Speaker
Description
Establishment of stable and reproducible mass-production method for high performance REBCO coated conductors (CCs) is now crucial issue to realize practical applications of REBCO CCs in industrial scale for such application as compact fusion reactor. Critical current, Ic, of the wire is one of the most important material parameters in the CCs. The steepness of the current-voltage (I-V) transition, which can be characterized by the power index of the I-V curve (so called n-index), is also relevant from both practical and fundamental point of view, however, there is very limited study on the influence of process conditions on the n-index. In this study, we demonstrate a new data-driven approach to accelerate the development of the manufacturing method of CC wires. We have had succeeded in measuring Ic and n-indices as a function of position with high-throughput in reel-to-reel manner in a long REBCO coated conductors. This measurement method is further coupled with machine learning based regression analysis by using combinatorial long length samples of REBCO CCs fabricated by pulsed laser deposition (PLD) to clarify the influence of PLD conditions on both Ic and n-index of the REBCO wire. We have succeeded in predicting Ic and n-index by the ML model from the deposition conditions. This data-driven approach allows us to understand the influence of the control parameters in the PLD process in detail and to optimize the deposition conditions efficiently in industrial scale production systems.
Acknowledgements: This work was supported by Moonshot R&D - MILLENNIA Program Grant Number JPMJMS24A2 and JSPS KAKENHI Grant Number JP24H00320.
