Pahoehoe to `a`a transition of Hawaiian lavas: an experimental study

Abstract

International audienceBasaltic lavas collected at the Muliwai a Pele lava channel, built during 1974 as part of Mauna Ulu’s eruption on Kilauea’s east rift zone, have been studied to investigate the effect of cooling and crystallization on the rheological properties of the lava. We have quantified the viscosity-strain-rate dependence of lava during cooling and crystallization, using concentric cylinder viscometry. We measured the viscosity of the crystal-free liquid between 1600 and 1230 °C, where we observed a deviation from the expected viscosity trend, marking the liquidus. We then made rheology measurements at subliquidus temperatures of 1207, 1203, 1183, 1176, and 1169 °C, varying the applied strain rates at each temperature. While the crystal-free liquid behaved as a Newtonian fluid, crystallization changed the rheological response to pseudo-plastic behavior, even at the lowest crystal volume fraction of 0.025. Pseudo-plastic behavior was observed down to a temperature of 1183 °C, with a crystal fraction of 0.15. Between 1183 and 1176 °C, the two-phase suspension transitioned from a power-law fluid to a Herschel-Bulkley fluid. At temperatures of 1176 and 1169 °C, with crystal fractions of 0.33 and 0.42, respectively, we observed lobate surface textures on the experimental samples, which remained preserved until the end of the experiments. Measurements at these temperatures indicated yield strengths of 82 ± 16 and 238 ± 18 Pa, respectively. The yield strength resulted from the development of an interconnected crystal network of diopside and enstatite by 1176 °C. By 1169 °C, diopside and plagioclase microcrystals had also appeared, and the effective viscosity was between 2000 and 5000 Pa s, depending on the strain rate. Further cooling to 1164 °C resulted in a rapid viscosity increase, to an effective viscosity in excess of 105 Pa s that exceeded the measurement range of our apparatus. The yield strength varies with crystallinity in an exponential fashion, with yield strength in Pa given by σ y = 1.25e12.93Φ c, where Φ c is the crystal volume fraction. The physical effect of crystals on the relative viscosity of magma was assessed by removing the effects of changing residual liquid viscosity, due to changing composition and temperature. To do this, we analyzed, synthesized, and measured the most evolved residual liquid from the subliquidus experiments. The effect of crystals was best captured by the Einstein-Roscoe equation for polydisperse spherical inclusions. We also measured the viscosity of the same crystal-liquid mixtures at low temperatures and strain rates using parallel-plate viscometry. The effect of crystals on magma viscosity was slightly greater at low strain rates, in agreement with theoretical models, although no single model reproduced these results well. In our experiments, the transition from pahoehoe to `a`a occurred between 1200 and 1170 °C, at viscosities between 100 and 1000 Pa s, depending on strain rate

Rheology of lava flows on Mercury: An analog experimental study

“…IcP‐HCT and enstatite basalt have an NBO/ T of 1.05 and 1.00, respectively, and show similar viscosities, which are about half a log unit below that of the NVP at liquidus temperatures. All four compositions are more viscous than a terrestrial ocean island basalt from Kilauea [ Sehlke et al , ], with NVP‐Na consistently about 0.7 log units more viscous at all temperatures compared to the terrestrial liquid composition. The other Mercury analog basalts are two log units more viscous than Kilauea basalt at ~700°C.…”

“…Liquid viscosities of the northern volcanic plains with low sodium content (NVP) composition in light blue, high sodium (NVP‐Na) in dark blue, intercrater plains‐heavily cratered terrain (IcP‐HCT) in orange, and enstatite basalt (EB) in green are compared to the viscosities of a terrestrial Hawaiian basalt composition (grey) from Sehlke et al []. Data points are fitted by VFT equations and thickness represents the root‐mean‐square deviation (RMSD) of the VFT fit to the data points.…”

“…Because our focus was on lava flows, experiments were done at 1 atm, and our analog compositions did not contain sulfur or other volatiles. We investigated the crystallization history and mineralogy of the magnesian basalts, and its effect on lava viscosity and yield strength, and compared the results to the study of a terrestrial ocean island basalt [ Sehlke et al , ]. Our results will contribute to the general understanding of mineralogy and rheological behavior of the lava flows on Mercury, providing modeling parameters for viscosity and yield strength over a wide range of temperature and crystal content.…”

Induced thermoluminescence as a method for dating recent volcanism: Eastern Snake River Plain, Idaho, USA

“…The sudden increase in microlites in the lava about 4 to 5 km away from the vent has been identified as the cause of a sudden increase in viscosity and thus a change in the surface texture and morphology of the flow. Thus, low microlite lavas generally have a lower viscosity and exhibit pahoehoe surface textures, while high‐microlite lavas have higher viscosity with `a`a surface textures at the same strain rate [ Sehlke et al ., ]. In this case, the scatter in our data would be due to sampling different places in a flow with variable amounts of primary feldspar on each volcano.…”

Simulating the thermorheological evolution of channel-contained lava: FLOWGO and its implementation in EXCEL

“…VTF viscosity relationships are becoming increasingly available, in order to cover most basaltic compositional scenarios (e.g. Whittington et al 2000;Giordano & Dingwell 2003;Bouhifd et al 2004;Villeneuve et al 2008;Sehlke et al 2014). However, this library still needs to be expanded so as to cover all possible eventualities.…”