Verification of the NIMSTELL code via application to linear magnetohydrodynamic instabilities in stellarators

A new simulation tool for macroscale dynamics in stellarators, NIMSTELL [Sovinec et al., “Semi-implicit stellarator magnetohydrodynamics (MHD) with nodal spectral elements,” arXiv:2606.28613 (2026)], is verified through benchmarking with the time-dependent code JOREK [Nikulsin et al., Phys. Plasmas 29, 063901 (2022)] and the eigenvalue code CASTOR3D [E. Strumberger and S. Günter, Nucl. Fusion 57, 016032 (2016)]. The comparisons establish NIMSTELL's accuracy and reliability in representative configurations, where analytical results are not tractable. Results from all three codes are assessed for linear tearing and ballooning instabilities in W7-A and tearing in a quasi-axisymmetric (QA) configuration. NIMSTELL and CASTOR3D are also applied to global non-resonant interchange in a quasi-helically symmetric (QH) configuration. The computed NIMSTELL and CASTOR3D growth rates agree to the level of 1.5% and 3% for tearing modes in W7-A and QA configurations, respectively, to 5.8% for a W7-A ballooning mode, and to 1.6% for a global interchange mode in QH configurations. While NIMSTELL results also agree with JOREK for a majority of the benchmark cases, the numerical discrepancies are somewhat larger than those with respect to CASTOR3D in specific parameter regimes. The discrepancies are largely ascribed to the use of reduced-MHD modeling in the JOREK computations. In addition to benchmarking, the nonlinear capability of NIMSTELL is demonstrated by modeling the loss of pressure due to global interchange modes in the QH configuration. The successful verification presented in this work provides support for applying NIMSTELL to physically relevant stability analyses of current experiments and future designs.