25–27 Mar 2026
Orto Botanico Università di Padova / Area della Ricerca CNR Padova
Europe/Rome timezone

Three-dimensional modeling of RFX-mod2 experiment for magnetic control of exotic tokamak plasma regimes

26 Mar 2026, 11:45
15m
Auditorium (Orto botanico)

Auditorium

Orto botanico

Oral 2. Electromagnetic Diagnostics

Speaker

Andrea CORBIOLI (University of Padua/RFX Consortium)

Description

Three dimensional (3D) effects play a crucial role in the stability and control of magnetohydrodynamic (MHD) instabilities in toroidal confinement devices, particularly when the plasma operates in naturally unstable regimes such as “low-$q$” (Lq) and “ultra-low-$q$” (ULq) configurations. Plasmas in such “exotic'' regimes are intrinsically unstable to ideal MHD modes and the control of resistive wall modes (RWMs) is mandatory for the experiment success.
Non axisymmetric features of the surrounding conducting structures and of the control coils can significantly modify the MHD plasma stability and the effectiveness of active control schemes. As a consequence, accurate 3D modeling is essential to correctly predict the mode growth rate and its spatial structure, evaluate passive stabilization provided by the surrounding conducting structures, and design robust feedback control strategies.

In this contribution, the main non axisymmetric features of RFX-mod2, a major upgrade of the former RFX-mod, are presented and the solutions adopted for their modeling are described.
One of these solutions has been integrated into the CarMa code, which couples the linearized MHD solver MARS-F with the CARIDDI code, that solves the eddy current problem in a three dimensional formulation. This coupling enables a consistent description of the dynamics of the RWM for an upper single null plasma equilibrium with $q_a < 2$, leading to a state space representation of the system.
The resulting state space model, following an order reduction, has been implemented in a Matlab-Simulink application for dynamic modeling of the instability evolution and control.
A nominal full state-feedback controller based on pole placement has been sketched, generally demonstrating RWM control in RFX-mod2 with model-based techniques. As an upgrade, a model-predictive controller has been designed and applied. The obtained results indicate the feasibility of shaped low-$q$ experiments with active control of RWM in this device.

Author

Andrea CORBIOLI (University of Padua/RFX Consortium)

Co-authors

Domenico ABATE (Consorzio RFX) Fabio VILLONE (DIETI, CREATE, Università degli Studi di Napoli Federico II) Giuseppe MARCHIORI (Consorzio RFX) Leonardo PIGATTO (Consorzio RFX) Matteo BROMBIN (Consorzio RFX)

Presentation materials