Sintering dynamics of fine-grained rhyolitic obsidian particles from Hrafntinnuhryggur (Krafla, Iceland) with implications for silicic volcanic eruptions

Abstract

Sintering – or welding – is a key process in volcanic eruptions and controls the formation of welded ignimbrites, obsidian pyroclasts in volcanic conduits, and possibly also silicic lavas. Here, we study the sintering behaviour of packs of fine-grained particles of rhyolitic obsidian subjected to different temperature pathways at atmospheric pressure, with a focus on the evolution of the total porosity of the sintering pack and material microtexture. We collect high-resolution continuous in situ data for obsidian sintering and compare our results with the ‘vented bubble model’ – a versatile model for viscous sintering kinetics. This model accounts for syn-sintering degassing and outgassing of dissolved H₂O, which affects the particle viscosity. We also account for polydisperse particle size distributions, and arbitrary thermal history – i.e. any heating or cooling pathway and/or isothermal conditions. We find that the model performs well for fine particles sieved to $\lesssim 63\mathrm{\mu m}$. For particles $>63\mathrm{\mu m}$, sintering changes rate compared with the model and finally occurs more slowly than the model prediction. We explore this deviation by defining a capillary Peclet number $\mathrm{Pc}$ which balances the rates of diffusive loss of H₂O from the particles with rates of sintering; particles that are relatively large compared with the diffusive lengthscale (here $>63\mathrm{\mu m}$) have large $\mathrm{Pc}\gtrsim 10$ and therefore it is likely that deviations from the model are associated with substantial intra-clast gradients in H₂O, which translate to viscosity gradients. However, the efficacy of the model for relatively small particles and across a range of conditions demonstrates its general applicability to natural scenarios in which relatively small obsidian particles ($\le 63\mathrm{\mu m}$) are deposited hot, and weld together to form variably dense deposits. After model validation, we apply this model to the case of sintering at Hrafntinnuhryggur (Krafla, Iceland) where a ridge of obsidian is interpreted to have formed through sintering of fine hot particles during a rhyolitic fissure eruption. In this application, we discuss the effects of intra-grain vesiculation and nanolite crystal precipitation, and what role those additional process would play in sintering. Using these results, we propose a sintering timescale map for obsidian sintering at rhyolite volcanoes, which will be useful for understanding silicic volcanic eruptions.