Hill, I.A., Taira, A., Firth, J.V., et al, 1993 Proceedings of the Ocean Drilling Program, Scientific Results, Vol. 131 10. COMPUTED TOMOGRAPHY (CT) SCAN IMAGE ANALYSIS OF SITE 808 CORES: STRUCTURAL AND PHYSICAL PROPERTY IMPLICATIONS 1 Wonn Soh, 2 Tim Byrne, 3 Asahiko Taira, 4 and Atsushi Kono 5 ABSTRACT X-ray computed tomography (CT) is a promising tool that yields data useful for understanding the fine-scale density structure of partly lithified and tectonically deformed sediments. We conducted 21 CT scans of ODP Leg 131 sediments, including whole-round cores and thin-section chips, obtained from the toe of the Nankai accretionary prism. The samples range from highly deformed pieces from the frontal thrust and décollement to homogeneous and essentially undeformed sediments above the frontal thrust and beneath the décollement. In the CT images, kink-like deformation bands and faults are recognized as obvious bright seams, bands, or stripes with relatively high linear attenuation coefficients. The differences in linear attenuation coefficients relative to the matrix range from 0.021 cm 2 /g (kink-like deformation band) to 0.038 cm 2 /g (fault). These data suggest a 0.10 g/cm 3 to 0.18 g/cm 3 increase in bulk density within the deformation structures, and they appear to be 13% and 33% more compacted than the nondeformed matrix, respectively. In contrast to the samples from the frontal thrust zone, CT images of the décollement sample exhibit relatively homogeneous textures. The attenuation coefficient of the sample of the décollement indicates bulk density and porosity values of 2.45 g/cm 3 and 18%, respectively. The sample, hence, is approximately 50% more compacted than the sediment outside the décollement zone. INTRODUCTION Densification of sediments in accretionary prisms has been attrib- uted to porosity reduction caused by differential stresses resulting from plate convergence. The process of densification plays an impor- tant role in dewatering and fluid migration of an accretionary prism. Few studies have investigated these processes (cf. Moore et al., 1986; Agar et al., 1989). To better understand these consolidation and dewatering processes, we applied X-ray computed tomography (CT) to sediments obtained from the Nankai accretionary prism during Leg 131 of the Ocean Drilling Program. In this paper, we present CT scan images of the sediments of the Nankai accretionary prism that range from deformed to undeformed. The actual mode of the densification is estimated from the linear attenuation coefficient as well as from detailed microscopic-scale textural studies. We focus particularly on two types of structures, kink-like deformation bands and small-scale faults. Core-scale and microscopic-to-submicroscopic descriptions of these structures are provided in Taira et al. (1992), Maltman et al. (this volume), and Byrne et al. (this volume). BACKGROUND OF COMPUTED TOMOGRAPHY SCANNING An X-ray CT scanner is a device for reconstruction of an image of an object penetrated by X-rays, using the X-ray attenuation coef- ficients calculated from absorption or scattering. Although several published articles present the CT theory (e.g., Brooks and DiChiro, 1975, 1976; Iwai, 1979), we include a brief discussion of CT theory. The principal behind computed tomography is based on the mathe- matical theory established by J. Radon in 1917. He showed that two- or three-dimensional images of an object can be reconstructed from 1 Hill, I.A., Taira, A., Firth, J.V., et al., 1993. Proc. ODP, Sci. Results, 131: College Station, TX (Ocean Drilling Program). Institute of Geosciences, Shizuoka University, Japan. Department of Geology and Geophysics, University of Connecticut, Storrs, CT 06269, U.S.A. 4 Ocean Research Institute, University of Tokyo, Japan. 5 Technical Research Center, Japan National Oil Corporation, Japan. an infinite number of its projection data. The image reconstruction of an object is performed with a permissible limit of error in a CT scanner because an infinite number of the projection data of an actual sample is available. A collimated X-ray beam of intensity I o , as a result of passing through a sample of material of thickness D, yields a linearly atten- uated beam of intensity /, behind the sample. The relationship is shown below: = / o exp(-μD), (1) where the linear attenuation coefficient is μ. A CT scan image of the object is obtained as a map of linear attenuation coefficients for any desired section (slice) of the sample. When the sample material is heterogeneous in composition and density over the distance D, as in a sample of real sediment, particularly for a deformed sample, the attenuation coefficient will vary in the region of imaging. A more general expression is (2) = / o exp(-| μ(x)dx), where x is the distance from the X-ray source which varies between 0 and D; sample thickness. Numerous beams are passed through the sam- ple at various angles (0°-180°), thus, the distribution of the attenuation coefficients at discrete points within the sample can be determined. The CT scan system used in this study is a JACK-320 (Toshiba TOSCANNER 23201 system) housed in Technology Research Cen- ter, Japan National Oil Corporation (JNOC), Chiba. It is an upgraded second-generation rotation traverse type CT scanner. The JACK-320, designed especially for scanning lithified (hard) rock samples, uses a strong (320 kVp) peak energy of radiation (Table 1). The JACK-320 scanner converts the linear attenuation coefficients into correspond- ing numerical values (CT value). The raw data of the linear attenuation coefficients are displayed on a 512 × 512 matrix on a gray-level viewing system. Additional adjustment of the CT value level creates the most suitable image. Zones of high linear attenuation coefficients are displayed as bright zones on the CT scan image. We reconverted the CT values into linear attenuation coefficients. In the conversion, an attenuation coefficient of water (reference material) of 0.137 cm 2 /g was adopted, assuming that 200 kV (65% of 135