Discrete 3D Digital Image Correlation (DIC) using particle reconnaissance

This bachelor project is done for the study of strain localisation in geomaterials. The conventional material testing only describes the fabric behaviour globally. In order to create better understanding of how sand/rock particles move under compression, micro X-ray CT-scans are made of specimens from the material in question. An X-ray CT-scan alone however, is not enough to properly analyse the fabric movement. For this purpose, Digital Image Correlation (DIC) is applied to obtain 3D images of the fabric movement. With this thesis research has been done to an alternative way of performing discrete 3d digital image correlation. While previous methods have attempted to use complex vector fields, an easier more straightforward way of detecting particles was developed. This method has been called ‘particle reconnaissance’. This might also be called refinding particles using their shape parameters such as volume, 3d area, moments of inertia, Eigen values and particle specific Cartesian coordinates. The ultimate goal was to refind all grains in a research mounting and visualize how they had moved. To develop and test this method the programs Avizo 6.0, Excel and Matlab were used. An attempt has been made to digitally transform the grains using their transformation matrix. Some trouble with the reconstruction of the rotation matrix from the inertia tensor and/or trouble with understanding how the principal axis in Avizo were defined has complicated this matter.Geo-EngineeringCivil Engineering and Geoscience