Experimental validation of a multiphysics modeling for a magnetocaloric bench

Abstract

• A new multiphysics model is applied to a magnetocaloric regenerator in long AMR sequences. • It is based on a magnetostatic reluctance network modeling of the entire AMR system. • Fully adiabatic AMR cycles are imposed to gadolinium plate regenerator in a highly effective electromagnet air gap at cold ambient temperature. • The experimental evolution of the fluid temperature span is well reproduced by the model. • A characteristic AMR time constant is proposed, using a heat transfer scale analysis. The numerical simulation results provided by a new multiphysics model of a magnetocaloric regenerator undergoing active magnetic regeneration (AMR) cycles inside an electromagnet air gap are compared with the experimental behavior of a similar built-in regenerator composed of gadolinium parallel plates. This multiphysics model is based on the coupling of an original semi-analytical magnetostatic model (reluctance network model) with a magnetocaloric model and thermo-fluidic model. The main objective of this work is to provide the first experimental validation of the multiphysics model by focusing on the production of the fluid temperature span by the magnetocaloric regenerator during long sequences of successive AMR cycles with an initial temperature below the Curie temperature of gadolinium. The case study offers a more extensive understanding of the thermal behavior of the regenerator under fully adiabatic conditions by theoretically and experimentally investigating the thermal inertia that governs this behavior. During this test, a trapezoidal alternating fluid flow was produced inside the regenerator by a controlled hydraulic cylinder that drove the calibrated magnetic field pulses generated by a highly effective electromagnet. An active magnetic refrigeration time constant of the magnetocaloric system is proposed based on a scaling analysis of the overall heat transfer during the transition to the stationary state. It effectively reflects the numerical and experimental behaviors. The numerical results are in good agreement with the experimental measurements and can facilitate further research into the estimation of magnetocaloric heating and cooling power as well as optimization of the design of plate regenerators of different magnetocaloric materials.