[en] A model of both equiaxed and columnar dendritic growth was developed that incorporates thermal, solutal and fluid flow effects in either two or three dimensions. The model solves the momentum, mass and energy transport equations, including phase change. An imposed anisotropy algorithm, combined with a modified projection method solution of the Navier–Stokes equations, allows a relative coarse mesh and hence excellent computational efficiency. The model was used to study the effect of dimensionality (2D versus 3D) on dendritic growth with and without convection. The influence of forced convection on unconstrained equiaxed growth was studied first. In 3D, the upstream boundary layer is much thinner with a lower concentration than in 2D. This increases tip undercooling, accelerating upstream tip growth and promoting secondary branching. The influence of natural convection on constrained, columnar dendritic, growth was then studied. The 2D flow is blocked by the primary dendrite arms (which are effectively plates), while the 3D flow can wrap around the primaries. This change in flow strongly alters solute distribution and consequently the developing dendritic microstructure. 3D simulations are required to correctly predict unconstrained solidification microstructures

[en] A multiscale model was developed to simulate the formation of Fe-rich intermetallics and pores in quaternary Al-Si-Cu-Fe alloys. At the microscale, the multicomponent diffusion equations were solved for multiphase (liquid-solid-gas) materials via a finite difference framework to predict microstructure formation. A fast and robust decentered plate algorithm was developed to simulate the strong anisotropy of the solid/liquid interfacial energy for the Fe-rich intermetallic phase. The growth of porosity was controlled by local pressure drop due to solidification and interactions with surrounding solid phases, in addition to hydrogen diffusion. The microscale model was implemented as a subroutine in a commercial finite element package, producing a coupled multiscale model. This allows the influence of varying casting conditions on the Fe-rich intermetallics, the pores, and their interactions to be predicted. Synchrotron x-ray tomography experiments were performed to validate the model by comparing the three-dimensional morphology and size distribution of Fe-rich intermetallics as a function of Fe content. Large platelike Fe-rich β intermetallics were successfully simulated by the multiscale model and their influence on pore size distribution in shape castings was predicted as a function of casting conditions.

[en] During the twin roll casting of Al alloys, the interdendritic liquid may flow as the two solidification fronts are compressed together between the rolls. This can lead to defects such as centerline segregation. To understand the flow properties of the interdendritic liquid, samples of Al-12 wt.% Cu were solidified directionally in a Bridgman furnace and quenched to capture the growing columnar dendritic structures. The quenched samples were scanned using a laboratory X-ray microtomography (XMT) unit to obtain the 3D structure with a voxel resolution of 7.2 μm. Image analysis was used to separate the Al dendrite from the interdendritic Al-Al₂Cu eutectic. Flow between the dendrites was simulated by solving the Stokes equation to calculate the permeability tensor as a function of the fraction solid. The results were compared to prior experimental measurements and calculations using synchrotron tomography observations of equiaxed structures. Elasto-plastic finite element (FE) simulations were performed on the dendritic structures to determine flow stress behavior as a function of fraction solid. It was found that the standard approximations for the reduction in flow stress in the semi-solid have a variation in excess of 100% from that calculated using the true structure. Therefore, it is critical to simulate the actual dendrite for effective flow stress determination

[en] We present a novel iterative reconstruction method applied to in situ x-ray synchrotron tomographic data of dendrite formation during the solidification of magnesium alloy. Frequently, fast dynamic imaging projection data are undersampled, noisy, of poor contrast and can contain various acquisition artifacts. Direct reconstruction methods are not suitable and iterative reconstruction techniques must be adapted to the existing data features. Normally, an accurate modelling of the objective function can guarantee a better reconstruction. In this work, we design a special cost function where the data fidelity term is based on the Group-Huber functional to minimize ring artifacts and the regularization term is a higher-order variational penalty. We show that the total variation penalty is unsuitable for some cases and higher-order regularization functionals can ensure a better fit to the expected properties of the data. Additionally, we highlight the importance of 3D regularization over 2D for the problematic data. The proposed method shows a promising performance dealing with angular undersampled noisy dynamic data with ring artifacts. (paper)

[en] Highlights: • A novel upscaling method is proposed to quantify pores from the nm-to mm-scale. • Three advanced 3D imaging techniques are applied across five distinct scales. • A two-step analysis prior to upscaling ensures the images are representative. • Four types of pores are recognized at nanoscale and then all upscaled to mm-scale. • Upscaled porosity differs by less than 10% compared to measured helium porosity. -- Abstract: Microstructures and pore systems in shales are key to understanding the role of shale in many energy applications. This study proposes a novel multi-stage upscaling procedure to comprehensively investigate the heterogeneous and complex microstructures and pore systems in a laminated and microfractured shale, utilising 3D multi-scale imaging data. Five imaging techniques were used for characterisation from sub-nanoscale to macroscale (core-scale), spanning four orders of magnitude. Image data collected using X-ray tomography, Focused Ion Beam, and Electron Tomography techniques range in voxel size from 0.6 nm to 13 μm. Prior to upscaling, a novel two-step analysis was performed to ensure sub-samples were representative. Following this, a three-step procedure, based on homogenising descriptors and computed volume coefficients, was used to upscale the quantified microstructure and pore system. At the highest resolution (nanoscale), four distinct pore types were identified. At the sub-micron scale equations were derived for three pore-associated phases. At the microscale, the volume coefficients were recalculated to upscale the pore system to the millimetre- scale. The accuracy of the upscaling methodology was verified, predicting the total porosity within 7.2% discrepancy. The results provide a unique perspective to understand heterogeneous rock types, breaking though prior scale limitations in the pore system.

[en] In situ synchrotron radiography of Fe-rich intermetallic formation was performed during the solidification of an Al-7.5Si-3.5Cu-0.8Fe (wt.%) alloy. Growth kinetics was quantified by segmenting the Fe-rich intermetallic phases and nucleation temperatures were determined by extrapolating to zero size. Fe-rich β-intermetallics nucleated between 550 and 570 deg. C, growing initially at an instantaneous tip velocity of 100 μm s^-1, and slowing to 10 μm s^-1 towards the end of growth. Final plate size was controlled by local Fe concentration and α-Al impingement

[en] Rapid developments in photon-counting and energy-discriminating detectors have the potential to provide an additional spectral dimension to conventional x-ray grayscale imaging. Reconstructed spectroscopic tomographic data can be used to distinguish individual materials by characteristic absorption peaks. The acquired energy-binned data, however, suffer from low signal-to-noise ratio, acquisition artifacts, and frequently angular undersampled conditions. New regularized iterative reconstruction methods have the potential to produce higher quality images and since energy channels are mutually correlated it can be advantageous to exploit this additional knowledge. In this paper, we propose a novel method which jointly reconstructs all energy channels while imposing a strong structural correlation. The core of the proposed algorithm is to employ a variational framework of parallel level sets to encourage joint smoothing directions. In particular, the method selects reference channels from which to propagate structure in an adaptive and stochastic way while preferring channels with a high data signal-to-noise ratio. The method is compared with current state-of-the-art multi-channel reconstruction techniques including channel-wise total variation and correlative total nuclear variation regularization. Realistic simulation experiments demonstrate the performance improvements achievable by using correlative regularization methods. (paper)

[en] The structure of a fly ash/aluminum syntactic foam was characterized using X-ray microcomputed tomography. Microstructural features, such as size, morphology and distribution of fly ash microballoons, were quantitatively related to the deformation behavior using interrupted compression tests. The syntactic foam was found to exhibit four distinct stages of deformation: (i) elastic; (ii) dispersed collapse of the fly ash microballoons; (iii) unification of the dispersed collapse region forming a band of densification, usually normal to the loading axis, via localized plasticity; and (iv) final densification. The tomographic observations show that deformation is initially controlled by the fragmentation and collapse of dispersed individual fly ash particles. After significant strain, sufficient microballoons have collapsed, localizing the stress and hence damage, and forming densification bands through severe plastic deformation of the aluminum matrix. A peak stress of 73 MPa and an energy absorption capacity of 27 MJ m^-3 at 47% strain were obtained.

[en] X-ray in-line phase contrast tomography holds great promise for the quantitative analysis of soft materials. However, its applications have been limited, so far, by the fact that direct methods based on the transport-of-intensity equation and the contrast transfer function are sensitive to noise and applicable only to limited types of samples. Here, we propose an iterative method based on the Gerchberg-Saxton algorithm (R. W. Gerchberg and W. O. Saxton, Optik 35, 237 (1972)), but overcoming its slow convergence by an acceleration technique, named random signed feedback, which shows an excellent performance, both in numerical simulation and tomographic experiment, of discriminating various polymers even when using 53 keV synchrotron X-rays.

[en] In recent years, the addition of the slag phase to numerical models of the Continuous Casting (CC) process has opened the door to a whole new range of predictions. These include the estimation of slag infiltration and powder consumption (lubrication), heat transfer and cooling through the cooper mould (solidification) and investigating the effect of operational parameters such as mould oscillation and powder composition on surface quality / defect formation. This work presents 2D and 3D CC models capable of describing the dynamic behaviour of the liquid/solid slag in both the shell mould-gap and bed as well as its effects on heat extraction and shell formation. The present paper also illustrates the application of the model to a variety of casters and the challenges faced during its implementation. The model attained good agreement on the prediction of mould temperatures and shell thicknesses as well as slag film formation and heat flux variations during the casting sequence. The effect of different oscillation strategies (sinusoidal and non-sinusoidal) was explored in order to enhance powder consumption and surface quality. Furthermore, the modelling approach allows one to predict the conditions leading to irregular shell growth and uneven lubrication; these are responsible for defects such as, stickers, cracking and, in the worst case scenario, to breakouts. Possible mechanisms for defect formation are presented together with strategies to enhance process stability and productivity of the CC machine.