Intrusion and extrusion of fluidized sand in experiments on granular materials and in the North Sea
Résumé
Injected sand has long been a subject of attention, but in the last few years, as the quality of seismic surveys has improved, more examples have been discovered. In the North Sea, allochthonous sand bodies are almost pervasive within Tertiary sediment of the Hordaland Group, where they form dykes, sills, mounds, or extrusive bodies. In shape, these are similar to igneous bodies. However, their mechanisms of formation are somewhat different. Magma, even if it is of low viscosity, can migrate no more than a short distance through pore space, before it freezes. In contrast, water or gas can migrate over much longer distances without undergoing significant changes in viscosity. To investigate mechanisms of sand injection, we have resorted to small-scale physical models, which in principle are dynamically similar to their natural prototypes. In the models, the solid framework is a granular material, such as silica powder, fine sand, or glass micro-spheres. Although the first of these materials is cohesive, whereas the other two are non-cohesive, all obey a Mohr-Coulomb law for failure. As a pore fluid, we have used compressed air. This obeys Darcy's law for fluid flow through a granular material, up to a threshold velocity, at which the material fluidizes. In a typical experiment, the model consists of horizontal layers of various granular materials, resting on a diffusing bed. Compressed air migrates upwards through the diffusing bed and the overlying model. Hypodermic needles, connected to U-tubes, monitor the fluid pressure at various points within the model. In a cohesive layer, when the fluid pressure reaches a critical value, which is equal to the weight of the overburden plus the tensile strength of the material, horizontal fractures appear. If this happens at a local source of weakness, the fractures may deviate from the horizontal and the layer may bend. If there is an underlying non-cohesive layer, for example of glass micro-spheres, this material may fluidize and inject the overlying open fractures. The resulting sand bodies are sills, low-angle dykes, or laccoliths. In a lithostatic setting, the sand dykes tend to have conical shapes, whereas an additional horizontal compression makes them more cylindrical. Where the fluidized sand reaches the top surface of the model, it extrudes through vents. At this stage, inertial forces become significant and dynamic similarity no longer strictly holds. In conclusion, the sand bodies in our experiments are very similar in shape to those that occur in the North Sea. We therefore believe that the experiments provide good insights into the mechanisms of intrusion and extrusion that operate in nature.