The response of submerged structures subject to loading by underwater blast waves is governed by complex interactions between the moving or deforming structure and the surrounding fluid and these phenomena need to be thoroughly understood in order to design structural components against underwater blast. This thesis has addressed the response of simple structural systems to blast loading in shallow or deep water environment. Analytical models have been developed to examine the one-dimensional response of both water-backed and air-backed submerged rigid plates, supported by linear springs and loaded by underwater shock waves. Cavitation phenomena as well as the effect of initial static fluid pressure are explicitly included in the models and their predictions were found in excellent agreement with detailed FE simulations. Then, a novel experimental apparatus has been developed, to reproduce controlled blast loading in initially pressurised liquids. It consists of a transparent water shock tube and allows observing the structural response as well as the propagation of cavitation fronts initiated by fluid-structure interaction in a blast event. This experimental technique was then employed to explore the one-dimensional response of monolithic plates, sandwich panels and double-walled structures subject to loading by underwater shock waves. The performed experiments provide great visual insight into the cavitation process and the experimental measurements were found to be in good agreement with analytical predictions and dynamic FE results. Finally, underwater blast loading of circular elastic plates has been investigated by theoretically modelling the main phenomena of dynamic plate deformation and fluid-structure interaction. In addition, underwater shock experiments have been performed on circular composite plates and the obtained measurements were found in good correlation with the corresponding analytical predictions. The validated analytical models were then used to determine the optimal designs of circular elastic plates which maximise the resistance to underwater blast.

Human dental tissues are fascinating nano-structured hierarchical materials that combine organic and mineral phases in an intricate and ingenious way to obtain remarkable combinations of mechanical strength, thermal endurance, wear resistance and chemical stability. Attempts to imitate and emulate this performance have been made since time immemorial, in order to provide replacement (e.g. in dental prosthodontics) or to develop artificial materials with similar characteristics (e.g. light armour). The key objectives of the present project are to understand the structure-property relationships that underlie the integrity of natural materials, human dental tissues in particular, and the multi-scale architecture of mineralized tissues and its evolution under thermal treatment and mechanical loading. The final objective is to derive ideas for designing and manufacturing novel artificial materials serving biomimetic purposes. The objectives are achieved using the combination of a range of characterization techniques, with particular attention paid to the synchrotron X-ray scattering (Small- and Wide-Angle X-ray Scattering, SAXS and WAXS) and imaging techniques (Micro Computed Tomography), as well as microscopy techniques such as Environmental Scanning Electron Microscopy (ESEM), Transmission Electron Microscopy (TEM) and Atomic Force Microscopy (AFM). Mechanical properties were characterized by nanoindentation and photoelasticity; and thermal analysis was carried out via thermogravimetric analysis (TGA). Experimental observations were critically examined and matched by advanced numerical simulation of the tissue under thermal-mechanical loading. SAXS and WAXS provided the initial basis for elucidating the structure-property relationships in human dentine and enamel through in situ experimentation. Four principal types of experiment were used to examine the thermal and mechanical behaviour of the hierarchical structure of human dental tissue and contributed to the Chapters of this thesis: (i) In situ elastic strain evolution under loading within the hydroxyapatite (HAp) in both dentine and enamel. An improved multi-scale Eshelby inclusion model was proposed taking into account the two-level hierarchical structure, and was validated against the experimental strain evaluation data. The achieved agreement indicates that the multi-scale model accurately reflects the structural arrangement of human dental tissue and its response to applied forces. (ii) The morphology of the dentine-enamel junction (DEJ) was examined by a range of techniques, including X-ray imaging and diffraction. The transition of mechanical properties across the DEJ was evaluated by the high resolution mapping and in situ compression measurement, followed by a brief description of the thermal behaviour of DEJ. The results show that DEJ is a narrow band of material with graded structure and mechanical properties, rather than a discrete interface. (iii) Further investigation regarding the thermo-mechanical structure-property relationships in human dental tissues was carried out by nanoindentation mapping of the nano-mechanical properties in ex situ thermally treated dental tissues. (iv) In order to understand the details of the thermal behaviour, in situ heat treatment was carried out on both human dental tissues and synthetic HAp crystallites. For the first time the in situ ultrastructural alteration of natural and synthetic HAp crystallites was captured in these experiments. The results presented in this thesis contribute to the fundamental understanding of the structure-property integrity mechanisms of natural materials, human dental tissues in particular. These results were reported in several first author publications in peer-reviewed journals, conference proceedings, and a book chapter.

Thesis (MEng)--Stellenbosch University, 2014. / ENGLISH ABSTRACT: In this thesis we set out to design an algorithm for solving the all-pairs fixed-radius nearest neighbours search problem for a massively parallel heterogeneous system. The all-pairs search problem is stated as follows: Given a set of N points in d-dimensional space, find all pairs of points within a horizon distance of one another. This search is required by any nonlocal or meshless numerical modelling method to construct the neighbour list of each mesh point in the problem domain. Therefore, this work is applicable to a wide variety of fields, ranging from molecular dynamics to pattern recognition and geographical information systems. Here we focus on nonlocal solid mechanics methods. The basic method of solving the all-pairs search is to calculate, for each mesh point, the distance to each other mesh point and compare with the horizon value to determine if the points are neighbours. This can be a very computationally intensive procedure, especially if the neighbourhood needs to be updated at every time step to account for changes in material configuration. The problem also becomes more complex if the analysis is done in parallel. Furthermore, GPU computing has become very popular in the last decade. Most of the fastest supercomputers in the world today employ GPU processors as accelerators to CPU processors. It is also believed that the next-generation exascale supercomputers will be heterogeneous. Therefore the focus is on how to develop a neighbour searching algorithm that will take advantage of next-generation hardware. In this thesis we propose a CPU - multi GPU algorithm, which is an extension of the fixed-grid method, for the fixed-radius nearest neighbours search on massively parallel systems. / AFRIKAANSE OPSOMMING: In hierdie tesis het ons die ontwerp van ’n algoritme vir die oplossing van die alle-pare vaste-radius naaste bure soektog probleem vir groot skaal parallele heterogene stelsels aangepak. Die alle-pare soektog probleem is as volg gestel: Gegewe ’n stel van N punte in d-dimensionele ruimte, vind al die pare van punte wat binne ’n horison afstand van mekaar af is. Die soektog word deur enige nie-lokale of roosterlose numeriese metode benodig om die bure-lys van alle rooster-punte in die probleem te kry. Daarom is hierdie werk van toepassing op ’n wye verskeidenheid van velde, wat wissel van molekulêre dinamika tot patroon herkenning en geografiese inligtingstelsels. Hier is ons fokus op nie-lokale soliede meganika metodes. Die basiese metode vir die oplossing van die alle-pare soektog is om vir elke rooster-punt, die afstand na elke ander rooster-punt te bereken en te vergelyk met die horison lente, om dus so te bepaal of die punte bure is. Dit kan ’n baie berekenings intensiewe proses wees, veral as die probleem by elke stap opgedateer moet word om die veranderinge in die materiaal konfigurasie daar te stel. Die probleem word ook baie meer kompleks as die analise in parallel gedoen word. Verder het GVE’s (Grafiese verwerkings eenhede) baie gewild geword in die afgelope dekade. Die meeste van die vinnigste superrekenaars in die wêreld vandag gebruik GVE’s as versnellers te same met SVE’s (Sentrale verwerkings eenhede). Dit is ook van mening dat die volgende generasie exa-skaal superrekenaars GVE’s sal implementeer. Daarom is die fokus op hoe om ’n bure-lys soektog algoritme te ontwikkel wat gebruik sal maak van die volgende generasie hardeware. In hierdie tesis stel ons ’n SVE - veelvoudige GVE algoritme voor, wat ’n verlenging van die vaste-rooster metode is, vir die vaste-radius naaste bure soektog op groot skaal parallele stelsels.

This thesis is devoted to the numerical and experimental investigation of residual stress and deformation in polycrystalline metallic alloys at different scales. The emphasis in the current study is placed on establishing the connection between the simulation of deformation by the Finite Element (FE) method and experimental characterisation by synchrotron X-Ray Diffraction (XRD). Of particular importance is the interpretation of modelling results and their validation by careful comparison with experimental data. The concept of eigenstrain was used extensively throughout the report to study the residual elastic strain distributions and their sources. A pseudo-thermal strain FE procedure was used systematically to simulate the residual stress states in samples and engineering components of different shape and dimensionality. The case of 1-D strain variation was considered using the example of a plastically bent bar. The direct and inverse problems of eigenstrain analysis were solved, and validated experimentally by the use of XRD and EDM slitting methods. A novel 2-D discrete inverse eigenstrain algorithm was proposed and implemented to reconstruct the residual stress distribution in a worn rail head. The link between the residual stress and deformation history was studied via thermo-mechanical modelling of the Linear Friction Welding (LFW) process. To advance the understanding of polycrystalline deformation behaviour across the scales, a crystal plasticity model was employed to simulate the elastic-plastic deformation behaviour of Ti-6Al-4V alloy. A post-processor was developed to extract the average elastic strains for orientation-specific grain groups and to compare them with XRD data. A “peak constructor” post-processor was developed that utilised the knowledge of both the elastic strain and dislocation density. In a further development step, a strain gradient crystal plasticity formulation was adopted to account for the local dislocation evolution. Intra-granular deformation analysis was carried out and micro-beam Laue experimental diffraction technique was used for validation. Thus, local lattice arrangement was studied at the microscopic, intragranular scale. Special attention was paid to the phenomenon of Laue spot “streaking”, indicative of the local lattice misorientation caused by dislocation activity during deformation. The results presented in this thesis contributed to the fundamental understanding of the residual stress and deformation in polycrystalline metallic alloys and lead to more than 20 publications in peer-reviewed journals and conference proceedings, which are listed in the Appendix.

The currently operating Generation II Advanced Gas-Cooled Reactors (AGR) in the nuclear power stations in the UK, mainly built in the 1960s and 1970s, are approaching their designed life. Besides the development of the new generation of reactors, the government is also seeking to extend the life of some AGRs. Creep and failure properties of Type 316H austenitic stainless steels used in some components of AGR at elevated temperature are under investigation in EDF Energy Ltd. However, the current empirical creep models used and examined in EDF Energy have deficiency and demonstrate poor agreement with the experimental data in the operational complex thermal/mechanical conditions. The overall objective of the present research is to improve our general understanding of the creep behaviour of Type 316H stainless steels under various conditions by undertaking theoretical studies and developing a physically based multiscale state variable model taking into account the evolution of different microstructural elements and a range of different internal mechanisms in order to make realistic life prediction. A detailed review shows that different microstructural elements are responsible for the internal deformation mechanisms for engineering alloys such as 316H stainless steels. These include the strengthening effects, associated with forest dislocation junctions, solute atoms and precipitates, and softening effects, associated with recovery of dislocation structure and coarsening of precipitates. All the mechanisms involve interactions between dislocations and different types of obstacles. Thus change in the microstructural state will lead to the change in materials' internal state and influence the mechanical/creep property. Based on these understandings, a multiscale self-consistent model for a polycrystalline material is established, consisting of continuum, crystal plasticity framework and dislocation link length model that allows the detailed dislocation distribution structure and its evolution during deformation to be incorporated. The model captures the interaction between individual slip planes (self- and latent hardening) and between individual grains and the surrounding matrix (plastic mismatch, leading to the residual stress). The state variables associated with all the microstructure elements are identified as the mean spacing between each type of obstacles. The evolution of these state variables are described in a number of physical processes, including the dislocation multiplication and climb-controlled network coarsening and the phase transformation (nucleation, growth and coarsening of different phases). The enhancements to the deformation kinetics at elevated temperature are also presented. Further, several simulations are carried out to validate the established model and further evaluate and interpret various available data measured for 316H stainless steels. Specimens are divided into two groups, respectively ex-service plus laboratory aged (EXLA) with a considerable population of precipitates and solution treated (ST) where precipitates are not present. For the EXLA specimens, the model is used to evaluate the microscopic lattice response, either parallel or perpendicular to the loading direction, subjected to uniaxial tensile and/or compressive loading at ambient temperature, and macroscopic Bauschinger effect, taking into account the effect of pre-loading and pre-crept history. For the ST specimens, the model is used to evaluate the phase transformation in the specimen head volume subjected to pure thermal ageing, and multiple secondary stages observed during uniaxial tensile creep in the specimen gauge volume at various temperatures and stresses. The results and analysis in this thesis improve the fundamental understanding of the relationship between the evolution of microstructure and the creep behaviour of the material. They are also beneficial to the assessment of materials' internal state and further investigation of deformation mechanism for a broader range of temperature and stress.

Predictive models of structures containing frictional joints presently suffer from poor descriptions of interface behaviour at the joints. This thesis aims to address this shortfall by furthering the physical understanding of parameters affecting interface behaviour such as friction and contact stiffness. Aspects of friction and contact stiffness relevant to the characterisation of fretting joints are investigated by a combined modelling and experimental approach. Friction and wear behaviour in gross-slip fretting are investigated by in-line and rotational fretting tests. New 3D topography parameters are found to be useful in the analysis of surfaces during fretting. Wear-scar shape is found to be dependent on material. A phenomenon whereby friction increases during the gross-slip phase of individual cycles is found to be due to wear-scar interaction primarily through the interference of local features distributed over the contact area. These features are similar in size to the applied fretting stroke. A simple model to explain the behaviour is put forward which shows that wear-scar shape determines the form of the friction variation. A finite-element (FE) model of the interaction of an elastic-plastic asperity junction is used to predict sliding friction coefficients. The modelling differs from previous work by: permitting greater asperity overlaps, enforcing an interface shear strength, and allowing material failure. The results are also used to predict friction coefficients for a stochastic rough surface. The magnitudes of the predicted friction coefficients are generally representative of experimental measurements. Results suggest that friction arises from both plasticity and tangential interface adhesion. Contact stiffness is studied for both fretting and non-fretting. A technique to isolate the true interface stiffness from results derived from load-deflection data is developed by comparing experimental and FE results. In the fretting wear case, comparison of tangential contact stiffness results in the literature with FE results reveals an interface whose compliance dominates the response to the extent that stiffness is proportional to contact area. In fretting tests such as this, wear debris is thought to be a factor contributing to high interface compliance. Non-fretting experiments performed here show that, at higher pressures, interface domination is reduced as the contact approaches the smooth case. Experiments are performed where contact stiffness is measured simultaneously by both ultrasound and digital image correlation. The effect of normal and tangential loading upon the contact stiffness (normal and tangential) is investigated. Experimental evidence showing that ultrasound measures an ‘unloading’ stiffness while DIC measures a ‘loading’ stiffness is obtained for the case of tangential loading where the ‘DIC stiffness’ decreases with increasing tangential load whereas the ‘ultrasound stiffness’ remains approximately constant. On average, ultrasound gives magnitudes 3.5 and 2.5 times stiffer than the DIC results for the normal and tangential stiffness cases, respectively. The difference in magnitudes can largely be physically explained, and is relatively small considering the significant differences between the techniques. Therefore, both methods can claim to give valid measurements of contact stiffness – though each has its own limitations which are outlined herein.

Capillaries in tumours are often severely buckled (in a plane perpendicular to the axis) and / or chaotic in their direction. We develop a model of these phenomena using nonlinear solid mechanics. Our model focusses on the immediate surrounding of a capillary. The vessel and surrounding tissue are modelled as concentric annulii. The growth is dependent on the concentration of a nutrient (oxygen) diffusing from the vessel into the tumour interstitium. The stress is modelled using a multiplicative decomposition of the deformation gradient F=F_e F_g. The stress is determined by substituting the elastic deformation gradient F_e (which gives the deformation gradient from the hypothetical configuration to the current configuration) into a hyperelastic constitutive model as per classical solid mechanics. We use a Blatz-Ko model, parameterised using uniaxial compression experiments. The entire system is in quasi-static equilibrium, with the divergence of the stress tensor equal to zero. We determine the onset of buckling using a linear stability analysis. We then investigate the postbuckling behaviour by introducing higher order perturbations in the deformation and growth before using the Fredholm Alternative to obtain the magnitude of the buckle. Our results demonstrate that the growth-induced stresses are sufficient for the capillary to buckle in the absence of external loading and / or constraints. Planar buckling usually occurs after 2-5 times the cellular proliferation timescale. Buckles with axial variation almost always go unstable after planar buckles. Buckles of fine wavelength are initially preferred by the system, but over time buckles of large wavelength become energetically more favourable. The tumoural hoop stress T_{ThetaTheta} is the most invariant (Eulerian) variable at the time of buckling: it is typically of the order of the tumoural Young's Modulus when this occurs.

The core aim of this research project was to improve understanding of the effects of microstructure and crystallographic texture on the high strain rate plastic deformation behaviour of the industrially important Titanium alloy, Ti-6Al-4V. To facilitate this study, four rolled plates of Ti-6Al-4V, with varying thermo-mechanical processing histories, were provided by TIMET Corp., the world’s largest supplier of Titanium product. To determine the nature of each plate’s microstructure and the crystallographic texture of the dominant α phase, the four Ti-6Al-4V plates were microstructurally characterised using techniques such as optical microscopy and electron backscatter diffraction (EBSD). To determine the effects of the measured microstructures and textures on the strain rate dependent plastic deformation behaviour of the four Ti-6Al-4V plates, uniaxial compression and tension tests were carried out in the three orthogonal material orientations at quasi-static (10^-3 s^-1) and high strain rates (10^3 s^-1) using a standard electro-mechanical test device and split-Hopkinson pressure bars (SHPB), respectively. To provide further understanding of the effects of microstructure and texture on the plastic deformation behaviour of Ti-6Al-4V, this time under complex impact loading conditions, the classic Taylor impact experiment was adapted to include an optical measurement and geometry reconstruction technique. A novel experimental setup was designed that consists of an ultra-high speed camera and mirror arrangement, allowing the Taylor impact specimen to be viewed from multiple angles during the experiment. Using the previously mentioned optical measurement and geometry reconstruction technique, it was then possible to gain valuable, previously unobtainable, data on the deformation history of Taylor impact specimens in-situ, such as the major/minor axes of the anisotropically deforming elliptical specimen cross-sections as a function of time and axial position, true strain as a function of time and axial position, and the true strain rate as a function of axial position. The technique was verified by testing a specimen cut from the in-plane material orientation of a clock-rolled high purity Zirconium plate. The output measurements from a post-deformation image frame were compared with measurements of the recovered specimen made using a coordinate measurement machine (CMM), with analysis showing excellent agreement between the two techniques. The experiment was then carried out on specimens cut from the two orthogonal in-plane material orientations of one of the four Ti-6Al-4V plates. Analysis of the data from these experiments gave significant insight into the plastic deformation behaviour of macroscopically textured Ti-6Al-4V under complex impact loading. Recovered Ti-6Al-4V specimens from the outlined Taylor impact experiments were then sectioned along specific planes and microstructurally characterised using EBSD, with comparisons made between the pre and post-deformation microstructures. From this analysis, and the previously discussed geometry reconstruction technique, insight was gained into the effects of micro-texture on the general anisotropic plastic deformation behaviour of Ti-6Al- 4V plate materials and in particular the role of micro-texture on the formation of deformation twins. Finally, the understanding gained from these experiments, and a detailed review of the literature, was used to inform a novel, physically based material modelling framework, capable of capturing the effects of microstructure and texture on the strain rate and temperature dependent plastic deformation behaviour of Ti-6Al-4V. The model was implemented in the computational software package, MATLAB, and verified by comparison with the mechanical characterisation results from one of the Ti-6Al-4V plates. A number of frameworks are discussed for implementing the new Ti-6Al-4V model within finite element (FE) analysis software packages, such as ABAQUS, LS-DYNA and DEFORM. It is hoped that the new Ti-6Al-4V model can be used to optimise the design of Ti-6Al-4V components and structures for impact loading scenarios.

The primary aim of this research was to develop and improve the experimental method and data interpretation for strain measurements using diffraction methods to gain a better understanding of micromechanical deformation and anisotropy of lattice strain response. Substantial part of the research was devoted to the development of the laboratory high energy X-ray diffractometer (HEXameter) for bulk residual strain evaluation. White beam energy dispersive X-ray diffraction was chosen as the principal diffraction mode due to its extreme efficiency in utilising X-ray flux and its ability to capture large segments of diffraction patterns. The specimens that have been examined were real engineering components, mechanically deformed specimens and thermally treated specimens, ranging from dynamic in-situ measurements to ex-situ materials engineering. For the real engineering components, a wedge coupon from the trailing edge of a Ti64 wide fan blade and a turbine combustion casing were examined. Among the mechanically deformed specimens that have been measured were shot-peened steel plates, elasto-plastically bent bars of Mg alloy and cold expanded Al disks. Amongst the thermally deformed specimens, laser-formed steel plates, thermal spray coatings, a manual inert gas weld of Al plates, a friction stir weld of Al plates and Ni tubes and a quenched Ni superalloy cylinder used for strain tomography were studied. In-situ loading experiments have also been carried out, such as experiments on pointwise mapping of grain orientation and strain using the 3DXRD microscope at the ESRF and in-situ loading experiments on titanium alloy, rheo-diecast and high pressure diecast Mg alloy, IN718 Ni-base superalloy and Al2024 aluminium alloy. Experimental results from X-Ray diffraction and strain tomography were used to achieve a better understanding of the material properties. Some results were compared with polycrystal Finite Element model predictions. Amongst the most prominent research achievements are the development on the HEXameter laboratory instrument, including: (i) the development of special collimation systems for the detectors and the source tube; (ii) the development of a twin-detector setup (that allows for simultaneous determination of strain in two mutually orthogonal directions); (iii) improved alignment procedures for better performance; and (iv) the adaptation of instrumentation for efficient scanning of both large and small components, that included choosing and adapting translation devices, programming of the translation system and designing sample mounting procedures. In this research several approaches to data treatment were investigated. Quantitative phase analysis, single peak fitting (using custom Matlab routines and GSAS) and full pattern fitting (with individual pattern data refined by GSAS and batch refinement done by invoking GSAS via a Matlab routine) were applied. Different Matlab routines were written for specific experimental setups; and various analysis methods were selected and used for refinement depending on the requirements of the measurement results interpretation. 16 papers were published, ensuring that the results of this thesis are readily available to other researchers in the field.

L'application du procédé SLM est limitée par la difficulté à contrôler le procédé. Son application aux céramiques est particulièrement difficile en raison de leur faible absorption au laser et de leur faible résistance au choc thermique. La maîtrise de ce procédé nécessite une compréhension complète du transfert de chaleur, de la dynamique des fluides et de la mécanique des solides. Dans ce travail, nous proposons un modèle numérique pour la simulation du procédé SLM appliqué aux céramiques. Le modèle est développé à l'échelle du cordon et avec l'hypothèse d'un lit de poudre continu. Il est basé sur la méthode level set et l'homogénéisation multiphasique, avec laquelle nous sommes capables de suivre l'évolution de l'interface gaz/matière et les transformations de phase. La simulation dévelopée permet d'étudier l'influence des propriétés du matériau et des paramètres du procédé sur la température, la forme du bain liquide, la dynamique des fluides et la mécanique des solides. En dehors de la puissance du laser et de la vitesse de balayage, l'absorption du matériau est également importante pour la thermique et la forme du bain liquide. Avec la dynamique des fluides, la forme convexe du cordon est obtenue sous tension de surface. Les gouttelettes liquides se forment lors de la fusion de la poudre et créent une instabilité du bain. Ceci entraîne une irrégularité du cordon après solidification. L'effet Marangoni, provoqué par le gradient surfacique de la tension de surface, est étudié. Son influence sur la répartition de la température, la forme du bain liquide et la régularité du cordon est évoquée. Cet effet peut lisser la surface du cordon avec ∂γ/∂T négatif. En augmentant la vitesse de balayage, la surface du cordon devient plus irrégulière. L'effet de « balling » est reproduit avec une vitesse de balayage élevée. Cela peut être utile pour trouver le régime donnant une forme de cordon régulière étant données la puissance et la vitesse du laser. Le défaut de fissuration est délétère dans la fabrication additive. L'utilisation d'un laser auxiliaire peut aider à éviter ce défaut en diminuant la contrainte de traction maximale. Le mode de fonctionnement de ce laser auxiliaire reste un sujet intéressant à étudier et quelques pistes ont été données par les simulations présentées. Le modèle est validé par la comparaison de la forme du bain liquide avec des expériences dans différentes conditions de procédé. Les simulations peuvent également révéler la tendance de variation de la surface du cordon dans certains cas. Par la simulation de la déposition de cordons multiples, l'influence de taux de recouvrement sur la surface d'une couche, la température et l'évolution de contrainte est soulignée. / The application of SLM process is limited by the difficulty of process control. Its application to ceramics is especially challengeable due to their weak absorption to laser and weak resistance to thermal shock. The mastery of this process requires a full understanding of heat transfer, fluid dynamics in melt pool and solid mechanics. In this work, we propose a numerical model for the simulation of SLM process applied to ceramics. The model is developed at the track scale and with the assumption of continuous powder bed. It is based on level set method and multiphase homogenization, with which we are able to follow the evolution of gas/material interface and phase transformation. Simulations are performed to study the influence of material properties and process parameters on temperature, melt pool shape, fluid dynamics and solid mechanics. Apart from the laser power and scanning speed, material absorption is also found to be important to the thermal behavior and the melt pool shape. With the fluid dynamics, convex shape of track cross section is achieved under surface tension. Besides that, liquid droplets collapsing formed by the melting of powder create melt pool instability when falling, thus leading to track irregularity after solidification. The Marangoni effect, caused by surface tension gradient at gas/material interface, is investigated. Its influence on temperature distribution, melt pool shape and track regularity is recognized. One interesting finding is the smoothing effect of track surface with negative ∂γ/∂T. When combine surface tension with scanning speed, track surface becomes more irregular with the increase of scanning speed. The well-known balling effect is reproduced with high scanning speed. This can be helpful to find the regime for regular track shape with given laser power and scanning speed. Cracking defect is deleterious in additive manufacturing. The use of an auxiliary laser can help to avoid this defect by decreasing the maximum tensile stress. The process mode of this auxiliary laser remains an interesting subject to be studied and some guidelines have been given by the presented simulations. The model is validated by the comparison of melt pool shape with experiments under different process conditions. Simulations can also reveal the tendency of track surface variation for certain cases. By the application to multi-track deposition, the influence of hatch distance on layer surface, temperature and stress evolution is emphasized.