Image capture and analysis is now widely used throughout the research group in a variety of applications. In its most straightforward form, we use miniature cameras onboard the centrifuge to provide a qualitative commentary on modelling events. More sophisticated applications involve heavy processing and analysis to track pre-failure strains, pollutant concentrations, particle breakage and fabric anisotropy. The flexibility of modern CCD cameras combined with the non-contact nature of their measurement means they are often preferred over conventional types of transducer.
A displacement measurement system based on Particle Image Velocimetry (PIV) combined with close-range photogrammetry has been developed. This allows the non-contact measurement of deformation at many thousands of points within the field of view during modelling events and element tests. The precision of the system exceeds that available using conventional photography combined with target markers, and is comparable to local strain transducers.
Close-up digital photography of soil grains within a glass-sided oedometer allows fabric changes and particle breakage to be observed during loading. Post-test resin impregnation is also carried out to allow observation of particle orientation and breakage within tested samples using both scanning electron microscopy and optical microscopy. Sophisticated image analysis software provides automatic recognition of grain outlines and allows subsequent shape analysis.
Other image analysis software is used to study images of pollutant flow. The use of light filters and channel separation permits the identification of multiple phases, and their concentrations. Megapixel images can be divided into a large number of measurement stations, offering a more efficient method of data gathering than conventional discrete transducers.