Space–time digital image analysis for granular flows

Digital image analysis (DIA) groups the post-processing methods that study the kinematics recorded by a digital camera. The DIA may be compromised by the flow constraints that could lead to an arbitrary definition of the region of interest and arbitrary selection of the representative sequence of frames. This paper presents a space–time technique that is able to synthesise a sequence of frames, abstract the flow features and identify the representative domain in space and time. The space–time technique is implemented in the study of free-surface dry sand flows down an inclined plane in a geotechnical centrifuge. The effect of orientation of the inclined plane in a centrifugal acceleration field is investigated. This implementation enables a standardised study of granular flows, delimiting the particle image velocimetry analysis to a representative domain in the flow sequence. The resultant space–time representation successfully identifies the effective flow width, and allows the representation of transient flow processes such as the flow-front evolution and the initiation of deposition on the inclined plane. Furthermore, it is found that the orientation of the sand flows in the centrifugal acceleration field may contribute to the occurrence of material deposition at slope angles higher than the material friction angle.