The flow of crystal-bearing magmas.

Many volcanoes erupt magma slowly and non-explosively for a long period of time. The magma is often piled up as a 'dome' that sits at the top of the volcano. It is quite common for the eruption at such volcanoes suddenly to switch to a highly-explosive type of eruption, which is much more dangerous. At the moment we cannot predict accurately when such a switch will occur, which is a problem for the people who live near such volcanoes. We now think that the switchings are caused mainly by effects of crystals growing in the magma. The crystals make the magma more viscous, that is, more 'sticky', so it doesn't flow so easily. The magma becomes rigid and forms a plug which blocks the path for the rising magma. The blockage causes the pressure to rise in the volcano. Eventually, the pressure becomes so great that the plug is destroyed and a violent explosion occurs. If we want to be able to describe this process accurately, then we need to know how crystals in the magma change the way the magma flows. The project that we are planning here will address this problem. We will conduct a series of laboratory experiments in which we will produce mixtures of solid particles (in place of crystals) in a liquid. We will measure the viscosity of these mixtures (i.e. how easily they flow or how 'sticky' they are) using an instrument called a 'rheometer'. The particular rheometer we will use has a microscope attached to it, so we will be able to see how the particles are moving. It is particularly important to be able to see this because the particles move around as the mixture flows. As the particles move, they sometimes line up, or form clumps and this affects the viscosity. We think it will be important to see what is happening to the particles inside the mixture as it flows so that we can relate this to the measured viscosity. We will use the results of the laboratory experiments to produce equations that describe the viscosity of the mixtures so that it will be possible to calculate the viscosity of the particle/liquid mixture, rather than measure it. This is interesting because, if we manage to do this, we will have understood something new about the way magma flows. We will then use the equations we have produced to calculate the viscosity of a magma that contains crystals in computer models that try to predict how magmas erupt and especially the switchings from slow, non-explosive to violent, highly-explosive eruptions. This will allow us to see what controls the behaviour and, will help to make the outputs from the numerical models useful for the prediction of volcanic eruptions.