The effect of pores (fluid-filled vs. drained) on magma rheology

The presence of pore space strongly affects the rheological behavior of magma and thus influences all volcanic processes (pre-, syn- and post- eruptive). The effects of porosity on magma rheology are, however, unresolved and subject to debate. Here, we present new high-temperature experiments designed to constrain the rheological properties of variably porous melts (0.09–0.66 fractional porosity) deforming at high temperature (750–800 ◦C) and low strain rates (10 4 — 10 7 s 1). The starting materials are cylindrical cores of natural vesicle- and crystalfree rhyolitic obsidian from Krafla (Iceland) initially containing 0.114 wt% of dissolved H2O. Our experiments comprise two steps. First, cores are heated above the glass transition temperature (700 ◦C) to 900–1050 ◦C; second, the cores are deformed at lower temperatures (750 or 800 ◦C) under a constant low load (1.5 N). We have employed two different strategies for the second step: i) samples are deformed in situ directly after foaming (single-stage, SS); or ii) samples are quenched then reheated and deformed at 750 ◦C after 15 days repose at room conditions (double-stage, DS). Our experiments provide data that inform on the effects of porosity on the viscosity of natural rhyolitic deposits (e.g., ignimbrites, lavas, domes). Discordant results between SS and DS experiments (~0.6 log10 Pa s for 0.5 fractional porosity) suggest that the rheology of porous volcanic materials depends on whether pore spaces are isolated, fluid-filled bubbles (e.g., magmas in the conduit) or are interconnected, drained voids (e.g., domes, lavas, pyroclastic deposits).