Ghassane Ayad

Sequential and global optimization in powder injection molding processing

The work is focused on the overall optimization associated to powder injection molding (PIM). The PIM process includes four main stages, from the mixture of the powders and binders to the final sintering stage. Injection and sintering stages are considered to be the most important for optimization, as they mostly affect the final quality of the produced components. The injection stage shapes the green parts but initiates powders segregation that will be inherited and amplified by the sintering stage to finally appear in the resulting products. One first introduces an optimization loop based on the surfaces response method to minimize the powder segregation. Then the results are transferred to a sintering optimization loop applied through an experimentally calibrated thermo-mechanical creep model to predict the shrinkage and density contours on the final parts. The overall optimization combines both optimizers based on the developed simulation tools to provide a realistic way to improve the PIM process design accounting the different processing stages.

Optimisation de la chaîne complete du procédé de moulage par injection métallique

Cet article est consacré à l’identification paramétrique et à l’optimisation des phases d’injection et de densification par diffusion à l’état solide du procédé de moulage par injection métallique. La phase d’injection est modélisée et simulée grâce à un modèle d’écoulement biphasique, tandis que les mécanismes de densification par diffusion à l’état solide sont représentés par un modèle macroscopique. La qualité des résultats numériques reposant sur les modèles physiques dépend directement de l’identification des parameters matériaux pour chacune des étapes. On propose dans l’article une combinaison entre l’optimisation paramétrique de la phase d’injection, l’identification paramétrique du modèle de diffusion et enfin l’optimisation de la phase de densification afin de déterminer de manière optimale les dimensions d’empreintes de moule, de façon à obtenir des composants sans défauts, aux dimensions et propriétés mécaniques souhaitées.

An optimization strategy for the determination of material and process parameters to avoid segregation defects during metal injection powder

An explicit 3D software “FEAPIM” has been developed at LMARC to perform efficiently the injection simulation to predict and simulate a segregation effect in the mixture flows during MIM, and segregation is the most defect in metal injection molding. In order to reduce it on suggest optimization strategy.A cost function is first defined to account it, and optimization strategy has been developed using the Design Of Experiments (DOE) to find the most influent parameters. Four sensitive parameters are found : Powder volume fraction, Interaction coefficient, Powder density, and Binder density. In order to decrease the computational cost associated to optimization, the response surface, using the Moving Least Square Approximation (MLSA), is used to approximate the cost function. Then a genetic algorithm is coupled with this response surface to obtain the optimal values in reasonable time. The optimization strategy proposed in this paper has been applied to the tensile test. Optimization results are compared to...