The chemical binders used to make sand molds and cores thermally decompose and release gas when subjected to the high temperature conditions in sand casting processes. Computational models that predict the evolution of the binder gas are being introduced into casting simulations in order to better predict and eliminate gas defects in metal castings. These models require knowledge of the evolved binder gas mass and molecular weight as a function of temperature, but available gas evolution data are limited. In the present study, the mass and molecular weight of gas evolved from PUNB bonded sand are measured as a function of temperature for use with binder gas models. Thermogravimetric analysis of bonded sand is employed to measure the binder gas mass evolution as a function of temperature for heating rates experienced in molds and cores during casting. The volume and pressure of gas evolved from bonded sand are measured as a function of temperature in a specially designed quartz manometer during heating and cooling in a furnace. The results from these experiments are combined with the ideal gas law to determine the binder gas molecular weight as a function of temperature. Thermogravimetric analysis reveals that the PUNB binder significantly decomposes when heated to elevated temperatures, and the PUNB binder gas mass evolution is not strongly influenced by heating rate. During heating of PUNB bonded sand at a rate of 2°C/min, the binder gas molecular weight rapidly decreases from 375 g/mol at 115°C to 99.8 g/mol at 200°C. The molecular weight is relatively constant until 270°C, after which it decreases to 47.7 g/mol at 550°C. The molecular weight then steeply decreases to 30.3 g/mol at 585°C and then steeply increases to 47.2 g/mol at 630°C, where it remains constant until 750°C. Above 750°C, the binder gas molecular weight gradually decreases to 33.3 g/mol at 898°C. The present measurements are consistent with the molecular weights calculated using the binder gas composition data from previous studies. The binder gas is composed of incondensable gases above 709°C, and the binder gas partially condenses during cooling at 165°C if the bonded sand is previously heated below 507°C.

Foundries today use temporary molds made from silica sand with a resin bonding agent to hold a form until the metal is poured. With the aid of computer simulations, the molds are designed to produce good castings with minimal pattern iterations by calculating cooling and porosity. Stress analysis simulations are being developed using the current software, but the known mechanical properties for the sand mold are minimal and incomplete. This study measures the elastic modulus of bonded sand as a function of temperature to obtain baseline data for the model. Following ASTM standards, a three point bend test is used to measure the elastic modulus of chemically bonded sand as a function of temperature to better understand the complex nature of the mold as it undergoes heating and cooling. Multiple measurements of the elastic modulus of PUNB bonded silica sand are performed from room temperature to 500°C in a nitrogen atmosphere to capture the changes in the elastic modulus under heating. It is found that for an intermediate heating rate of 8°C/min, the elastic modulus decreases steeply from a room temperature value of about 3,600 MPa to 600 MPa at 125°C. Between 125°C and 250°C, the elastic modulus is relatively constant. Above 250°C, it increases to 1,200 MPa at 280°C and then decreases again to 800 MPa at 350°C. Above 350°C, the elastic modulus increases linearly with temperature until it reaches 2,200 MPa at 500°C. At approximately 500°C, the strength of the bonded sand vanishes. At a given temperature above 125°C, the elastic modulus can vary by more than a factor of two depending on the heating rate. Furthermore, the elastic modulus agrees with previous steady state temperature measurement literature when specimens are held at a constant temperature until the elastic modulus reach steady. It is also found that the addition of black iron oxide has no effect on the elastic modulus, whereas solvent removal before a test increases the stiffness of the bonded sand at temperatures below 150°C.

The goal of this work is to develop and characterize a manufacturing process that is able to create metal matrix composites with complex cellular geometries. The novel manufacturing method uses two distinct additive manufacturing processes: i) fabrication of patternless molds for cellular metal castings and ii) printing an advanced cellular ceramic for embedding in a metal matrix. However, while the use of AM greatly improves the freedom in the design of MMCs, it is important to identify the constraints imposed by the process and its process relationships. First, the author investigates potential differences in material properties (microstructure, porosity, mechanical strength) of A356 — T6 castings resulting from two different commercially available Binder Jetting media and traditional 'no-bake' silica sand. It was determined that they yielded statistically equivalent results in four of the seven tests performed: dendrite arm spacing, porosity, surface roughness, and tensile strength. They differed in sand tensile strength, hardness, and density. Additionally, two critical sources of process constraints on part geometry are examined: (i) depowdering unbound material from intricate casting channels and (ii) metal flow and solidification distances through complex mold geometries. A Taguchi Design of Experiments is used to determine the relationships of important independent variables of each constraint. For depowdering, a minimum cleaning diameter of 3 mm was determined along with an equation relating cleaning distance as a function of channel diameter. Furthermore, for metal flow, choke diameter was found to be significantly significant variable. Finally, the author presents methods to process complex ceramic structure from precursor powders via Binder Jetting AM technology to incorporate into a bonded sand mold and the subsequently casted metal matrix. Through sintering experiments, a sintering temperature of 1375 °C was established for the ceramic insert (78% cordierite). Upon printing and sintering the ceramic, three point bend tests showed the MMCs had less strength than the matrix material likely due to the relatively high porosity developed in the body. Additionally, it was found that the ceramic metal interface had minimal mechanical interlocking and chemical bonding limiting the strength of the final MMCs. / Ph. D.

Afin d’optimiser le procédé de production de culasses automobiles en aluminium et d’améliorer sa fiabilité, une meilleure connaissance du comportement des noyaux de fonderie est nécessaire. Les noyaux sont des matériaux composés de grains de sable liés par une résine en faible proportion, et servent à mouler les conduits et cavités intérieurs des pièces métalliques. Au cours de leur cycle de vie, les noyaux sont amenés à subir de fortes températures et des sollicitations mécaniques complexes. L’objectif de cette étude est de caractériser le comportement thermomécanique des noyaux sous diverses sollicitations selon le taux de liant, la température et surtout l’état de dégradation de ce liant. En particulier, les travaux de cette thèse s’attardent sur les mécanismes d’endommagement et de fissuration des noyaux, dont la compréhension est la clé pour optimiser l’étape de débourrage qui consiste en la fracturation et l’élimination des noyaux. Les résultats de cette thèse serviront par ailleurs à nourrir un modèle numérique simulant le débourrage des pièces. Différents modes de chargements mécaniques ont été étudiés : flexion, compression, cisaillement, push-out, fatigue, fluage ou encore compression œdométrique. Ils correspondent à différents types de sollicitations, avec ou sans confinement du noyau, et permettent de caractériser soit le cœur ou la surface du matériau. L’endommagement peut être suivi grâce à la réalisation de cycles de décharge-recharge. Dans tous les cas, le comportement mécanique du noyau est fortement dépendant des propriétés et de l’état du liant. Ainsi, la dégradation thermique induite par la coulée de l’aluminium liquide modifie et dégrade nettement les propriétés du noyau. Des parallèles peuvent ainsi être dressés entre les propriétés du liant, le comportement mécanique du matériau sable/résine et l’endommagement microstructural observé après rupture. / Better knowledge about the mechanical behavior of foundry sand cores is required in order to optimize the aluminum cylinder head production process. Sand cores allow the casting of complex shape metallic parts and are made of sand grains, bound with a resin in low proportion. Sand cores are subjected to high temperatures and complex mechanical load during the production process. This study aims at characterizing the thermomechanical behavior of sand cores under various loads depending on the binder proportion, the temperature and mostly the binder thermal degradation. Particularly, we focus on the cores damage and fracture mechanisms. The understanding of such phenomena could lead to an optimization of the decoring step, consisting in the fracture and removing of the core from the metallic part. Indeed, the results of this Ph.D. thesis will be used to implement a numerical model of the decoring process. Different kinds of mechanical loads have been studied: bending, compression, shear, push-out, fatigue, creep or also oedometric tests. They correspond to different kinds of load, with or without confining pressure and allow a characterization of the bulk or the surface of the material. The core damage is followed by mechanical tests with unload-reload cycles sequences. For all the cases, the core mechanical behavior is highly dependent on the binder properties and thermal degradation. For example, the thermal degradation induced by the aluminum casting modifies and degrades significantly the core properties. Analogies could be drawn between the binder properties, bonded-sand cores mechanical behavior and the evolution of the fracture surfaces.