The ITER magnet coils are wound from Cable-In-Conduit Conductors (CICC) made up of superconducting and copper strands assembled into a multistage, rope-type cable inserted into a conduit of butt-welded austenitic steel tubes. The conductors for the Toroidal Field (TF) and Central Solenoid (CS) coils require about 500 tons of Nb3Sn strands while the Poloidal Field (PF) and Correction Coil (CC) conductors need around 250 tons of Nb-Ti strands. The required amount of Nb3Sn strands far exceeds pre-existing industrial capacity and calls for a significant worldwide production scale up. After explaining the in-kind procurement sharing of the various conductor types among the six ITER Domestic Agencies (DA) involved: China, Europe, Japan, South Korea, Russia, and the United States, we detail the technical requirements defined by the ITER International Fusion Energy Organization (IO), and we present a brief status of ongoing productions. The most advanced production is that for the TF conductors, where all six DAs have qualified suppliers and about 50% of the required strands have been produced and registered into the web-based conductor database developed by the IO.

Status of ITER Conductor Development and Production

We have developed a model that describes the transverse load degradation in Nb3Sn CICCs, based on strand and cable properties, and that is capable of predicting how such degradation can be prevented. The Nb3Sn cable in conduit conductors (CICCs) for the International Thermonuclear Experimental Reactor (ITER) show a significant degradation in their performance with increasing electromagnetic load. Not only do the differences in the thermal contraction of the composite materials affect the critical current and temperature margin, but mostly electromagnetic forces cause significant transverse strand contact and bending strain in the Nb3Sn layers. Here, we present the model for transverse electro-magnetic load optimization (TEMLOP) and report the first results of computations for the ITER type of conductors, based on the measured properties of the internal tin strand used for the toroidal field model coil (TFMC). As input, the model uses data describing the behaviour of single strands under periodic bending and contact loads, measured with the TARSIS set-up, enabling a discrimination in performance reduction per specific load and strand type. The most important conclusion of the model computations is that the problem of the severe degradation of large CICCs can be drastically and straightforwardly improved by increasing the pitch length of subsequent cabling stages. It is the first time that an increase of the pitches has been proposed and no experimental data are available yet to confirm this beneficial outcome of the TEMLOP model. Larger pitch lengths will result in a more homogeneous distribution of the stresses and strains in the cable by significantly moderating the local peak stresses associated with the intermediate-length twist pitches. The twist pitch scheme of the present conductor layout turns out to be unfortunately close to a worst-case scenario. The model also makes clear that strand bending is the dominant mechanism causing degradation. The transverse load on strand crossings and line contacts, abbreviated as contact load, can locally reach 90 MPa but this occurs in the low field area of the conductor and does not play a significant role in the observed critical current degradation. The model gives an accurate description for the mechanical response of the strands to a transverse load, from layer to layer in the cable, in agreement with mechanical experiments performed on cables. It is possible to improve the ITER conductor design or the operation margin, mainly by a change in the cabling scheme. We also find that a lower cable void fraction and larger strand stiffness add to a further improvement of the conductor performance.

https://iopscience.iop.org/article/10.1088/0953-2048/19/9/011/pdf

Transverse load optimization in Nb3Sn CICC design; influence of cabling, void fraction and strand stiffness

Comprehensive measurements are reported of the critical current density (JC) of internal-tin and bronze-route Nb3Sn superconducting wires as a function of magnetic field (B?23?T), temperature (4.2?K ?T?12?K) and axial strain (?1.6%??I?0.40%). Electric field?temperature characteristics are shown to be equivalent to the standard electric field?current density characteristics to within an experimental uncertainty of ~20?mK, implying that JC can be described using thermodynamic variables. We report a new universal relation between normalized effective upper critical field (BC2*(0)) and strain that is valid over a large strain range for Nb3Sn wires characterized by high upper critical fields. A power-law relation between BC2*(0,?I) and TC*(?I) (the effective critical temperature) is observed with an exponent of ~2.2 for high-upper-critical-field Nb3Sn compared to the value ?3 for binary Nb3Sn. These data are consistent with microscopic theoretical predictions and suggest that uniaxial strain predominantly affects the phononic rather than the electronic properties of the material. The standard Summers scaling law predicts a weaker strain dependence than is observed. We propose a scaling law for JC(B,T,?I) based on microscopic theory and phenomenological scaling that is sufficiently general to describe materials with different impurity scattering rates and electron?phonon coupling strengths. It parametrizes complete datasets with a typical accuracy of ~4%, and provides reasonable predictions for the JC(B,T,?I) surface from partial datasets.

The scaling law for the strain dependence of the critical current density in Nb3Sn superconducting wires

The ITER toroidal field model coil project

Mechanical and magnetic load effects in Nb3Sn cable-in-conduit conductors

Testing of the five International Thermonuclear Experimental Reactor (ITER) model coils is nearly complete and it is possible to produce an overall summary of the conductor performance assessment. The interpretation is supported by a large database of tests on components (especially strand) and by analysis of the two main phenomena that influence the strand performance in the cable (current non-uniformity and mechanical loads). Individually, interpretation of individual coils is difficult but taken together the results enable conclusions to be drawn about the expected behaviour of conductors in the ITER coils. The conductor design has successfully resolved many of the problems of stability and pulsed current behaviour associated with large multi-strand cables, and the performance is better than expected in these areas. The current sharing behaviour is below expectations, probably due to an extra unexpected strain effect in the Nb3Sn strands arising from the local magnetic loads, and the paper concentrates on this issue. Design modifications are required to recover the performance and updated design criteria to implement these are proposed.

N Mitchell

Papers

Summary, assessment and implications of the ITER Model Coil test results