Associated People

J. Schröder, O. Gutfleisch (TU Darmstadt), R. Niekamp, M. Grönewald (TU Darmstadt)
 

Abstract

The collaborative research project SFB/TRR 270 "Hysteresis design of magnetic materials for efficient energy conversion" - HoMMage for short - is dedicated to the research of new magnetic materials for energy conversion. This includes both hard permanent magnets with maximized hysteresis, as well as soft magnets with minimized hysteresis. Both types have different applications, the former in the field of wind generators or electromobility, the latter in magnetic cooling using the magnetocaloric effect. What all these applications have in common is that they benefit from new magnetic materials that have been improved for the respective application. Therefore, within the SFB/TRR 270 researchers from different disciplines such as materials science, physics, chemistry or production engineering, distributed over several working groups at five locations, are looking for new innovative magnetic materials and ways to process them.
In addition to the classical paradigms, this multidisciplinary joint project is pursuing the data-driven approach to find new magnetic materials with improved properties. The INF subproject provides support in this regard, especially in questions of research data management(RDM), so that among other things the generated research results can be stored and exchanged in a centralized and structured way.
As a central service project within the SFB/TRR 270, INF aims to enable sustainable management of research data according to the FAIR principle for all stakeholders across sites. In an abstract view, besides purely infrastructural challenges, the most important ones are: complexity, heterogeneity and size of data. One of the first tasks was to define plans for data collection and management (so-called data management plans, DMPs). An electronic laboratory record book (ELB) based on free software (FLOSS) is available and will be continuously extended. Large parts of the resulting research data can be loaded and shared via it and can be found via a user interface according to basic criteria. Via an interface, the content is available for automated data conversion and reduction.
In conjunction with a storage infrastructure and interfaces for particularly large research data, it is part of the necessary toolkit to analyze the collected research data and metadata using machine learning algorithms to discover new material. One other essential task for INF is to provide training to all members of SFB/TRR 270 on topics related to FDM through regular workshops and consultation, and to raise awareness of possible improvements in data handling. To prevent silo thinking, the integration of data and tools into the broader FDM landscape is also an important issue. Many challenges are posed to RDM within an interdisciplinary collaborative project.
In addition to technical requirements and incompatibilities in the integration of existing infrastructure, collaboration between the technically heterogeneous groups also generates new, and in some cases specialized, needs. The INF project within the SFB/TRR 270 HoMMage aims at meeting these challenges with a combination of self-managed and centralized infrastructure, as well as the task to provide knowledge and support to the researchers within the network for a better RDM, and the claim of a scientific reuse in the context of artificial intelligence. Thus, with the help of structured research data and modern computer methods, a successful contribution can be made to the development of new promising magnetic materials can be delivered.
 

References

Tolle, Kristin M.; Tansley, D. Stewart W.; Hey, Anthony J. G. (August 2011). "The Fourth Paradigm: Data-Intensive Scientific Discovery [Point of View]". Proceedings of the IEEE. 99 (8): 1334–1337. doi:10.1109/JPROC.2011.2155130

C. Draxl, M. Scheffler, NOMAD: The FAIR concept for big data-driven materials science, MRS Bulletin, 43, 676-682, 2018. doi:10.48550/arXiv.1805.05039

M. D. Wilkinson, et al., The FAIR Guiding Principles for scientific data management and stewardship, Sci. Data 3, 2016. doi:10.1038/sdata.2016.18