Sphéroidisation des poudres dans un réacteur à plasma

A theoretical and experimental study is described in this thesis on the development and use of a d.c. plasma torch for processing of powders. The theoretical criteria for the development of a d.c. plasma torch is laid out. Guidelines are proposed for the design of an efficient d.c. plasma torch. The study demonstrates that most of the energy is lost at the anode, while only 2-5% of the total plasma power is lost at the cathode. The anode lasses can be minimized by reducing the size of anode. Other parameters investigated were the plasma gas flow rate, the electrode spacing, the axial or tangential injection and addition of small percentage of diatomic gases. The developed d.c. torch is employed for the spheroidization of refractory materials such as nickel and alumina powders. The objective of the experimental study is to explore the effect of various parameters such as the particle size, the particle flow rate, the mode of injection whether single or multiport, the injection velocity and the power level as well as the composition of the plasma gas, Ar/N2 or N2/H2. It is established that the nature and flow rate of plasma gas and power input to the plasma contribute significantly to the spheroidization efficiency. The physical properties like specific heat, melting and boiling points, heat of fusion affect markedly the processing. Initial powder feed size and size range may undergo a shift in the final product due to particle evaporation. Smaller size particles and narrow size range are suggested or better efficiencies. Single and multiport injection results show that the former gives adequate efficiency over the range investigated. Loading ratio affects the spheroidization efficiency markedly. An attempt is made to develop a mathematical model for the simulation of the plasma jet. For the simulation of the plasma jet in the absence of particles, the temperature and flow fields are computed using the mixing length turbulent model and GENMIX code. Three cases are studied for argon/nitrogen, nitrogen/hydrogen and argon/hydrogen plasma jets. The predicted results are in good agreement with the available experimental data. The model is extended for the calculation of particle trajectories and thermal histories by taking into account the plasma­particle inter-action effects. Theoretical calculations demonstrate that the processing efficiency and particle evaporation depends on the initial injection velocity, loading ratio, and power level. The experimental data obtained in Chapter 3 agrees well with the theoretical simulations. The thesis is concluded in the fifth chapter by enumerating the principal contributions of this work and suggestions for future research on the subject.