Influence of Scanning Velocity and Gap Distance on Magnetic Flux Leakage Measurement

Transcription

1 118 ECTI TRANSACTIONS ON ELECTRICAL ENG., ELECTRONICS, AND COMMUNICATIONS VOL.5, NO.1 February 2007 Influence of Scanning Velocity and Gap Distance on Magnetic Flux Leakage Measurement Noppadon Sumyong 1, Asa Prateepasen 2, and Pakorn Kaewtrakulpong 3, Non-members ABSTRACT This paper presents the development of a magnetic flux leakage instrument. Mechanical driving parts were introduced to improve the stability of the instrument during the movement and to control constant scanning velocity. The effect on scanning velocity and gap distance between defects and coil sensors of the instrument were studied. In our experiments, various types of artificial defects were employed to investigate performance of the instrument. The results showed that the signal amplitude decreased when the gap distances and the scanning velocities increased. However, interaction effect between both parameters was also exhibited. Characteristic of the defect on the signal output was also revealed. Keywords: wire rope, magnetic flux leakage, gap distance, scanning velocity 1. INTRODUCTION Steel wire ropes are important parts of large structures such as elevators and cable-strayed bridges. Defects in the ropes may lead to catastrophes; therefore, preventive maintenance is normally allocated to detect the problem. Common degradations of wire rope are wear, fatigue, corrosion and abrasion. In the maintenance process, Non-Destructive Testing (NDT) techniques are often employed to inspect deterioration of the wire rope regularly. The techniques include visual testing, magnetic flux leakage inspection, radiographic testing and imaging technique [1]. The most economical and simple technique is the visual test whereas the magnetic flux leakage method is a more automatic and robust technique [2]. In previous research [3-6], the instrument using coil sensors to measure magnetic flux leakage for evaluating the deterioration of wire ropes was designed and constructed. The approach was based on applying Manuscript received on July 31, 2006 ; revised on November 17, ,3 The authors are with Department of Instrumentation and Control System Engineering King Mongkut s University of Technology Thonburi, Bangkok 10140, Thailand. noppadon.sum@kmutt.ac.th, pakorn.kae@kmutt.ac.th 2 The author is with Department of Production Engineering,King Mongkut s University of Technology Thonburi, Bangkok 10140, Thailand. iasaasen@kmutt.ac.th high magnetization into a wire rope specimen. In metal loss region, high variation of the main flux occurs due to the flux leaking into the air. The coil sensors located inside of the external case of the instrument, which were placed around the wire rope, were used to detect leakage flux at the defect locations by moving the sensor along the length of the wire rope. The signals detected were used to identify the flaws of the wire rope [7-8]. Experiments were performed to verify performance of the transducer on a wire rope specimen of diameter 38 mm. Results showed that the instrument could be employed to inspect Localized Faults (LF) and to evaluate quantitative Loss of Metallic cross-sectional Area (LMA). Repeatability, resolution and sensitivity of the signals were also assessed to show the performances of the instrument. In this paper, the instrument was developed and the influence factors on the variation of corresponding signal amplitudes, i.e. the scanning velocity of the instrument and the gap between the cable and the coil sensor were investigated. 2. THEORIES In this section, structure of the wire rope and the basic principle of the instrument will be explained Wire Rope Structure Wire rope as shown in Fig. 1, is a metal rope in its strongest form. It consists of a group of strands laid helically around a core. The strands of a wire rope consist of a number of individual wires laid around a central wire Principle of Magnetic Flux Leakage Principle of operation of magnetic flux leakage instrument is illustrated in Fig. 2. A direct current is normally passed into copper solenoid (magnetization unit) to generate a constant magnetizing flux throughout the length of the rope specimen. When the instrument moves pass discontinuities in the wire rope, the magnetic flux leakage can be detected by a group of coil sensors. The sensors in this research were constructed from 40 AWG coil in a printedcircuit-shape and were most sensitive to the flux di-

2 Influence of Scanning Velocity and Gap Distance on Magnetic Flux Leakage Measurement 119 changing magnetic field and is written as Fig.1: Basic components of a standard wire rope rection perpendicular to their apertures. In Fig. 3, v d(b A) dt (2) The magnetic field density is constant in the region of no defect; therefore, no voltage is induced. The flaws on the other hand, produce non-uniform density which generates induced voltage at terminals of the coil. The output voltage of an induction coil is directly proportional to the instrument scanning velocity, in the axial direction (dz/dt). v d(b A) dz dt dt v dz dt (3) (4) Fig.2: Basic components of a standard wire rope 3. EXPERIMENTAL SETUP In this work, our modified main-flux equipment has been designed and constructed for steel wire rope inspections in Fig. 4. The modification includes redesigning a new external case of the instrument and modifying mechanical driving parts such as roller wheels, gear system. An adjustable direct current motor was introduced to drag and control scanning velocity. The signal was conditioned and displayed on a digital oscilloscope. the magnetic field density (B) in axial direction produced by a solenoid at any point P can be approximated by [9] B = µni 2 { } (L z p ) [a 2 + (L z p ) 2 ] + z p 1/2 (a 2 + zp) 2 1/2 (1) where L: length of the solenoid. n: number of turns per unit length. µ: permeability of the wire rope. I: current in amperes a: average radial of the solenoid. z p : distance from point P to edge of the solenoid. Fig.3: Axial induction of finite solenoid The output voltage (v) is induced across the ends of the coil terminals while it is moving through a Fig.4: Structure of modified main-flux equipment A copper solenoid was used as the magnetizing unit wrapping around the external case of the instrument. Its coil consists of 240 turns and has a cross sectional area of 6x2 mm2. In each experiment, a direct current of 50 amperes was applied to induce magnetic field density at the centre point of the sensor of more than 20,000 gauss which also covered the distance of approximately one meter from the centre to the edges of the solenoid. The amount of the magnetic field density computed from equation (1) is also verified by a finite element program namely the Finite Element Method Magnetics (FEMM) [10]. A variable speed DC motor was used to drive the instrument at various scanning velocities to investigate their effects on the output of the sensor. Four scanning velocities, 0.5, 0.7, 0.9 and 1.3 m/s were set for each experiment in this work. The steel cable used in this work consists of 8 outside and 4 inside strands. Each strand comprises of 19 individual wires. The gap distance which is the distance between the surfaces of the cable and the coil is

3 120 ECTI TRANSACTIONS ON ELECTRICAL ENG., ELECTRONICS, AND COMMUNICATIONS VOL.5, NO.1 February 2007 varied to examine their influence on the output voltage. Instead of changing the internal diameter of the instrument external case, the cables of different diameter were used to alter the gap distance. The inside diameter of the instrument case is 60 mm. Three sizes of cable diameter: 40, 45 and 50 mm, which produce the gap distance of 20, 15, and 10 mm respectively. Artificial defects, rectangular shapes from sawing, were made to each of the cable as illustrated in Fig. 5. In the figure, A-D are artificial defects with a constant depth of 3 mm and a distance of 20 mm apart for each pair. Their widths are 3, 6, 5, and 4 mm respectively. Defect items E-H have a 1 mm constant width and a depth of 5, 4, 3, and 2 mm respectively. An equal interval of each pair of consecutive defects is 10 mm. The smallest defect (item H with 1mm width and 2 mm. depth) was used to scrutinize the sensitivity of the instrument. Items I-J are subsurface defects. These 5 mm wide defects were located at 5 mm under the surface. Fig.5: Diagram of locations of artificial defects 4. EXPERIMENTAL RESULTS AND DIS- CUSSION The performance of the instrument as well as the effects on scanning velocity and gap distance were investigated and explained as follow Resolution The signal amplitude displayed by the oscilloscope can identify the defect items E, F and G as shown in Fig. 6. From the result, it can be seen that the instrument can be used to classify three defects which located 10 mm from each other. with better spatial resolution than 10 mm. Fig.6: The signal amplitude of the defect items E, F and G 4. 2 Sensitivity In Fig. 7, the result showed that the instrument was able to detect the smallest defect H(1 2mm.) Peak amplitude of the defect was noticeable in the oscilloscope display. It should be noted that the last peak is not corresponding to the artificial defects. The signal amplitude is usually shown at the start and the end point of the inspection. Fig.7: item H Signal amplitude obtained from the defect 4. 3 Display of Corresponding Signal Amplitude The signal amplitude received from all defect items can be displayed on the oscilloscope. Typical signal amplitude along the cable of length 2.0 meter obtained using the scanning velocity of 0.5 m/s and the cable size of 40 mm was demonstrated in Fig. 5. It can be clearly seen that magnetic flux leaked from all defects could be detected by the instrument. Characteristic of each peak is based on the width and depth of the defect. This is due to different amount of magnetic field density leaked from each defect as presented in equation (3). Fig.8: Typical signal amplitude obtained from 0.5 m/s scanning velocity on a 40 mm cable diameter 4. 4 Analysis of Variance The correlation between the variations of the peak amplitude and the scanning velocity together with the gap distance were analyzed based on the analysis of variance (ANOVA) using MINITAB software. The results showed that the scanning velocity, the gap distance and the interaction between the scanning velocity and the gap distance were significant parameters with the R-square value [11] of 96.27% Influence of the Scanning Velocity and Gap Distance on Signal Amplitude Fig. 9 illustrates the result of varying cable diameter and scanning speed. It can be seen that the signal amplitude significantly decreases when the gap

4 Influence of Scanning Velocity and Gap Distance on Magnetic Flux Leakage Measurement 121 distance increases. This is based on the basic principle that the detected leakage magnetic flux density decreases with distance. at various scanning velocity is examined and depicted in Fig. 11. The result shows that the depth of defect has greater effect on the signal amplitude. The amplitudes of the defect items A-D (the same depth but different widths) are similar, whereas the signal amplitudes of the defect items E-H (the same width but different depths) decrease significantly with depth. Fig.9: Relation between the cable diameter and the signal amplitude at various scanning velocity Fig. 10 exhibits the same result as the previous figure using different set of variables. It can be seen that the signal amplitude decreases with the scanning velocity. As expressed in the equations (2) to (4), the signal amplitude should increase with the scanning velocity. However, the induced electromagnetic force called Lorentz force [9] generated by the coils is acting in the opposite direction of the magnetic field density, the total of magnetic field density reduces. Due to an interaction effect between the scanning velocity and the gap distance, as analyzed by MINITAB software, the signal amplitude of the cable diameter at 40 mm is slightly changed. This may be caused by both factors. In addition, the results of sensitivity and resolution shown in Fig. 6 and 7 can be seen from all levels of cable gap and instrument velocity used in this experiment. The signal amplitude of each defect Fig.11: The signal amplitude of each defect at various scanning velocity 5. CONCLUSION In this research, a magnetic flux leakage instrument was developed. The effect on different scanning velocities and gap distances on signal amplitude of defects were studied. Various types (locations, widths and depths) of artificial defects were employed to investigate the performance of the instrument. In our experiments, signal amplitude obtained from the flux leakage from the defects were detected by a group of coil sensors and the variation of the magnetic field density was converted into voltage and displayed on the oscilloscope. The results can be concluded that the signal amplitude decreases when the gap distances and the scanning velocities increase. However, interaction effect between both parameters was also exhibited. Signal strength of the output was also increased with the depth of the defect. 6. ACKNOWLEDGEMENT The authors would like to acknowledge King Mongkut s University of Technology and National Research Council of Thailand for financial support as well as Mr. Chedpong Jomdecha for his recommendation. Fig.10: Relation between the scanning velocity and signal amplitude at various cable diameters 7. REFERENCES References [1] T.Moriya., K.Tsukada, and K.Hanasaki, A magnetic method for evaluation of deterioration of large diameter wire ropes, Proceeding of WC- NDT conference, Roma, pp.474, [2] A.Haller, Wire cable testing using high resolution magnetic induction, NDTnet, Vol.3, No.2, 1998 [3] H.R.Weischedel, The Inspection of Wire Ropes in Service: A critical review, Material Evaluation, Vol.43, No.13, pp , 1985.

5 122 ECTI TRANSACTIONS ON ELECTRICAL ENG., ELECTRONICS, AND COMMUNICATIONS VOL.5, NO.1 February 2007 [4] E.A.N.Whitehead, Method of Obtaining an Electrical Signal Proportional to the Cross- Sectional Area of a Magnetic Tube or Rod, UK Patent 913,780 (December 1962). [5] C.Jianxing,andG.Wei, ThePrinciple andapplication of a New Technique for Detecting Wire Rope Defects, IEEE conference on Industrial technology, pp , 1996 [6] C.Jomdecha, A.Prateepasen, and W.Methong, Coil Sensors for Wire Rope Inspection using Magnetic Flux Leakage Instrument, Proceeding of ICCAS2002, Muju Resort, Jeonbuk, Korea,pp , October 16-19, 2002 [7] C.Jomdecha, A.Prateepasen, and W.Methong, Characterization of Wire Rope Defects from Magnetic Flux Leakage Signals, Thammasat International Journal of Science and Technology, Vol. 8, No. 1, pp , 2003 [8] S.Mukhopadhyay, and G.P.Srivastava, Characterisation of Metal Loss Defects from Magnetic Flux Leakage Signals with Discrete Wavelet Transform, NDTE International, Vol.33, pp.57-65, [9] R.K.Wangsness, Electromagnetic field, John Wiley Sons Inc., pp , 1979 [10] D.C. Meeker, Finite element method magnetics v3.1, user s manual, [11] McClave, T. James Statistics, pearson education International, pp , 2006 Noppadon Sumyong was born in Phisanulok, Thailand in He received B.Sc. in physics from Naresuan University, Thailand in He obtained M.Eng. in industrial metrology engineering from King Mongkut s University of Technology Thonburi (KMUTT), Thailand in Asa Prateepasen was born in Bangkok, Thailand in He received B.Eng and M.Eng. in production engineering and mechanical engineering from King Mongkut s University of Technology Thonburi (KMUTT), Thailand in 1984 and 1992 respectively. He received Ph.D. degree in systems engineering from Brunel University, UK in Pakorn Kaewtrakulpong was born in Bangkok, Thailand in He received his B.Eng. in the electrical engineering from King Mongkut s University of Technology Thonburi, Thailand in He received M.Sc. and Ph.D. in industrial measurement systems (metrology) and computer vision from Brunel University, UK in 1998 and 2002 respectively.