Rope or cable fault detection method and system

The method and system improve fault detection in ropes by using a computational model trained with varied fault scenarios to accurately differentiate between single and multiple faults, enhancing the accuracy of fault identification and reducing unnecessary actions on ropes.

WO2024240985A1PCT designated stage expired Publication Date: 2024-11-28KONECRANES GLOBAL OY
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
PCT/FI2024/050239
Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
Priority Date
2023-05-19
Filing Date
2024-05-17
Publication Date
2024-11-28

AI Technical Summary

Technical Problem

Existing methods for detecting faults in ropes or cables struggle to accurately differentiate between single and multiple faults at the same or similar axial position, often due to limited resolution and small magnitude differences in magnetic flux leakage, which can lead to inaccurate detection and potential misuse of ropes.

Method used

A method and system that utilize a computational model trained with data from test ropes with intentionally created faults at various axial positions, using magnetic flux leakage measurements and machine learning algorithms to distinguish between single and multiple faults, allowing for precise detection of faults spaced less than the rope's width dimension.

Benefits of technology

The system enables accurate detection of multiple faults at close axial positions, improving fault resolution and preventing unnecessary rope discarding or repair, by correlating magnetic flux leakage data with a trained model, thus enhancing the reliability of fault identification.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure FI2024050239_28112024_PF_FP_ABST
    Figure FI2024050239_28112024_PF_FP_ABST
Patent Text Reader

Abstract

The present invention relates to a method of detecting a fault in a rope or cable comprising providing a test rope (26) comprising at least two faults (88), the faults (88) provided in the same position or spaced by an axial distance along the axial length of the rope (26) less than a width dimension of the rope. The method further comprises measuring the magnetic flux leakage of said test rope (26); inputting magnetic flux leakage data indicative of the faults in the test rope into a computational model; measuring the magnetic flux leakage of a further rope (62); and determining if the further rope (62) comprises one or more fault therein using said computational model and the measured magnetic flux leakage of the further rope (62).
Need to check novelty before this filing date? Find Prior Art

Description

[0001] Rope or Cable Fault Detection Method and System

[0002] The present disclosure relates to a method and system of modelling and detecting faults in a wire rope or cable.

[0003] Background of the Invention

[0004] A prior art method of detecting faults in a rope is described with reference to figure 1 . The rope 2 comprises a fault 4 in one or more component wire 6 of the rope 4. The fault 4 can be detected using magnetic flux leakage. As shown in figure 1 , the location of the fault 4 is indicated via a peak 8 in the measured magnetic flux leakage (i.e. the peak extending toward the right hand side of the page). Suitable filtering and / or software analysis can be used to correlate such peak to a fault 4, thereby indicating the position of the fault 4 within the rope 2. In accordance with ISO 4309:2017, the number and / or position of the faults may be used to determine the condition of the rope 4, and indicate whether such a rope 4 should be discarded or repaired.

[0005] The inventor has found numerous problems with the prior art arrangement. As shown in the centre figure, where two faults are provided in the same axial position along the length of the rope, the magnetic flux leakage may be larger in magnitude. However, it is often difficult to differentiate between a single fault and multiple faults in the same or similar axial position, as such a magnitude different may be small, and can be dependent on many factors. In general, special resolution of fault along the axial length of the rope is limited to the thickness of the rope (i.e. if a rope is 25mm thick, then faults with less than 25mm axial spacing may not be detectable). Given that multiple faults may occur in a similar axial position (e.g. due to similar loading conditions or due to transfer of load to adjacent wires after breakage of a given wire), it is important that multiple faults are detected accurately. The present invention aims to overcome or ameliorate one or more or the above problems, in particular to better detect multiple faults at a same or similar axial position in a rope.

[0006] Statement of Invention

[0007] According to a first aspect, there is provided: a method of detecting a fault in a rope or cable comprising: providing a test rope comprising at least two faults, the faults provided in the same axial position or spaced by an axial distance along the axial length of the rope less than a width dimension of the rope; measuring the magnetic flux leakage of said test rope; inputting magnetic flux leakage data indicative of the faults in the test rope into a computational model; measuring the magnetic flux leakage of a further rope; determining if the further rope comprises one or more fault therein using said computational model and the measured magnetic flux leakage of the further rope.

[0008] The method may comprise comparing the measured magnetic flux leakage with the computational model. The method may comprise determining if the further rope comprises two or more faults provided in the same position or spaced by an axial distance along the axial length of the rope less than a width dimension of the rope. The faults may be overlapping.

[0009] The method may comprise providing a plurality of test ropes. The test ropes may differ in at least one variable. The method may comprise inputting the magnetic flux leakage data of the respective test ropes into the computational model.

[0010] The rope may comprise one or more strands. The strands may comprise one or more wires or filaments.

[0011] The variable may comprise one or more of: a strand construction; a number of strands in the rope; a number of wires in each strand; and / or the diameter of the rope. The variable may comprise a number of faults at a given axial position. The variable may comprise: a number of faults in a given strand and / or a number of faults distributed across respective strands.

[0012] The method may comprise applying a load to the test rope during measurement of the magnetic flux leakage of the test rope. The load may be representative of real world conditions.

[0013] The method may comprise pre-processing the magnetic flux leakage data. The pre-processing may comprise filtering one or more frequency thereof. The preprocessing may comprise denoising the magnetic flux leakage data.

[0014] The method may comprise extracting data indicative of a fault from the magnetic flux leakage data. The method may comprise extracting data relating to one or more peak from the magnetic flux leakage data. The data indicative of a fault (e.g. a peak) may be paired or correlated with one or more known fault.

[0015] Data indicative of a fault may comprise a peak in the measured magnetic flux leakage data. One or more characteristic of the peak may be determined and input into the computational model. The characteristic may comprise one or more of: an amplitude; width; area; and / or characteristic shape of the peak.

[0016] The computational model may comprise one or more of: machine learn ing / artificial intelligence model; artificial neural network; and / or deep learning model. The computational model may comprise one or more of: a linear; non-linear; or parametric model. A plurality of computational models may be provided. The plurality of computational models may be provided in accordance with one or more parameter of the rope.

[0017] The method may comprise measuring the position of the sensor along the rope. The position data may be correlated with magnetic flux leakage data indicative of a fault. The position data may be correlated with a peak in the magnetic flux leakage data. Where a position is not provided, one or more peak and the known fault data may be manually correlated / paired.

[0018] The method may comprise providing an indication upon detection of a fault. The method may comprise providing an indication of the position of the fault and / or number of faults at a position. The method may comprise counting a number or type of faults. The method may comprise determining if the number / type of faults is outside a predetermined threshold / parameter. The indication may be provided in the event the number / type of faults is outside said predetermined threshold / parameter.

[0019] The method may comprise transmitting measured magnetic flux leakage data and / or indicative of a fault to a remote device.

[0020] The magnetic flux leakages measurements indicative of a fault in the further rope may be input into the computational model.

[0021] Measurement of the further rope may be provided in situ. Determination of the fault may be provided in real time.

[0022] According to a further aspect there is provided: a computer program or computer readable medium comprising program instructions which, when executed by the computer, cause the computer to carry out a computer process implementing the method according to any preceding claim.

[0023] According to a further aspect, there is provided: a method according to any aspect, comprising program instructions which, when executed by the computer, cause the computer to carry out a computer process to receive magnetic flux leakage data indicative of the faults in the test rope into a computational model to determine if the further rope comprises one or more fault therein using said computational model and the measured magnetic flux leakage of the further rope. According to a further aspect there is provided: A system for detecting a fault in a rope or cable comprising: a testing apparatus for supporting a test rope comprising at least two faults, the faults provided in the same axial position or spaced by an axial distance along the axial length of the rope less than a width dimension of the rope; a magnetic flux leakage sensor for measuring the magnetic flux leakage of said test rope; a controller configured to receive magnetic flux leakage data indicative of the faults in the test rope into a computational model; a sensor configured to measure the magnetic flux leakage of a further rope; and a controller configured to determine if the further rope comprises one or more fault therein using said computational model and the measured magnetic flux leakage of the further rope.

[0024] The testing apparatus may comprise a clamping means configured to apply a tensile load the test rope.

[0025] Any aspect of the invention may be combined with any other aspect of the invention where practicable.

[0026] Description

[0027] Embodiments of the present invention are described below, by way of example only, with reference to the accompanying drawings:

[0028] Figure 1 shows a prior art fault detection system;

[0029] Figure 2 shows a schematic overview of a fault detection system; Figure 3 shows a schematic view of a test rig for the system; Figure 4 shows a schematic view of a training step for the system; Figures 5A-H show schematic views of strands for training the system;

[0030] Figure 6 shows an example of a measured magnetic flux leakage; Figure 7 shows a schematic view of a measurement apparatus for the system;

[0031] Figure 8 shows a schematic view of a measurement step for the system;

[0032] Figure 9 shows a schematic view of a fault detection user interface.

[0033] An overview of a rope or cable fault detection system 10 is shown in figure 2. The system 10 is configured to detect faults (i.e. fractures, cracks, discontinuities, degradations, corrosion, or other features that may compromise the structural integrity of the rope / cable). The rope or cable is primarily constructed from a metallic or conductive material. In some embodiments, the rope or cable may comprise polymeric or non-conductive materials components.

[0034] In a first step 12, a test rope is provided. The test rope comprises a number of predetermined faults (i.e. the faults may be intentionally created). At least some of the faults are provided at the same axial position. In the next step 14, the magnetic flux leakage is measured along the length of the test rope. In the next step 16, the magnetic flux leakage measurements are used to create a model that can be used to detect faults.

[0035] In step 18, the magnetic flux leakage of a desired rope (i.e. a rope that requires real world testing) is performed. In step 20, the magnetic flux leakage is compared with the model created in step 16. In the next step 22, such a comparison is used to determine the presence of one or more fault in the rope, in particular whether two faults are present at the same axial position. The user can the take appropriate action with respect to the rope (e.g. record errors, repair or discarding thereof).

[0036] Steps 12, 14 and 16 provide a training step for a fault detection model. The training step is described in detail in figures 3-6. A rope measuring unit 24 (i.e. test rig) for measuring the magnetic flux leakages of the test rope 26 is shown in figure 3. The rig 24 comprises a clamp like arrangement 28 for supporting the test rope 26. The clamp 28 is configured to hold the test rope 26 under tension to emulate real-world loading conditions. The tension may be greater than 1000N or 10,000N.

[0037] A magnetic flux leakage sensor 30 is mounted about the rope 26 to measure magnetic flux leakage therefrom. The sensor 30 comprises any suitable arrangement for measuring magnetic flux leakage, for example, a Hall sensor and / or magnetic coil. The sensor 30 may comprise the sensor as described WO201 5071538, incorporated herein by reference.

[0038] The position of the sensor 30 may be effected by a driving unit 32. The driving unit 32 may comprise a motor or the like. The motor may be connected to pulley, cable or screw drive to allow movement thereof. In some embodiments, the motor may be provided on the sensor 30 unit. In some embodiments, rope 26 may be configured to move relative to the sensor 30 (the sensor may remain static).

[0039] A position sensor 34 may be used to determine the position of the sensor 30 relative to the rope 26. The detected magnetic flux leakage can therefore be correlated with the position of sensor 30. The position sensor 34 may be incorporated into the driving unit 32 (i.e. position along the rope 26 is estimated using the number of revolutions thereof).

[0040] A power source 36 provides power to the motor 32, sensor 30, position sensor 34 and / or computing unit 38. The power source is typically mains power. In some embodiments, the power source could be a battery, portable generator or the like.

[0041] The computing unit 38 is operatively connected to sensor 30 and / or the position sensor 34. The computing unit 38 is configured to store and / or process such measurements, as will be described later. The computing unit 38 comprises a controller / processor. The computing unit 38 may comprise RAM, non-volatile memory and / or other hardware / software to provide data processing.

[0042] The training process is described with reference to figure 4. In step 40, the fault model is initialised. Typically, this includes pre-populating the model with a number of desired variables or other parameters on which the training based is based. The variables may comprise one or more properties of the rope, for example, one or more of:

[0043] • The construction of the strand of the rope. This may comprise the composition (i.e. material), positional arrangement (e.g. number of layers or core structure), or the lay of the strand.

[0044] • The number of the strands (e.g. in total or per layer).

[0045] • The number / construction of the filaments, strings, wires or other subcomponent of the strands.

[0046] • The diameter and / or length and / or pitch of the rope.

[0047] • The age of the rope and / or prior operating conditions thereof.

[0048] • The load applied to the rope.

[0049] The variable may comprise one or properties the measured magnetic flux leakage. The variables typically comprise relating to a peak indicative of a fault, however, the variables may also include any feature of the measured magnetic flux leakage. The variables may comprise one or more of:

[0050] • Amplitude of the peak(s).

[0051] • Width of the peak(s).

[0052] • Area of the peak(s).

[0053] • An envelope or shape of said envelope of the peak(s). The shape may define a characteristic shape of the envelope (e.g. whether the envelope is square type or bell type curve).

[0054] • Raw signal data. This may allow for further processing at a later time.

[0055] • Any time dependent characteristic of the peak(s).

[0056] • A noise / background level and / or noise / signal ratio. The variable may comprise data indicative or one or more environmental condition, for example, temperature or humidity of the test environment. Other variables may include data indicative of the time / date on which a training test is performed.

[0057] In step 42, a test rope 26 is prepared. The test rope 26 comprises a number of predetermined faults. The faults may be provided by any suitable means. For example, incisions or cut-outs may be made into the rope or strand. This allows for controlled distribution / size of the faults. Alternatively, ropes with known defects from prior use may be used. This may allow for defects more representative of real-world use. It can be appreciated that test ropes from multiple sources may be used.

[0058] The defects are typically provided in internal positions within the rope (i.e. not visually inspectable). For example, defects may be provided on a core of the rope, or any suitable internal strand / wire. Defects may be provided on an external portion of the rope to aid with training (i.e. detection of multiple defects).

[0059] Generally, a plurality of test ropes 26 or fault arrangements are provided. This increases the amount of training data, thereby improving the quality of the model. The test ropes 26 may comprise one or more of:

[0060] • A different number of faults at same axial position. For example, a first rope comprises a single fault in a first wire, a second rope comprises two faults and so on.

[0061] • A different number of wire faults at same position located on the same strand. For example, a first rope comprises one fault in a given strand, a second rope comprises two faults in the given strand and so on.

[0062] • A different number of wire faults at same position located on different, respective strands. For example, a first rope comprises one fault in a first strand, a second rope comprises one fault in a first strand and a second fault in a second strand, and so on.

[0063] • Combinations of wire breaks at the same position in the same strand and distributed over different strands.

[0064] • Different number of strands comprising faults.

[0065] • Faults provided at different axial positions.

[0066] • The spacing of faults at different axial positions.

[0067] • Faults of differing axial length and / or different severity (e.g. how much of the wire is severed).

[0068] • Different types of ropes (e.g. composition, positional arrangement, or lay).

[0069] • Depth of the fault(s).

[0070] • Varying number of strands and / or wires, and / or width of the rope.

[0071] Additionally, identifiers or other metadata may be stored for each rope. For example, unique ID or reference number may be assigned to a specific rope. This allows identification of a specific rope, for example, to allow retesting thereof.

[0072] Examples ropes used for training data are provided below. Each set of defects could be provided on a single rope (i.e. axially spaced), or provided in separate test pieces.

[0073] Example 1 (figure 5A) - six broken wires next to each other (axially adjacent) on one strand. Only a single strand 86 is shown, with the positions of a fault 88 shown in respective wires 90. Example 2 (fig. 5B) - six broken wires next to each other on one strand. Example 3 (fig. 5C) - two broken wires next to each other on one strand.

[0074] Example 4 (fig. 5D) - one broken wire next to each other on two adjacent strands 86A,B.

[0075] Example 5 (fig. 5E) - one broken wire next to each other on two adjacent strands.

[0076] Example 6 (fig. 5F) - one broken wire next to each other on two adjacent strands.

[0077] The same process for examples 4-6 can be used with increasing number of adjacent strands. For example, one wire fault in three adjacent strands, two wires fault in three adjacent strands and so on, and then one wire fault in four adjacent strands, two wires fault in four adjacent strands and so on. A single wire fault in each strand may be used for any strand number between 1 and 10 for a given rope. For example, one wire fault in five adjacent strands (fig. 5G) or six adjacent strands (fig. 5H) may be used.

[0078] The test rope 26 is loaded into the rig 24 and tension is applied to the rope 26. The sensor 30 is placed onto / around the test rope 26.

[0079] In step 44, the magnetic flux leakage of the test rope 26 is measured using the sensor 30. The measurement signals are received and / stored by the computing unit 38. The motor 32 effects movement of the sensor 30 along the axial length of the test rope 26. Typically, measurements of the magnetic flux leakage along the test tope 26 are continuous or otherwise have a small sampling interval.

[0080] The magnetic flux leakage is therefore determined continuously along the length of the rope 26 (i.e. the sensor 30 scans along the length of the rope 26). The sensor 30 may scan all of a desired portion of the length of the rope 26. The sensor 30 may make multiple passes along the length of the rope 26. Passes may be made in one or both directions along the rope 26 (i.e. in an oscillating fashion).

[0081] Simultaneously, the position sensor 34 provides position data to the computing unit 38.

[0082] In the next step 46, pre-processing is applied to signal received from the sensor 30. Such pre-processing is primarily concerned with improving signal quality. For example, the pre-processing may comprise denoising the signal. Denoising may be performed using a low-pass filter.

[0083] In the next step 48, the sensor data is analysed to detect faults using a peak detection algorithm. An example of a measured magnetic flux leakage as a function of position is shown in figure 6. The measured magnetic flux leakage comprises a number of peaks 56 indicative of a fault. An algorithm is used to computationally detect said peaks, and record data relating to said peaks. In broad terms, the algorithm is configured to analyse the amplitude and / width of said peaks, and determine if the peak correlates to a fault. Such algorithms are known. Said data may be appended with the position data from the position sensor 34.

[0084] In the next step 50, the data relating to the peaks 56 is extracted. One or more variables (i.e. features thereof) relating to the peaks is determined. The amplitude, width, characteristic shape, or other variable discussed above are determ ined / extracted.

[0085] In the next step 52, a given peak is paired / synchronised with a known fault. Where a position sensor 34 is used, this process may be performed automatically. For example, information on the fault and the position thereof are input to the system, and the computing unit 38 uses the position sensor 34 to assign the detected peak to the fault input at the given position. Any other variables relating to the fault and / or the rope are input to the system.

[0086] Where a sensor is not provided, pairing may be provided manually. For example, if a given number of faults are provided in the test rope 26, then the same number of peaks should also be provided. The system and / or the user can then pair each peak with the known fault in sequence.

[0087] In the final step 54, the collected training data is input into a fault detection computational model. The training data includes variables relating the test rope 26 and the associated variables relating to the measured peak data of said test rope 26. The input data is analysed and processed to provide a model that may be used to detect faults in given rope. This process is repeated for a number of test ropes 26 with differing variables, thus providing measured data across said differing variables. The computational model uses the training data to determine or estimate the characteristic magnetic flux leakages peak for a given fault and rope type. This characteristic peak can then be used to estimate the fault type for a rope of a given rope type.

[0088] In particular, the training data included test ropes 26 which comprise multiple faults that are in the same, similar or overlapping axial position along the rope. The faults may be spaced by less than the width / diameter of the rope 26, which are typically not accurately resolvable in prior art solutions. The ropes 26 may comprise defects which are axially spaced by less than or equal to 5mm; preferably, less than or equal to 3mm; preferably, less than or equal to 1 mm. The system is therefore especially configured to detect such faults and distinguish between multiple and singular faults.

[0089] The model may be created using any suitable technique, for example, one or more of: linear regression; multivariate regression; parametric (non-linear) models; machine learning / artificial intelligence; least squares methods; optimization methods; artificial neural networks; or other suitable methods. In some embodiments, a deep learning model or other machine learning model may be used. In such embodiments, steps 48 and 50 may not be required, and the raw data or raw peak data may be used. Such system may inherently assign greater significance to the measured peaks, thus explicit processing thereof may not be required.

[0090] Some or all of the processing and / or analysis in the above step may be performed by the computing unit 38. In some embodiments, a further or remote processing unit may be used (e.g. a remote server or supercomputer). In some embodiments, processing may be performed in the Cloud. It can be appreciated that the exact hardware components of the system are not pertinent to the invention at hand.

[0091] A single model may be used across all variables (i.e. across all rope or fault types). In other embodiments, multiple models may be used for certain variable. For example, different models may be used for different rope types.

[0092] Once the model has been suitable trained, then model may be used to detect faults in real world situations. Such fault detection may be performed in situ (i.e. where ropes are still in active use). For example, the rope may be provided on a crane, elevator, winch, aerial lift (e.g. cable car) or fixed structure (e.g. a bridge). Additionally or alternatively, detection may be performed “offline”, for example, in dedicated testing environment.

[0093] The apparatus 58 to perform the real world testing is shown in figure 7. The apparatus comprises a sensor arrangement 60 configured to be mounted to a rope 62. The sensor 60 comprises a magnetic flux leakage sensor as previously described. The sensor 60 unit may comprise a propulsion means to effect movement along the rope 62 or may be moved by an external apparatus. A position sensor 64 may indicate the position of the sensor 60 as previously described. The position sensor 64 may comprise an encoder attached to rope, drum, gearbox axle, or motor. The position sensor 64 may comprise a draw wire sensor configured to measure distance between end of a rope drum and the rope leaving the drum, from which rope position can be deduced. The position sensor 64 may comprise a crane PLC, or drum motor inverter.

[0094] The sensor 60 is operatively connected to a computing unit 66. The computing unit 66 may be provided onboard the sensor 60, or may be remote therefrom. A power source 61 powers the sensor 60 and the computing unit 66.

[0095] The sensor 60 and / or computing unit 66 are operatively connected to an indicator 68. The indicator 68 is configured to indicate a fault to a user. The indicator 68 may comprise a visual (e.g. LED), audible and / or tactile indicator. The indicator 68 may comprise a “traffic light” type system. For example, a green light may indicative a no fault detected; amber may indicate one or more fault within an acceptable tolerance; and red indicates a fault outside said acceptable tolerance.

[0096] The sensor 60 and / or computing unit 66 are operatively connected to a communication interface 70. This allows transmission of the measured data or fault indication to remote device. Analysis of the magnetic flux leakage may be performed by the remote computer. The remote computer may also receive indication of fault, as described with respect the indicator 68. The remote computer may comprise a mobile / cellular phone. The user may therefore directly receive measurement / fault data. The communication interface 70 may be wired or wireless.

[0097] The rope 62 can be measured with one or multiple sensors 60. The sensors 60 may at different positions of the rope 62 to get measurement coverage of whole of the rope 62. Typically, the sensors 60 are provided below the rope drum and rope pullies, when rope reeving has multiple falls. Multiple sensors 60 may be used to get full coverage when needed, depending on the normal work cycle movements of the hook / rope. One or multiple ropes can be measured simultaneously (e.g. in cases the crane has several ropes).

[0098] The process of detecting a fault in the rope 62 is described with reference to figure 8. In the first step 72, the sensor 60 is mounted to otherwise placed on the desired rope 62. In step 74, the magnetic flux leakage is determined along the desired length of the rope 62. In step 76, the magnetic flux leakage signal is pre-processed. In step 78, the peaks are detected and the features thereof are extracted in step 80. Steps 74-80 are substantially the same as for the training step and will not be described in detail.

[0099] In step 82, the raw data or extracted peak data is used to detect a fault. Firstly, the relevant variables for the rope 62 are input (for example, number of strands, lay, rope thickness, material properties etc.). The variables may be input via configuration file, user text interface, and / or user visual interface. The variable may be input via the remote device. The system uses said variables and the raw data or extracted peak data to determine the presence of the fault within the rope 62 using the model created during the training regime. The system may also determine the number of faults within a given area, thus allowing detection of multiple faults in a given axial position. If a non-integer number of faults are detected, then the number of faults may be rounded to the nearest integer. The testing procedure may be performed in accordance with ISO 4309:2017 Annex C. The number and / or severity of the fault may be calculated to provide a fault parameter. If the fault parameter is above a certain threshold, then action is required.

[0100] In step 84, the fault is indicated to the user via the indicator 68 or via the remote device. For example, the user may receive an SMS, push notification or email. The positions and / or number of faults may be indicated to the user. An example of such an interface is shown in figure 9. A number of a predetermined messages may be provided to the user depending on the fault parameter, for example “no fault detection”, “rope replacement required” etc.

[0101] Raw data, fault data and / or position data may be stored for later retrieval. For example, the data may be imported into a software program to allow visual representation of the fault for a given rope. For example, the faults and / or number of faults could be mapped along a digital representation of the rope at their respective positions.

[0102] Generally, fault detection is performed in real time of near real time. This allows direct and fast detection of faults in the rope. The system can therefore easily be used in situ or in the field. The fault detection process can be repeated periodically, for example, at least once a day.

[0103] If a fault is detected, then appropriate action may be taken. Where a fault is detected, said rope may be used in the training in the training data. The position and / or type of the faults may be visually inspected to ensure accurate fault detection.

[0104] The start and end position of the rope 62 (i.e. the positions where the rope starts / end or where measurements are started / ended). The start / end positions are recorded with the rope data and associated with the specific rope reference number. If a rope is discarded or otherwise deemed unsuitable for use, it can be further investigated, for example, manually with the reference information. A rope measurement method and system (i.e. model) can be trained based on discarded and re-studied ropes.

[0105] In some embodiments, the test rope may comprise a digital model of a rope. The digital model comprises faults as previously described. The test rope may be subjected to a physics-based simulation to simulate loading conditions. Computational methods (e.g. FEM or other physical simulation) can be used to calculate magnetic flux leakage. This may allow a large number of rope variations with relative ease. However, it can be appreciated that using physical models allows more accurate recreation of real world conditions (e.g. real world manufacturing or use conditions), and therefore generally provide more accurate results for real world use.

[0106] In some embodiments, the test rig sensor 30 may be used to test the desire rope 62. However, using a separate training rig 24 and in situ test apparatus 58 allows a more flexible configuration.

[0107] The present system allows more accurate determination of a fault in a rope. The present system is particularly beneficial in detecting multiple faults at a same or similar axial position in the rope. In particular, the present system allows detection of faults axially spaced by less than the width of the rope.

[0108] Use of multiple rope types in the training data helps to increase the universality of the system, thereby allowing fault detection in a wider range of the ropes. The use of multiple training ropes with faults at the same / similar or closely spaced axial position allows the system to gain greater accuracy in detecting said faults, and discriminating between singular and multiple faults (i.e. the peaks in magnetic flux leakages are more accurately characterised). As a consequence, multiple faults can be identified, and appropriate action can be taken. Conversely, this allows the system to more accurately determine a singular fault (e.g. where an erroneously large peak is provided) to avoid unnecessary repair or discarding of a rope etc.

Claims

Claims:1 . A method of detecting a fault in a rope or cable comprising: providing a test rope (26) comprising at least two faults (88), the faults (88) provided in the same axial position or spaced by an axial distance along the axial length of the rope (26) less than a width dimension of the rope (26); measuring the magnetic flux leakage of said test rope (26); inputting magnetic flux leakage data indicative of the faults in the test rope (26) into a computational model; measuring the magnetic flux leakage of a further rope (62); determining if the further rope (62) comprises one or more fault therein using said computational model and the measured magnetic flux leakage of the further rope.

2. A method according to claim 1 , comprising determining if the further rope (62) comprises two or more faults (88) provided in the same axial position or spaced by an axial distance along the axial length of the rope less than a width dimension of the rope (62).

3. A method according to any preceding claim, comprising providing a plurality of test ropes (26), the test ropes (26) differing in at least one variable, and inputting the magnetic flux leakage data of the respective test ropes (26) into the computational model.

4. A method according to claim 3, where the variable comprises one or more of: a strand construction; a number of strands (86) in the rope (26); a number of wires (90) in each strand (86); and / or the diameter of the rope (26).

5. A method according to claim 3 or 4, where the variable comprises: a number of faults (88) in a given strand (86) and / or a number of faults (88) distributed across respective strands (86).

6. A method according to any preceding claim, comprising applying a load to the test rope (26) during measurement of the magnetic flux leakage of the test rope.

7. A method according to any preceding claim, where data indicative of a fault (88) comprises a peak (56) in the measured magnetic flux leakage data.

8. A method according to claim 7, where one or more characteristic of the peak (56) is determined and input into the computational model, the characteristic comprising one or more of: an amplitude; width; area; and / or characteristic shape of the peak.

9. A method according to any preceding claim, comprising providing an indication upon detection of a fault (88), preferably, providing an indication of the position of the fault and / or number of faults at a position.

10. A method according to any preceding claim, comprising transmitting measured magnetic flux leakage data and / or indicative of a fault to a remote device.

11. A method according to any preceding claim, where the magnetic flux leakage measurements indicative of a fault (88) in the further rope (62) are input into the computational model.

12. A method according to any preceding claim, where measurement of the further rope (62) is provided in situ, and determination of the fault (88) is provided in real time.

13. A computer program or computer readable medium comprising program instructions which, when executed by the computer, cause the computer to carry out a computer process implementing the method according to any preceding claim.

14. A system for detecting a fault in a rope or cable comprising: a testing apparatus (24) for supporting a test rope (26) comprising at least two faults (88), the faults (88) provided in the same axial position or spaced by an axial distance along the axial length of the rope (26) less than a width dimension of the rope (26);a magnetic flux leakage sensor (30) for measuring the magnetic flux leakage of said test rope (26); a controller (38) configured to receive magnetic flux leakage data indicative of the faults (88) in the test rope (26) into a computational model; a sensor (60) configured to measure the magnetic flux leakage of a further rope (62); and a controller (66) configured to determine if the further rope (62) comprises one or more fault therein using said computational model and the measured magnetic flux leakage of the further rope (62).

15. A system according to claim 15, where the testing apparatus (24) comprises a clamping means (28) configured to apply a tensile load the test rope.

Citation Information

Patent Citations

  • Apparatus and arrangement for monitoring a condition of an elongated ferrous object having a longitudinal axis

    WO2015071538A1

  • Method and apparatus for detecting elevator rope degradation using electrical energy

    US20040046540A1

  • Nondestructive testing method and device for detecting and distinguishing internal defect and external defect of wire rope

    US20220187246A1

  • Apparatus for increasing linear resolution of electromagnetic wire rope testing

    US5565771A

  • Magnetic non-destructive method and apparatus for measurement of cross sectional area and detection of local flaws in elongated ferrous objects in response to longitudinally spaced sensors in an inter-pole area

    US6492808B1