Show Abstract ▼

Abstract: Accurate predictions of the size and morphology of microstructural features, including defects such as porosity, are essential for predicting the performance of engineering components. Although several multiscale approaches exist in the literature, including direct simulations and volume-averaged models, their predictions are limited due to large computational times and relatively low accuracy. This work utilises transfer learning to link the macroscopic field variable distributions to the mesoscale, in order to estimate sub-grid microstructural defects. Specifically, the model parameters are corrected using experimental measurements of sub-grid scale defects. The proposed methodology is illustrated for predicting porosity in an aluminium alloy automotive component produced using high pressure die casting. The model uses a physics-based localised porosity model for combined gas and shrinkage porosity to train an artificial neural network. This trained machine learning model is subsequently re-trained using macroscale field variables and experimental X-ray microtomography porosity measurements from industrial component made using different process conditions. An unseen region of the same component is used for further testing of the performance of the model. The results show good prediction of pore size distribution and location. These results are then used to determine component fatigue life. Thus, a full process-structure-property model is established. The framework has the potential to be applied to a large class of problems involving predictions of microstructural features over entire macroscopic components.

Open Access

Show Abstract ▼

Abstract: Growth kinetics and orientation selection play a significant role in microstructure evolution during metal solidification, while gravity-induced convection adds significant complexity to the process. In-situ, time-resolved X-ray imaging of solidifying grain-refined Al–20 wt.% Cu alloy onboard the MASER-13 sounding rocket enabled the study of equiaxed dendrite growth under diffusion-controlled conditions, eliminating the influence of gravity. A machine learning-enabled analytical pipeline was developed to extract and evaluate the spatiotemporal behaviour of a large number of individual dendrites, including their growth characteristics, rotations and interactions. Post-flight synchrotron X-ray computed tomography and electron backscatter diffraction were used to reconstruct the three-dimensional dendrite structure with embedded details of crystallographic orientations. Correlated data analysis confirmed that most dendrites grew along directions parallel to the {100} plane under highly isothermal, diffusion-controlled conditions. However, growth along atypical directions was also observed, even in this simplified regime. The benchmark data revealed variation in dendrite arm evolution, influenced by local grain interactions and crystallographic orientation selection. It is shown that the equiaxed grains have random crystallographic orientations and evidence suggests that these survive from shortly after nucleation in the bulk liquid under microgravity conditions. The data processing protocols demonstrated here highlight the potential of integrating advanced experimental techniques with modern data science approaches to analyse solidification microstructure formation in metallic alloys under terrestrial and microgravity conditions.

Open Access

Show Abstract ▼

Abstract: This research examines the dynamics of reactive CO2 transport in carbonate rock, focusing on the impact of carbonic acid-induced formation damage. We provide real-time visualization of these processes by employing four-dimensional (4D) high-resolution synchrotron imaging at the I13 beamline hosted at the Diamond Light Source. We visualize and quantify the temporal effects of reactive CO2 transport at the pore scale in carbonate rock. The experiment involved injecting CO2-saturated brine through the sample with in situ scanning to track the different stages of chemical dissolution. Analysis of the images shows a channelled dissolution pattern which corresponds with a gradual increase in porosity due to pore structure changes. Pore network models were generated from the segmented images to carry out a sequence of drainage and imbibition simulations. The result demonstrated that reduced capillary entry pressure with increased pore connectivity after dissolution. Furthermore, the trapping efficiency was quantified to predict a slight decrease in dissolution as the pores become broader and better connected.

Show Abstract ▼

Abstract: The drying process of Li-ion batteries contributes significantly to both their manufacturing cost and electrochemical and mechanical performance. Mud cracking is a mechanical defect which occurs spontaneously as a result of capillary pressure during the final stages of the drying process. This thesis investigates the causes of mud cracking, the mechanism of their formation and their impact on ion transport and electrochemical performance. At the core of the work is the observation that mud cracks are qualitatively similar to engineered vertical channels, which have been introduced into Li-ion battery electrodes by a variety of manufacturing methods in order to enhance ion transport and improve capacity at high charge rates or in thick electrodes. It was shown that cracks enhance potential ion flux through the electrode thickness to a greater degree than by increasing porosity by the same volume. Cracked electrodes of 70 – 130 μm dry thickness were shown to have improved discharge capacities at rates of 1 – 4 C, compared with uncracked electrodes of equivalent thickness - as much as 68 % higher capacity retention at 4C. These performance enhancements were related to 3D crack network structures determined by X-ray computed tomography imaging, which analysis showed that the arrangement of pore channels is of importance in determining the efficacy of vertical porosity in improving electrode performance. Image-based electrochemical modelling was investigated as a tool for better understanding experimental electrochemical data, and this work highlighted the value of multi-modal model validation, including literature synchrotron studies. The formation of mud cracks during the electrode drying process was studied using in situ X-ray computed tomography. This imaging showed the microstructural evolution of battery electrodes during the drying process with unprecedented clarity and resolution, as well as the relationship between crack growth and local microstructural features. Cracking intensity and morphology were shown to be strongly influenced by coating thickness, in agreement with prior literature, as was the delamination which occurs with severe cracking. Cracks were also shown to nucleate where air bubbles are present in the slurry, and this mode of crack growth was shown to differ markedly from cracks in bubble-free coatings.

Open Access

Show Abstract ▼

Abstract: Avoiding lithium dendrites at the lithium/ceramic electrolyte interface and, as a result, avoiding cell short circuit when plating at practical current densities remains a significant challenge for all-solid-state batteries. Typically, values are limited to around 1 mA cm−2, even, for example, for garnets with a relative density of >99%. It is not obvious that simply densifying ceramic electrolytes will deliver high plating currents. Here we show that plating currents of 9 mA cm−2 can be achieved without dendrite formation, by densifying argyrodite, Li6PS5Cl, to 99%. Changes in the microstructure of Li6PS5Cl on densification from 83 to 99% were determined by focused ion beam-scanning electron microscopy tomography and used to calculate their effect on the critical current density (CCD). Modelling shows that not all changes in microstructure with densification act to increase CCD. Whereas smaller pores and shorter cracks increase CCD, lower pore population and narrower cracks act to decrease CCD. Calculations show that the former changes dominate over the latter, predicating an overall increase in CCD, as observed experimentally.