The literature concerning the injection moulding of engineering ceramics has been reviewed. This indicated that a number of claims had been made for the successful use of different organic binders during moulding and their removal prior to sintering. However, many of the claims were not supported by detailed/exact eScperimental evidence as to powder-binder compositions, moulding conditions, moulded properties, debinding times/cycles, or details of the structure and properties of the solid ceramic bodies produced. From the available information it was clear that there were few systematic and scientific investigations concerning the understanding of each stage of the injection moulding process. The present research programme has been carried out in two phases as follows. The first phase was concerned with the reinvestigation and re-evaluation of binder systems claimed to be successful for the injection moulding of alumina ceramics. The binders re-investigated included the thermoplastic-based binders such as polystyrene, polyacetal and atactic polypropylene and the water-based methylcellulose (Rivers) binder system. Alumina was chosen as the main powder to be investigated due to its simple handling and, highest applications amongst ceramic materials and on the basis that there is incomplete published work for almost every step of the injection moulding process. During the first stage of this work the optimum properties such as powder-binder compositions, mixing and moulding conditions, debinding properties, green and sintered densities provided by each binder system were determined. The results of these investigations showed that all the previous (re-evaluated) binder systems had major limitations and disadvantages. These included low volume loading (64 % maximum) of the alumina powder resulting in rather low sintered densities (96 % maximum-of theoretical density) and very long debinding times in the case of the thermoplastic-based binders. it ry low alumina volume loading (55 % maximum resulting in a 94 % . sintered theoretical density) and long moulding cycle time (- 5 min) along with adhesion and distortion problems during demoulding occurred in the case of the water-based methylcellulose binder system. Further work did not appear worthwhile. The newly developed binder systems have been used with a number of other powders such as zirconia, silicon nitride, silicon carbide, tungsten carbide-6 weight % cobalt and iron-2 weight % nickel, to establish- whether injection moulding is feasible. Optimum properties such as powder volume loadings, mixing, moulding, demoulding, moulded densities, debinding and some sintered density results showed that these new binder systems can also be used successfully for the injection moulding of other ceramic and metallic powders, although a fuller evaluation of the properties such as optimum sintered densities and mechanical properties is required.

Indiana University-Purdue University Indianapolis (IUPUI) / Sintering is a primary particulate manufacturing technology to provide densification and strength for ceramics and many metals. A persistent problem in this manufacturing technology has been to maintain the quality of the manufactured parts. This can be attributed to the various sources of uncertainty present during the manufacturing process. In this work, a two-particle phase-field model has been analyzed which simulates microstructure evolution during the solid-state sintering process. The sources of uncertainty have been considered as the two input parameters surface diffusivity and inter-particle distance. The response quantity of interest (QOI) has been selected as the size of the neck region that develops between the two particles. Two different cases with equal and unequal sized particles were studied. It was observed that the neck size increased with increasing surface diffusivity and decreased with increasing inter-particle distance irrespective of particle size. Sensitivity analysis found that the inter-particle distance has more influence on variation in neck size than that of surface diffusivity. The machine-learning algorithm Gaussian Process Regression was used to create the surrogate model of the QOI. Bayesian Optimization method was used to find optimal values of the input parameters. For equal-sized particles, optimization using Probability of Improvement provided optimal values of surface diffusivity and inter-particle distance as 23.8268 and 40.0001, respectively. The Expected Improvement as an acquisition function gave optimal values 23.9874 and 40.7428, respectively. For unequal sized particles, optimal design values from Probability of Improvement were 23.9700 and 33.3005 for surface diffusivity and inter-particle distance, respectively, while those from Expected Improvement were 23.9893 and 33.9627. The optimization results from the two different acquisition functions seemed to be in good agreement with each other. The results also validated the fact that surface diffusivity should be higher and inter-particle distance should be lower for achieving larger neck size and better mechanical properties of the material.

Sintering is a primary particulate manufacturing technology to provide densification and strength for ceramics and many metals. A persistent problem in this manufacturing technology has been to maintain the quality of the manufactured parts. This can be attributed to the various sources of uncertainty present during the manufacturing process. In this work, a two-particle phase-field model has been analyzed which simulates microstructure evolution during the solid-state sintering process. The sources of uncertainty have been considered as the two input parameters surface diffusivity and inter-particle distance. The response quantity of interest (QOI) has been selected as the size of the neck region that develops between the two particles. Two different cases with equal and unequal sized particles were studied. It was observed that the neck size increased with increasing surface diffusivity and decreased with increasing inter-particle distance irrespective of particle size. Sensitivity analysis found that the inter-particle distance has more influence on variation in neck size than that of surface diffusivity. The machine-learning algorithm Gaussian Process Regression was used to create the surrogate model of the QOI. Bayesian Optimization method was used to find optimal values of the input parameters. For equal-sized particles, optimization using Probability of Improvement provided optimal values of surface diffusivity and inter-particle distance as 23.8268 and 40.0001, respectively. The Expected Improvement as an acquisition function gave optimal values 23.9874 and 40.7428, respectively. For unequal sized particles, optimal design values from Probability of Improvement were 23.9700 and 33.3005 for surface diffusivity and inter-particle distance, respectively, while those from Expected Improvement were 23.9893 and 33.9627. The optimization results from the two different acquisition functions seemed to be in good agreement with each other. The results also validated the fact that surface diffusivity should be higher and inter-particle distance should be lower for achieving larger neck size and better mechanical properties of the material.

Die Arbeit analysiert das Lasermikrosintern von Metallen in seiner Gesamtheit durch kalkulierbare Modelle und Formulierungen. Hierfür wird der Sintervorgang in relevante Prozessschritte untergliedert. Die darauf aufbauenden Berechnungen der Prozessparameter werden in Form analytischer Ansätze und durch numerische Simulation analysiert. Anfänglich werden die Modelle der Einzelschritte und deren Resultate werden auf Übereinstimmung mit den experimentellen Daten und Beobachtungen geprüft. Hierbei dienen die Modelle, Algorithmen und analytischen Beziehungen als nachhaltiges Instrumentarium für Analysen und Beschreibungen dieses und ähnlicher Prozesse. Mithilfe der entwickelten Algorithmen werden die bisherigen Vorstellungen zum Prozessmechanismus verifiziert, korrigiert, ergänzt und offene Fragen zur Prozessmechanik beantwortet. Abschließend werden die mechanistischen Modelle der Einzelschritte zur Diskussion experimenteller Ergebnisse und beobachteter Prozessphänomene in Form einer Gesamtinterpretation der betreffenden Lasermikrosinterregime zusammengeführt. Zunächst werden die Prozessbedingungen und Prozessbeschreibungen sowie die Modelle zur Strahlungsabsorption in Pulvern in Arbeiten fremder Autoren vorgestellt. Zur Beschreibung der Ursache und Wirkung laserinduzierter Plasmen, ein besonders beim Lasermikrosintern bedeutsames Phänomen, wird auf eigene frühere Arbeiten zurückgegriffen. Aus den besonderen Bedingungen des Lasermikrosinterns wird die formale Gliederung des Prozessverlaufs in drei separate Dissipationsphasen abgeleitet. Die drei relevanten Reaktionsphasen sind hierbei, die Absorption der Strahlung, die Wärmeausbreitung und Schmelzerzeugung sowie die Wärmeleitung im Pulverbett. Die ablaufenden Einzelmechanismen als Folge der gepulsten Laserstrahlung werden durch geeignete Näherungen beschrieben und berechnet. Die Resultate werden mit grundlegender Aussage mit experimentellen Beispielen verglichen und verifiziert. Die Betrachtung unterschiedlicher Regime des Lasermikrosinterns erfolgt, gestützt auf die Ergebnisse der Kalkulationen der energetisch-thermodynamischen Synthese. Es werden Aussagen zu den regimespezifischen Prozessmerkmalen getroffen und mit den Beobachtungen verglichen. Oder, es werden regimespezifische Beobachtungen mit Hilfe der im vorhergehenden Abschnitt entwickelten prozessanalytischen Werkzeuge interpretiert. Auswirkungen von Parametervariationen auf den Gesamtprozess werden im Hinblick auf das Optimierungspotential diskutiert.

Le SPS (Spark plasma sintering) ou frittage flash est une technique innovante de compaction de poudre. Ce procédé fait intervenir le courant électrique pour chauffer l'échantillon en appliquant simultanément une pression. Grâce à la vitesse de chauffage, le procédé SPS apparaît comme étant une technologie prometteuse dans le secteur aéronautique servant à produire des matériaux denses à microstructure fine, composés par des intermétalliques difficiles à fondre, à former et à usiner avec les procédés conventionnels. Cependant, la fabrication de formes complexes est problématique à cause des hétérogénéités en densité qui peuvent apparaître lors de la compaction et qui proviennent de la distribution de la température et des contraintes dans la poudre compactée. La distribution du courant, de la température et des contraintes, ainsi que leurs différents effets, font l'objet d'une large étude, étant responsables de l'homogénéité de la microstructure. Une modélisation numérique 3D du procédé est réalisée, dans le cadre de la librairie CimLib. Elle englobe trois problèmes physiques fortement couplés : électrique, thermique et mécanique. Nous utilisons une approche monolithique qui consiste à résoudre une équation pour chaque problème sur un maillage unique représentant outils et poudre. Tout d'abord le couplage électrique-thermique est modélisé et les simulations numériques sont validées. Une loi de comportement viscoplastique compressible s'appuyant sur un modèle d'Abouaf est utilisée pour modéliser la densification de la poudre de TiAl. Ce modèle est validé par plusieurs cas tests de compaction de poudre dans un contexte lagrangien puis eulérien avant de passer à une simulation complète de couplage électrique-thermique-mécanique. Dans ce contexte monolithique, nous développons un modèle pour prendre en compte les effets du frottement entre la poudre et le moule. Enfin, la loi de comportement utilisée est identifiée pour la poudre intermétallique de TiAl. Le frittage par SPS d'échantillons de différentes tailles est simulé. Les résultats en termes de distribution de densité et déplacement sont validés grâce à une comparaison avec l'expérience. / Spark plasma sintering process (SPS) is a breakthrough technology for producing high quality sintered materials. An electric current is applied simultaneously with a vertical load to sinter the powder placed in a graphite mould. Joule effect leads to high heating rates which are favorable for enhancing the microstructure and physical properties. However, manufacturing complex shapes is problematic due to heterogeneities in density distribution that may appeari during compaction. For that reason, the development of a numerical model to predict sintering is necessary. The model should help controlling temperature and stress distributions, which are responsible for the microstructure homogeneity. A 3D numerical model is developed to ensure a predictive tool for SPS using CimLib, a code developed at CEMEF. The numerical model presents three physical problems strongly coupled: an electric problem, a thermal problem and a mechanical problem. A monolithic approach is used which consists in solving one equation for each problem using one unique mesh for tools and powder. First the electric thermal coupling is modeled and the numerical simulations are validated by comparison with commercial codes. A viscoplastic compressible law based on Abouaf model is implemented to model the densification of TiAl powder. This model is validated by comparing the numerical results of different compaction tests with analytic solutions using a Lagrangian and Eulerian framework. Then a fully coupled electric-thermal-mechanical simulation is carried out. In the monolithic framework, a model is developed to take into account friction effects between powder and mould. Finally, the parameters of the selected material law are identified for TiAl powder using our numerical model and SPS experiments. Sintering of different samples is then simulated. Results are compared with the experiments in terms of density distribution and displacement.