Nonisothermal viscous sintering of volcanic ash

Abstract

Volcanic ash is often deposited in a hot state. Volcanic ash containing glass, deposited above the glass transition interval, has the potential to sinter viscously both to itself (particle‐particle) and to exposed surfaces. Here we constrain the kinetics of this process experimentally under nonisothermal conditions using standard glasses. In the absence of external load, this process is dominantly driven by surface relaxation. In such cases the sintering process is rate limited by the melt viscosity, the size of the particles and the melt‐vapor interfacial tension. We propose a polydisperse continuum model that describes the transition from a packing of particles to a dense pore‐free melt and evaluate its efficacy in describing the kinetics of volcanic viscous sintering. We apply our model to viscous sintering scenarios for cooling crystal‐poor rhyolitic ash using the 2008 eruption of Chaitén volcano as a case example. We predict that moderate linear cooling rates of > 0.1°C min−1 can result in the common observation of incomplete sintering and the preservation of pore networks.

Topological inversions in coalescing granular media control fluid-flow regimes

“…These are mounted on the synchrotronsource tomography beam line (TOMCAT) [18] and heated by lasers [19] at ∼5 K min −1 to isothermal temperatures T 0 above the glass transition interval T g , yielding highviscosity droplets; experiments at different T 0 yield different viscosities. The glass beads do not lose mass, and do not show liquid-liquid immiscibility or crystallize on the experimental time scales or at the temperatures tested, which has been confirmed previously by differential scanning calorimetry and thermogravimetry [2]. The glass transition onset, determined on samples of the same size and at 5 K min −1 , is T g = 810 K (where we take the onset of the transition as a threshold temperature; see Fig.…”

“…A second transition, at which the pore phase becomes fully isolated (φ p /φ = 0), occurs at a consistent critical dimensionless timet ≈ 1.4. We term this transition 2 . We hypothesize that these transitions, which are microstructural in origin, influence bulk sintering dynamics and associated permeability evolution.…”

“…The spatial resolution of the x-ray computed 3D tomography is 1.9 μm, which permits accurate quantification of pores 6 μm radius [20], smaller than [2]. (c) The size distribution of glass beads, which become viscous droplets at high temperature, and the predicted pore size a i .…”

“…These include planetesimal formation [1], welding in shallow volcanic interiors [2][3][4], formation of deposits of large meteorite impacts and volcanic supereruptions [5], adhesion of volcanic ash in commercial jet engines [6,7], and industrial fabrication of ceramics and glasses [8][9][10][11]. In most sintering scenarios of practical interest, the viscosity of the droplets is high and the droplets are small, such that the Ohnesorge and Eötvös numbers are high and low, respectively.…”

“…In such systems the sintering time scale is found to scale with the initial droplet radius. However, for more complex heterogeneous and random systems of many droplets [15], the change of the surface area is not easily calculated, leading to the development of approximate descriptions of sintering, based on bulk geometric parameters such as pore number density and bulk porosity [2,8,16,17]. In those systems the sintering time scale is found to scale with the droplet size at early times, consistent with the models of simple droplet coalescence, but with the size of the interdroplet pore spaces at longer times, when the pore space is approximated as spherical bubbles in liquid shells [16,17].…”