On December 4, 2025, Marco Acosta, a student in the Group of Applied Superconductivity, successfully defended his master’s thesis. Supervised by Prof. Carmine Senatore and co-supervised by Francesco Lonardo, his thesis was titled “Deep Learning-Based Grain Segmentation and Characterization of Nb3Sn Swiss Roll Superconducting Wires”. Dr. Simon Hopkins from CERN served as a member of the jury that assessed the work.
Marco’s research combined multiple advanced approaches: the manufacturing of Nb3Sn prototype wires, including an innovative method for implementing internal oxidation in jelly-rolled Nb3Sn filaments, as well as low temperature/high field measurements and microstructure characterization. A key component of his work was the development of a deep learning-based tool to determine the grain size distribution, an important factor as the critical current density in Nb3Sn is inversely proportional to the grain size. This tool will be used for ongoing analysis of Nb3Sn wires developed within the group and will eventually be shared with the broader research community as a standardized method for grain size determination. Congratulations to Marco for this impressive achievement!