Available online at ScienceDirect. Physics Procedia 88 (2017 )

Transcription

1 Available online at ScienceDirect Physics Procedia 88 (2017 ) th International Topical Meeting on Neutron Radiography, Beijing, China, 4-8 September 2016 Crystallographic analysis of a Japanese sword by using Bragg edge transmission spectroscopy Yoshinori Shiota a, Hiroyuki Hasemi b, Yoshiaki Kiyanagi a* a Nagoya University, Furo-cho, Chikusa-ku, Nagoya ,Japan b Hokkaido University, Kita 1, Nishi 8, Kita-ku, Sapporo , Japan Abstract Neutron imaging using a pulsed neutron source can give crystallographic information over wide area of a sample by analysing position dependent transmission spectra. With the use of a Bragg edge imaging method we non-destructively obtained crystallographic information of a Japanese sword, signed by Bishu Osafune Norimitsu, in order to know position dependent crystallographic characteristics and to check usefulness of the method for the Japanese sword investigation. Strong texture appeared on the back side. On the other hand in the middle area almost isotropic feature appeared and edge side showed feature between them. Rather isotropic area in the centre area gradually reduced from the grip side to the tip side. The crystallite size was smaller near the edge and became larger towards the back side. The smaller crystallite size will be due to quenching around the edge and this trend disappeared in the grip (nakago) area. The larger crystallite size will be due to strong hammering. Coarse grains were also observed directly as transmission images with the use of a high spatial resolution detector. The spatial distribution of the grains was not uniform but the reason have not been understood. Furthermore, a white area around a tip area was proved to be a void by looking at the Brag edge transmission spectra. This void may be formed during forging process of two kinds of steel. It is suggested that consideration on differences in the texture and the crystallite size depending on position will give information to clarify the manufacturing process, and Bragg edge analysis will be a profitable tool for research of Japanese sword The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license 2017 The Authors. Published by Elsevier B.V. ( Peer-review under under responsibility responsibility of the of organizing the organizing committee committee of ITMNR-8 of ITMNR-8. Keywords: Japanese sword, pulsed neutron imaging, Bragg edge, texture, crystallite size * Corresponding author. Tel.: ; fax: address: kiyanagi@phi.phys.nagoya-u.ac.jp The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license ( Peer-review under responsibility of the organizing committee of ITMNR-8 doi: /j.phpro

2 Yoshinori Shiota et al. / Physics Procedia 88 ( 2017 ) Introduction Japanese swords are interesting artefacts due to its sophisticated metallurgical structure as well as artistic value. Detailed manufacturing process was not clear since it was not transferred as a written document. Metallurgical information is useful to speculate the manufacturing process and a region produced. Metallurgical researches have been performed in destructive way (Tanimura 1980) by observing cut planes by a microscope, an EBSD method and so on. However, it is very difficult to apply the destructive method to many of swords and also valuable ones. Therefore, it is highly desired to investigate the Japanese swords in a non-destructive way. Neutron has strong penetration power and ability to investigate crystallographic information. Several studies have been performed using traditional neutron imaging and diffraction (Salvemini et al and Grazzi et al. 2009). Pulsed neutron imaging with the use of spectroscopic analysis of transmission spectra can give spatial dependent crystallographic information such as crystal lattice plane spacing, crystalline orientation, crystallite size and crystalline phase by analysing the Bragg-edges (Kiyanagi et al. 2012). Spectroscopic analysis of the Bragg-edge transmission spectrum through a bulk metal is performed by using the RITS code based on the Rietveld theory (Sato et al. 2011). The method will be very powerful tool to investigate the Japanese swords non-destructively and it is expected to extend a systematic study. We applied the method to broken Japanese swords. We obtained useful information indicating difference in crystallite size and texture (preferred orientation) depending on age and area of the swords (Kino et al and Nagashima et al. 2014). However, the information did not cover whole area of a sword. So far the spectroscopic analysis of the Bragg edge transmission spectra has not been applied to a whole area of a Japanese sword, and it is necessary to see change of the crystallographic characteristics depending on positions of the sword that reflects making process or treatment. Therefore, neutron transmission measurements of a Japanese sword (Bishu Osafune Norimitsu) were carried out using 2D neutron detectors at BL10/MLF/J-PARC. We obtained texture and crystallite size maps to know position dependent crystallographic characteristics and check usefulness of the method for non-destructive Japanese sword investigation. 2. Experimental The experiments were performed at NOBORU beamline at J-PARC/MLF in Japan. We used two position sensitive detectors, GEM (Uno et al. 2012) and MCP (Tremsin et al. 2013) detectors. The GEM detector has a 0.8 mm pixel size and 10cm 2 active area, which was used for large area measurements. The MCP detector has a mm pixel size and an active area of 28 mm in diameter, which was used for small area with high spatial resolution measurements. The sword sample was privately supplied by Dr. Hirota at Nagoya University, which has a sign of Bishu Osafune Norimitsu. A photo of the sword is shown in Fig. 1. The length of a blade area is 45.5, a short type sword. Nominal designations used in this paper are presented in the figure. The sword maker was working around A.D.1430 in Bishu, which corresponds to present Okayama prefecture in Japan. Therefore, the sword was originally made around this age but it was refurbished afterward probably due to a fire. The age of refurbishing was not clear. Fig. 1. Photo of a sample Japanese sword signed by Bishu Osafune Norimitsu.

3 130 Yoshinori Shiota et al. / Physics Procedia 88 ( 2017 ) Results and discussions Fig. 2 shows a transmission image of the sword obtained by the GEM detector. The positions where transmission image were obtained are shown in the photo of the sword: a tip area, two middle areas and one area around a boundary between the blade area and the nakago area (hilt area). The image indicates smooth change of thickness of the sword. Fig.3 is examples of the transmission spectra at two different positions around the tip area. They have typical Bragg edges corresponding to ferrite although the transmission rate is different due to thickness. We can see a little difference in the shape around the (110) Bragg edge reflecting different crystallographic characteristics, namely, crystallite size and texture. Fig. 2. Transmission images of 4 parts of the Japanese sword. The areas are indicated in the sword photo. Fig. 3. Transmission spectra around a tip area. By analyzing the transmission spectra we obtained the texture and the crystallite size distributions (Sato et al. 2011). Fig. 4 shows the crystallite size distribution calculated from the neutron extinction effect. Information on crystallites with a size up to few m are obtained by the Bragg edge analysis. The crystallite size is smaller around the edge area and becomes larger around center to back area. Around the tip larger crystallite size appears at the back but near the tip the size becomes smaller, which indicates a making process of the sword. Fig.5 shows the texture distribution as indicated by the coefficient of March-Dollase-function for the (110) plane of steel lattice, therefore, it means the selective orientation of (110) lattice plane. In the texture indication a value of 1 corresponds to the random texture (isotropic), and the smaller the value becomes, the stronger texture (un-isotropy) becomes. Strong texture regions exist at the back area over whole blade area other than the nakago. Around the center the texture is rather weak, namely, almost isotropic. Around edge the texture is between those of two areas. The strong texture and the large crystallite size at the back side will be due to flattering by hammering. As in the present iron manufacturing rolling makes the iron texture stronger and the crystallite size larger. Rather weak texture around the edge area will be due to quenching. The weak texture region around center will be due to less hammering process. Such crystallographic characteristics seem to change smoothly and reasonably as a sword, and it suggest the refurbish would be done according to traditional making process.

4 Yoshinori Shiota et al. / Physics Procedia 88 ( 2017 ) Fig. 4. Crystallite size mapping at four regions of the sward. Fig. 5. Texture mapping at four regions of the sward. Coefficient value 1: isotropic, less than 1: un-isotropic. Coarse grains were observed by the MCP detector due to its higher spatial distribution. Fig. 6 shows transmission images of five areas of the sword. The areas are indicated in the sword photo. (a) is around the tip area, (b) around 1/3 of the blade area, (c) around center of the sword, (d) around boundary between the blade and the nakago, and (e) around center of the nakago. In the figure black spots appear especially in Fig. 6 (b). The transmission spectra are shown in Fig. 7 for a ferrite area and a black spot. The spectrum at a ferrite area does not have small dips as shown in the spectrum at the black spot. The dips come from neutron diffraction at coarse grains as like a single crystal since its lattice plane orientation aligned in the grain. Corresponding to the transmission spectra the black spots appears and disappears depending on the neutron wavelength as shown in Fig. 8. It is not clear now the black spots are inclusion or simply a large ferrite grain. To obtain such information we need the dip analysis of the coarse grain. Fig. 6. Neutron transmission images at 5 positions of the sword to see coarse grains around 0.35 nm.

5 132 Yoshinori Shiota et al. / Physics Procedia 88 ( 2017 ) Fig. 7. Transmission spectra obtained at a smooth area in (a) and at a black spot in (b). Fig. 8. Wavelength dependent transmission spectra at seven wavelengths. Furthermore, we observed a white long and thin line at the tip area as shown in the left figure of Fig. 9. The line may be produced due to vacancy or some inclusion. To consider the reason we deduced transmission spectra across the white line for 15 positions from left to right, namely, an edge side to a back side. The transmission spectra are shown in the right figure of Fig. 9. The transmission patterns are almost the same and the transmission rate simply decreases. To evaluate the change of transmission numerically, we plotted the averaged transmission of 15 positions. Fig. 10 shows the spatial distribution of averaged transmission at the white area. At the region from position 7 to 10, corresponding almost to the white line, the transmission rates clearly increase. They indicate that there is no inclusion and the white line is due to a void. The void would be created during forge welding of outer and inner steels, so in the void gasses around A.D may be included. Fig. 9. A white line observed around the tip area, and transmission spectra at 15 regions across the white area.

6 Yoshinori Shiota et al. / Physics Procedia 88 ( 2017 ) Transmission Position Fig. 10. Spatial dependence of the averaged transmission at the white area. The transmission are averaged over the TOF range of 3.5 ~ 15.2 msec (in wavelength: ~ nm). 4. Summary The spectroscopic analysis of the Bragg edge spectra has been performed to observe the crystallographic change in a whole area of a Japanese sword. Change of crystallite size and texture were observed depending on position. The crystallite size information was obtained by the Bragg edge analysis up to a size of few mm, and over 100 mm crystallite size was observed by direct transmission measurements using a high spatial distribution detector. The results will offer clues to speculate making process. It is indicated that the pulsed neutron imaging using spectroscopic analysis is very useful to study crystallographic structure of Japanese swords non-destructively. The method will open a way for a systematic study of the Japanese swords since one of main reasons that prevent such a systematic study is the destructive manner so far used. Acknowledgements The authors thank Dr. A. S. Tremsin very much for his support in the MCP detector measurement. This work was supported by a Grant-in-Aid for Scientific Research (S) from Japan Society for the Promotion of Science (No ). References Grazzi, F., Bartoli, L., Civita, F., Zoppi, M., Neutron diffraction characterization of Japanese artworks of Tokugawa age. Analytical and Bioanalytical Chemistry 395, Kino, K., Ayukawa, N., Kiyanagi, Y., Uchida, T., Uno, S., Grazzi, F., Scherillo, A., Analysis of crystallographic structure of a Japanese sword by the pulsed neutron transmission method. Physics Procedia 43, Kiyanagi, Y., Sato, H., Kamiyama, T., Shinohara, T., A new imaging method using pulsed neutron sources for visualizing structural and dynamical information. J. Phys. Conf. Ser. 340, Nagashima, S., Shiota, Y., Sato, H.,Kamiyama, T., Ohnuma, M., Kiyanagi, Y., Imaging of Crystalline Structural Information of Japanese Swords by Pulsed Neutron Transmission Spectroscopy. Physics Procedia, 60, Salvemini, F., Grazzi, F., Peetermans, S., Civita, F., Franci, R., Hartmann, S., Lehmann, E., Zoppi, M., Quantitative characterization of Japanese ancient swords through energy-resolved neutron imaging. J. Analytical Atomic Spectrometry Sato, H., Kamiyama, T., Kiyanagi, Y., Mater. Trans. 2011, 52, Tanimura, H., Development of the Japanese Sword. Journal of Metals, 32, Tremsin, A. S., Vallerga, V., McPhate, J. B., Siegmund, O. H. W., Raffanti, R., High Resolution Photon Counting With MCP-Timepix Quad Parallel Readout Operating at KHz Frame Rates. IEEE Trans. Nucl. Sci. 60, Uchida, T., Ikeno, M., Koike, T., Miyama, K., Murakami, T., Nakano, E., Ohwada, H., Sekimoto, M. Shoji, M. Tanaka, M. Uno, S., Wada, M., 2010 IEEE Nuclear Science Symposium Conference Record N Uno, S., T. Uchida, T., Sekimoto, M., Murakami, T., Miyama, K., Shoji, M., Nakano, E., Koike, T., Development of a two-dimensional gaseous detector for energy-selective neutron radiography. Physics Procedia 37,