To understand how sand behave mechanically is important for everything from the design of safe foundations for buildings and infrastructure, to the prediction of geohazards such as landslide and earthquake liquefaction. Sand is characterized as a granular material, meaning that it consists of a collection of distinct particles interacting with each other through physical contact. What is special with granular materials is that they can behave as solids, fluids or gas, depending on the state. Imagine that you take a cup of sand (solid), tilt it so the sand starts to get out (fluid flow) and finally put it upside down so all the sand grains fall out (gas). Adding to the complexity, the grains have many different properties depending on shape, size, roughness and stiffness where all these combined have an effect on the material behaviour.
To understand the fundamental behaviour of sand it is necessary to investigate it on the micro scale, meaning on the grain level. In this project we will utilize novel experimental and numerical technology that allows us to observe and model sand grains down to a resolution of about 10 mikrometers, which is 1/100 of a millimeter. Observation of displacements and rotations of individual grains is done through X-ray Computed Tomography (XRCT) at different stages in a laboratory test. The CT images then serve as input to realistic computational modelling of sand at the grain scale, where each grain's shape is accurately represented.
More specifically, the objective of the project is to use the technology mentioned above to understand how the sand behave when you subject it to a "cyclic" load. With cyclic loading we mean that the load repeatedly changes direction. This is a typical condition for sand below the foundation of an offshore structure, where the cyclic load is caused by waves. Another example is the shaking of the ground that occurs during an earthquake. It is well known that sand behaves different under cyclic and static loading, but not exactly how. As a consequence there are currently no precise models to predict for example how the foundation of an offshore wind turbine will behave during its lifetime.
The hypothesis is that a "cyclic memory" is stored in the so-called fabric of the material, defined as the particle arrangement and contact network between the grains, the particle sizes, shapes and the distributions. In this project we will quantify fabric descriptors from the experiments and numerical models and study how they evolve during cyclic loading. Finding a relation between the evolution of fabric and the behaviour of the sand will likely lead to better engineering models for sand.