The properties of many important natural and engineered substances are governed by complex structure and composition at a range of length scales. Examples of complex natural materials are bone and tissue in animals; stems, leaves and wood from plants, and soil and rock from the earth. Engineered complex materials include foams, semiconductors, concrete, fibre composites and 3D-printed components. These materials influence almost all aspects of our life, but their complex and hierarchical structures mean that their properties – mechanical, thermal or transport – can be fiendishly difficult to predict.
X-ray micro-tomography (XMCT) has emerged in recent decades as a non-destructive microscopy technique able to produce extraordinarily detailed maps of the internal structure of complex materials. The Department of Materials Physics is a world leader in XMCT science and technology, having spun off a successful company, and also created CTLab, the National Laboratory for X-ray Micro-Computed Tomography. The group leads the ARC Training Centre for M3D Innovation, which is training a cohort of PhD students and ECRs in the emerging discipline of digital materials, and collaborating with industry partners who provide direct mentoring on applied projects.
The X-ray tomography CTLab instruments enable us to see the complex structures inside samples with unprecedented clarity; this enables the properties to be modelled in a way that is unimaginable with 2D imaging. In addition, because tomography is non-destructive we can directly observe changes occurring within an evolving sample - such as a plant taking up water; the curing of negative-carbon cement; the crushing of mineral ores; the trapping of CO2 bubbles in the pores of a rock core, developing biomimetic materials or an additively-manufactured component deforming under stress. These are just some of the application areas that the group is actively exploring.
While the primary goal of our research is in solving problems in application areas, we often need to develop new techniques - in imaging or analysis - along the way. One technique of particular interest today is machine learning, and particularly deep learning, that offers the possibility of automated analysis and even decision-making from tomographic data sets.
This group works closely with the Applied topology group.