Multiscale Approach and Role of Validation in the Thermal-Hydraulic Modeling of the ITER Superconducting Magnets

The ITER superconducting (SC) magnets require the forced flow of supercritical He (SHe) at
$\sim$
4.5 K and
$\sim$
0.5 MPa, giving thermal-hydraulics (TH) a key role in the multiphysics arena of SC magnets. Here we introduce a multiscale approach to the TH modeling of ITER magnets, based on the fact that the TH relevant space scales range from the 10 to 100 m of magnet size/cable-in-conduit conductors (CICC) length, down to the
$10^{-2}\ \hbox{m}$
of the transverse size of a CICC, while the relevant TH time scales also cover several orders of magnitude. On the “macro-scale”, the entire system (winding
$+$
structures
$+$
cryogenic circuit) is considered; this requires the treatment of the “meso-scale”, where single CICC are treated, weakly thermally coupled inside a winding as needed. The constitutive relations needed by the 1D meso-scale models, i.e., friction factors and heat transfer coefficients, may in turn be derived analyzing a limited portion of the CICC on the “micro-scale”, with detailed 2D–3D computational thermal-fluid-dynamics models. At each scale, the different issues related to code development, benchmarking/validation, and application are considered in the paper. The choice of developing a code in-house is compared to the commercial codes and/or freeware. The reciprocal benefits obtained from these codes by the ITER magnet R&D program (which led, e.g., to the realization and test of Model and Insert coils, as well as many short samples), and vice versa, are discussed. Several examples of the multiscale approach to the TH modeling of SC magnets will be presented in the paper, based on the experience developed during the last 15 years within our group, in collaboration with laboratories in the EU, Japan, Russia, South Korea, and the U.S. It is argued that the intrinsic modularity of the multiscale approach leads to significant benefits. It is also argued that the effort towards verification and validation of the existing TH models of the ITER SC magnets has been rather limited so far, sometimes notwithstanding the existence of a significant experimental database; therefore, it is recommended to launch a systematic initiative in that direction, with particular attention to the assessment of the predictive capabilities of the existing TH codes. While these capabilities are going to be more and more relevant for the ITER nuclear device, for operation and safety studies in particular, there is at this time hardly any evidence of such predictive capabilities in the published literature.