Estimation of constitutive parameters for powder pressing by inverse modelling

Powder metallurgy processes are used in many material technologies for manufacturing of a wide range of industrial parts. Products such as components for cars, cemented carbides and high-speed steels for mechanical cutting, magnets and soft magnetic materials, bearings and refractory metals are made from powder. These parts are manufactured by powder die pressing followed by sintering of the resulting green body in a furnace. Traditionally, experience-based methods have been used to design and adapt the processing variables for optimal performance. Cost savings can be made if the tool design can be based on reliable predictive numerical simulations of the powder compaction process. Computer modelling could aid process and design engineers in selecting and optimizing the best pressing route for many industrial components. The aim of the present work has been to develop an efficient way to determine the necessary constitutive model parameters of the numerical models by means of inverse modelling. An experiment for establishing input data to the inverse problem has been designed and validated. The objective function is formed based on the discrepancy in force–displacement data between the numerical model prediction and the experiment. Minimization of the objective function with respect to the material parameters is performed using an in-house optimization software shell which is built on a modified Nelder–Mead simplex method also known as the subplex method. The completed analyses show that the proposed approach can readily be used to determine material parameters.