Soil structure interaction in transparent synthetic soils using digital image correlation

Model tests to measure three-dimensional deformation patterns within a soil mass are presently limited by the fact that soil sensors do not provide a continuous image of the measured continuum. Additionally, soil sensors exhibit static and dynamic characteristics, which are different from those of the surrounding soils, and therefore can change the response of the measured continuum. The fundamental premise of this research is that transparent synthetic soil surrogates can be used to overcome these difficulties. The materials described in this research can provide a truly transparent mass, thus permitting complete light intrusion. This property is utilized for the study of spatial soil response in transparent synthetic soil models using an optical technique. This research presents (1) two families of transparent materials that can be customized for modeling sand and clay; and (2) a new imaging technique for measuring spatial deformations in transparent synthetic soil models.; Transparent synthetic soil models are formed using transparent particles made of amorphous silica and pore fluid of matching refractive index. The geotechnical properties for two families of amorphous silica were investigated: (1) amorphous silica powders, with aggregates size range 1.6–150 μm, for modeling clay, and (2) amorphous silica gel with aggregates size range 0.30–5 mm, for modeling sand. The geotechnical investigation included shear strength, permeability, and consolidation. All the investigated materials have the same refractive index, so that they can be combined in a stratified soil model.; Five optical techniques were investigated for use in imaging spatial deformations. These methods are: target tracking, cross tomography, photoelasticity, interferometry, and Digital Image Correlation (DIC). The investigation indicated that DIC was the most promising technique. A system consisting of a laser source and a line-generating lens was used to optically slice the transparent synthetic soil models. A digital camera was used to capture image before and after deformation. Next DIC was performed by comparing the two images using a cross correlation function. The cross correlation function is applied on interrogated images in small square windows. The deformation is detected by identifying the peak value of the cross correlation function of each window. A Matlab® algorithm was applied for the DIC analysis. The accuracy of the DIC technique was evaluated based on a scheme of predefined digital movement of synthetic soil images. Finally, an example model consisting of strip footing on loose silica gel is presented. The spatial deformations in the model were evaluated using the proposed methodology and were compared to finite element analysis results.