Qualification of NSTX-U Inner TF Bundle Using Hi-Fidelity Models. (17th November 2021)

• Record Type: Journal Article
• Authors: Zhai, Yuhu; Willard, Thomas; Neumeyer, Charles; Gerhardt, Stefan; Titus, Peter; Pauley, Mark; Raftopoulos, Steve; Ellis, Robert; Hawryluk, Richard
• Is Part Of: Fusion science and technology. Volume 77:Number 7/8(2021)
• Page Start: 658
• Page End: 675
• Publication Date: 2021-11-17
• Subjects: Spherical tokamaks -- fusion magnets -- hi-fidelity models; Fusion reactors -- Periodicals; Nuclear fusion -- Periodicals; Fusion reactors; Nuclear fusion; Periodicals; 621.48405
• DOI: 10.1080/15361055.2021.1906150
• Languages: English
• ISSNs: 1536-1055

Abstract

Abstract: The National Spherical Torus eXperiment Upgrade (NSTX-U) Recovery Project provides a national facility for the study of plasma confinement, heating, and current drive in a low-aspect-ratio, spherical torus configuration. For the NSTX-U Recovery Project, key system capabilities must be achieved in order to meet operational goals. The inner toroidal field (TF) bundle is part of the center stack with 36 pie-shaped wedges (TF inner legs) bonded into a cylindrical bundle. The inner bundle is connected to outer legs via flex straps, and an ohmic heating (OH) coil is wound around the inner TF bundle, with a layer of aquapour in between acting as a winding surface between TF-OH coils. The TF bundle for the Recovery Project underwent two expert reviews to gain an improved understanding of issues identified from the extent of condition and the magnet design verification and validation review; this was to ensure confidence in the ability of the TF-OH center stack assembly to meet physics requirements for operations. A hi-fidelity, finite element model was developed with appropriate detail for multiphysics analysis of the bundle and to provide a baseline model to use as a standard against which the allowable stresses were compared. The model was also used to understand mechanical behavior and to assess implications of hypothetical delamination to the machine operation. Normal tensile stress at the bundle stub was the result of a nonuniform resistive current distribution where a relatively cold region was formed when the TF coil was pulsed. The worst case occurred at the end of the pulse when the maximum temperature gradient was established at the bundle ends. The hot spot was in compression, and the cold region was in tension. The compressive stress, which had a maximum at the bundle stub, enhanced the insulation shear capability. The in-plane torsional shear stress peaked at the end corner of the inner leg outer diameter at end of flat top. The insulation tensile and shear stresses were extracted from the hi-fidelity model using orthotropic material properties derived from first principles, and these theoretical data were verified via material tests. Load path probes were used to quantify the redistribution of torsional loads between the bundle and the external load path for a hypothetical extreme delamination condition. The extent of possible delamination was limited to regions within the radius of the flex straps. There is no significant effect on the bundle inner leg copper stress. The differential movement between the flex straps in the hypothetical extreme delamination condition is far less than the toroidal gaps between them, so there is no risk of closing these electrical gaps between the bundle turns. There is thus no impact on machine operation or even hypothetical delamination to an extreme delamination beyond the simulation-predicted delamination boundary.