Imaging peat using neutron and X-ray CT
More Info
expand_more
Abstract
Peat is formed by biochemical processes and the accumulation of the soil depends on the aerobic and anaerobic conditions. The soil covers five to eight percent of the land surface of the earth. For this thesis a peat coming from a site called Markermeeer, which is a lake between the provinces Noord-Holland and Flevoland in the Netherlands, is used. The fabric of peat consists of a fine dark peat material and fibres. The peat material is made of water and decomposed plant tissue whereas the fibres consist of fragments of (semi-)degraded wood, stems, branches and grass. Furthermore the cell walls of the fibres consists of a primary and a secondary wall which are mostly made of cellulose and lignin. Knowing the fabric of wet peat is of importance to better understand the role of fibres in the unusual behavior of the soil. Previous studies on imaging wet peat with X-ray micro CT has proven to be difficult due to the indistinguishable linear attenuation coefficient of the peat organic matter and water. This coefficient is a function of the interactions of X-rays with the materials present in peat. Flushing a peat sample with lead(II)nitrate have shown to be the most successful approach to show the stems and branches in a peat sample using X-ray micro CT. However, according to \cite{kettridge2008x} this flushing approach showed that the focus was limited on the larger fibres present in the sample because the different materials in peat have almost the same linear attenuation coefficient and therefore this thesis also focuses on two other techniques to image the fabric of wet peat.
Neutron CT was performed on cylindrical peat samples. This technique uses a thermal neutron beam to image a sample and was available at the Reactor Institute Delft. The resolution of this imaging station is 150 $\mu$m. Since the cell walls and the peat material between the fibres in peat have both a large attenuation coefficient for neutrons, it was not possible to distinguish the fibres from the peat material between the fibres, using this technique. To gain a contrast heavy water was used to replace the water because it has a low attenuation coefficient. Flushing a peat sample with heavy water showed the most effective way to do so. The aim of this procedure was to get a contrast difference between the peat material between the fibres and the fibres itself. A triaxial setup was used to flush the sample with heavy water. Heavy water diffused in the water present in the peat inducing a decrease in attenuation of the peat material between the fibres. The reconstructed tomographic images were filtered using a 3D visualization program Avizo version 9.4. Only the air-filled fibres could be observed on the tomographic images whereas the water-filled fibres could not be observed. Samples of the same peat were scanned in a dry and wet state using a X-ray micro CT scanner present at the geoscience section of TU Delft. On the tomographic images of the wet sample, white halo were observed representing the cell walls of the fibres.
X-ray phase CT present at the Ghent University Centre for X-ray CT was performed on a peat sample coming from the same site. This method results in an image contrast using a large fixed distance for low absorbing materials like peat. The aim of this scan was to reveal more fibres than observed with the X-ray micro CT scan at TU Delft. However, edge enhancement did not occur because of the filtering needed during reconstruction of the raw data to visualize different structures in peat. \\ The white halo could be thresholded and filtered in avizo resulting in a 3D image of the rod-like fibres. These fibres were randomly orientated. On the micro CT scan of the sample in a dry state other fibrous structures than the rod-like fibres were observed. These fibres were not shown as white halo on the tomographic images of the wet samples.