IMprove the thermal models in order to take into acount radiation and conduction and the influence of the temperature on the parameter variations

Abstract

This project has as objective to improve the thermal model of Majid Souley thesis based on Multi-coil heating technology, specifically, understand what is the result of changing the value of the coefficient of thermal convection from a constant value to an expression dependent from temperature. It also aspires to obtain graphical results from a simplified model of two coils in the area of magnetic and thermal results. For obtaining these results, some programs as PSIM® and ANSYS® will be used. PSIM® uses simulation to plot results. In order of doing these simulations, it uses equations that have to been deduced before, so the first step will be to introduce thermal equations to the model. These equations have been already written in Majid Souley thesis but some of them have been changed. ANSYS® model has been created from zero and using some 2-D examples of the problem a 3-D model has been made. The results obtained shown that, in the case of PSIM® model, hav to be improved because of there is a lack of information between the first point and the center of the disk, and resolving that type of problem is out from this project knowledge. So, the results obtained have been distorted by this lack of information, but without taking this into account, PSIM® model has better results with a constant value of the coefficient of thermal convection. With ANSYS®, promising results have been obtained in the case of magnetic flux density across the metal. The results of the distribution of magnetic flux density corresponds with the expected from theory but the distribution of temperature has slightly differences of what is supposed to be. Also, some results about structural movements have been obtained, and they must be deeper studied. To conclude, there is not any reason to reject thermal model used in PSIM® program but, solving this lack of information will verify if using a coefficient of thermal convection dependent from temperature the model is better or not. On the other hand, with the results obtained, the model is better with a constant value of the coefficient of thermal convection. With ANSYS®, the solution given can be a support for future models, although that the distribution of temperatures has to be improved, but with that model other interesting results can be obtained, like interior stress of the metal as a result of the magnetic field.