• 专利附录
  • 专利说明
  • 权利要求
  • 类似技术
  • 同族专利
  • 引用文献
  • 法律状态
  • 优先权
    • 专利摘要:
    A method for the early detection of a crack (64) in a wheel set (10) for a rail vehicle comprises generating a measurement signal by a fixedly mounted measuring sensor (15) due to a vibration of the wheel set (10). A shift in the natural frequency of a higher-order mode shape of the wheel set (10) is determined from the measurement signal in order to determine or assess the crack (64), the vibration that occurs during operation of the wheel set (10) being used to excite (65) the mode shape becomes.
    公开号:CH715397A2
    申请号:CH00596/19
    申请日:2019-05-06
    公开日:2020-03-31
    发明作者:Bättig Bruno
    申请人:Vibro Consult Ag；
    IPC主号:
    专利说明:

    background
    The present invention relates to a method and a device for the early detection of a crack in a wheel set for a rail vehicle.
    The history of the railroad knows many events where defective wheelsets led to tragic personal accidents, to consequential damage to rail vehicles, track systems and in driving operations. Wheelsets are safety-critical components and are designed, maintained and controlled with correspondingly high requirements.
    Proven measures to achieve high reliability are oversizing and age-related and increasingly also condition-dependent replacement of critical components.
    With the technical development new control possibilities came up. With regard to the crack testing of shafts - which is the focus of this invention - these were, for example, optical, metallurgical, magnetic and radiological methods. The disadvantages of these methods are the dismantling and assembly effort, the lack of accessibility, the restriction to service intervals, the limited informative value (for example, the depth of the crack), etc., as well as the generally great effort.
    With the advances in vibration measurement technology, monitoring of rotating parts is nowadays easily possible and is also used successfully in mechanical engineering, usually to monitor the vibration amplitudes. The monitoring of rolling bearings and gears has also proven itself, but this does not yet represent the general state of the art in rail technology. The crack monitoring of shafts is more demanding, since cracks only change the running behavior of the shafts at an advanced stage.
    State of the art
    [0006] Methods and devices which evaluate vibration signals measured on wheel set bearings or bogies are known in rail technology. The assessment criterion for determining defects is the signal strength, i.e. the vibration amplitude (DE 19 837 554 A1).
    The comparison of measured and historical spectra (frequency profiles) is known from DE 10 022 684 A1, DE 19 827 271 A1 and WO 00/60 322 A1, which corresponds to a frequency-wise comparison of the vibration amplitudes.
    [0008] From DE 10 062 602 B4, the comparison between expected and measured signal components which are harmonic to the wheel speed is known. Here too, vibration amplitudes are compared.
    [0009] The above devices have in common the vibration amplitudes as an assessment criterion, that is to say the vibration strength of the signal. The challenge posed by the large dynamic range of operational vibrations is tackled with different types of signal processing before the assessment.
    Another approach is known from WO 2010/057 623 A2. A sensor mounted on the shaft rotates with the shaft and transmits measurement data to the rail vehicle by radio. By monitoring the natural resonance frequency of the shaft, damage or varying loads on the shaft can be determined. Cracking is mentioned in this document, for example.The document shows how a measuring device rotating with the shaft can be mounted at all, can be supplied with sufficient energy and how its measurement data can be transmitted. These paths are laborious and restrictive.In the method with a rotating transducer, the vibration transducer and crack are at a fixed angle to each other and rotate synchronously with each other. As a result, the transducer always measures the shaft at the same angle, which means that the crack-related angle of rotation dependency of a natural frequency of the shaft cannot be recognized.
    Measures to protect the shaft, such as a tire tread placed under the sensor, limit its frequency response and thus the evaluation of higher-order eigenmodes. The evaluation of higher-order natural forms also requires additional computing power and thus more energy, while lower natural frequencies are difficult to distinguish from drive-typical stimuli.
    Signal components typical of cracks are weak compared to the strongly varying operating vibrations and must be measured, separated and processed for assessment. There is therefore a need for a structurally simple and amplitude-independent signal evaluation for monitoring cracks in wheel sets.
    Object of the invention
    It is the object of the invention to develop a method to detect shaft cracks in wheel sets of rail vehicles at an early stage, so that broken axles can be avoided with high reliability.
    The object of the invention is also to provide a device for monitoring or diagnosing the condition of a wheelset, for example a shaft, which makes it possible to detect a defect before a possible failure of the wheelset can occur with an increased load.
    Description of the invention
    The object of the invention is achieved by a method for the early detection of a crack in a wheel set according to claim 1. The object is also achieved by a device for monitoring or condition diagnosis of a wheel set according to claim 10. Advantageous embodiments of the devices are the subject of Claims 11 to 13. Preferred uses of the devices are the subject of claims 14 to 15.
    [0016] Damage to wheels, wheelset bearings or gearbox damage are noticeable through abnormal driving noises and deteriorated running behavior can also be determined with known vibration sensor-based monitoring systems. Cracks in waves are more difficult to see. Cracks often only become noticeable when they are big enough that the external shape of the wheelset and thus the running behavior change. The last phase up to the claim is progressive and comparatively short.
    The present method is based on the fact that a crack changes internal properties of the shaft and this in turn changes its rigidity. Eigenforms (also called natural vibration mode or vibration mode) and the associated natural frequencies are such properties. When the shaft is excited to vibrate, the natural frequencies can be determined by means of a vibration sensor.
    The basic formula applicable to undamped, mechanical vibrations:
    with: fE natural frequency in [Hz] c spring constant (here: stiffness) in [N / m] m mass in [kg]shows the essential parameters that determine the natural frequency, namely stiffness and mass. For a given wheelset geometry, a given wheelset mass distribution and a certain order of the eigenmode:
    There is thus a direct relationship between natural frequency and rigidity.
    In the usually rotationally symmetrical shafts, the stiffness is the same with respect to all angles of rotation. With a crack, the shaft usually loses its rotational symmetry and the rigidity becomes dependent on the angle of rotation. This behavior can be used to detect cracks.
    If the term "for example" is used in the following description, this term refers to exemplary embodiments and / or embodiments, which is not necessarily to be understood as a more preferred application of the teaching of the invention. In a similar manner, the terms “preferably”, “preferred” are to be understood by referring to an example from a set of exemplary embodiments and / or embodiments, which is not necessarily to be understood as a preferred application of the teaching of the invention. Accordingly, the terms “for example”, “preferably” or “preferred” can relate to a plurality of exemplary embodiments and / or embodiments.
    The following detailed description contains various exemplary embodiments of the method according to the invention and the device according to the invention for monitoring or diagnosing the condition of a wheel set. The description of a specific device of a specific method is to be regarded as exemplary only. In the specification and claims, the terms "include", "comprise", "have" are interpreted as "including, but not limited to".
    [0023] In the following, the term crack size is used for one or more cracks or the sum of the crack sizes of the cracks.
    If the term natural frequency or resonance frequency is used in the following, this means the natural resonance frequency of the natural shape of a certain order.
    In the following, the term shift in connection with the natural frequency means a single or multiple change in the natural frequency. The natural frequency takes on at least two different frequencies one after the other. This includes, for example, commuting between two different frequencies or changing between a previous and a current frequency. In the same context, the term “difference” means the degree of change or the difference between the different frequencies. The difference can arise, for example, from changing the shaft rotation angle or from changing the shaft properties over time.
    The term wheelset also includes parts of the wheelset, for example the shaft or the wheels.
    For the term eigenmode, the synonymous terms natural oscillation form, oscillation mode or mode are also used.
    The term order is used to number the various eigenmodes. In the one-dimensional approach, the fundamental oscillation has the first order and the lowest natural frequency. Higher orders are characterized by higher natural frequencies.
    A vibration sensor, also called a vibration sensor, is understood to mean a measurement sensor for the measured variable vibration displacement or vibration speed or vibration acceleration. This invention uses a "fixed mount vibration sensor". This means that it is mounted, for example, on the wheel bearing housing or on the gear housing and does not rotate with the wheel set.
    The method according to the invention is used for the early detection of a crack in a wheel set for a rail vehicle. A permanently mounted sensor generates a measurement signal due to a vibration in the wheelset. The measurement signal is used to determine a shift in a natural frequency of a higher-order natural shape of the wheelset to determine or assess the crack, with the vibration that occurs during operation of the wheelset being used to excite the higher natural shape.
    Early detection is to be understood here in particular as the early detection of the existence or change of a crack, also referred to as crack formation. Assessment of crack formation means recording the characteristics of the crack, including the location of the crack on the wheelset, the depth of the crack in the wheelset, the length of the crack on the surface or inside the wheelset, the width of the gap, which is formed by the crack, and the change over time can include at least one of the aforementioned characteristics. Simplified and generalized, it is about the early detection of the change in the number, position or shape of the crack (s).
    When driving, the wheelset is exposed to vibrations which are caused by itself, by the vehicle or, above all, by the track (for example rail bumps, rail defects, switches, cornering).
    Very weak to very strong vibrations occur during driving, which generate vibration signals with small (e.g. 2) or large (e.g. 2000) amplitudes. The dynamic range, i.e. the amplitude ratio of the vibration signals, is 1000 (2000/2) for this hypothetical example. A large dynamic range poses a challenge for the measurement technology and evaluation used. Some methods from the prior art record the size of the amplitude of the vibration signals, which are subject to the large dynamic range. No conclusion about a change in the frequency of the oscillation signal can be derived from a change in the amplitude. Vibration amplitudes that occur when the wheelset is in operation are used to excite the eigenmode. The necessary measurements of the displacement of the natural frequencies are possible in a large dynamic range of vibration amplitudes, which are caused by the driving operation. In particular, the method makes it possible to dispense with the monitoring of the magnitude of oscillation amplitudes.
    A feature of the invention is that the vibration sensor is firmly mounted. In this arrangement, the shaft - and with it any crack - rotates in relation to the transducer. The preferred measurement direction is also vertical, in which the strongest excitations occur. As a result, the sensor scans all wheel rotation angles equally with every wheel rotation. A feature of a crack that can be evaluated with this method is that the observed natural frequency of the shaft changes periodically and synchronously with the wheel rotation angle.
    When driving, the vibration signals are composed of different frequency components. Part of this is frequency components typical of drives, which are due to wear and tear or construction and are not directly related to frequency components typical of risk. The sources of such drive-typical frequency components are, for example, wheel flat spots, rollers of the wheel bearings or teeth of the transmission gears. A feature of the invention is that, for crack detection, it evaluates higher-order natural frequencies which, in terms of frequency, are above the frequency components typical of the drive.
    In one embodiment, in a rotationally symmetrical wheel set, the shift of the natural frequencies from the natural frequency rotational symmetry with respect to the wheel rotation angle can be used for crack assessment. In particular, the method can be used for rotationally symmetrical wheel sets for crack monitoring while driving or in the factory.
    In another exemplary embodiment, the shift in the natural frequencies of higher orders compared to previous measurements can be used for crack assessment in a wheelset. In particular, the method can be used for wheel sets for crack monitoring during driving or in the factory.
    According to one exemplary embodiment, eigenforms from the 5th order can provide ratios between useful and interference signal components which are suitable for evaluation. According to one embodiment, a higher order eigenmode can be used for crack detection, the resonance frequency of which is in the upper 10% of the measured spectrum in terms of power. According to one embodiment, a higher order eigenmode can be used for crack detection, the resonance frequency of which is in the frequency range above the drive-typical excitations.
    In particular, by means of the method according to one of the exemplary embodiments, the existence of a crack in the wheel set can be determined without the position or location of the crack having to be known. A crack can be detected at any position or location in the wheelset. The localization of the crack can - if of interest - take place afterwards. With a method according to one of the exemplary embodiments, all types of cracks can be determined, since all of them reduce the rigidity. Crack types include cracks in any spatial position and orientation or cracks due to bending or torsional overload with their typical appearances.
    In particular, cracks can be recognized by means of the method according to one of the exemplary embodiments without plastic deformations having to have already occurred on the wheelset.
    Changes in the wave through cracks result in shifts in the natural frequencies. The larger the crack and the more cracks there are, the more natural frequencies shift in the spectrum and can be detected with this method.
    With the shift of the natural frequency, be it as a function of the wheel rotation angle or as a function of time, a measure is available for the quantitative estimation of the crack size, as well as for the observation and prognosis of the course of the crack size over time, i.e. the crack development.
    According to one exemplary embodiment, a wheel set can be checked in the factory, the natural shape being excited by a stationary, mechanical testing device. According to this exemplary embodiment, vibrations that are triggered by a stationary, mechanical test device are used to excite the eigenmode. If the natural shape is excited at different angles of rotation, the rotational asymmetry of the natural frequency can be used to monitor cracks. A temporal change in the natural frequency compared to previous measurements can also be used to monitor cracks. The method can thus be used to monitor cracks while driving or in the factory.
    A method according to one of the preceding embodiments has the advantage that a wheel set can be monitored simultaneously in the current state for change in the rotational asymmetry of the natural frequency or for change in the state compared to previous measurements by shifting the natural frequency over time. This means that two independent criteria are available for assessing the condition of the wheelset.
    A method according to one of the preceding exemplary embodiments has the advantage that it already works with a single vibration sensor per wheel set.
    A method according to one of the preceding exemplary embodiments has the advantage that a measurement signal from a vibration sensor can be used for the simultaneous monitoring of various eigenmodes for changes typical of cracks.
    If the natural frequencies are subject to known or learnable influences due to boundary conditions, the method can be additionally improved by compensating the influences in terms of sensitivity, security, reaction time, susceptibility to false alarms, etc. Examples of influences are wheel wear and ambient temperature.
    A device for monitoring or diagnosing the condition of a wheelset for cracks contains a vibration sensor which is permanently mounted on the rail vehicle and which, when mechanically stimulated, emits a vibration signal to the monitoring unit, which determines the signal ΔfEN of a natural frequency of a higher order eigenmode. The signal ΔfEN is a measure of the current weakening of the wheelset due to crack formation. The state of the wheelset with regard to strength can thus be assessed using the signal ΔfEN.
    To assess the condition of the wheelset, this method uses the shifts in the natural frequencies of natural forms in contrast to the monitoring of vibration amplitudes by means of amplitude limit values, which is applied to all or to selected frequency components.
    The vibration sensor can be attached to or close to the wheel set, but is not co-rotating. Advantageous mounting locations are on the stationary housings of wheel bearings or gear bearings.
    A variety of common oscillation displacement, oscillation speed and oscillation acceleration sensors of different measurement principles can be used as the oscillation sensor.
    In the monitoring unit, the vibration signal is analyzed in a computer using generally known mathematical and / or numerical methods. The mathematical and / or numerical methods contain in particular at least one element from the group of Fourier transformation, filtering, normalization, averaging and weighting.
    In the case of a round shaft - provided that the shaft consists of a homogeneous material - the rigidity is rotationally symmetrical with respect to the axis of rotation, that is, constant over the circumference. Cracks are usually not rotationally symmetrical and cause different natural frequencies for an eigenmode of a certain order, which depending on the direction of the excitation in relation to the circumferential position, respectively. are the wheel rotation angle ϕ, as Fig. 5a shows. The rotational asymmetry of the natural frequencies, for example of order N, can be determined and used for crack assessment. The criterion can be, for example, the difference ΔfEN between the maximum and minimum natural frequency fENmax, fENmin. With this assessment, material inhomogeneities, especially cracks, can be determined, even if no historical natural frequencies are known, such as with new or first-time tested shafts.
    In particular, the state of the shaft can be ascertained and / or assessed without knowledge of an earlier state. In particular, the device can be used to detect and / or assess cracks without knowledge of a previous state.
    According to one embodiment, the vibration sensor can be connected to a pre-filter, a pre-filtered vibration signal can be generated in the pre-filter, the pre-filtered vibration signal being digitizable in an analog / digital converter, whereby a digitized vibration signal S is available.
    According to one exemplary embodiment, the oscillation signal S can be broken down into a plurality of segments, for example into segments a to z, in a segmentation unit. According to this exemplary embodiment, the segmentation takes place as a function of the wheel angle. The segments can each be transformed into the spectral forms by means of a Fourier transformation in order to calculate useful spectra from them in subsequent signal processing. The natural frequencies, for example fENa to fENz, of a certain eigenform of a higher order can be determined from the useful frequencies with an associated resonance frequency measurement, of which the highest determined natural frequency fENmax and the lowest determined natural frequency fENmin can be determined in a selector and from this the natural frequency can be determined in a subtractor -Difference ΔfEN = fENmax - fENmin can be calculated. In particular, the natural frequency difference ΔfEN is a measure of the number or size of the existing cracks and can be used to assess the condition of the wheelset.
    Signal processing can contain filtering, normalization, averaging, weighting.
    At a minimum, one eigenform is monitored. Each wheel set can have several natural modes suitable for monitoring, each with its characteristic natural frequency. A monitoring unit can be used to simultaneously monitor several natural forms measured by a sensor, in particular higher and / or lower order natural forms, for example for plausibility checking or verification of the assessment, for increasing safety, reliability or sensitivity, for improving the reaction time or for avoiding False positives. According to one embodiment, a natural frequency difference ΔfEM for another higher order eigenmode, for example identified by the index M, can be calculated from the segments and used in a manner analogous to ΔfEN for assessing the wheelset.
    In a device according to another exemplary embodiment, the natural frequency fEN 'of a higher-order natural form N can be determined in the monitoring unit from the digitized vibration signal S in a resonance frequency determination. A difference ΔfEN ́ between the natural frequency fEN ́ and a previously determined reference natural frequency fRN ́ of the same order can be determined. The shift ΔfEN1 is a measure of the weakening of the wheelset that has taken place in the meantime due to the formation of cracks. Thus, the condition of the wheelset with regard to strength can be assessed by shifting the natural frequency ΔfEN ́.
    In the determination of the resonance frequency, the digitized oscillation signal S can be transformed into the spectral form by means of Fourier transformation, with a useful spectrum being calculable in a subsequent signal processing.
    Signal conditioning can contain filtering, normalization, averaging, weighting.
    Known or learned influences of boundary conditions can be correctable on the natural frequencies fEN 'before the comparison with the reference natural frequency fRN'. A boundary condition can be, for example, the wheel set temperature or the wheel wear. For example, it is known that the modulus of elasticity of steel decreases with increasing temperature. If fEN ́ and fRN ́ are measured at different wheel set temperatures, the calculated shift in natural frequency ΔfEN ́ does not exactly represent the actual crack formation. Since the temperature influence on the natural frequency is known, the uncorrected natural frequency can be corrected up or down by the temperature influence, which results in the corrected natural frequency fEN ́, whereby fEN ́ and fRN ́ are based on the same reference temperature and the calculated ΔfEN ́ the measure the crack formation better represented.
    At a minimum, one eigenform is monitored. Each wheel set can have several natural modes suitable for monitoring, each with its characteristic natural frequency. According to one embodiment, a natural frequency difference ΔfEM 'for another higher-order eigenmode, for example identified by the index M, can be calculated from the signal S and used in a manner analogous to ΔfEN, ΔfEM or ΔfEN1 to assess the wheelset.
    The difference, which corresponds to the shift in the natural frequency ΔfEN ', is a measure of the severity of the damage. The damage can include, for example, a crack, in particular a crack that has formed since the start of monitoring. The shift in the natural frequency ΔfEN ́ can be used to assess a condition, for example the strength, of the wheelset. The condition of the wheelset can thus be assessed by shifting the natural frequency ΔfEN ́. The state is understood to mean the extent or progress of the weakening through the crack or cracks. This means that the condition shows the functionality, for example the remaining strength, of the wheelset.
    The device according to one embodiment can monitor the wheelset with a single vibration sensor.
    The device according to each of the exemplary embodiments can be expanded to monitor other types of damage such as wheel damage, wheel set bearing or gear damage, derailment, track damage, etc. without further hardware.
    The device can be expanded with one or more further sensors to form a holistic monitoring of wheelsets, bogies, vehicles or tracks.
    The device according to one of the exemplary embodiments can include monitoring of its own functionality. In particular, a defect such as a malfunction or a functional failure can impair functionality.
    An apparatus can comprise one or more components. In particular, a component can be part of a method or an apparatus or an application example. In particular, a component can form an assessment of the condition of the wheelset that is used as an alternative or in addition to other components. For example, a device can monitor a natural frequency difference ΔfE and at the same time monitor the shift in the natural frequency ΔfE '.
    The method is used in particular to prevent axle breakages.
    The method is particularly suitable for continuous crack monitoring while driving. A single vibration sensor is sufficient for this, which is permanently attached at a suitable point on or in the immediate vicinity of the wheel set and whose signal is analyzed by a monitoring unit. The natural frequencies of the wheelset are determined using the operational vibrations and monitored for frequency shifts. One advantage of this application is the control of highly dynamic vibration amplitudes, such as those that occur in rail traffic.
    The device according to one of the exemplary embodiments can thus be used to monitor a wheel set of rail vehicles for a type of damage from the group of cracks, wheel damage, wheel set bearing or gear damage, derailment.
    The device according to one of the preceding exemplary embodiments is used, for example, for a rail vehicle in operation or for checking wheelsets in the factory.
    As an alternative or in addition, the method can be used for individual measurements in the factory, for example during checks in the maintenance interval of the wheelset. Temporary or permanent sensors can be used. Known methods such as impact, use of shakers and test drives can be used for mechanical excitation of the natural frequencies.
    The method is therefore suitable for combination with further monitoring methods in the sense of expansion or as backup functions. Examples of known processes from the machine industry or railway technology are; Monitoring of bearings, gears, broken wheels, broken axles, derailments or tracks.
    Brief description of the drawings
    The methods according to the invention, as well as associated devices, are illustrated below with the aid of a few exemplary embodiments. FIG. 1 shows a schematic structure of a device according to an exemplary embodiment, wherein two different evaluation methods can be used, FIG. 2 shows details of block 25, which occurs several times in FIG. 1, for determining the natural frequencies, FIG. 3 shows a wheel set 10 which has a sensor 15 is monitored, the wave center 61 at rest and with two of the possible eigenmodes 62N and 62M as well as a possible crack 64, FIG. 4 the general case of a natural frequency shift ΔfE of the eigenmodes of the Nth and Mth order in the spectrum, FIG. 5a a polar diagram of the rotationally asymmetrical natural frequency fEN with a torn shaft, FIG. 5b a diagram for determining the rotational asymmetry of the natural frequency ΔfEN according to assembly 50 in FIG. 1.
    Detailed description of the drawings
    Some elements are shown in dashed lines in the figures. These are options that are possible as an alternative or in addition to the initial variant shown in solid lines.
    1 shows a schematic structure of an exemplary device, consisting of the monitored wheel set 10, the monitoring unit 20 and the reporting unit 40.
    The wheelset 10 consists of the wheels 11a, 11b, the shaft 12, the wheelset bearings 13a, 13b and the gearbox 14. It stands on the rails 18a, 18b.
    A vibration sensor 15 is fixedly mounted, for example on the wheelset bearing 13a. An alternative mounting location 16 for the vibration sensor is on the gearbox 14. Other locations at which wheelset natural frequencies can be measured are also possible. All vibration sensors are mounted on fixed parts of the wheel set and do not rotate with the shaft or the wheels, so that simple power supply and data communication via cable is possible. The additional sensor 17 is explained below with option 55.
    The vibration sensor 15 is connected to the monitoring unit 20. The signal of the vibration sensor 15 is pre-filtered in the pre-filter 21 and digitized in the analog / digital converter 22, whereby the vibration signal S is produced.
    The assembly 50 (associated diagrams, FIGS. 5a and 5b) shows an exemplary embodiment for a device by means of which the rotational asymmetry of a natural frequency can be determined for crack assessment. A prerequisite for this exemplary embodiment when used in driving operation is a wheel rotation angle detection 24 in which the wheel rotation angle ϕ is measured or derived from the vibration signal S. In assembly 50, the vibration signal S is fed to the segmentation 23 and is converted into the angular resp. Time segments a to z (a = 1, z ≥ 3, here as an example: z = 8). In the natural frequency determination 25 (details in FIG. 2), the natural frequency of a natural shape with a higher order (here designated as N) is determined from each segment. Orders with N = 5 or higher have proven advantageous. For example, the natural frequency fENa of the Nth order is determined from segment a. The natural frequencies fENb to fENz are also determined. The highest and lowest natural frequencies fENmax, fENmin are selected in the minimum / maximum value selector 26 and the natural frequency difference ΔfEN is calculated in the subtracter 27. ΔfEN is a measure of the number and size of the cracks in the shaft.
    A reporting unit 40 shows different types of display for the natural frequency difference ΔfEN, for example as a direct display 41 in [Hz]. Optionally, ΔfEN can also be visualized as a “crack index” in [%] in display 42 or in trend display 43 according to application-specific scaling 30. The optional assessment 31 compares ΔfEN with frequency limit values and actuates the traffic light 44. In analogy to a traffic light, the visual displays can be red, yellow or green, for example. The red display signals danger, the yellow display signals an advance warning and the green display signals that the wheelset is free of damage, in particular free of cracks.
    Option 51 shows in summary how the natural frequency difference ΔfEM of a further order (here, for example, designated as M) can be determined from segments a to z.
    Analogously to the natural frequency difference ΔfEN, ΔfEM is displayed by the reporting unit 40.
    The method and the associated device can also be used to control wheelsets in the factory. The inspection can take place on the stationary rail vehicle or on the dismantled wheelset. At several different wheel rotation angles, corresponding to segments a to z, the shaft can be excited to vibrate with a test device and the natural frequencies can be measured. In order to assess the condition of the shaft, natural frequency differences ΔfEN, ΔfEM of the higher orders N, M are determined.
    The assembly 52 shows an exemplary embodiment of a device by means of which a shift in a natural frequency can be evaluated for crack assessment. The signals in module 52 are marked with an apostrophe (́) to distinguish them from module 50.The natural frequency fEN 'of an eigenform with a higher order N is determined from the vibration signal S in the natural frequency determination 25 (details in FIG. 2). Orders with N = 5 or higher have proven advantageous. The reference value memory 28 outputs a natural frequency reference value fRN 'of the same order N. This can be a fixed value or contain variable influence corrections. Ideally, it represents the healthy wheelset under standard boundary conditions and is based, for example, on a natural frequency measurement on the healthy wheelset when new. The shift in the natural frequency ΔfEN '= fRN - fEN1 is calculated in the subtracter 27. ΔfEN ́ is a measure of the number and size of the cracks in the shaft.Analogous to the natural frequency difference ΔfEN from assembly 50, ΔfEN ́ is displayed by the reporting unit 40.
    Option 53 shows in summary how the displacement of the natural frequency ΔfEM 'of a further order (here, for example, designated as M) can be determined from the vibration signal S in the same way. Analogous to the natural frequency difference ΔfEN from assembly 50, ΔfEM ́ is displayed by the reporting unit 40.
    Option 54 shows in summary how the wheel set can be monitored in a simple manner with the same sensor 15 for other types of damage. For example, the signal amplitude of the oscillation signal S can be monitored in order to determine defects in the wheel, axle box, gearbox and axle as well as derailment. This monitoring is possible without further sensors, as the vibrations spread over the entire wheel set. If abnormally strong vibrations are measured, the engine driver is alerted. This type of monitoring is well known and is used to prevent or reduce consequential damage. Further monitoring evaluates, for example, wheelset bearing or transmission-typical vibration components or vibration patterns in the vibration signal S, which report defects or the onset of wear and tear at an early stage so that the vehicle can be taken out of service as planned. These processes are also known and are used successfully in larger machines. The interesting thing about option 54 is that it works without additional hardware, uses the monitoring infrastructure that is already in place and supplements the crack monitoring system with all-in-one monitoring.
    Option 55 shows how further sensors can be connected to the monitoring unit 20. This can be, for example, a vibration sensor that monitors a further wheelset in a similar way to sensor 15, so that a bogie monitoring or vehicle monitoring results from the wheelset monitoring. However, the sensor does not necessarily have to be a vibration sensor; it can also be a temperature sensor, for example, which triggers an alarm message when the wheelset bearing 13a / 13b is overheating.What is interesting about option 55 is that the protected area can be expanded cost-effectively with little additional sensor technology.
    Option 32 is a system monitor and detects defects in the device and reports abnormalities or malfunctions. Examples of defects are a failure of the power supply, a defective vibration sensor or a computer failure.
    FIG. 2 shows the structure of the natural frequency determination block 25, as it is used several times in FIG. Fig. 2 is shown and described for determining the natural frequency of the order N, but applies equally to other orders such as the order M.The input signal from 25 is an oscillation signal in the time domain that comes as one of the segments a to z from the segmentation 23 or as a complete oscillation signal S directly from the analog / digital converter 22. The input signal is transformed into the frequency range by means of Fourier transformation 251. In the signal processing 252, the useful spectrum of the selected order N (for example N = 6 for the 7-node oscillation) is determined by filtering, normalization, averaging, weighting, etc. and fed to the resonance frequency determination 253, which uses the current, uncorrected Natural frequency determined. Without the optional correction module 254, the uncorrected natural frequency is output unchanged as the current natural frequency. Depending on the input signal of block 25, this can be fENa ... fENz ́ or FEN ́. The optional correction module 254 is used when the uncorrected natural frequency is subject to known (or learnable) influences. For example, the temperature-compensated natural frequency fEN, which is independent of temperature fluctuations, can be calculated from the uncorrected natural frequency fEN with the aid of a temperature signal y.
    FIG. 3 shows - greatly exaggerated - two possible eigenmodes 62N and 62M of the excited wave which can be used for crack monitoring. In the idle state, the eigenforms cannot be measured because there is no excitation and the wave center is stationary, shown as line 61.
    When traveling over rail joints and bumps, switches, etc., the wheelset 10 is shaken by suggestions 65 from the rail. These excitations can have very weak to very strong amplitudes. The excitations are desirable here, as they stimulate the wave to vibrate in its natural forms, so that the vibration signal can be analyzed. A crack 64 reveals its presence by shifting the monitored natural frequencies down the spectrum.One advantage of the invention is that a permanently mounted sensor can continuously scan the circumference of the passing shaft for cracks: A crack changes its position both with respect to the vibration sensor 15 and its measuring direction, as well as with respect to the excitation 65 and its direction of force. The associated changes in the vibration behavior of the shaft can be recorded and evaluated.
    4 shows normalized resonance curves of the eigenforms of the orders N and M in the spectrum. The amplitudes are standardized in order to control the large dynamic range of the excitations.
    The abscissa values are labeled twice: The upper labels (fENmin, fENmax, fEMmin, fEMmax) show the variant in which the resonance frequency shift is monitored as a function of the angle of rotation, corresponding to assembly 50 in FIG. 1.The lower values (fEN ́, fRN ́, fEM ́, fRM ́) show the variant in which the resonance frequency shift is monitored as a function of time, corresponding to assembly 52 in Fig. 1.
    Curves 67N and 67M could be from a wave without a crack, while 66N and 66M have a crack. The associated shifts in the resonance frequencies by ΔfEN, ΔfEM, respectively. ΔfEN ́, fEM ́ makes use of this invention.
    5a shows a polar diagram of the rotationally asymmetrical natural frequency fEN as a function of the wheel rotation angle ϕ in the case of a cracked shaft. As a result of the crack, the stiffness and the natural frequency are usually rotationally asymmetrical. The polar representation shows the natural frequency as an ellipse with the semi-axes fENmax and fENmin and the difference ΔfEN, which corresponds to the difference between the maximum (fENmax) and minimum (fENmin) measured natural frequency. The method uses the difference between the semi-axes, i.e. the natural frequency difference ΔfEN, to assess cracks.
    FIG. 5b shows a Cartesian diagram of the rotationally asymmetrical natural frequency fEN as a function of the wheel rotation angle ϕ when the shaft is torn. The device according to the variant shown in FIG. 1, assembly 50 divides a full revolution of the wheel into segments. The natural frequency (here: fENa to fENz) is determined for each segment and the natural frequency difference ΔfEN is calculated from this.
    It is obvious to a person skilled in the art that many further variants are possible in addition to the exemplary embodiments described without deviating from the inventive concept. The subject matter of the invention is therefore not restricted by the preceding description and is determined by the scope of protection which is defined by the claims. The broadest possible reading of the claims is authoritative for the interpretation of the claims or the description. In particular, the terms “contain” or “contain” are to be interpreted in such a way that they refer to elements, components or steps in a non-exclusive sense, which is intended to indicate that the elements, components or steps can be present or can be used, that they can be combined with other elements, components or steps that are not explicitly mentioned. When the claims relate to an element or component from a group which may consist of A, B, C to N elements or components, this formulation is to be interpreted in such a way that only a single element of that group is required, and not one Combination of A and N, B and N, or any other combination of two or more elements or components of this group.
    List of reference symbols
    10 wheelset 11a, 11b wheels 12 wheelset shaft 13a, 13b wheelset bearings 14 transmission (if available) 15 vibration sensor (different functional principles are possible) 16 one of the alternative mounting locations for vibration sensor 15 17 optional, additional vibration sensor 18a, 18b rails 20 monitoring unit 21 Pre-filter for pre-filtering the vibration signal 22 Analog / digital converter 23 Segmentation unit for ϕ-dependent segmentation of S 24 Determination of the wheel rotation angle ϕ from S or by direct measurement 25 Determination of the resonance frequency of an eigenmode from the signal segments a to z or the complete signal S Internal details of 25 see in Fig. 2 251 Fourier transformation 252 Signal processing 253 Determination of resonance frequency 254 Optional correction of the natural frequency by means of y 26 Minimum / maximum value selector 27 Subtractor 28 Long-term memory of a previously determined reference natural frequency fRN ́ 30 Optional scaling of ΔfEN in user-friendly measures units 31 optional assessment of ΔfEN for controlling the traffic light 44 32 optional system monitoring 40 reporting unit with examples for the further use of ΔfEN, ΔfEM, ΔfEN ́ or ΔfEM ́ 41 direct display of ΔfEN in [Hz], e.g. 12 Hz 42 display of ΔfEN in [ %] of the permissible value, for example 78% 43 Trend display to visualize the temporal progression of ΔfEN 44 Possible crack display for the train driver as a traffic light (red-yellow-green) 50 Assembly group for determining ΔfEN 51 Optional determination (s) of ΔfEM further mode shapes 52 alternative / supplementary assembly for the determination of ΔfEN ́ 53 optional device (s) of further eigenmodes ΔfEM ́ 54 optional monitoring (s) for further types of damage 55 optional monitoring (s) for further sensors for vibrations or for other physical quantities 61 shaft center (axis of rotation) at rest 62N, 62M possible eigenmodes (course of the shaft center in the case of resonance, in a greatly exaggerated representation) 64 possible crack 65 possible excitation of the eigenmodes from the rail 66N, 66M resonance curves for a torn shaft 67N, 67M resonance curves for the same shaft without a crack a, b, ... z indices of the angle resp. Time segments of SN, M indices of the Nth, Mth eigenmode f frequency fE eigenfrequency of an eigenmode fEN, fEM eigenfrequencies of the Nth and Mth eigenmodes fENa, ..., fENz eigenfrequencies (of the Nth eigenmode) of Segments a, b, ... z fENmin, fENmax minimum and maximum natural frequencies (of the Nth order) ΔfE natural frequency difference of a natural shape ΔfEN, ΔfEM natural frequency differences of the orders N and M fRN ́ reference natural frequency natural shape (of the N- th order) ΔfEN * natural frequency shift of the eigenmode (of the Nth order) S digitized vibration signal y known or learned influence on a natural frequency ϕ wheel rotation angle
    权利要求:
    Claims (15)
    [1]
    1. A method for the early detection of a crack in a wheel set (10) for a rail vehicle, wherein a permanently mounted measuring sensor (15) that does not co-rotate with the wheel set generates a measurement signal due to a vibration of the wheel set (10), characterized in that from the Measuring signal, a shift in the natural frequency of a higher-order mode shape of the wheelset (10) is determined as a measure for determining or assessing the crack, the vibration which is emitted by the driving mode or by a test device on the wheelset (10) being used to excite the mode shape becomes.
    [2]
    2. The method according to claim 1, wherein the wheel set (10) is rotationally symmetrical, the shift of the natural frequency from the natural frequency rotational symmetry with respect to the wheel rotation angle being used as a measure for determining or assessing the crack.
    [3]
    3. The method according to any one of claims 1 or 2, wherein the shift of the natural frequency is used as a time change compared to the previously determined natural frequency as a measure for determining or assessing the crack.
    [4]
    4. The method according to any one of the preceding claims, wherein an intrinsic shape of a higher order is used for crack detection, the resonance frequency of which lies in the frequency range above the drive-typical excitations.
    [5]
    5. The method according to any one of the preceding claims, wherein the test device is a factory-located, fixed device which emits mechanical vibrations to the wheel set (10) as a test object.
    [6]
    6. The method according to any one of the preceding claims, wherein at the same time eigenmodes of different orders for seamless coverage of the wheel set (10) can be evaluated in the axial direction.
    [7]
    7. The method according to any one of the preceding claims, wherein the crack is detected before plastic deformations occur on the wheel set (10).
    [8]
    8. The method according to any one of claims 1, 3 to 7, wherein the degree of damage to the wheel set (10) can be determined without knowledge of an earlier state.
    [9]
    9. The method according to any one of the preceding claims, wherein a crack at any position or direction in or on the wheelset (10) can be detected.
    [10]
    10. Device for carrying out a method according to one of the preceding claims for monitoring or condition diagnosis of a wheel set (10), comprising a vibration sensor (15) which does not co-rotate with the wheel set and which is set up to emit a vibration signal to a monitoring unit (20) upon excitation , characterized in that in the monitoring unit (20) the natural frequency difference of an intrinsic shape of the higher-order wheel set can be determined and is used to determine or assess the crack.
    [11]
    11. The device according to claim 10, wherein the natural frequency difference in a first assembly (50) as the difference ΔfE between the maximum and the minimum, depending on the wheel angle measured natural frequency (fEmax, fEmin) or in a second assembly (52) the difference ΔfE ́ to a previously determined natural frequency of the same order can be used.
    [12]
    12. The device according to one of claims 10 or 11, wherein a wheel set can be monitored with a single sensor.
    [13]
    13. The device according to one of the preceding claims 10 to 12, wherein the vibration sensor (15) is set up in such a way that different eigenmodes can be monitored simultaneously.
    [14]
    14. Use of the device according to one of the preceding claims 10 to 13 for monitoring a wheel set (10) or bogie or vehicle or track system for a type of damage from the group of cracks, wheel damage, wheel set bearing or gear damage, derailment, track damage.
    [15]
    15. Use of the device according to one of the preceding claims 10 to 13 for monitoring a rail vehicle in operation or for checking a rail vehicle which is out of operation by means of a usually stationary test device, the test device emitting mechanical vibrations to excite natural vibrations on the wheel set as a test object.
    类似技术:
    公开号 | 公开日 | 专利标题
    EP1423827B1|2008-06-11|Device and method for the early recognition and prediction of unit damage
    DE102007024065B3|2009-01-02|Device and method for fault monitoring of chassis components of rail vehicles
    EP2730906A2|2014-05-14|Device and method for monitoring the state of a roller bearing
    EP2083338B1|2019-06-19|Process and device for monitoring a machine
    EP2937560B1|2018-06-06|Wind energy system diagnosis device for generator components
    DE102012213709A1|2014-02-06|A method for detecting a fault of a motor assembly with an electric machine and engine control unit
    EP3019845A1|2016-05-18|Method and arrangement for monitoring an industrial device such as, for example, a machine or a system
    DE102011089101B4|2018-03-29|Method and device for detecting a fault in a drive train of a vehicle
    EP2131178A1|2009-12-09|Diagnostic method for at least one ball bearing, in particular for an angular ball bearing, a corresponding diagnostic system and the use of such a diagnostic system
    EP2115412A2|2009-11-11|Method and device for monitoring a drive train having a highly flexible coupling
    EP3517927B1|2020-11-18|Method and device for early detection of a crack in a wheelset for a rail vehicle
    WO2015043619A1|2015-04-02|Method and device for monitoring the state of a drive system comprising an electrical drive unit
    EP3206933B1|2021-07-14|State diagnosis of rail vehicle wheels
    DE102015218941A1|2017-03-30|Method for detecting a failure of an acceleration sensor and measuring system
    DE102014203720A1|2015-09-03|Method and device for condition testing of a drive train of a traction vehicle
    DE102011083858A1|2013-04-04|Method and arrangement for monitoring gears in operation
    DE102013205353B4|2020-06-18|Process for maintaining a machine in the tobacco processing industry, and machine set up accordingly
    DE102019108569A1|2019-11-21|Device for monitoring the state of a traction device in a traction mechanism
    DE102015210911A1|2016-12-15|Method and device for detecting changes in an electrically driven drive
    DE102013211635A1|2014-12-24|Method and system for monitoring wear of a clutch in a motor vehicle
    DE102014226935A1|2015-12-17|Method for monitoring a wheel axle of a vehicle
    DE102019210795A1|2021-01-28|Stress wave transmission and method for stress wave transmission
    DE102017116839A1|2019-01-31|Method for fault monitoring for a vehicle, associated equipment, associated system and vehicle
    DE102016222660A1|2018-05-17|Method and device for detecting changes in an electrically driven drive
    DE102019201568A1|2020-08-13|Device and method for operating a rail vehicle
    同族专利:
    公开号 | 公开日
    EP3517927B1|2020-11-18|
    EP3517927A1|2019-07-31|
    引用文献:
    公开号 | 申请日 | 公开日 | 申请人 | 专利标题
    DE10062602B4|2000-12-12|2006-02-23|Db Fernverkehr Ag|Method and device for monitoring the behavior of rail vehicles and for diagnosing components of rail vehicles|
    US7640139B2|2004-10-18|2009-12-29|Nsk Ltd.|Abnormality diagnosing system for mechanical equipment|
    DE102009020428A1|2008-11-19|2010-05-20|Eureka Navigation Solutions Ag|Device and method for a rail vehicle|
    CN104364629B|2012-05-23|2017-08-29|国际电子机械公司|The track component inspection analyzed based on resonance signal|
    法律状态:
    2020-11-30| AZW| Rejection (application)|
    优先权:
    申请号 | 申请日 | 专利标题
    EP18153248|2018-01-24|
    EP18196364.6A|EP3517927B1|2018-01-24|2018-09-24|Method and device for early detection of a crack in a wheelset for a rail vehicle|
    [返回顶部]