Influence of layering and Curie temperature uncertainty on the performance of multilayer active magnetic regenerators

Tomc, Urban (author), Peixer, Guilherme Fidelis (author), Bahl, Christian Robert Haffenden (author), Nielsen, Kaspar K. (author), Lozano, Jaime A. (author), Barbosa, Jader R. (author), Kitanovski, Andrej (author)

Magnetic refrigeration is a promising alternative to traditional vapor compression systems, with potential efficiency and environmental sustainability advantages. However, the narrow operational temperature range of magnetocaloric materials (MCMs) and their reliance on rare-earth elements remain key challenges. Multilayer active magnetic regenerators (AMRs) address the temperature range limitations by combining multiple magnetocaloric layers, each with different Curie temperatures. This study investigates the impact of statistical deviations in Curie temperatures on the performance of multilayer AMRs, specifically using second-order La-Fe-Co-Si materials. A 1D multilayer AMR numerical model is developed to simulate the effects of Curie temperature variability, with radial basis function neural networks employed to efficiently predict performance. The results indicate that although increasing the number of MCM layers enhances cooling power and coefficient of performance (COP), Curie temperature uncertainties significantly degrade the AMR performance. The likelihood of achieving cooling targets diminishes as the number of MCM layers increases, particularly for standard deviations exceeding 1 K. These findings emphasize the importance of accounting for Curie temperature uncertainties in AMR design. Moreover, enhancing the manufacturing precision of the Curie temperatures of MCMs is essential for improving the performance and commercialization of magnetocaloric technology.

• Language: English
• Keywords: active magnetic regenerator, magnetocaloric properties, Curie temperature, neurual networks, numerical modelling, machine learning, multilayers, numerical models
• Work type: Article
• Typology: 1.01 - Original Scientific Article
• Organization: FS - Faculty of Mechanical Engineering
• Publication status: Published
• Publication version: Version of Record
• Year: 2025
• Number of pages: 21 str.
• Numbering: Vol. 35, iss. 52, art. e24282
• PID: 20.500.12556/RUL-177850
• UDC: 537:536
• ISSN on article: 1616-3028
• DOI: 10.1002/adfm.202424282
• COBISS.SI-ID: 249789699
• Publication date in RUL: 09.01.2026

Record is a part of a journal

• Title: Advanced functional materials
• Shortened title: Adv. funct. mater.
• Publisher: Wiley-VCH
• ISSN: 1616-3028
• COBISS.SI-ID: 23505413

Licences

• License: CC BY-NC-ND 4.0, Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International
• Link: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Description: The most restrictive Creative Commons license. This only allows people to download and share the work for no commercial gain and for no other purposes.

Secondary language

• Language: Slovenian
• Keywords: aktivni magnetni regenerator, magnetokalorične lastnosti, Curiejeva temperatura, nevronske mreže, numerični modeli