STABILITY OF MELT-RICH CHANNELS IN EARTH'S MANTLE : HIGH PRESSURE AND TEMPERATURE EXPERIMENTS ON OLIVINE, CHROMITE AND MORB

High-temperature, high-pressure annealing experiments on sheared partially molten rocks of olivine + chromite + 4 vol.% mid-ocean ridge basalt investigate the role of surface tension on melt distribution in mantle rocks. In torsion deformation with constant twist rate of 0.15 mrad/sec at 1473 K and 300 MPa confining pressure, stress causes melt to segregate into a network of melt-rich bands. Once the stress is removed, surface tension acts to redistribute melt homogeneously. A 4-hour static anneal at 1473 K and 150-200 MPa confining pressure dissipates bands and doubles the bandwidth. Melt has an important influence on mantle flow since even a small amount of melt weakens the mantle. Experiments indicate that, in deforming partially molten rocks, melt segregation occurs in response to an applied stress. Melt-rich areas also experience significant deformation, becoming zones of localized deformation. Highly permeable channels resulting from stress driven melt segregation may provide a mechanism for rapid buoyancy-driven transport of melt from within the mantle to the surface allowing melt to extrude from a planet's interior. Melt transport is important in many geological processes, such as the formation of oceanic crust at mid-ocean ridges and volcanoes overlying spreading centers, subduction zones and hot spots. This process of stress driven melt segregation has implications for these melt transport areas. At mid-ocean ridges, the erupted basaltic melt (MORB) is not in equilibrium with the upper mantle peridotites through which it travels. Chemical isolation of the melt during transport can occur in highly permeable channels such as those produced by deformation of partially molten rocks. Field evidence for channels can be seen in ophiolites where dunites have been interpreted as dissolution channels in chemical equilibrium with MORB but not with surrounding mantle peridotites. Extrapolating experimental results up to Earth time and length scales will contribute to the understanding of the time scales necessary to rehomogenize these structures after melt segregation occurs and will inform discussion of mantle formation on other planetary bodies by adding to our fundamental knowledge of multi-phase flow.