Research On Numerical Simulation Of Powder Sintering Process Based On Thermo-Mechano-Diffusional Coupling Model

Powder metallurgy is a process in which metal powder and non-metal powder are used as raw materials and formed by sintering.At present,it is not only widely used in the production of various common metal and ceramic materials,but also has been greatly developed in the manufacture of various metal-based and ceramic-based composite materials,single crystals,and other special materials.At the same time,due to the development of new powder sintering technologies such as current-assisted sintering and spark plasma sintering,powder metallurgy has broad application prospects in frontier fields such as heterogeneous material connection and cracks repair.These sintering technologies have greatly expanded the field of traditional material preparation.In this process,the combination of a numerical simulation study of sintering mechanism and prediction of sintering densification is of great significance to the development and application of the sintering process.From an energy point of view,powder sintering is a process of system energy minimization.The phase-field method simulation of the sintering process is exactly a method to minimize the free energy of the volume with the Cahn-Hilliard equation and Allen-Cahn equation as the basic governing equations.In this paper,the basic theoretical framework of the coupled phase-field method is used to explore and apply the powder sintering in detail as follows.(1)The thermo-mechano-diffusional coupled governing equations based on the phase-field model are derived.The governing equations of thermal-force-diffusion coupling based on the phase-field model are derived.Starting from the energy conservation equation,entropy equation,and mass conservation equation,the full differential form of the free energy density function is given,and the generalized constitutive relationship between the free energy density function and each physical field variable is derived through the generalized Gibbs-Duhem formula.Finally,on this basis,the partial differential governing equations under the non-equilibrium thermo-mechano-diffusional coupled form are derived.(2)A multi-physics coupling calculation program was developed to solve the thermal-mechano-diffusional coupled governing equation and realize the numerical simulation of hot-press sintering.Taking the finite element as the solution method of partial differential control equations,an algorithm framework for the solution of the thermal-mechano-diffusional multi-physical coupling governing equations is designed,and the numerical solution of the control equations is obtained by using the self-developed C++ multi-physical coupling program.The microstructure evolution under pressure-assisted sintering was simulated,and the influence of the strain energy stored in the sintering neck and the coupling energy formed by the interaction of force and diffusion on the sintered microstructure evolution was analyzed.(3)Combined with machine learning,a sintering transfer learning model is proposed to solve the problem of densification prediction of difficult-to-sinter materials,and the problem of sintering trend prediction in the case of a lack of samples in special sintering scenarios is discussed,such as the preparation of difficult-to-sinter material like Si C.At the same time,the feasibility of our multi-physics coupling program to provide some data samples for the machine learning method for small sample data was discussed.