Thirty-Sixth Canadian Geotechnical Colloquium: Advances in visualization of geotechnical processes through digital image correlation

Abstract

Digital image correlation (DIC) is an image-processing technique that calculates fields of incremental displacement by comparing two digital images and locating numerous small regions in both images to high subpixel accuracy. This technique is particularly well suited to the visualization of geotechnical failure processes such as the plastic collapse of a shallow foundation or the evolution of failure within a physical model landslide as it can yield full-field displacements at high spatial and temporal resolution. The adoption of digital image correlation methods by the geotechnical engineering community over the past 15 years has therefore provided researchers with a transformative tool for the visualization of failure mechanisms and the quantification of soil and soil-structure interaction behaviour within physical model experiments. The objective of this Canadian Geotechnical Colloquium is to provide an updated review of the factors that affect accuracy and precision of the technique and to highlight selected recent advances and emerging uses of DIC in geotechnical engineering applications with particular emphasis on geotechnical physical modelling and field monitoring.Key words: digital image correlation, particle image velocimetry, physical modelling, centrifuge testing, field monitoring. Résumé :La corrélation d'images digitales (CID) est une technique de traitement d'images qui permet de calculer les champs de déplacements incrémentaux en comparant deux images digitales et en y localisant les nombreuses petites zones avec une précision élevée, inférieure au pixel. Cette technique est particulièrement bien adaptée à la visualisation des processus de rupture en géotechnique, tels que l'effondrement plastique d'une fondation superficielle ou l'évolution de la rupture dans un glissement de terrain physiquement modélisé, car elle permet de générer des déplacements plein champ, à forte résolution spatiale et temporelle. L'adoption des techniques de corrélation d'images digitales par la communauté des ingénieurs géotech-niques au cours de 15 dernières années fournit ainsi aux chercheurs un nouvel outil de visualisation des mécanismes de rupture et de quantification du comportement du sol et des interactions sol-structure lors des expériences de modélisation physique. L'objectif du Colloque canadien de géotechnique est d'effectuer un examen actualisé des facteurs des facteurs qui influent sur l'exactitude et la précision de la technique et de présenter les perfectionnements récents et les nouvelles utilisations de la CID en ingénierie géotechnique, en insistant notamment sur la modélisation physique en géotechnique et la surveillance sur le terrain. [Traduit par la Rédaction] Mots-clés : corrélation d'images digitales, vélocimétrie par images de particules, modélisation physique, essais en centrifugeuse, surveillance sur le terrain.

Quantification of Optical Clarity of Transparent Soil Using the Modulation Transfer Function

“…Along with the quality of the sub-pixel interpolation scheme, one of the most important factors influencing the precision and accuracy of DIC is image texture (e.g., Take 2015). Image texture describes the degree of variation in pixel intensity values within an image subset.…”

“…3). Following the procedure described in detail by Take (2015), each point of light that makes up the speckle pattern of Fig. 13 has a non-integer pixel position of the centre and a shape defined by a Gaussian intensity profile of size of 10 pixels, a maximum intensity I max out of a maximum 8-bit scale of 255.…”

Performance assessment of peat rail subgrade before and after mass stabilization

“…DIC determines the change in position of a specified subset of pixels between two images with sub-pixel accuracy (e.g., White et al 2003;Take 2015). The amount of movement can be converted from pixel space to a physical displacement by applying a scale factor to the data.…”

Improved image-based deformation measurement for geotechnical applications

“…Measurement of soil deformations in geotechnical models using particle image velocimetry (PIV) techniques (Adrian 1991) -also known as digital image correlation (DIC) (Sutton et al 1983 -has become routine experimental practice in many geotechnical research laboratories over the past 15 years (White et al 2001(White et al , 2003Rechenmacher and Finno 2004;Iskander 2010;Hall 2012;Take 2015). The capabilities of the algorithms and analysis techniques that underpin PIV-DIC have also improved significantly over the same timeframe.…”