Methane (CH4) hydrate dissociation and CH4 release are potential geohazards currently investigated using X-ray computed tomography (XCT). Image segmentation is an important data processing step for this type of research. However, it is often time consuming, computing resource-intensive, operator-dependent, and tailored for each XCT dataset due to differences in greyscale contrast. In this paper, an investigation is carried out using U-Nets, a class of Convolutional Neural Network, to segment synchrotron XCT images of CH4-bearing sand during hydrate formation, and extract porosity and CH4 gas saturation. Three U-Net deployments previously untried for this task are assessed: (1) a bespoke 3D hierarchical method, (2) a 2D multi-label, multi-axis method and (3) RootPainter, a 2D U-Net application with interactive corrections. U-Nets are trained using small, targeted hand-annotated datasets to reduce operator time. It was found that the segmentation accuracy of all three methods surpass mainstream watershed and thresholding techniques. Accuracy slightly reduces in low-contrast data, which affects volume fraction measurements, but errors are small compared with gravimetric methods. Moreover, U-Net models trained on low-contrast images can be used to segment higher-contrast datasets, without further training. This demonstrates model portability, which can expedite the segmentation of large datasets over short timespans.
1 November 2024 in International Journal of Impact Engineering
The morphology and architecture of structures formed by sediment-dwelling invertebrates, such as excavations or burrows, are often assumed to be characteristic of a given species, consistent across a range of environmental conditions, and used to categorise species contributions to ecosystem functioning. However, very few investigations use non-invasive high-resolution techniques capable of determining fine scale variations in burrow form and complexity, or consider whether or not the form of the burrow is context dependent. Here, we provide replicate high-resolution micro-focus computed tomography data for the complete burrow systems of the Opheliid polychaete, Armandia cirrhosa, across a range of salinity and habitat conditions. These data provide reference models which can be used by ecologists investigating intraspecific variation in species traits and organism-sediment interactions and, more generally, by those tasked with pattern and shape recognition of objects that are morphologically highly variable and which adjust their architecture with changing circumstance or context.
Low- and medium-density White Chalk often destructures during civil engineering works, forming a putty that exhibits problematic low strengths in the short term but may build-up strength with time. The underlying mechanisms that control the mechanical performance of the material are not well understood. This paper explores the prospect of developing a framework to characterise the behaviour of chalk putty. To this end, triaxial tests were carried out using low- and medium-density reconstituted chalks produced by different methods. Using void ratio against mean effective stress results, a unique critical state line (CSL) is proposed, regardless of parent rock origin or the crushing process used to generate the putty material. The CSL was also found to satisfactorily fit independent test data by various authors, even though each independent sample material may have had different ‘intrinsic’ CSLs. This was suggested to stem from the similar particle characteristics of the chalks used. An assessment of the ageing effect indicates increases in strength as a result of densification due to consolidation and secondary compression, however, the end-of-test state could still be described using the CSL.