An open source DNS solver for the simulation of Active Magnetocaloric Regenerative cycles. (August 2018)

• Record Type: Journal Article
• Authors: Mugica, Ibai; Poncet, Sébastien; Bouchard, Jonathan
• Is Part Of: Applied thermal engineering. Volume 141(2018)
• Page Start: 600
• Page End: 616
• Publication Date: 2018-08
• Subjects: Magnetic refrigeration -- Multiphysics solver -- Active Magnetocaloric Regenerative -- Porous medium -- Parallel computing; Heat engineering -- Periodicals; Heating -- Equipment and supplies -- Periodicals; Periodicals; 621.40205
• DOI: 10.1016/j.applthermaleng.2018.06.007
• Languages: English
• ISSNs: 1359-4311

Abstract

Highlights: The separate solution of the constituent physical phenomena was validated. Decoupling the temperature from magnetic and velocity fields gave a better runtime. The developed 3D solver shows very close agreement to a parallel plate prototype. The speed-up of the solver is analyzed to provide the user a wall time estimation. Abstract: This paper describes the development of a new versatile multiphysics solver, whose final objective is to characterize Active Magnetocaloric Regenerative (AMR) cycles. The description is followed by a thorough validation divided in two parts. First, the separate simulations of the constituent physics are validated (porous geometry generation, fluid flow, Conjugated Heat Transfer and magnetic field). Then, the whole AMR code is compared to a measured prototype. The results are in all cases close to published numerical or experimental data. Also, two coupling mechanisms are studied, the temperature dependency of the viscosity of the Heat Transfer Fluid (HTF), and the influence of the magnetic permeability on the internal magnetic field of Gadolinium metal. Both effects prove to be negligible to compute the temperature field of a Magnetic Regenerator (MR). Moreover, by neglecting these effects the magnetic and flow fields of the MR can be computed beforehand, which reduces substantially the computational load at run time. Finally, the speed-up of the parallel multiphysics solver is evaluated.