Perspective on the Application of Evolutionary Algorithms for the Design of Magnetocaloric Devices
Author: Michel Risser
Video: Perspective on the Application of Evolutionary Algorithms for the Design of Magnetocaloric Devices - Video
The magnetocaloric effect consists in an adiabatic temperature change induced by a variation of an applied magnetic field on magnetic material. The physical phenomenon makes possible the design of magneto caloric system for heat pumping or energy conversion with better efficiencies than conventional gaz compression technologies. This new technology, in complete breakthrough with the traditional devices, has been under development for about 40 years by the scientific community and industrial companies. The magnetocaloric effect is a multiphysical and multiscale phenomenon which is interconnecting magnetic, thermal and hydraulic behaviors. One design today consists in the application on a porous matrix of a variable magnetic field in synchrony with an alternating coolant fluid motion to create a thermal gradient in a porous matrix. To improve the efficiency, this matrix can be built with a cascade of Curie temperature in order to obtain a multi-stage magnetothermodynamic cycles. Therefore, an industrial magnetocaloric device is a complex system. Several contradictions occur between the different subsystems driven with interdependent physical phenomena, hence today conventional designing methods fail to reach the optimum configuration. We will present how the evolutionary algorithms can help us to break the limit of Human designing methods.