square of the magnitude of the component (or the product of the component with the complex conjugate of the component) is determined. However, the determination in step S31 is always resolved in both dimensions (i.e. both in time t and frequency f).

[0122] In step S32, the evaluation device 17 selects a predefined frequency band FB. The selected frequency band FB extends from a lower cutoff frequency f1 to an upper cutoff frequency f2. For the sake of clarity, it is pointed out here that the term “cutoff frequency” in the context of the present invention has nothing to do with signal filters and the corresponding cutoff frequencies. The term “cutoff frequency” refers only to the lower or upper limit of the frequency band FB.

[0123] In a subsequent step S33, the evaluation device 17 determines the energy content of the selected frequency band FB as a function of time t. It thus forms the integral of the energy content E over the frequency f from the lower cutoff frequency f1 to the upper cutoff frequency f2. Since, as already mentioned, it is assumed within the context of the present invention that the spectral decomposition FZ directly specifies the energy content for the respective time point t and the respective frequency f, the integral of the spectral decomposition FZ itself can be formed in step S33.

[0124] In an optional step S34, the evaluation device 17 can then filter the determined energy content E, in particular, perform a temporal averaging. If the filtering is carried out, the filtering is based on a considerably shorter period than the period for which the vibration data SD was acquired and hence for which the spectral decomposition FZ is also determined. For example, if the difference between the end time t2 and the start time t1 is 5 minutes, filtering can occur for a time frame of 1 second, for example. The numerical values mentioned are purely exemplary, but show the principle.

[0125] In step S35, the evaluation device 17 checks whether it has performed step S33 and, if applicable, step S34 for all frequency bands FB. If this is not the case, the evaluation device 17 returns to step S32. In the repeated execution of step S32 the evaluation device 17 selects a different frequency band FB for which it has not yet performed step S33 and step S34, if applicable.

[0126] Otherwise, the evaluation device 17 proceeds to step S36, in which the further evaluation is carried out. This evaluation is based on the previously determined energy content values E. In particular, the evaluation device 17 in step S36 uses not only the energy content E determined during the last execution of step S33, but also the energy content values E determined for all frequency bands FB.

[0127] The procedure according to FIG. 12 returns the temporal characteristic of the energy content E for each of the frequency bands FB. A possible temporal characteristic is shown in FIG. 13 for a single frequency band FB.

[0128] The number of frequency bands FB for which the energy content E is determined as a function of time t can be determined as required. At a minimum, only a single frequency band FB is used. As a rule, however, the energy content E is determined for a plurality of frequency bands FB. The number of frequency bands FB is usually between three and eight. If the energy content E is determined for a plurality of frequency bands FB, the frequency bands FB are usually additionally separate from each other (disjoint). This means that they do not overlap. If, purely as an example, one of the frequency bands FB extends from a lower cutoff frequency f1 of 100 Hz to an upper cutoff frequency f2 of 200 Hz, this frequency range is, so to speak, a prohibited zone for all other frequency bands FB. A different frequency band FB that has an upper cutoff frequency f2 above 200 Hz—for example, 500 Hz—can thus have a lower cutoff frequency f1 of, for example, 300 Hz. The lower cutoff frequency f1 of this other frequency band FB can also be less than 300 Hz. Under no circumstances though should the cutoff frequency f1 of this other frequency band FB be less than 200 Hz. However, the limiting case in which the upper cutoff frequency f2 of a frequency band FB is equal to the lower cutoff frequency f1 of another frequency band FB is still just permissible.

[0129] In a concrete implementation of the present invention, experiments were carried out with four frequency bands FB. In this experiment, one frequency band FB ranged from 7 kHz to 11 kHz, another frequency band FB from 11 kHz to 15 kHz, another frequency band FB from 30 kHz to 37 kHz, and another frequency band FB from 42 kHz to 49 kHz. The first and the second frequency bands FB mentioned are therefore exactly adjacent to each other, while the second frequency band FB and the two other frequency bands FB are separated from each other.

[0130] Possible implementations of step S36 are explained below in conjunction with the remaining figures.

[0131] In a further possible configuration of step S36 of FIG. 12, the evaluation device 17 can select one of the frequency bands FB in step S41, as illustrated in FIG. 14. Within the selected frequency band FB, the evaluation device 17 determines a mean value El of the energy content values E and a maximum value E2 of the energy content values E in step S42. In a step S43, the evaluation device 17 forms the difference 5E between the maximum value E2 and the mean value E1 of the energy content values E. In step S44, the evaluation device 17 then determines a preliminary status z for the selected frequency range FB. This determination is based on the difference 5E. For example, in step S44, the evaluation device 17 may determine the value “good” as the preliminary status z if the difference 5E is below a predetermined limit, and otherwise determine the value “poor” as the preliminary status z. The pre-determined limit can be relatively small. In particular, it can be between 1% and 10% of the mean value E1, for example between 2% and 5%.

[0132] In step S45, the evaluation device 17 checks whether it has performed steps S42 to S44 for all frequency bands FB. If this is not the case, the evaluation device 17 returns to step S41. In the repeated execution of step S41 the evaluation device 17 selects a different frequency band FB for which it has not yet performed steps S42 to S44.

[0133] Otherwise, the evaluation device 17 proceeds to step S46, in which the further evaluation is carried out. In particular, in step S46 the evaluation device 17 determines the status Z using the preliminary statuses z. For example, it is possible that as the status Z the evaluation device 17 determines [0134] the value “good” if a sufficient number—in particular all—of the preliminary statuses z are “good”, [0135] the value “poor” if a sufficient number—in particular all—of the preliminary statuses z are “poor”, and [0136] otherwise, the value “fair”.

[0137] Other procedures are also possible, for example, that the status Z is determined solely on the basis of the number of poor preliminary statuses z. It is also possible to determine exclusively the value “good” or the value “poor” as the status Z.

[0138] In a further possible configuration of step S36 of FIG. 12, the evaluation device 17 can select one of the frequency bands FB in step S51—analogous to step S41 of FIG. 14—as illustrated in FIG. 15. Within the selected frequency band FB, in step S52—analogous to step S42 of FIG. 14—the evaluation device 17 determines a mean value E1 of the energy content values E and a maximum value E2 of the energy content values E. In step S53, the evaluation device 17 determines an intermediate value E3 between the maximum value E2 and the mean value E1 of the energy content values E. For example, the evaluation device 17 can form the sum of maximum value E2 and mean value E1 and divide this sum by 2.

[0139] The evaluation device 17 then analyzes the temporal characteristic of the determined energy content values E. In concrete terms, the evaluation device 17 sets a start time point t0 to the start time t1 in a step S54. In step S55, the evaluation device 17 then determines a start time point ta. The start time point ta is the time point at which, starting from the start time point t0, the energy content E exceeds the intermediate value E3 for the first time. Starting from the determined start time point ta, the evaluation device 17 then determines an end time point te in step S56. The end time point te is the time point at which, starting from the start time point ta, the energy content E falls below the mean value E1 again for the first time. The evaluation device 17 then determines a resulting time point tr in step S57. The resulting time point tr is the time point that lies within the time interval from the start time point ta to the end time point te and has the maximum energy content E within this time interval.

[0140] In step S58, the evaluation device 17 checks whether it should repeat steps S55 to S58, starting from the previous end time point te as the new start time point ta. If this is the case, the evaluation unit 17 proceeds to step S59. In step S59, the evaluation device 17 sets the previous end time point te as the new start time point ta. It then returns to step S55. At the end of the process, all the resulting time points tr within the respective frequency band FB are thereby determined.

[0141] When the determination of the resulting time points tr for the frequency band FB currently being analyzed is complete, the evaluation device 17 proceeds to step S60. In step S61, the evaluation device 17 determines the time intervals δt between directly consecutive resulting time points tr. Based on the time intervals δt, the evaluation device 17 then determines a period TR for the resulting time points tr in step S62. For example, the evaluation device 17 can form the mean value of the time intervals δt and treat this mean value as the period TR. The evaluation device 17 can therefore assume that the energy content E within the selected frequency band FB varies on average with the period TR.

[0142] In step S62, the evaluation device 17 checks whether it has already performed steps S52 to S61 for all frequency bands FB. If this is not the case, the evaluation device 17 returns to step S51. In the repeated execution of step S51 the evaluation device 17 selects a different frequency band FB for which it has not yet performed steps S52 to S61.

[0143] Otherwise, the evaluation device 17 proceeds to step S63 in which the further evaluation is carried out. This evaluation is based on the previously determined periods TR. In particular, it is to be expected that the poorer the status Z of the roller bearing 10, the more consistent are the determined periods TR. The evaluation device 17 can therefore check, for example, how close to each other the periods TR determined for the different frequency bands FB are located. If the determined periods TR are close to each other, the status Z is assigned the value “poor”. If the determined periods TR are far apart, the status Z is assigned the value “good”. If neither case is present, the status Z is assigned the value of “fair”.

[0144] In a further possible configuration of step S36 of FIG. 12, the evaluation device 17 can select one of the frequency bands FB in step S71—analogous to step S41 of FIG. 14—as illustrated in FIG. 16. Within the selected frequency band FB, the evaluation device 17 performs a frequency analysis of the determined energy content E in step S72. For example, in step S72 the evaluation device 17 can calculate a Fourier transformation of the energy content E. FIG. 17 and FIG. 18 show purely exemplary possible results of such a Fourier transformation, i.e. the Fourier transform of possible energy content values E of a frequency band FB. The numbers entered on the abscissa are normalized to the rotation speed of the strand-guiding roller 6 in question and are therefore expressed in terms of the rotation speed of the strand-guiding roller 6 under analysis.

[0145] If, as shown in FIG. 17, appreciable amplitudes occur at the rotation rate of the strand-guiding roller 6 in question and integer multiples thereof, this can be interpreted as an indication that the status Z of the roller bearing 10 is poor. The value “poor” can therefore be assumed as the preliminary status z for the frequency band FB in question, provided that such appreciable amplitudes occur. If, on the other hand, as shown in FIG. 18, no appreciable amplitudes are present, the value “good” can be assumed as the preliminary status z for the frequency band FB in question. The evaluation device 17 can therefore determine the corresponding components in step S73, compare the components with limit values, and determine the preliminary status z for the selected frequency band FB depending on the comparison.

[0146] In step S74, the evaluation device 17 checks whether it has already performed steps S72 and S73 for all frequency bands FB. If this is not the case, the evaluation device 17 returns to step S71. In the repeated execution of step S71 the evaluation device 17 selects a different frequency band FB for which it has not yet performed steps S72 and S73.

[0147] Otherwise, the evaluation device 17 proceeds to step S75 in which the further evaluation is carried out. In particular, in step S75 the evaluation device 17 determines the status Z using the preliminary statuses z. For example, it is possible that, analogous to step S46 of FIG. 14, the evaluation device 17 determines as the status Z [0148] the value “good” if a sufficient number—in particular all—of the preliminary statuses z are “good”, [0149] the value “poor” if a sufficient number—in particular all—of the preliminary statuses z are “poor”, and [0150] otherwise, the value “fair”.

[0151] Here, also, different procedures are possible, analogous to FIG. 14. In particular, the status Z can be determined exclusively based on the number of poor preliminary statuses z.

[0152] The present invention has been explained above in conjunction with a single sensor 15 and a single sensor device 14. However, the present invention can also be embodied if a plurality of sensors 15 and/or a plurality of sensor devices 14 are present for a single strand-guiding roller 6. The present invention is also quite generally embodied for all strand-guiding rollers 6. Furthermore, the invention has been explained in connection with an embodiment in which the vibration data SD is transmitted to the evaluation device 17 immediately after the sensor signals SS have been acquired. It is also conceivable, however, to store the vibration data SD within the sensor device 14 first and only transmit them to the evaluation device 17 later. For example, the sensor device 14 can detect and store a measurement sequence at specific time intervals—see the explanations given for FIG. 7. In this case, for example, the evaluation device 17 can be briefly connected to the interface 16 during a downtime of the continuous casting line 1 and the vibration data SD of a plurality of measurement sequences can be read out cumulatively.

[0153] The present invention has many advantages. Firstly, signal acquisition can be carried out in a simple and reliable manner, even during operation of the continuous casting line 1. Furthermore, an evaluation of the acquired sensor signals SS or the vibration data SD derived from them can be carried out in a reliable manner. Furthermore, both the signal acquisition and the signal evaluation can be automated. In the case of an immediate transmission of the vibration data SD to the evaluation device 17, problems occurring in the continuous casting line 6 or its roller bearings 10 can be detected immediately, even during operation of the continuous casting line 1. Damaged and worn strand-guiding rollers 6 can be easily identified and localized. During operation of the continuous casting line 1, maintenance measures can be scheduled which should be carried out at the next shutdown of the continuous casting line. The availability and reliability of the continuous casting line 1 are increased. If a plurality of vibration sensors 15 are used for a single strand-guiding roller 6, it may also be possible to localize within a single strand-guiding roller 6 which roller bearing 10 is defective or worn.

[0154] Although the invention has been illustrated and described in greater detail by means of the preferred exemplary embodiment, the invention is not restricted by the examples disclosed and other variants can be derived therefrom by the person skilled in the art without departing from the scope of protection of the invention.

LIST OF REFERENCE SIGNS

[0155] 1 continuous casting line

[0156] 2 continuous casting mold

[0157] 3 immersion tube

[0158] 4 liquid metal

[0159] 5 metal strand

[0160] 6 strand-guiding rollers

[0161] 7 gas cutting machine

[0162] 8 strand-guiding sections

[0163] 9 actuating devices

[0164] 10 roller bearing

[0165] 11 cooling water

[0166] 12 rotary feed-throughs

[0167] 13 supply lines

[0168] 14 sensor device

[0169] 15 vibration sensor

[0170] 16 interface

[0171] 17 evaluation device

[0172] 18 analog-to-digital converter

[0173] 19 energy source

[0174] 20 control device

[0175] 21 filter

[0176] 22 amplifier

[0177] 23 timer

[0178] 24 operator

[0179] 25 monitoring system

[0180] 26 memory device

[0181] A, A′ activation commands

[0182] BZ operating state

[0183] E energy content

[0184] E1 mean value of the energy content values

[0185] E2 maximum value of the energy content values

[0186] E3 intermediate value

[0187] f frequency

[0188] f1, f2 cutoff frequencies

[0189] FB frequency bands

[0190] FZ spectral decomposition

[0191] G1, G2 groups of strand-guiding rollers

[0192] LIM1, LIM2 limits

[0193] M message

[0194] MAX maximum of spectral decomposition

[0195] OFF passive state

[0196] ON activated state

[0197] P, P′ passivation commands

[0198] S1 to S75 steps

[0199] SD vibration data

[0200] SS sensor signal

[0201] t time, or time points

[0202] t0 start time point

[0203] t1 start time

[0204] t2 end time

[0205] ta start time point

[0206] te end time point

[0207] tr resulting time point

[0208] T sampling time

[0209] T′, T″ time periods

[0210] TR period

[0211] v extraction speed

[0212] Z status

[0213] z preliminary statuses

[0214] δE difference of energy values

[0215] δt time intervals