Previous   |    Next

doi:10.2204/iodp.proc.308.213.2009

Methodology

Whole-core handling and preparation

During Expedition 308, the coring techniques included APC and XCB systems. These standard coring systems and their characteristics are summarized in Technical Note 31 of the Ocean Drilling Program (Graber et al., 2002). The sample was not extruded from the core liner on board the drilling ship. Whole-core samples were capped and sealed in wax to maintain natural saturation during refrigerated shipping and storage.

Radiography

Samples were X-rayed at MIT’s radiography facility using a procedure similar to that described by American Society for Testing and Materials (ASTM) D4452 (ASTM International, 2003) in order to assess the sample quality, general material type, presence of inclusions, and variation in macrofabric (e.g., uniform, stratified, etc.). This information was used to help select the most appropriate and highest quality material for testing.

Figure F4 (Sauls, 1984) shows a schematic of the procedure for radiographing soil that is in 3 inch diameter steel tubes typically 2–3 ft long. Since the tubes are cylindrical, X-rays that strike the center of the tube (point A, Fig. F4) must travel through 0.2 inches of steel and 2.8 inches of soil, whereas those hitting point B penetrate much less soil. Therefore, aluminum plates of varying thickness are positioned in front of the specimen such that all X-rays will penetrate an approximately equal mass. Vertical lines in the photograph are caused by abrupt changes in the thickness at the edges of these aluminum plates and the black background results from lead shielding placed around the tube to reduce scattered radiation. Lead numbers and letters are attached at 1 inch intervals along the length of the tube to provide distance reference marks. At MIT, the tubes are radiographed in 10 inch segments to minimize the divergence of the X-ray beam. Each segment is exposed for about 5 min. The radiation is generated by a metal ceramic, double focus, beryllium X-ray tube that is excited by a Philips MG151-160 kV constant potential high-voltage generator. The radiographs are generated at full capacity of the system (160 kV and 3.8 mA).

The radiograph image is recorded on plate film. These negatives (much like those used in a doctor’s office) are very difficult to work with in the laboratory. Therefore, the image is printed onto paper, yielding a photographic positive. On the positive print, dense objects appear white and voids appear black. The final product is slightly larger (~5%) than the actual size of the soil sample.

The image produced on the radiograph is an integration of all the material along the line from the radiograph source to the film. Changes in darkness depend on the relative absorption capacity of the materials being penetrated (i.e., steel, soil, air, shells, etc.). As a result, some features do not cause a sufficient change in absorption capacity and hence cannot be seen on the photograph. Other features are only visible when X-rays penetrate at the correct orientation. For example, an inclined crack filled with air within the sample will not be seen unless the X-ray path is parallel to the crack orientation. In general, changes in absorption capacity (absorption capacity is generally equated to density) as small as 5% can be observed.

Because of database issues and Fugro’s methods of data archiving, we were unable to obtain specific information with regard to their equipment and methodology. For our qualitative study of the radiographs, this information would have been useful but is not critical for our research.

X-ray CT scans at PSU

X-ray CT imaging was performed by Abraham Grader at CQI. The machine used was a medical-based scanner, HD350. CQI used 130 kV and 100 mA energy settings and scanned all the cores at a resolution of ~0.2 mm × 0.2 mm × 0.2 mm. Volumetric imaging of the sample was completed by X-ray CT (Universal Systems Inc.) as a sequence of stacked sections orthogonal to the long axis of the core (Yasuhara at al., 2006). Each CT slide image has two orthogonal longitudinal views through the center of the cores. The color scale is between 1000 (black) and 1500 (white).

Top of page   |    Previous   |    Next