BEYOND EXPLOSIVE MFCI: THE HYDRODYNAMICS OF MAGMA-WATER MIXING AND ITS EFFECT ON FRAGMENTATION AND ERUPTIVE VIOLENCE

Pyroclastic products of hydromagmatic eruptions tend to be finer-grained, more variable in vesicularity, and contain more abundant blocky, poorly vesicular fragments than those of dry magmatic eruptions. This fine-scale fragmentation is thought to accelerate heat transfer from magma to water, which in turn drives explosive steam expansion. Partly for these reasons, volcanologists frequently use the term “explosion” to describe violent magma-water mixing; and usually attribute such explosions to so-called molten fuel-coolant interactions (MFCI’s), in which the passage of a shock wave through a coarse dispersal of liquid magma in water collapses a layer of insulating steam between the materials. In the laboratory, MFCI’s produce violent explosions with powerful shock waves. In nature, however, the best-observed examples of magma-water mixing that generate fine tephra, at Surtsey and Kilauea, involve steady, turbulent jets that generate no shock waves. In these events, fragmentation may be controlled by turbulence rather than MFCI-type explosions. In industrial applications, turbulence intensity, or the fraction of a flow’s total kinetic energy contained in fine-scale eddies, is increased to reduce droplet size in sprays and ensure fine-scale mixing in ventilators and fuel injectors. With increasing Reynolds number, turbulence intensity increases and the size of the smallest eddies (which control the scale of mixing) decreases. I hypothesize that turbulent magma-water jets promote fragmentation by hydrodynamically breaking magma apart within eddies, and by increasing the frequency at which chilled margins on magma clasts grow and peel off. I am studying these process experimentally by injecting an analogue of magma (a glass-forming liquid containing sucrose, glucose and water) into an analogue of water (liquid nitrogen), under flow regimes that range from laminar to highly turbulent. Preliminary results suggest that average fragment size decreases with increasing injection rate (i.e. with increasing turbulence), and that clast shapes range from dominantly droplet- or coarse-thread (Pele’s hair) morphology under low injection rate, to more complex and fracture-dominated morphology under high injection rate.