Numerical and experimental cross analysis of the fluid solid transition in granular geomaterials

Geomaterials can behave like solids or like fluids depending on the mechanical constrains they undergo. Such solid to fluid transitions are observed in numerous environmental situations such as avalanches and landslides, but also in industrial context when considering concrete casting for instance.

In the case of a landslide for instance, solid-fluid transitions occur when the soil is fluidized into mud and starts running down the slope, but also along dead zones behind obstacles where mud transits into solid regime again, and at the toe of the slope when the mud stops flowing. The solid-fluid transition phenomenon is therefore of prime importance to define the flowing volume, the impact force of the flow on structures, and the run out distance.

Yet, currently used mechanical and numerical models only give a rough description of this transition phenomenon and fail to give an accurate estimation of the above mentioned data, mainly because the physical processes at play in these materials remain poorly understood.

In this project, the dynamics of the solid-fluid transition is studied on a simplified experimental system with the objective of developing and validating a numerical tool capable of simulating the transition phenomenon in a unified framework.

To better understand the internal dynamics of fluidization, solid-fluid transition is experimentally obtained on an inclined plane using a transparent model material, and studied through PIV1 method.

The numerical tool ellipsis based on FEMLIP2 is used to simulate the experiments, and development efforts are devoted to the implementation of proper constitutive laws and of inertial terms.

Finally, our model is confronted to more realistic and complex geomaterials, using the experimental setup as a rheometer. The final goal is to infer appropriate constitutive laws for describing the solid-fluid transition, and to calibrate the associated parameters by inverse analysis using ellipsis.

1 Particle Image Velocimetry

2 Finite Element Method with Lagrangian Integration Points