Key Engineering Materials Online: 25-11-15 ISSN: 1662-9795, Vols. 297-3, pp 221-226 doi:1.428/www.scientific.net/kem.297-3.221 25 Trans Tech Publications, Switzerland Ultrasonic Transmission Characteristics of Continuous Casting Slab for Medium Carbon Steel Jun Youn Lee 1,a, Soon Bok Lee 1,b and Jae Kyung Yi 2,c 1 Department of Mechanical Engineering, Korea Advanced Institute of Science & Technology, Daejeon, Republic of Korea 2 Instrument & Control Research Group, POSCO, Pohang, Gyeongbuk, Republic of Korea a leewnsdus@kaist.ac.kr, b sblee@kaist.ac.kr, c jkyi@posco.co,kr Keywords: longitudinal wave, Rayleigh wave, transmission characteristics, continuous casting slab Abstract. Until now, surface defects of continuous casting slab have been removed by the enforced surface scarfing to produce high quality steel materials. An evaluation technique for surface and internal defects of slab is required to enhance the production of medium carbon steels and acquire defect-map. Accordingly as a preliminary step, longitudinal wave testing and Rayleigh wave testing were carried out on slab specimens of medium carbon steel to get basic transmission characteristics of ultrasonic waves. This research provides as basic data for on-line defect estimation using a laser ultrasonic or EMAT in non-contact ultrasonic detecting techniques in future. Introduction Currently continuous casting process for steel making is recommended strongly for its good productivity and high actual incoming rate. The enforced surface scarfing work is being adopted for removing surface defects in manufacturing fields in order to ensure intermediate-materials for high quality steels. And a defect evaluation technique on the surface layer of slab is needed to increase an actual incoming rate of slab and to acquire a defect map. In the case of continuous casting structure, it consists of many enormous and coarse grains and includes many porosities and defects. Therefore the vibrating characteristic of stress wave is very poor and the attenuation of ultrasonic is extremely severe. But considering a poor surface state and temperature problem of continuous casting steel, a non-destructive technique based on the stress wave is proper due to its excellent transmission characteristics. Until now, although many studies related to examine material characteristics using the ultrasonic method have been carried out and are being progressed [1, 2], there is no report published about an application of the ultrasonic method on continuous casting steels. Accordingly this study was carried out for the estimation of ultrasonic transmission characteristics on a continuous casting slab of medium carbon steel. Ultrasonic transmission testing method & specimens Tests were carried out for estimating ultrasonic transmission characteristics of longitudinal wave and Rayleigh wave. The pulse-echo method was adopted using contact transducers for measurements. Longitudinal wave tests were performed at several frequencies in the range of to 2 and Rayleigh wave tests were performed by a sender-receiver method with the mode conversion of longitudinal wave using a wedge at 1, 2.25,. Ultrasonic signals were sampled by a digital oscilloscope which is possibly applicable up to 2.5GS/s sampling rate. Then the contact testing method shows many errors due to an energy loss of ultrasonic occurred at the interface between a transducer and a material and the couplant effect. Therefore to minimize these errors all the measurements were carried out under a uniform contact force where it was maintained to 1Kgf. The specimens were made from medium carbon steel and produced by the continuous casting process. The chemical composition of specimens is shown in Table 1. Actually thickness of the casting slab produced by the continuous casting process is up to 25mm. It is not easy to carry out All rights reserved. No part of contents of this paper may be reproduced or transmitted in any form or by any means without the written permission of Trans Tech Publications, www.ttp.net. (#6982651, Pennsylvania State University, University Park, USA-19/9/16,4:1:33)
222 Advances in Fracture and Strength measurement tests on an original slab in the laboratory because of its heavy weights. In addition it is impossible to perform ultrasonic testing on a thick original slab because the attenuation of ultrasonic is very severe inside this material. Therefore one small part was partially separated from a real slab and was divided again into seven parts in the direction of thickness shown in Fig. 1. Finally all the specimens were machined to have the dimension of 35mm 9mm 3mm and surface polished to neglect the effect of surface state. To estimate ultrasonic transmission characteristics at the whole positions in the thickness direction, longitudinal wave tests and Rayleigh wave tests on all the specimens were performed and their signals were analyzed and compared. Fig. 1 Specimen sampling from a real continuous casting slab & numbering Table 1 Chemical composition of material (unit: weight%) C Si Mn P S.98.256.132.11.4 There are many surface defects at the surface layer of the real continuous casting slab. The basic data is needed to decide the working depth of scarfing in order to increase the actual incoming rate of continuous casting steel in the industry. Four types of artificial defect of several sizes were machined and their attenuation was measured on respective defects. Four types of defects are circular defect, vertical line defect, horizontal line defect and closed defect. And the shape and sizes of the defects are shown in Fig. 2 and Table 2. Sender defect receiver (a) Circular (c) Horizontal line (d) closed (b) Vertical line Fig. 2 Shapes of defects Φ5mm Φ1mm Kind-of- defect Circular Vertical line Horizontal line closed Table 2 List of defect sizes Size of defect 1.5mm, 3.mm depth 1.5mm, 3.mm, 5.mm, 1.mm, 15.mm diameter mm, 1.5mm, 2.mm depth 3.mm, 7.mm, 1.mm, 15.mm length 1.5mm, 3.mm depth 3.mm, 7.mm, 1.mm, 15.mm length mm,1.5mm, 3.mm, 5.mm depth Longitudinal wave testing Longitudinal wave tests were carried out to estimate transmission characteristics of longitudinal wave on continuous casting structures in the frequency range of MHz to 2. To reduce error occurred during measurement, average values of repetitive experiments were taken as data. In relative high frequency range of 15 to 2 much higher energy has to be applied to carry out tests because of a violent attenuation. Then a main beam of ultrasonic is so large that signals acquired
Key Engineering Materials Vols. 297-3 223 contain more errors. In this case the adoption of delay transducers can prevent this problem. Moreover the attenuation of longitudinal wave is very severe because there are so many porosities and defects in continuous casting structures. For these reasons higher frequency part of high frequency waves disappears throughout continuous casting structures and the peak frequency reveals actually 1MHz in round values despite tests practiced with 15, 2, 2 transducers. The contact method was impossible to be applied to cases of more than 1. The immersion method might be more appropriate in a high frequency region. On the other hand it was possible to achieve longitudinal wave tests at MHz to 1MHz without disturbance. The shapes of signal could be clearly observed in a time-domain obtained by experiments at each frequency. Fig. 3 shows the first reflected echo signal of longitudinal wave in the frequency of obtained at the 2.5GS/s sampling rate. From this signal, it was found that the test result does not embrace any noise and distortion. This means that longitudinal wave test is suitable for this study. And the fact that the attenuation of high frequency is violent can also be demonstrated by the FFT analysis of time-domain signals shown in Fig. 4. The frequency reveals about 1MHz after traveling a 1mm distance inside a slab specimen but the higher frequency part of signal vanishes quite after traveling a 2mm distance. It shows that in a higher frequency region the attenuation occurs fast and the peak frequency shifts to lower frequency during the attenuation. 5 2mm traveling 1mm traveling amplitude (volt) Amplitude.2.1-5. 6.x1-6 7.x1-6 8.x1-6 9.x1-6 x1-5 time (s) Fig. 3 First reflected echo of longitudinal wave at in time domain 3.x1 6 6.x1 6 9.x1 6 1.2x1 7 1.5x1 7 1.8x1 7 Frequency (Hz) Fig. 4 FFT analysis according to traveling distances at 1MHz 8 16 vertical axis (cm) 6 4 165 16 155 2 15 145 14 5 1 15 2 h o riz o n ta l a x is (c m ) Fig. 5 Distribution of attenuation coefficients of specimen 1 at In the case of continuous casting slab, there exist some variations of ultrasonic characteristics according to measuring positions due to its coarse grains and many internal defects. The variation could be demonstrated through a distribution of attenuation coefficients of longitudinal wave at in the specimen 1 shown in Fig. 5. This result shows that continuous casting structures have a complicate arrangement of coarse grains so that longitudinal wave is affected by the state of internal structure of materials during propagation. The attenuation coefficients of longitudinal wave on a continuous casting slab were obtained. The trend of attenuation coefficients shows an increase with the increase of frequency as shown in Fig. 6 and Papadakis [3] showed the similar trend. Then attenuation coefficients show slack slopes in the
224 Advances in Fracture and Strength range of less than but slopes increase drastically in the range of more than. The attenuation coefficients of all the specimens were compared mutually according to positions in the direction of thickness at each frequency to find a trend of attenuation coefficients as shown in Fig. 7. As a result, attenuation coefficients reveal no definite trend in the direction of thickness and appear rise and fall repeatedly. One reason is that a continuous casting slab which is composed of enormous coarse grains was divided arbitrarily in the direction of thickness for testing convenience. Moreover the variation of attenuation coefficient is getting larger as the increase of frequency, which shows that the selection of higher frequency is better to improve a resolution of detection for investigating microstructures and defects inside materials. Attenuation coefficient (db/m) 16 14 12 1 8 6 4 2 specimen1 specimen2 specimen3 specimen4 specimen5 specimen6 specimen7 2 4 6 8 1 Frequency (MHz) Fig. 6 Attenuation coefficients of longitudinal wave according to a variation of frequency Attenuation coefficient (db/m) 14 12 1 8 6 4 2 MHz 3. 1MHz 1 2 3 4 5 6 7 Number of specimen Fig. 7 Attenuation coefficients of Rayleigh wave according to the specimen number Rayleigh wave testing Rayleigh wave testing was performed in the frequency of 1, 2.25 and. This testing was also performed on all of seven specimens in the same manner of longitudinal wave testing and additionally on the upper and lower surface of each specimen. Then upper and lower surface of each specimen is named A and B respectively. The shape of signals in time-domain was clearly observed as shown in Fig. 8. The attenuation coefficients of Rayleigh wave were calculated by exponential decay curve fitting of data points sampled at several traveling distances. The attenuation coefficients of all the specimens are shown in Fig. 9 and their trends exhibit repetitions of rise and fall similarly with trends in case of longitudinal wave. This result also means that the internal structure of continuous casting slab scattered randomly in an intricate form. But as shown in the result, the attenuation coefficient of is smaller than that of. In general attenuation coefficients increase with the increase of frequency but the result shows the opposition trend. The cause could be figured out by the FFT analysis of time-domain signals at each frequency shown in Fig. 1. In the case of relative higher frequency part of signal disappears almost inside a material and its peak frequency reveals actually about 68MHz. In other words the experimental result acquired by a transducer is not a result of but 68MHz. Fig. 11 shows attenuation coefficients calculated from the result of FFT analysis. The difference of attenuation coefficients between and gets smaller than that calculated from time-domain signals because there is a little difference of frequency characteristic between and whose maximum amplitude occurred in frequency domain at.763mhz and 68MHz respectively. It s not due to a problem of the transducer used in this research since this transducer shows or so in the characteristic estimation and experiments on other materials. The cause can be a drastic disappearance of relative higher frequency part of as a material characteristic of continuous casting structures as the cases higher than 1MHz of longitudinal wave.
Key Engineering Materials Vols. 297-3 225 amplitude (volt) 2-2 attenuation coefficient (db/m) 14 12 1 8 6 4 8.x1-5 8.4x1-5 8.8x1-5 time (s) Fig. 8 Time domain signal of Rayleigh wave at 1-A1-B2-A2-B3-A3-B4-A4-B5-A5-B6-A6-B7-A7-B specimen number Fig. 9 Attenuation coefficients of Rayleigh wave according to the specimen number amplitude (volt). attenuation coefficient (db/m) 16 14 12 1 8 6 4 1x1 6 2x1 6 3x1 6 frequency (Hz) Fig. 1 FFT analysis of Rayleigh signals at each frequency 1A 1B 2A 2B 3A 3B 4A 4B 5A 5B 6A 6B 7A 7B specimen number Fig. 11 Attenuation coefficients calculated from the result of FFT analysis There are so many surface defects in a real continuous casing slab. Improvement in the resolution applied to investigate defects is a key factor closely related to the actual incoming rate in the continuous casting process. Therefore a study on Rayleigh wave transmission testing on defects can give a beneficial effect on producing steels economically. Reflection, scattering and other phenomenon occur and lead to an energy loss of Rayleigh wave when Rayleigh wave hits a defect during propagation. Rayleigh wave tests were carried out with specimens having four types of artificial defects of several sizes. Test results show that the loss distributes from.2db to 1.82dB in the case of circular defects, from 1.59dB to 13.26dB in the case of vertical line defects, from.79db to 1.65dB in the case of horizontal line defects, from.9db to 3.39dB in the case of closed defects. The loss increases with the increase of defect size and the increase of frequency in the case of circular defects and vertical line defects. In the case of horizontal line defects, it s not easy to discriminate a loss occurred by defects because its defect size acting as an actual defect is too small by.8mm equivalent to the thickness of defects. Accordingly the loss caused by a defect was covered with an error bound formed during experiments. For this reason a definite trend could not be observed in connection with defect size and frequency (Fig. 15). On the other hand because closed defects were machined on the backside of material, the loss increases as the depth decreases and the frequency decreases. The losses induced by four types of defects were compared each other. The loss of circular defects is lager than that of closed defect except the mm depth closed defect shown in Fig. 12. Then it was found that the energy density of Rayleigh wave is getting larger as it gets close to the surface of media during propagation. Other comparisons are shown in Fig. 13, Fig. 14 and Fig. 15. The loss of circular defects is less than that of vertical line defects but larger than that of horizontal line defects. This implies that although the sizes of defects are same (where the size of defect is designated to a section area perpendicular to the direction of ultrasonic propagation) circular defects cause smaller losses than vertical line defects due to diffraction phenomenon of ultrasonic.
226 Advances in Fracture and Strength Consequently the degree of ultrasonic loss on the same size defects appears to vertical line defect > circular defect > horizontal line defect as the order of loss amount. 3.5 3. 2.5 2. 1.5. - 1 2 3 4 5 depth of defect (mm) circular circular2.25mhz circular closed closed closed Fig. 12 Circular defect vs. closed defect 3.5 3. 2.5 2. 1.5 1.2 1.4 1.6 1.8 2. 2.2 2.4 2.6 2.8 3. 3.2 depth of defect (mm) circular circular circular vertical vertical vertical Fig. 13 Circular defect vs. vertical line defect 12 11 1 9 8 7 1mm circular defect 1mm vertical defect 2.5 2. 1.5 circular circular circular horizontal horizontal horizontal 6 5 1 2 3 4 5 frequency (MHz) Fig. 14 Circular defect vs. vertical line defect. 2 4 6 8 1 12 14 16 size (mm) Fig. 15 1mm circular defect vs. horizontal line defect Conclusions Transmission tests were performed with longitudinal wave and Rayleigh wave on specimens of continuous casting slab whose structures consists of enormous and coarse grains and intricate microstructures. It was found that the attenuation coefficient varies heavily according to measuring positions of the specimen made of continuous casting structures and higher frequency part disappears suddenly above the critical frequency. And the fact that the energy density of Rayleigh wave is concentrated near the surface of material was confirmed from a comparison of the loss between circular defects and closed defects. Finally the loss of Rayleigh wave appears in the order as vertical line defect > circular defect > horizontal line defect. References [1] D.N. Collins and W. Alcheikh: J. Mater. Process Tech. (1995), p. 85 [2] W. Orlowicz and Z. Opiekun: Theor. Appl. Fract. Mec. Vol. 22 (1995), p. 6 [3] E.P. Papadakis: J. Acoust. Soc. Am. (1965), p. 711