PhD in super-resolution X-ray CT and multiscale modelling of horticultural products

Horticultural products such as fruit are composed of different tissues, composed of cells and intercellular spaces. Fruit tissues are very heterogenous at the micro scale and the microscopic architecture determines to a large extent the fruit response to postharvest storage conditions that are aimed to control a minimal respiration rate of the fruit by modifying O2 and CO2 gas concentrations. As there are currently few experimental methods available to investigate in vivo respiratory gas transport processes at the cellular level of fruit, in silico modelling is essential. In this context, resolving the three-dimensional microstructure and its distribution across the fruit is required for simulating respiratory gas exchange in fruit. Visualization of the 3D fruit microstructure can be achieved by non-destructive imaging methods, such as X-ray computed tomography. However, the cost and effort to obtain sufficiently high image resolution of the porous microstructure, and capture the spatial heterogeneity of the microstructure across a fruit is large. Today, multiscale modelling is applied, in which effective transport parameters of the macroscale fruit model are calculated from simulations with a limited number of representative microscale tissue models obtained from high resolution X-ray CT, in which the spatial heterogeneity is parameterized based on, for example, porosity maps. In this PhD project, you will investigate alternative approaches that can effectively resolve the spatial heterogeneity using super-resolution deep learning methods applied to low resolution images of intact fruit. In addition, you will develop and apply more efficient models, such as pore network and lattice-Boltzmann methods to solve the respiration-diffusion problem at the microscale of the porous tissues.