Can a dike “feel” a free surface?

Abstract:

This study uses analogue experiments to examine the various factors that could influence the dynamic or arrest of a dike prior to an eruption. Dike propagation has been well-studied from the perspective of analogue experiments and it has been shown how various competing forces interact (buoyancy, viscosity, elastic properties, fracture toughness,…). It has been observed that dikes experience acceleration as they approach a free surface, which theoretically is due to increasing stress intensity as distance to the free surface decreases. Other unrelated studies analysed the constant-volume propagation of low-surface-energy liquids (e.g. hydrophilic liquids) in a semi-infinite medium and found the propagation rate to decelerate, as liquid gradually transferred from the head region to the tail and buoyancy abated. In contrast, high-surface-energy fluids transfer negligible quantities of their volume into the tail region, essentially maintain their volume, and experience no such deceleration.

This study focuses on the aforementioned factors influencing dike propagation, acceleration due to a free surface and deceleration due to head volume loss, to determine what their combined effect is on magma eruptability. As previous studies used hydrophobic liquids to investigate free-surface effects, and as magma is expected to have a low surface energy with the crust (they are both silicate), hydrophilic liquids (e.g. glucose syrup) are used in this study. Each experiment is arranged to have a free surface condition occupying half of the apparatus, and a no-slip condition occupying the other half, which allows for side-by-side comparison. The results are analysed to determine the minimum volume needed for a dike to erupt; failure to erupt implies that the dike would instead come to arrest in the crust.