The structure and rheology of crystal mushes

Lead Research Organisation: University of Cambridge
Department Name: Earth Sciences


As molten rock (magma) cools, it crystallizes and a 'crystal mush' forms on the margins of the magma body. The mush contains a mixture of crystals and volcanic liquid. Continued cooling steadily converts the remaining liquid to crystals, and the mush layer grows until the entire body is solid. The way this mush layer responds to stress (its strength, or 'rheology') controls volcanic processes on all scales, from the evolution of large magma chambers under super-volcanoes, to the eruption of magma at mid-ocean ridges, the emplacement of lava flows, and the dynamics and explosivity of hazardous volcanic eruptions. An understanding of mush rheology is therefore vital if we are to understand many volcanic processes. Knowing how mixtures of liquid and particles respond to external stresses is also important to a wide range of problems, including making ice-cream, pouring concrete and understanding the behaviour of mud-flows.

Much progress has been made in understanding the rheology of magmas with few suspended crystals (>50% liquid) and of rocks containing very little melt (<5% liquid). However, relatively little has been done to investigate the rheology of mushes with approximately 10-50% liquid, and it is difficult to scale up or down from previous studies because the rheology changes strongly. This intermediate case is important because many volcanic processes involve rocks with these intermediate liquid contents. For example, we don't know how important crystal mush compaction is in controlling basalt magma evolution, because of a lack of suitable data on rheology. We propose to address this gap by investigating the bulk rheology of crystal mushes with intermediate liquid contents, by combining experimental results with observations on the structure of natural crystal mushes.

The rheology of a mush depends on its crystal-scale structure. For example, the size and shape of the particles has an effect on how rigid and strong the mush is. We will therefore focus on quantifying the mush structure, which will also help us to link together natural and experimental results. Firstly we will describe and quantify the microstructure of crystal mushes that we can be sure haven't been deformed, using natural examples of gabbro. Once we know the structure of a typical gabbro mush, we will design simple experiments using low-temperature analogue materials that mimic the gabbro mush. These experiments will show us how the mush structure changes when it is deformed and how various parameters (e.g. grain size, shape and the amount of liquid) affect the mush strength and the way it deforms. We will finally examine natural rocks that have been deformed, in order to calibrate our results and determine the importance of processes such as compaction. In this way we will build a quantitative understanding of rheology during cooling and crystallisation of magma. The results will have broad applicability for other areas of Earth science, and will also be relevant to a range of problems in chemical engineering, food processing and metallurgy.

Planned Impact

We have identified a number of non-academic end-users who may benefit from the proposed research:

1. Our proposed research and methodologies have potential for incorporation into mining exploration methods. An improved understanding of the processes operating during the late stages of magma solidification, when evolved fluid migration can lead to ore deposition, will be helpful to companies involved in exploration for deposits of base and precious metals. Government agencies (including MineralsUK, the Centre for Sustainable Mineral Development at the British Geological Survey) and overseas research organisations (e.g. GEUS, Denmark) may benefit for the same reasons. We will promote these results at fora attended by industry representatives (such as the MDSG)

2. Many industrial applications rely on an understanding of the way crystal-rich slurries behave. One important example of this is the making of frozen food products such as ice-cream. Our discussions with the research team at Unilever show that they, and other companies working in this area, will benefit from the general physical constraints our research will place on mush rheology.

3. Our results may have direct relevance to understanding certain features of silicic volcanic rocks. Wider, less direct benefits are therefore possible for government agencies, volcano observatories and the hazard and risk analysis sector, both within and outside the UK, in terms of potential future integration of geophysical monitoring with petrology.

4. In the UK public sector, we anticipate general interest from local museums, science clubs and science festivals. Museums may be interested in the broad background behind our work, in our results, and in our field and experimental methods.

5. UK School groups (including both teaching staff and pupils) would benefit broadly from exposure to wide-ranging materials relating to earth science, living in the Arctic, the rock cycle and volcanic activity. Our proposed work is of general relevance in various every-day materials, including foods and paints. We also anticipate interest from schools and the general public for these reasons.

6. The post-doctoral researcher employed on the project will benefit from training in a range of techniques, from teaching, presentation and communication skills, to experimental and analytical methods, quantitative skills, and field skills. The PDRA will be well-suited to a career within any profession requiring a numerate, analytical employee with good communication skills. We also anticipate that the project will lead to several final-year undergraduate or PhD studentship projects during the course of the three years; these students will also benefit from training and exposure to an active research environment including external collaboration.


10 25 50
Description I have discovered that the shape of the mineral plagioclase, a dominant component of solidified basaltic magma, is a sensitive function of the time for the rock to crystallise, with long thin crystals in fast-cooled rock and more stubby grains in slow-cooled rock. This finding is critical to understanding the rheological behaviour of crystal mushy layers as the shape of the particles controls the mush physical properties.
A further finding is that the magma intrusion (the Shiant Isles Main Sill) we had identified as a key target for our work on determining the structure of simple settled crystal accumulations records a more complex history than has previously been envisaged. The base of the intrusion is covered with a thick layer of olivine crystals that settled out of the magma. The lower parts of this accumulation are formed of loose aggregates of grains that become smaller upwards. The upper parts of the accumulation are much more strongly clustered, and formed of well-sintered olivine aggregates that become coarser upwards. This can only be explained by the action of vigorous convection in the intrusion, that hindered the settliing of the olivine grains.
Experimental work has been undertaken to examine the packings of particles of a range of shapes (mimicking those of plagioclase) formed during simple settling in a viscous fluid. The packings become less dense as the particle shape diverges from a simple cube.
Exploitation Route Our findings of plagioclase grain shape can be used as a simple-to-use and robust speedometer for basalt solidification.

The observations in the Shiant Isles Main Sill point to the possiblity of using grain size and extent of sintering as an indicator for strong convection.

We have developed our plagioclase work into a project (recently funded by NERC) to examine the variation of grain shape as a function of convective velocity.

The experimental work is going to be imaged in 3D using X-ray tomography to develop a detailed understanding of packings.
Sectors Agriculture, Food and Drink,Environment

Description School visit 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach Regional
Primary Audience Schools
Results and Impact 50 children attended a talk on geology as an undergraduate topic of study. This is part of our on-going recruitment efforts at Cambridge
Year(s) Of Engagement Activity 2014
Description radio broadcast 
Form Of Engagement Activity A broadcast e.g. TV/radio/film/podcast (other than news/press)
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Public/other audiences
Results and Impact I was part of a panel of scientists talking about volcanoes and the impact of their eruptions for the BBC Naked Scientists radio programme
Year(s) Of Engagement Activity 2015
Description talk to a general audience as part of the Cambridge Science Festival 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach Local
Primary Audience Public/other audiences
Results and Impact A talk as part of the Cambridge Science Festival, introducing a general audience to optical microscopy as used by petrologists
Year(s) Of Engagement Activity 2015