METHOD FOR RESONANCE ANALYSIS OF AN VIBRATION MACHINE

Description

[0001] The present invention relates to a method for resonance analysis of a vibrating machine.

[0002] Machines with a vibrating unit can be damaged by the self-generated vibration if the vibration of the vibrating unit is equal to or similar to the natural frequency of the machine. Due to the periodic recurring excitation, increasing amounts of energy are transferred to the system. This constructive interference stores the energy in the system until a resonance catastrophe occurs. In the worst case, this can destroy the machine.

[0003] To avoid this, it is necessary to know the frequency at which such a resonance catastrophe can occur. This is done using so-called resonance analyses.

[0004] Previously, resonance analyses involved causing the machine to vibrate with a brief external pulse, for example, by striking the machine with an impact hammer. Meanwhile, sensors record the vibration response to the single pulse. The pulse is a Dirac signal, i.e., a short-term signal superimposed by numerous vibrations. Using appropriate analysis methods, such as frequency analysis of the measurement, conclusions can be drawn about the machine's natural frequency.

[0005] To perform this vibration analysis, the vibration machine is stopped so that the vibration response is not superimposed with other vibration signals.

[0006] DE 10 2005 042085 A1 discloses a vibration measurement system for frequency-selective vibration measurement, particularly of low frequencies as are relevant in the field of automation and drive technology. It is proposed to couple a broadband transmitter structure, which is directly excited by the excitation signal to be determined, to a receiver structure via an electrostatic or inductive force. This force coupling results in amplitude modulation of a carrier signal exciting the receiver structure. The actual excitation signal can be extracted from the spectrum of the amplitude-modulated carrier signal, for example, by a suitable selection of the carrier signal frequency. To enable vibration analysis that is as immune to interference as possible, an interference signal caused, for example, by nozzle excitations, is largely eliminated from the amplitude-modulated carrier signal beforehand.

[0007] WO 2004 / 059399 A2 discloses a system for diagnosing or testing the condition of machines and their parts, specifically rotating machines and their components, such as aircraft and helicopter engines, transmissions, etc., based on vibration data. Vibrations are measured at various points on the machine during operation using sensors, and a so-called "vibration signature" is created. For aircraft engines, for example, different rotation speeds and other parameters are also taken into account. A so-called "baseline signature" is created for the "normal case," i.e., the intact condition of the machine. Deviations of the vibration signature data from the baseline signature in subsequent measurements can provide indications of irregularities, such as wear or damage to rotating parts, clogged nozzles, etc., and of potential failure risks.

[0008] Rahman et al., "Effectiveness of Impact-Synchronous Time Averaging in determination of dynamic characteristics of a rotor dynamic system", Measurement 44, (2011), 34-45, discloses a method in which a spectrum of the vibrations of a rotating system is recorded. The natural frequencies of the rotating system are also to be characterized. One problem addressed by Rahman et al. is the background and harmonics caused by the rotational forces, which can dominate the spectrum and mask the natural frequencies. Unlike non-rotating systems, such as a vibrating machine, the dynamic properties and thus the determined spectrum of the rotating systems addressed by Rahman et al. are also influenced by gyroscopic effects that depend on the rotational speed. A method developed by Rahman et al.The method presented, ISTA ("Impact-Synchronous Time Averaging"), is said to be suitable for eliminating the rotation speed-dependent components, their harmonics and the background from the determined response signal.

[0009] WO 2019 / 072462 A1 discloses a mobile device for detecting the state and operating parameters of vibrating machines with sensor units and an evaluation unit connected to the sensor units, wherein the measurement data detected by the sensor units can be transmitted wirelessly to the evaluation unit, and wherein each sensor unit is equipped with at least three acceleration sensors aligned orthogonally to one another and an integrated circuit for processing the measurement data detected by the sensor units.At least four sensor units form a sensor network, wherein the sensor units can be detachably fastened to the vibrating machine at a mutual distance with an indeterminate orientation, and a local coordinate system X1, Y1, Z1 is defined by the at least three acceleration sensors of a sensor unit, to whose spatial axes the local measurement data recorded in a sensor unit are related, and each sensor unit has a gravity sensor for recording the orientation of the local coordinate system X1, Y1, Z1 in space, and the evaluation unit has a device for transforming the local measurement data into a higher-level uniform coordinate system X0, Y0, Z0 taking into account the measurement data of the gravity sensor.

[0010] The object of the present invention is to carry out a resonance analysis of a vibrating machine which is simple and cost-effective and disturbs the operation as little as possible.

[0011] The object is achieved by the subject matter of the independent claims. Advantageous further developments and preferred embodiments form the subject matter of the dependent claims.

[0012] A method for resonance analysis of a vibrating machine, namely a vibrating screen or a vibrating conveyor, during operation comprises the following steps: Determining an operating vibration signal comprising the vibrations of the machine during regular operation, exciting the vibrating machine during operation with an excitation signal for exciting additional vibrations in the vibrating machine, wherein the excitation signal is a Dirac pulse, measuring a response vibration signal of the vibrating machine to the excitation during operation, determining a correction signal (9) by using the operating vibration signal (8), determining a subtraction signal by subtracting the correction signal from the response vibration signal, and frequency analyzing the subtraction signal to determine the natural frequencies of the vibrating machine, wherein the dominant frequencies of the subtraction signal correspond to the natural frequencies of the system in the frequency range of the excitation signal.

[0013] A vibrating machine is a machine that is intentionally set into vibration, at least in part, during operation, with this vibration oscillating at a predetermined frequency.

[0014] The vibrating machine has at least one natural frequency, which is characterized by the fact that if the frequency of the operating vibration is equal to the natural frequency, it can lead to a resonance catastrophe.

[0015] The operating vibration signal and the response vibration signal are measured by a vibration detection device, wherein the vibration detection device comprises at least one or more vibration sensors. These vibration sensors can be, for example, MEMS acceleration sensors, piezo sensors, or sensors from microphone technology. Sensors from microphone technology include, for example, pressure microphones, condenser microphones, moving coil microphones, ribbon microphones, or crystal microphones.

[0016] The determination of the subtraction signal as well as the frequency analysis of the subtraction signal is carried out on a computer, whereby computer is understood here in a general way and can include PCs, laptops, portable handheld devices such as smartphones and tablets, but also devices specially manufactured for this process.

[0017] The vibration machine can be excited manually or mechanically. The most common method is an impact hammer.

[0018] Previously, resonance analysis could not be performed during operation because the operating vibrations dominate any measurement and overlay the impulse responses of an excitation signal. Thus, in the evaluation, e.g., after a Fourier transformation of the signal, only the frequency of the operating vibration could be detected. Determining the vibration response was not possible in this case.

[0019] The inventors have recognized that the operating oscillations remain stable over time. In this context, "stable over time" means that the frequencies, amplitudes, and phases of the operating oscillation do not change by more than 10%, preferably no more than 5%, and especially no more than 1%, within, say, 10 minutes.

[0020] The operating vibration also depends on the loading condition of the vibrating machine. An unloaded vibrating machine exhibits essentially the same operating vibration over a long period of time, e.g., several weeks. If the loading condition changes, the vibration behavior of the vibrating machine can change. The vibration amplitude can change by up to 10%. It is therefore advisable to record the operating vibration signal and the response vibration signal under essentially the same loading condition.

[0021] Because the operating vibration is temporally stable, the operating vibration signal can be used as a correction signal and subtracted from the measurement signal, e.g., the response vibration signal. The remaining signal is the response vibration signal generated by the excitation signal.

[0022] If no excitation signal was transmitted to the vibrating machine, a zero line results after deducting the correction signal as a subtraction signal.

[0023] One advantage of this method is that it can be carried out during operation. Unlike with the prior art, the user does not have to shut down the machine being tested and can continue to use it during the measurement. The resonance analysis of the machine can also be carried out during regular operation, i.e. when the machine is loaded. Depending on the application of such a vibrating machine, the costs incurred due to downtime during a measurement can amount to several million euros. Many users therefore shy away from such an inspection and risk damaging the machine. In the worst case, this can endanger human lives.

[0024] Depending on the application location, other machines located in close proximity to the vibrating machine and operating at a certain frequency themselves can cause the vibrating machine under test to vibrate. Previously, these machines would also have had to be shut down, otherwise the measurement would be "contaminated." This means that the frequencies of the other machines would also appear in the analysis, potentially resulting in incorrect natural frequencies being determined for the vibrating machine under test. To prevent this, entire factory halls and production facilities have been shut down in order to test a single machine. This is uneconomical.

[0025] However, with the present method, the vibration of surrounding machines can also be included as part of the operating vibration signal in the correction signal and then subtracted from the response vibration signal. The only prerequisite for this is that the properties of the surrounding machines do not change, for example, by switching them on or off or by setting a different frequency.

[0026] Another advantage is that the machine's natural frequency is measured under realistic conditions. The inventors realized that when the machine is stationary, the natural frequencies differ slightly from the natural frequencies during operation. This is due, for example, to the fact that the machine in question is filled, and the filling material affects the natural frequencies.

[0027] Since vibration behavior can change due to loading, resonance analysis during regular operation is more realistic than conventional resonance analyses, in which the vibrating machine is analyzed in an unloaded state and out of operation. Preferably, the operating vibration signal is modified before being used as a correction signal. For this purpose, the correction signal can be modeled as a function of the operating vibration signal, for example, by fitting.

[0028] When generating the correction signal, the frequency with the strongest amplitude can be determined from the operating oscillation signal, which is referred to below as the operating frequency. The amplitude and phase of the operating frequency in the operating oscillation signal and the amplitude and phase of the harmonics of the operating frequencies in the operating oscillation signal are then determined. The correction signal is then generated from the amplitude and phase of the operating frequency and the amplitude and phase of the harmonics of the operating frequency.

[0029] However, the correction signal may also include signals from other machines in operation in the vicinity and / or around sidebands and / or spring / hard body resonances of the vibrating machine.

[0030] The operating oscillation signal essentially corresponds to a sine wave, with the sine wave superimposed by the sine waves of the harmonics. Therefore, it is useful if the correction signal is also represented as a sine wave of the operating frequency, superimposed with the sine waves of the harmonics. A sine function can be defined by the frequency f, amplitude A, and phase P. Thus, if the frequency, amplitude, and phase of the operating frequency and the harmonics are known, the correction signal can be generated using this information.

[0031] Preferably, the harmonics are one or more multiples of the operating frequency, preferably 1 / 2, 3 / 2, 1 / 4, 2, 3, 4, 5, 8, 16 and 32 times the operating frequency.

[0032] A multiple of the operating frequency can also be a fraction of the natural frequency, such as 1 / 2 or 1 / 4. These are then called subharmonics.

[0033] In the simplest case, the operating vibration signal can be graphically displayed on a display, and the parameters of the correction signal are adjusted so that it corresponds as closely as possible to the operating vibration signal. Alternatively, the subtraction signal can be generated continuously by subtracting the correction signal from a response vibration signal, which is continuously measured without any excitation vibration being applied to the vibrating machine. The parameters of the correction signal are then adjusted until a zero line is recognizable as the response vibration signal.

[0034] Preferably, the operating vibration signal is estimated by automatic parameter estimation and modified with the resulting parameters.

[0035] Methods for automatic parameter estimation, also known as fitting, are known. Such methods are particularly suitable when the signal to be analyzed is a periodic, sinusoidal signal. Such automatic parameter estimation can be implemented using a software product in which the resulting correction signal is mathematically very close to the operating vibration signal. This allows the operating vibration signal to be almost completely subtracted from the response vibration signal, resulting in only the response of the vibrating machine to the excitation signal.

[0036] Preferably, the function of the operating vibration signal is approximately given by: y t = ∑ n = 1 k A n sin n f t + P n , where n indicates the degree of the harmonic, k represents the highest harmonic, An is the amplitude of the harmonic, f is the operating frequency, t is the time and P n is the phase of the respective harmonic.

[0037] This mathematical function represents a sum over several sine functions.

[0038] Preferably, k is at least 1, 2 or 4 in order to obtain sufficient precision of the correction signal.

[0039] Preferably, k is a maximum of 64, 32 or 16 so that the computational effort is not too great.

[0040] If there are partial harmonics, which are also called subharmonics, such as 1 / 2 or 1 / 4, then the above formula is adjusted accordingly with a term for these partial harmonics.

[0041] It is also conceivable to add other terms having different frequencies to cover other machines in operation and / or to cover sidebands and / or spring / hard body resonances.

[0042] Alternatively, the correction signal can also be modeled on the operating vibration signal by recording the operating vibration signal over a predetermined period and using it as a correction signal.

[0043] This avoids the need for a mathematical description of the operating vibration signal and allows the original operating vibration signal to be used instead. The disadvantage of this could be that the required duration has to be estimated. On the other hand, if the recording is too short, the recorded signal could be reused.

[0044] By using the recorded signal, the operating vibration signal can be subtracted from the response vibration signal without any loss of quality due to mathematical adjustment.

[0045] Furthermore, this would not take into account sinusoidal signals.

[0046] The excitation signal is a Dirac pulse.

[0047] A Dirac pulse in the sense of the present invention is to be understood as an only approximately mathematical Dirac pulse, since it is physically impossible to generate a pure mathematical Dirac pulse.

[0048] In practice, such a Dirac pulse is generated by a blow from an impulse hammer.

[0049] A Dirac pulse is characterized by the fact that a pulse with the shortest possible duration is transmitted to the oscillating machine under investigation, whereby all frequencies are excited equally.

[0050] Frequencies that do not correspond to the natural frequency or a multiple thereof are immediately damped.

[0051] However, the natural frequencies and multiples thereof continue to oscillate under the influence of the Dirac pulse and can be measured.

[0052] In practice, however, such a Dirac pulse will only excite a certain spectrum of frequencies.

[0053] Preferably, the natural frequency is determined in a frequency range of greater than 0.01 Hz, preferably greater than 0.1 Hz and in particular greater than 1 Hz.

[0054] Frequencies that are lower here cannot be excited so easily by the Dirac pulse and are not of interest for the measurement, since the typical frequencies of the machine are above the limit specified here.

[0055] Preferably, the natural frequency is determined in a frequency range of less than 1 kHz, preferably of less than 500 Hz and in particular of less than 50 Hz.

[0056] The Dirac pulse will only slightly excite higher frequencies, and the operating frequency will not exceed this limit, so that natural frequencies above these values do not pose a danger to the machine.

[0057] Preferably, after the frequency analysis step, the operating frequency is changed and all steps are repeated.

[0058] By changing the operating frequency, it can be ensured that no new natural frequencies arise at different operating frequencies, or that existing natural frequencies are weighted differently. A user of a vibrating machine wants to be certain of a wide range of possible operating frequencies and not have to check whether it is safe every time they want to set a new operating frequency.

[0059] Preferably, an excitation vibration signal is transmitted to the vibrating machine at several points.

[0060] This ensures that natural frequencies of the vibrating machine are also detected, which might have been overlooked at other locations.

[0061] Preferably, a response vibration signal is measured at several locations.

[0062] This ensures that all natural frequencies of the vibrating machine are recorded within the predetermined range.

[0063] Preferably, the determination of an operating vibration signal, the excitation of the vibrating machine, and the measurement of a response vibration signal take place temporally superimposed and / or simultaneously. A response vibration signal is acquired that includes the operating vibration signal and the reaction to the excitation. The correction signal is generated from this response vibration signal, and the correction signal is then subtracted from the response vibration signal to obtain the subtraction signal. The subtraction signal thus roughly corresponds to a response vibration signal for a machine that is not in operation. The subtraction signal is then subjected to frequency analysis to determine the natural frequencies of the vibrating machine.

[0064] The use of a response vibration signal, which includes the operating vibration signal and the response to the excitation, simplifies the evaluation. The phases of the operating vibration and its harmonics can be determined using an FFT. Since the individual phases of the harmonics of the operating vibration signal and the individual phases of the harmonics of the response vibration signal are based on the same measurement of the response vibration signal, they have the same relative relationship. This eliminates the need to subsequently determine such a phase relationship between the response vibration signal and the operating vibration signal, for example, by means of phase adjustment, an additional FFT determination, or a phase determination by determining a time interval.

[0065] Although the individual steps can be carried out one after the other, if they are superimposed in time and / or carried out simultaneously, the operating vibration signal and the response vibration signal can be recorded in a single measurement.

[0066] Since the frequency of the operating oscillation clearly dominates, the impulse response does not influence the determination of the frequency.

[0067] A measuring system for measuring natural vibrations on a vibrating machine, which is designed to carry out one of the methods described above, comprises at least one vibration detection device and one evaluation device.

[0068] Such vibration detection devices comprise, for example, the sensors described above, and an evaluation device may be a computer as described above.

[0069] A vibrating screen is a machine used for screening. It separates a mixture of solids according to the size of the grains.

[0070] A vibratory conveyor is a machine for transporting bulk materials. The bulk material is moved by means of linear, circular, or elliptical vibrations.

[0071] Machines are also known that include both features.

[0072] Vibratory conveyors comprise a conveying device, usually a vibrating conveyor trough, a vibrating drive and an elastic suspension.

[0073] All components of the vibratory conveyor, except for the suspension, should be deformation-resistant. This allows vibration to be transmitted evenly along the vibratory conveyor trough. Furthermore, fatigue fractures are avoided as long as possible.

[0074] The invention is explained in more detail below with reference to the examples shown in the drawings. The drawings schematically show: Figure 1A side view of a vibrating machine with a vibration sensor and an evaluation device, Figure 2A frequency spectrum of a response vibration signal, Figure 3A response vibration signal and a correction signal superimposed, Figure 4A response vibration signal and a subtraction signal, Figure 5Frequency spectrum of a conventional vibration analysis of a vibrating machine in the switched-off state, Figure 6Frequency spectrum of a vibration analysis of the subtraction signal, and Figure 7Flow diagram of the method for resonance analysis of a vibrating machine during operation.

[0075] A system for resonance analysis of a vibrating machine 1 comprises a vibrating conveyor 2, a vibration detection device 3 and an evaluation device 4 ( Figure 1 ).

[0076] The vibratory conveyor 2 is designed for conveying piece goods, e.g. castings, and bulk material, e.g. sand or gravel. The vibratory conveyor 2 comprises a vibratory conveyor trough 5 and a vibratory drive 6.

[0077] The vibrating conveyor trough 5 is designed with a conveyor floor (not shown) arranged in a vibrating frame. The conveyor floor transports the bulk material. A vibrating drive 6 is connected to the vibrating frame and sets it in vibrating motion.

[0078] The oscillating drive 6 is aligned at a predetermined angle to the conveyor floor (not shown) and causes it to oscillate in a predetermined direction.

[0079] The oscillating drive 6 can be adjusted with regard to the oscillation frequency.

[0080] The vibrating conveyor trough 5 is arranged at a predetermined angle to the ground and to the vibrating drive 6.

[0081] The conveying behavior is influenced depending on the angle to the ground and the angle to the vibrating drive 6 as well as the frequency.

[0082] The oscillating drive 6 comprises an unbalanced motor, which in this embodiment is a three-phase motor with an adjustable unbalance weight (not shown) at one shaft end. The amplitude of the generated vibration can be changed by manually adjusting the unbalance. The frequency is determined by the motor speed.

[0083] In the present embodiment, two counter-rotating drives are used. A single motor would produce a circular motion rather than a linear oscillation.

[0084] In an alternative embodiment, an oscillating armature drive can also be used.

[0085] The vibration detection device 3 has at least one vibration sensor 7. In this exemplary embodiment, the vibration sensor 7 is an acceleration sensor that measures vibrations of the vibratory conveyor 2 in all three spatial axes. The vibration sensor 7 is connected to the evaluation device 4 via a radio link. Wired connections are also possible in principle.

[0086] The connection between the vibration detection device 3 and the evaluation device 4 can be made via radio, for example Bluetooth, WLAN, ZigBee, Z-Wave or via a mobile network, or can be cable-based, for example via a LAN network.

[0087] The vibration sensor 7 is connected to the vibratory conveyor 2 via a mounting device. This connection can be fixed (e.g., by screws) or detachable (e.g., by adhesive strips or a clamping mechanism).

[0088] The evaluation device 4 can be designed as a computer, conventional smartphone or tablet, which processes the data of the vibration sensor 7 recorded by a receiving module using an additional software application, also called an app.

[0089] The evaluation device 4 can also comprise a display device which optically outputs the acquired and further processed data.

[0090] The system for resonance analysis of a vibrating machine 1 further comprises an impact hammer 13, which is not shown here. In this exemplary embodiment, the impact hammer 13 has a force sensor that measures the transmitted impulse when impacting the vibrating conveyor 2.

[0091] The procedure for resonance analysis of a vibrating machine 2 during operation is explained below.

[0092] The procedure begins with step S1 ( Figure 7 ).

[0093] In the next step (S2), the operating vibration signal 8 is determined. For this purpose, the sensors of the vibration detection device 3 measure the movements of the vibrating machine 2, which are generated by the vibration generator 2. The data are then sent to the evaluation device 4.

[0094] In the evaluation device 4, the operating vibration signal 8 is transferred from the time domain to the frequency domain by a Fast Fourier Transformation. Figure 2 a single peak can be seen which corresponds to the operating frequency.

[0095] The amplitude and phase are determined in the period. In Figure 3 a temporal section of the operating vibration signal 8 can be seen, from which the amplitude and phase can be read.

[0096] Subsequently, the operating vibration signal 8 is modified and used as correction signal 9 (step S3) ( Figure 3 and 4 ).

[0097] A specific duration of the operating vibration signal 8 is approximated into a mathematical function using an automatic parameter estimation (Fit).

[0098] This mathematical function has the form: y t = ∑ n = 1 k A n sin n f t + P n , where n indicates the degree of the harmonic, k represents the highest harmonic, An is the amplitude of the harmonic, f is the operating frequency, t is the time and P n is the phase of the respective harmonic.

[0099] Alternatively, the operating vibration signal 8 can also be converted to the frequency domain using a Fast Fourier Transformation (FFT). The operating frequency can then be read from the frequency domain, as it is the dominant signal here. The amplitude and phase of the operating vibration signal 8 are derived from the original operating vibration signal 8.

[0100] Once the operating frequency, amplitude and phase have been determined, the correction signal 9 can be calculated using these parameters as well as the parameters of the harmonics, whose parameters were also determined by the fit.

[0101] If the operating vibration signal 8 has been fitted, the fit is used as correction signal 9.

[0102] If the parameters were determined manually in the time and frequency domain, the correction signal 9 also results from the mathematical function: y t = ∑ n = 1 k A n sin n f t + P n , where the parameters have the same meaning as described above and the respective specific parameters were used. In Figure 3Such a manual adjustment of the temporal signal is shown. The modified operating vibration signal 8 is smoothed compared to the original measured operating vibration signal or correction signal 9. Such a "smooth" sinusoidal signal can be easily represented by a formula, which simplifies further calculations.

[0103] This is followed by step S4 in which the vibrating machine 2 is excited with an excitation signal 10 during operation.

[0104] This is done by means of an impact hammer 13, which is struck on the vibrating machine 2.

[0105] A visual or acoustic signal tells the user when to strike. This signal also starts the measurement.

[0106] In the following step S5, a response vibration signal 11 is measured. The measurement procedure is similar to the measurement of the operating vibration signal 8 from step S2. Here, too, the vibrations are measured via sensors of the vibration detection device 3 and sent to the evaluation device 4.

[0107] In this embodiment, the measurement of the operating vibration signal 8 and the measurement of the response vibration signal 11 are carried out successively.

[0108] Thereafter, in step S6, a subtraction signal 12 is determined by subtracting the correction signal 9 from the response oscillation signal 11.

[0109] Here, the response vibration signal 11 is selected in the evaluation device 4, and the correction signal 9 is mathematically subtracted from it. Since the measurement of the operating vibration signal 8 and the measurement of the response vibration signal 11 are carried out consecutively, attention must be paid to the correct phase of the response vibration signal 11.

[0110] The correct phase is determined by adjusting a parameter that changes the phase until the generated subtraction signal 12 in the FFT space does not or only weakly detects the operating frequency.

[0111] As a result, only the reaction of the vibration machine 2 to the excitation signal 10 can be seen in the generated subtraction signal 12 (see Figure 4 ).

[0112] In the subsequent step S7, the subtraction signal 12 is converted into the frequency domain using a Fast Fourier Transformation. The peaks identified there correspond to the natural frequencies of the vibrating machine 2 ( Figure 6 ).

[0113] How Figure 6 shows, the determined natural frequencies correspond almost to the determined natural frequencies as determined using a method in which the vibrating machine is stopped without load ( Figure 5 ): resonance State-of-the-art measurement [rpm] Measurement according to the present method [rpm] Difference [rpm] Difference [%] 1 1267,5 1248,75 -18,75 -1,48 2 1087,5 1072,5 -15,00 -1,38 3 1758,75 1751,25 -7,50 -0,43 4 2085 2043,75 -41,25 -2,00 5 116,25 108,75 -7,50 -6,45 6 217,5 217,5 0 0 7 - 1987,5 - - 8 - 2977,5 - -

[0114] A deviation of between 0.4 and 6.4% is observed. It is striking that the deviations always follow the same direction, which suggests that the deviations are systematic deviations that arise from the operation of the machine. Another possible cause could be the machine's loading.

[0115] Peaks 5 and 6 are each resonances of the rigid body, and here a particularly high deviation can be observed.

[0116] Peaks 7 and 8 are artifacts and correspond to the first two harmonics.

[0117] The method ends with step S8.

[0118] Another possibility is that once a correction signal 9 has been generated, any signal that is measured is cleaned with the correction signal 9.

[0119] This results in a zero line being created without an excitation signal 10 and this can be immediately recognized when an excitation signal 10 is applied.

[0120] In an alternative embodiment, an impact hammer 13 is omitted. The excitation signal 10 is automatically generated by larger chunks of material falling onto the vibrating machine 2. This allows for long-term and repeated monitoring of the natural frequencies during operation.

[0121] In a further alternative embodiment, the measurement of the response vibration signal 10 is triggered by the hammer blow.

[0122] As described above, if the operating vibration signal 8 and the response vibration signal 11 are not measured at the same time, the phase of the operating vibration in the response vibration signal 11 and its harmonics must be known in order to subtract the correction signal 9 with the correct phase from this response vibration signal 11 to thereby obtain the subtraction signal 12.

[0123] In addition to the above-mentioned possibility of adjusting the phase of the response oscillation signal 11 until the signal of the operating oscillation is minimized, alternatives are listed below as to how the phases of the operating oscillation signal 8 and the response oscillation signal 11 can be aligned.

[0124] In a first alternative embodiment, the phase of the correction signal 9 is determined by a phase spectrum. This phase spectrum is generated by the FFT as explained in the alternative form of step S3 above.

[0125] Subsequently, the phases of the response oscillation signal 11 are also determined by an FFT.

[0126] The phase of the operating oscillation in the response oscillation signal 11 is determined as the preferred phase at the corresponding frequency.

[0127] The phases of the harmonics are again determined by the FFT of the response oscillation signal 11. Alternatively, when determining the phases of the harmonics in the operating oscillation signal 8, the relative relationship to the phase of the operating oscillation is measured as relative phases. This relative relationship is also retained in the response oscillation signal 11. If the phase of the operating oscillation in the response oscillation signal 11 is known as the preferred phase, the absolute phases of the harmonics are determined by adding the preferred phase and the relative phases.

[0128] For example, if the relative phase of the 1st harmonic is 90° and the preferential phase is 45°, the absolute phase of the 1st harmonic results from the addition of 90° and 45° to 135°.

[0129] Alternatively, the phase can be determined by the time interval between the measurement of the operating vibration signal 8 and the measurement of the response vibration signal 11. If the duration is known exactly, as are the phases of the operating vibration signal 8 at the end of the measurement, the phase at the beginning of the measurement of the response vibration signal 11 is determined by the respective frequency and duration.

[0130] For example, if the duration is 60 seconds, the frequency of the operating oscillation is 30 Hz and the operating oscillation ends in the operating oscillation signal 8 with a phase of 45°, the initial phase PAA in the response oscillation signal 11 is: P AA = 30 s * 30 Hz ° + 45 ° = 1800 ° + 45 ° = 0 ° + 45 ° = 45 °

[0131] In the simplest case, the measurements are performed directly one after the other. The time between measurements is then zero. The phase of the response oscillation signal 11 at the beginning of the measurement is equal to the phase of the operating oscillation signal 8 at the end of the measurement.

[0132] The measurement of the operating vibration signal 8 and the response vibration signal 11 can also be performed as a continuous measurement, which is then divided manually or automatically at the evaluation device 4. In this case, the time between the partial measurements is also zero.

[0133] A further embodiment is described below, wherein identical elements as in the first embodiment are provided with the same reference numerals. The above explanations apply to identical elements unless otherwise stated below.

[0134] This embodiment differs from the first embodiment in that the determination of an operating vibration signal 8, the excitation of the vibration machine 2 and the measurement of a response vibration signal 11 take place simultaneously.

[0135] Here, only one signal is measured, which includes the response vibration signal 11. It is measured while the excitation signal 10 excites the vibration machine 2.

[0136] The operating oscillation signal 8 will dominate the measurement. If the measured signal is Fourier transformed using an FFT, the operating oscillation frequency will be identified as the strongest signal. The harmonics can then be determined as multiples of this frequency, including subharmonics with, for example, half the frequency.

[0137] The frequencies excited by the excitation signal 10 are also included in this FFT, but are superimposed by the operating oscillation and its harmonics.

[0138] During the Fourier transformation, the amplitudes and phases of the operating oscillation and its harmonics are also determined.

[0139] From this information, the frequencies, the amplitudes and phases, the correction signal 9 is formed, as explained in the above embodiment in step S3.

[0140] Subsequently, analogous to step S6, the subtraction signal is generated.

[0141] The advantage here is that only one measurement is used. This simplifies the measurement process. Furthermore, it is no longer necessary to ensure that the phase of the response oscillation signal 11 is identical to the operating oscillation signal 8. By using a common measurement, the phase is automatically identical.

[0142] The methods described here not only allow resonance measurements to be performed during operation, but the resulting measurements are also more accurate because they more closely reflect reality. A systematic deviation of the method described here compared to the conventional method in off-load operation without a load is evident. However, since the vibrating machine is in loaded operation most of the time, it is very advantageous to know the natural frequencies during loaded operation rather than those during off-load operation. This can effectively prevent a resonance catastrophe.

[0143] The vibrating machine 2 can also be designed as a vibrating screen 2 instead of a vibrating conveyor 2. Due to the different screen layers, the loading condition can also change during operation. This can have an impact on the natural frequencies, even if these are usually undesirable. Changes can also occur due to changes in the bulk material properties of the plant process.

[0144] Any changes in the bulk material behavior will again have an impact on the resonance behavior. Reference symbol

[0145] 1System for resonance analysis of a vibrating machine 2Vibration machine 3Vibration detection device 4Evaluation device 5Vibration conveyor 6Vibration drive 7Vibration sensor 8 Operating vibration signal 9Correction signal 10Excitation signal 11Response vibration signal 12Subtraction signal 13 Impulse hammer

Claims

1. A method for resonance analysis of a vibration machine (2), in particular a vibrating screen or a vibrating conveyor (2), during operation, comprising the following steps: - determination of an operating vibration signal (8) comprising the vibrations of the vibration machine (2) in regular operation, - excitation of the vibration machine (2) during operation with an excitation signal (10) for exciting additional vibrations at the vibration machine (2), wherein the excitation signal is a Dirac pulse, - measuring a response vibration signal (11) of the vibration machine (2) to the excitation during operation, - determining a correction signal (9) by using the operating vibration signal (8), - determining a subtraction signal (12) by subtracting the correction signal (9) from the response vibration signal (11), and - frequency analysis of the subtraction signal (12) for determining the natural frequencies of the vibration machine, wherein the dominant frequencies of the subtraction signal (12) correspond to the natural frequencies of the system in the frequency range of the excitation signal (10).

2. The method according to claim 1, characterized in that the operating vibration signal is modified before it is used as a correction signal.

3. The method according to claim 2, characterized in that when generating the correction signal (9): - the frequency with the highest amplitude is determined from the operating vibration signal (8), hereinafter referred to as the operating frequency, - the amplitude and phase of the operating frequency are determined in the operating vibration signal (8), - the amplitude and phase of the harmonics of the operating frequency in the operating vibration signal (8) is determined, - the correction signal (9) is generated from the amplitude and phase of the operating frequency and from the amplitude and phase of the harmonics of the operating frequency.

4. The method according to claim 3, characterized in that the harmonics are one or more of the multiples of the operating frequency, preferably 1 / 2, 3 / 2, 1 / 3, 2 / 3, 2, 3, 4, 5, 8, 16, and 32 times the operating frequency.

5. The method according to any one of claims 2 to 4, characterized in that the operating vibration signal (8) is estimated by automatic parameter estimation and modified using the resulting parameters.

6. The method according to claim 1, characterized in that the correction signal (9) is modeled on the operating vibration signal (8) by taking up the operating vibration signal (8) over a predetermined duration and using it as the correction signal (9).

7. The method according to claim 1, characterized in that the operating vibration signal (8) is used as correction signal (9).

8. The method according to any one of claims 1 to 7, characterized in that the natural frequencies are determined in a frequency range of greater than 0.01 Hz, preferably of greater than 0.1 Hz and in particular of greater than 1 Hz.

9. The method according to any one of claims 1 to 8, characterized in that the natural frequencies are determined in a frequency range of up to less than 1 kHz, preferably up to less than 500 Hz and in particular up to less than 50 Hz.

10. The method according to any one of claims 1 to 9, characterized in that after the step of frequency analysis, the operating frequency is changed and all steps are repeated.

11. The method according to any one of claims 1 to 10, characterized in that an excitation vibration signal (10) is transmitted to the vibration machine at several points.

12. The method according to any one of claims 1 to 11, characterized in that a response vibration signal (11) is measured at several of points.

13. The method according to any one of claims 1 to 12, characterized in that the determination of an operating vibration signal (8), the excitation of the vibration machine (2) and the measurement of a response vibration signal (11) are superimposed in time and / or take place simultaneously.