Low-cost digital image correlation and strain measurement for geotechnical applications

Abstract

Particle image velocimetry (PIV), or digital image correlation (DIC), is a widely used technique to measure soil displacements and strains in small-scale geotechnical models. Arrays of single-board computers (SBCs) produced by Raspberry Pi, and their associated 8-MP cameras, are being used at the University of Cambridge to capture the images required for DIC analysis. This alternative to more expensive camera set-ups has numerous advantages. A single expensive and large camera can be replaced—at low cost—by multiple cameras, adding flexibility and affordability to any experimental set-up. Traditionally, the alignment of multiple cameras to each other and the referencing to a known coordinate system required painted or machined markers to be located on the observation windows through which the experiments are viewed. This can obstruct localised soil grain displacement measurements in those areas of the model where such markers are placed. To complement the Raspberry Pi camera system, a markerless calibration method was used during image acquisition. This paper outlines the set-up of four of these small computers and associated cameras, provides an overview of the use of the markerless referencing system and reviews two different experimental apparatus used to measure soil displacement and strain. When the cost of additional cabling, connectors and mounting hardware is considered for this system, the total cost to implement was approximately $125 USD per camera plus one-time costs of $175 USD for system peripherals, which represents outstanding value and enables practically all geotechnical laboratories to develop similar capabilities.