Key technique of a detection sensor for coal mine wire ropes

Availableonlineatwww.sciencedirectcomMININGScience directSCIENCE ANDTECHNOLOGYELSEVIERning Science and Technology 19(2009)0170-0175www.elsevier.com/locate/jcumtKey technique of a detection sensor for coal mine wire ropesWANG Hong-yao, XU Zhao, HUA Gang, TIAN Jie, ZHOU Bing-bing,LU Yan-hong, CHEN Feng-junSchool of information and Electrical Engineering, China University of Mining Technology, Xuzhou, Jiangsu 221008, ChinaSchool of Mechanical Electronic and Information Engineering, China University of Mining& Technology, Beijing 100083, ChinaSchool of information and Technologies J1], University of Sydney, NSW 2006, AustraliaAbstract: Wire ropes, employed extensively in coal mine hoists and transportation systems are subject to damage due to wear,corrosion and fatigue. The extent of damage and the carrying capacity of ropes are closely related to the sense of safety by staff andequipments. Magnetic flux leakage detection method (MFL), as an effective method, is these days widely used in detection of bro-ken strands of wire ropes In order to improve the accuracy of detection of flaws in wire ropes by magnetic flux leakage(MFL), theeffect of the distance between a sensor and the surface of a wire rope (i. e, lift-off)on detection by magnetic flux leakage was investigated. An analysis of the main principles for the choice of lift-off is described by us and a new method that improves thture of the detector is proposed from the point of view of the design of a magnetic circuit, to restrain the impact of fluctuations ofnsor lift-off. The effect of this kind of method is validated by simulation and computation. The results show that the detectionensitivity is markedly increased by this method. Furthermore, the signal-to-noise ratio( SNR)can be increased by over 28%. Thismethod will lend itself to offer reliable scientific information to optimize the structure of excitation devices and improve the accu-of mfl detectinKeywords: coal mine wire rope; magnetic flux leakage detection; lift-off1 Introductionsending relation between lift-off and detected signalLi et al. proposed that increasing lift-off could reduceo At present, magnetic flux leakage (MFL), a the effect of lift-off fluctuation on the premise that then-destructive testing technology, is one of the main sensor detection sensitivity satisfied the demand inmethods for detecting flaws in broken coal mine wire the level of detection, but the range of lift-off used inropesll-2l. The intensity of the magnetic flux leakage their research only extended 0.2-1.4 mm4.Actually,field at the location of broken wire flaws is closely many physical factors are likely to cause about a 2broken wires, the level of intensity of magnetized search is, on the whole, not important to practicalwire ropes and the distance between the surface of theapplications; moreover, they did not investigate howfactorie and the sensor that has lifted off, etc. These to increase lift-off further on the premise of achievingwire rwill simultaneously affect the sensitivity and a high level of detection sensitivity. In practical engi-reliability of detection. a suitable choice of these pa- neering applications for the detection of broken coalrameters is important in order to obtain optimum mine wire ropes, the interior of the sensor togethertesting results. During the detection of flaws by the with the plastic interior lining and radial directionsensor, the shake of the wire rope can cause a change shake of wire rope, when in use, causes usually aof lift-off when the wire rope passes the sensor. The lift-off, of not less than 2 mm, in the sensor. How tohange of lift-off, also called a fluctuation can be realize higher detection sensitivity, as well as use aregarded as a jamming signal, which can occasion larger lift-off to reduce the effect of lift-off fluctuaadverse effects to the testing results and needs there- tion, is a problem which has not, as yet, been solvedfore to be restrainedbut urgently needs to be overcomehe effect of the sensor lift-off on magnetic fluxInad .he relation between theleakage was studied by Singh et al sI, but it was only magn中国煤化工 broken wires andpresented in general terms to state that there is a de- theCNMHG the design pointReceived O2 September 2008: accepted 1l November 2008CorrespondingauthorTel:+86-516-83885993;E-mailaddresshongyaowang2004@163.comWANG Hong-yao et alKey technique of a detection sensor for coal mine wire ropesof view of magnetic circuits, and a excitation device the broken strand and the magnetic leakage fieldof the magnetic flux leakage detectors is made of a produced by the fracture will be generated by twopermanent magnet. The effect of lift-off on detection magnetic charges. In Fig. 2, the diameter of steel wireresults is simulated by a numerical computing method. Dw=l6 mm, the gap between the two fractures28 =25.4 mm and we assume that the magnetictuations can be reduced and restrained by improving charges are separatethe structure of the detector.tely located at P(10)andP( o) with separated point charges -Q and +Q2 Relation between lift-off and intensity of The magnetic leakage field generated by the fracturemagnetic leakage fieldof a single strand is an axis symmetry magnetic fieldTherefore, we only need to compute the distributionmagnetic leakage detection of flaws in broken wire the stly gnetic field of the fracture in the plane ofIn order to study the effect of sensor lift-off d onropes,we have simulated the computation on one netic charge analysis and the principle of magnetictypical broken wire flaw in the wire rope of a coal field superposition, we know that the magnetic inducminetion intensity B, generated by the two magneticThe sensor adopted in the detection system of charges at any point P(x, y)of the coordinates plane,damage to wire ropes is a Hall sensor, which has anumber of advantages, such as small bulk, high de- is equal to the vector addition of the magnetic fieldAs shown in Fig. I, it is a method of detecting broken poim ed by the magnetic chargese and +e at thistection sensitivity, a broad response bandwidth, etc.generatePermanentFractureFig. 1 Sketch map of detection methodFig. 2 Magnetic leakage field analysis model ofBecause the characteristics of magnetic leakagesingle strand fracturefield signals of exterior and inner broken strands of awire rope are similar, we have used the brokenAssume that the magnetic induction intensity Brstrands of the external layer as an example. The generated by the magnetic-charge of point Pi at pointstructural parameters adopted in our simulation model Pisare:a wire rope with a nominal diameter of 24.5 mm;a distance of 107 mm between the centers of twoB(=znpermanent magnets that are the two magnetic poles ofthe wire rope excitation device; a distance of 37 mmQ(x+/i+ ybetween the magnetic poles and the surface of the4w[(x+)+y2wire rope (i. e, the air gap between permanent magtion of broken strands: a fracture length of 25.4 mm; In Eq (1). B,, B, are the vector compo (1)nets and surface of the wire rope): 19 concentrated=B i+B, jbroken wires; a fracture depth of 6.97 mm at the locasensor lift-off ranging from 0.5 mm to 16 mm.The analytical method used by us is a magneticallyQx+lcharged model. First, the magnetic leakage field fromthe fracture of each single strand is analyzed andcomputed. Then we computed the magnetic leakagefield distribution of the 19 concentrated broken wires中国煤化工given the principle of magnetic field superpositionCNMHGyNand the structural characteristics of the wire rope. Asshown in Fig. 2, we assume a magnetic charge sepa-Similarly, the magnetic induction intensity B2 genrating the small distance from both fractured ends of erated by point P2 at the magnetic-charge at point PMining Science and TechnologoL.19 No. 2In order to calculate the magnetic induction inten-a()=Q4:Lr.Asity at point P,wemagnetic-charge Q and L, both of which change withQthe single strand diameter d, the fracture gap 2ashape of the fracture and the intensity of magnetize-tion. therefore, we can not calculate B, until theseB2ri+ B2,jThe central place of the magnetic charges is deter-In Eq. 3).mined by the shape of the fracture. As shown in3, the shape of fracture of the broken strand of theQwire rope, as the result of various causes, mainly con-Gx-1)+ysists of four categories. The method of looking for(4) the central place of the magnetic charges is as followsassume there is a solid ball at the fractured end of theB2xbroken strand and that this ball occupies a big enoughmetal volume next to the fractured end of the brokenstrand, then the gravity centre of the ball is the sameTherefore, the magnetic induction intensity B, ()at as the gravity centre of the fractured end of the bro-point P isken strand (a small part). Given this kind of methodthe central places of the magnetic charges corre-Bn()=B(r)+B2()(5) sponding to the four kinds of fractures are simultaneously given in Fig. 3x-IThe amount of magnetic charge 2, determined byB.(xy)=B1+B=[(x-+the intensity of magnetic induction in the steelBd ofcalculated as follows:Q=TB d /4x+)+yTherefore in order to determine the status of thefracture, we must first compute the distribution of theB, (,y)=B,, +B2,magnetic leakage field, generated by the fracture of4each single broken strand, using Eqs. (6)and (7)Then we compute the distribution of the magneticleakage field, generated by the concentrated brokenstrands, according to the principle of magnetic fieldsuperposition, While computing the magnetic leakage(r+0)+yfield, generated by the concentrated broken strands,yP(127,25)P(1270)P127,0.5)P1272.mmP2127,0.5ig. 3 Shape of fracture of broken strand and central places of corresponding magnetic chargeswe should pay particular attention to the relative lo- the axial component of the intensity of the magneticcation among each single strand and the relation be- leakage field reduces rapidly along with the increasetween the route direction of a single strand and the in lift-off. In order to improve the detection sensitivdirection of the magnetization of the magnetic field. ity for lift-off, it is still true: the smaller, the betterFrom that, we calculated the distribution of the mag- But on the premise that the sensor detection sensitiv-netic leakage field generated by each single strand ity satisfies the demand in the level of detection, wefracture after appropriate coordinate alternation and she中国煤化工 rror as a principlethen we carried out superposition in the same coordi- in thenate system. As shown in Fig 4, it is the relation be- thecNMHGreouen l from ohetween lift-off and the axial components of the inten- shake generated by the wire rope of the coal minesity of the magnetic leakage field. Fig. 4 shows that while running. There is no doubt that this kind of un-WANG Hong-yao et alKey technique of a detection sensor for coal mine wire ropesdulation can produce a negative effect on the meas- broken wire flaws on the surface of the wire rope isurement of the magnetic leakage field signal, which the superposition of the magnetic air coupling fieldresults in a relative detection error. If we continue to and flaws of the magnetic leakage field. Because thetake this model as an example, as shown in Fig. 5, we analysis of flaws are largely based on the magneticget a simulation result of the average detection error, leakage field of the flaws, all non-flaw signals arecaused by undulation, of +0.5 mm if it occurs at all equivalent to noise, including the magnetic air cou-under conditions of multi-lift-offpling field. Fig. 6 shows the action principle of theAs shown in Fig. 5, with the increase of lift-off, the magnetic air coupling field and the magnetic leakagedulation error caused by itself decreases rapidly. To field of the flawsreduce the negative effect, generated by lift-off fluc-The change of lift-off can have two effects. First oftuation, we should choose a lift-off as large as possi- all, as shown in Fig. 4, the magnetic leakage field ofble, on the premise of allowable detection sensitivity. flaws of broken strands detected by the sensor willdecrease along with an increase in lift-off. On theother hand, the smaller the lift-off, the larger themagnetic air coupling field measured by the sensor,thus decreasing the signal-to-noise ratio (SNR), i.e,the ratio of the intensity of the magnetic leakage field益and the intensity of the measured magnetic air coupling field. In addition, the magnetic leakage fieldsignal produced by small flaws in a broken strand is012345678910weak, so that, when lift-off increases, this weak signal is more difficult to identify, i.e., as the lift-offFig 4 Relation between axial component ofincreases, the detection sensitivity of the sensor to theleakage field intensity and lift-offsmall flaw in the broken strand is clearly loweredWire ropeMagnetic air coupling fieldFig 6 Sketch map of magnetic air coupling field and2345678910Fig 5 Average fluctuation error of different lift-off4 Methods of reducing the effect of lift-off3 Relation between lift-off and magneticBased on this general discussion, we can see thatair coupling fieldwhen lift-off is small, the change of detection senstivity is more acute, even a very small fluctuation inIn on-line systems of detecting flaws in wire ropes, lift-off would lead to a greater change of detectionbased on a magnetic leakage field, the measurement sensitivity. therefore, in order to achieve a detectionresults of a Hall sensor actually consist of two parts, i: sensitivity that does not change very much, even if amagnetic leakage field produced by broken strands touch of fluctuation of lift-off occurs, we need to inof a wire rope; i: a magnetic air coupling field. This crease the lift-off as much as possiblemagnetic air coupling field refers to the magneticHowever, an increase in lift-off will again lead tofield formed by poles of a wire rope detector through the reduction of detection sensitivity itself. when weair coupling. In turn, this magnetic air coupling field face a high level of detection sensitivity and in orderincludes a so-called "magnetic air direct coupling to increase lift-off as much as possible, we must try tofield", formed directly through the air coupling be- increase the magnetic leakage field, generated by thetween the magnetic poles and a so-called"magnetic flaw in the broken strand and reduce the magnetic airair gap leakage field"formed by leakage to the air coupl中国煤化工 y and the Snr obetween the magnetic poles through the air gap be- the mtween the surface of the wire rope and the permanent thatCNMHGwire rope is satu-magnets. Whether or not there are flaws in the wire rated, we can achieve these two goals through changrope, a magnetic air coupling field will be created. ing the distance between the two magnetic poles ofWhat is measured by the Hall sensor at the location of wire rope detector as well as the size of the air gapMining Science and TechnologyVol 19 No. 2between the magnetic poles and the surface of the clearly that the decrease in radial length of the air gapwire ropecan reduce the air gap of the magnetic leakage fieldIncreasing the distance between the two magnetic simultaneously, which also can lead to the shorteningpoles is of benefit in reducing the magnetic air cou- of the radial distance from the sensor to the pure iron,pling field, but it can also lead to a decrease in the i.e., move the sensor relative to the place of higherintensity of the magnetic field within the wire rope. In intensity of the"magnetic air coupling field". Ourorder to ensure a large enough intensity of the mag- experiments shows that the weak level of the air gapnetic leakage field to detect flaws in an actual detec- in the magnetic leakage field, caused by a shorteningtion system, the magnetic field within the wire rope of the radial length of the air gap, is far greater thanmust be saturated Our experiments show that, when the corresponding, raised level of the"direct mag-the magnetic field within the wire rope is saturated, netic air coupling field"at the location of the sensor.we should increase the distance between the two That is, the superposition results of these two kinds ofmagnetic poles moderately, which does not lead toeffects decrease the intensity of the magnetic air cou-obvious weakening of the signals of the flaws in the pling field at the location of the sensor, with themagnetic leakage field. Through such an adjustment, shortening of the radial length of the air gap. There-we can lower the intensity of the magnetic air cou- fore, when the speed of detection is low and given thepling field and enhance the SNR of the measured premise of ensuring the traveling capacity of the designal. Of course, if the distance between two mag- tector, we can properly shorten the radial length ofnetic poles were too big, it might lead to such inten- the air gap to improve the Snr of the measured sigsity that the magnetic field within the wire rope does nalnot become saturated. Therefore, when designing Fig 8 shows the Snr obtained for the same flawmagnetic circuits, we should maximize the distance under the condition of different radial gap lengths andbetween the two magnetic poles on the premise of different lift-offs of the sensor. The structural simulaensuring that the intensity of the magnetic field tion parameters are: depth of flaw 6.97 mm, lengthwithin the wire rope is saturated. Fig. 7 shows the 25.4 mm, width 6.97 mm, air gap radial lengths arerelation between lift-off and the SNR of flaw detec- 27, 35 and 40 mm respectively. From Fig. 7, we cantion signals under conditions of different distances see that for the same lift-off of the sensor, the SNRbetween the two magnetic poles and the same flaw. can be increased by reducing the radial length of theThe structural simulation parameters are as follows: air gap within a certain range 2).depth of flaws 6.97 mm, length 25.4 mm, width 6.97mm and the distance between the magnetic poles areAir gap radial length 27respectively, 107, 97 and 87 mm.ir gap radial length 35tance between two poles 97 mmce between two poles 87 mu00.51.0152025303.540051.01.520253.03.54.0Fig 8 Effect of different air gap radial lengths onft-off of sensorig. 7 Effect on lift-off of different distances betweenFig. 9 shows the contrast between the waveformwo magnetic polesfigure detected by a detection system made by IntronAs shown in Fig. 7, for the same lift-off of the Plus Ltd. and the simulation results under the samesensor, the larger the distance between the two mag-conditions. We achieved these detection results at thenetic poles, the larger the SNR of the measured signal. wire rope testing laboratory of the China Universitytwo magnetic poles also depends on the speed of de- During detection, the moving speed of the wire ropetection.Usually, for higher detection speeds(over 4 was 2 m/s, with one point sampled every 0.05 mm.m/s), wide distances between the two magnetic poles The temperature in the laboratory was 22"C In thismust be maintained in order to reduce the effect of detedthe speed of detection 9-ol.中国煤化工 een the two magmm, with a radialIn order to improve the SNR, we also can reduce lengtCNMH Gift-off 6 mm. Ththe size of the air gap between the magnetic poles and depth or tne taw in une broken wire flaw was 6.97the surface of the wire rope. From Fig. 6, we can see mm, its length 25. 4 mm and width 6.97 mm. FromWANG Hong-yao et akEy technique of a detection sensor for coal mine wire ropes175Fig. 9, it is seen that the simulation results are very detector is 107 mm, with a radial length of the air gapsimilar to the signals measured under actual condi- 27 mm and lift-off 4 mm). This method provides ations. The SnR of the simulation results was: basis for optimizing the structural design of the deL meaa ss =13. 1 dB, while the SNR of tector and improves the accuracy of flaw detect70and assessment of wire ropes.the signal measured under actual conditions was:63.5Acknowledgements=12.5 dB, only a 4.6% difThe authors thank Mr. Xu Zhao, Mr. Hua gangference from the measured results under actual condi-iss Tian Jie, Ms Li Huifen, Miss Lu Yanhong andons. During detection, this SNr can satisfy the re- Mr. Chen Fengjun for their help during the experi-tOrquirements of higher sensitivity.ments. The authors also thank University of Sydneyand for the support from the China ScholarshipCouncil50一 Simulation resultse Practical measure resulReferenc[1] Du ZY, Ruan J J, Peng Y, Yu S F, Zhang Y, Gan Y, LiTation of velocity effects on magnetflux leakage testing signals. IEEE Transactions Magnetics,2008,4(6):l642-1645[2] Wang H Y, Hua G Tian J. Research on detection device0500100015002000250030003500for broken wires of coal mine- hoist cable. Journal ofSample pointsity of mining& Technology, 2007, 17(3)376-381.Fig 9 Comparison of simulation results and waveform[3] Singh W S, Rao B P C, Vaidyanathan S, Jayakumar T,measured under actual conditionsRaj B Detection of leakage magnetic flux from near-side and far-side defects in carbon steel plates using agiant magneto-resistive sensor. Measurement Scienceand Technology, 2008( 19): 1[4] LiL M, Zhang JJ, LiZ x, ShaoXY.tic fluκ leakaNondestructive Testing, 1999, 19(6): 1nese)[5] Jun M A, Ge SR, Zhang D K. Distriof wine deformation within strands of wire ropeal of chinaUniversity of mining Technology18(3)475-Y S, Liang C.Fig. 10 Wire rope testing laboratory of China Universitynetic flux leakage signal of flaw inside pipe based onANSYS. Computer Measurement Control, 20055 Conclusions[7] Huang S L, Li L M, Zhang J J Develnetic flux leakage testing device forthNondestructive Testing, 1999, 21( 8)34369.39es In use.(In Chi-We have analyzed the relation between lift-off andmagnetic air coupling field and presented a method of [8] Gu w, Chu jx. a transducer made up of fluxgate sen-reducing the effect of sensor lift-off fluctuations byors for testing wire rope defects. IEEE Transactions oradjusting the structure of the detection system. Theinstrumentation and Measurement, 2002, 51(1): 120-124.computed numerical values and actual detection re- [9] Zhang D L, Cao YN, Wang C, Xu d G A new methodsults of the detector show that the error caused byof defects identification for wire rope based on threelift-off fluctuations will be reduced with an increasedimensional magnetic flux leakage. Journal of Physics:in lift-off. The detection sensitivity can be effectivelyConference Series, 2006(48): 334-338improved through increasing the distance between the [10] Tian G Y, Sophian A. Reduction of lift-off effects fortwo magnetic poles and reducing the radial length ofulsed eddy current NDT NDt E International, 2005,38(l1):319-324the air gap. As well, the SNR can be increased to over [11] Giguere S, Dubois JMS. Pulsed eddy current28%, that is16-12512.5-x100%=28%(According to中国煤化工the Fig. 7 and the Fig. 8, we can know that the SNR12CN MHGtve Testing Beijingcan achieve a level of 16 db under the conditions thatChinese)the distance between the two magnetic poles of the