Modelling by analytical approach the coupling losses of CICCs used in tokamaks remains a challenge to be reliable. This is usually done using either CPU consuming numerical approaches or heuristic models such as MPAS now used for ITER. Experimental measurements of AC losses are performed at CEA using magnetization method on several JT-60SA TF type samples with various void fractions (25%-36%). Influence of void fraction on coupling losses is hard to heuristically model yet. We choose to develop an experimental protocol in order to measure coupling losses in a range of frequency relevant to fusion operation domain. AC losses model as MPAS is confronted to our JOSEFA experimental data. Conclusion and lessons will be taken into account for future work.

In spite of their complex geometry, CICCs have to be modelled with a rather simple description for coupling losses when operating in transient regime. Difficulties to predict AC losses in superconducting cable have already been shown in previous models such as the new analytical one developed at CEA named COLISEUM (after COupling Losses analytIcal Staged cablEs Unified Model) and CEA heuristic one MPAS. In this paper, we present a parametric analysis for coupling losses in superconducting strands and cables subjected to time-varying transverse magnetic field by using the recently developed COLISEUM. This analysis aims at understanding trends of the model in a broad domain of investigation and their associated limits of application. We show that for a wide range of parameters, it is possible to reduce this model of four time constants to a smaller subset. This reduction brings simplifications to the current COLISEUM and enables it to be consistent with the number of time constants considered in the MPAS model from CEA.

In the framework of activities embedding magnet design and associated R&D activities and relying on Cable in Conduit Conductors (CICC) technology, the singularity of the concept can rise some challenges versus their modelling in operation. Indeed, CICC includes thousands of superconducting strands, twisted together under a multi-staged scheme that also includes deformation by cabling and compaction during manufacture. This causes the accurate predictability of strands position in CICC extremely difficult, while it can be a key element for modelling their performances in operation. As a matter of fact the coupling losses rely on the CICC capability to establish interstrand shielding currents, driven by the inter-strand contacts mapping which are only accessible in prediction via accurate 3D strand trajectories geometry. Same applies for prediction of CICC mechanical properties (deformation of Nb3Sn strands) and hydraulic properties (helium coolant force-flowed between the strands), making those investigations of high added-value. In this context, INFLPR installed a new set-up dedicated to micro-tomography that is able to examine CICCs with high resolution, allowing to get an overall overall 3D overview regarding picture of the strands location. CEA and INFLPR further developed a post-processing method to reconstruct the strands trajectories with high accuracy. The measurement and data analysis workflow was applied to two middle-size CICCs with variable void fraction from which statistics of contacts were issued. The obtained database was exploited to reconstruct equivalent 3D resistive network, in view of interpreting coupling losses tests with help of analytic CEA model (COLISEUM) based on multi-stage representation.The above applications using CICC topology database will be discussed and tentatively compared to experimental AC losses database conducted at CEA. The outcomes will be discussed and the subsequent guidelines for future work presented.

Predicting analytically the coupling losses generated in a cable for fusion magnets is still a significant challenge. Difficulties are related to the complex geometry of the system: several multi-strand stages embedded in one another with different twist pitches length, difficulty to model multiplets of strands, including compaction to the final shape. A two-stage analytical geometry based model (COLISEUM) has previously been developed at CEA. We try to extend it to any n-stage cables We detail here an iterative enhancement method to an n-stage model . We validated it against experimental data and shown that it is robust enough to fit our measured coupling losses. Finally, this upgraded model can be used to assess coupling losses in fusion nstage cables in a particularly precise way from only geometrical information and analytical tools .