Numerical investigation of melting and solidification processes in modified surface layers of metal at induction heating

Abstract

One of the perspective ways to improve the operational properties of parts of machines during induction treatment of their surfaces is the modification of the melt by specially prepared nanoscale particles of refractory compounds (carbides, nitrides, carbonitrides, etc.). This approach allows us to increase the number of crystallization centers and to refine the structural components of the solidified metal. The resulting high dispersity and homogeneity of crystalline grains favorably affect the quality of the treated surfaces. 3D numerical simulation of thermophysical processes in the modification of the surface layer of metal in a moving substrate was carried out. It is assumed that the surface of the substrate is covered with a layer of specially prepared nanoscale particles of a refractory compound, which, upon penetration into the melt, are uniformly distributed in it. The possibility of applying a high-frequency electromagnetic field of high power for heating and melting of a metal (iron) for the purpose of its subsequent modification is investigated. The distribution of electromagnetic energy in the metal is described by empirical formulas. Melting of the metal is considered in the Stefan approximation, and upon solidification it is assumed that all nanoparticles serve as centers for volume-sequential crystallization. Calculations were carried out with the following parameters: specific power p0 = 35 and 40 kW/cm2 at frequency f = 440 and 1200 kHz, the substrate velocity V = 0.5-2.5 cm/s, the nanoparticles’ size is 50 nm and concentration Np = 2.0 · 109 cm−3. Based on the results obtained in a quasi-stationary formulation, the distribution of the temperature field, the dimensions of the melting and crystallization zones, the change in the solid fraction in the two-phase zone, the area of the treated substrate surface, depending on the speed of its movement and induction heating characteristics were estimated.