Search

Active feedback stabilization of resistive wall modes in tokamaks is studied both analytically, using large aspect ratio theory, and by means of toroidal computations. Extensive studies show that robust stabilization, with respect to variations in plasma current, pressure and flow velocity, can be achieved with a simple control system using poloidal sensors inside the first wall. The required coil voltages are modest, even for the two-wall structure of a tokamak reactor. © 2002 American Institute of Physics. [ABSTRACT FROM AUTHOR]

Active feedback of nonaxisymmetric resistive wall modes in tokamaks is investigated using control theory. Control systems are designed to stabilize the resistive wall mode of toroidal mode number n=1 and meet certain performance specifications for a set of test equilibria. The response of the plasma and resistive wall is described by low order rational functions from an electromagnetic model [Y. Q. Liu et al., Phys. Plasmas 7, 3681 (2000)]. Simple coil arrangements are assumed, both for the sensor and feedback coil arrays, and the sensors detect the perturbed poloidal field. The active coils are modeled both as broad strips and as thin wires, and several different controllers: P (proportional), PD (proportional plus derivative) and H[sub ∞] are investigated. An important parameter is the ratio, τ, of control system response time to the resistive wall time, and the analysis shows the restrictions on this ratio for acceptable performance. For an equilibrium that exceeds the no-wall beta limit by 63%, good control with broad strips and a PD controller is possible for τ≲5.4, while thin-wire coils require τ≲2.1. H[sub ∞] controllers give some improvement for thin wires and about a factor of 2 increase in τ for broad strips. The upper limit in τ decreases with increasing pressure. A control system designed to stabilize a certain pressure generally works well at lower pressures. © 2000 American Institute of Physics. [ABSTRACT FROM AUTHOR]

Active feedback stabilization of pressure-driven modes in tokamaks is studied computationally in toroidal geometry. The stability problem is formulated in terms of open-loop transfer functions for fluxes in sensor coils resulting from currents in feedback coils. The transfer functions are computed by an extended version of the MARS stability code [A. Bondeson et al., Phys. Fluids B 4, 1889 (1992)] and can be accurately modeled by low order rational functions. In the present paper stability is analyzed for a system with an ideal amplifier (current control). It is shown that feedback with modest gain, and a single coil array poloidally, gives substantial stabilization for a range of coil shapes. Optimum design uses sensors for the poloidal field, located inside the resistive wall, in combination with rather wide feedback coils outside the wall. Typically, the feedback does not strongly modify the plasma-generated magnetic field perturbation. A future companion paper [C. M. Fransson et al., Phys. Plasmas (accepted for publication)] will apply control theory to study the limitations arising for finite time-constant of the amplifier-feedback coil circuit. © 2000 American Institute of Physics. [ABSTRACT FROM AUTHOR]