An advanced adaptive control algorithm is used to study the control of external kink modes that are excited in the HBT-EP tokamak and have a natural toroidal rotation frequency near 8 kHz. The controller's system model is parametrized and the parameters are adjusted in real-time to match the measured plasma evolution. Depending upon a programmed phase angle, active feedback control is shown to either suppress or amplify the controlled amplitude by an amount comparable to the variations observed over a variety of plasma discharges. With negative feedback, the maximum amplitude is reduced to 40% of the uncontrolled value. With positive feedback, the amplitude increases to 155%. Intermediate feedback phases lead to an acceleration or deceleration of the mode rotation frequency in addition to changes in mode amplitude. As the feedback gain increases, the level of both mode suppression and mode amplification also increases. However, for sufficiently high gain, kink mode suppression becomes limited by the excitation of an additional, slowly rotating 1.4 kHz mode having the same helical structure as the uncontrolled kink. High gain amplification is limited by disruptive loss of plasma current. Experimental results are compared with numerical simulations based on a single helicity model, and qualitative agreement is found.

Feedback control has become a crucial tool in the research on magnetic confinement of plasmas for achieving controlled nuclear fusion. We present the first experimental results from a novel feedback control system that, for the first time, employs a graphics processing unit (GPU) for microsecond-latency, real-time control computations. The system was tested on the HBT-EP tokamak using an adaptive control algorithm for control of rotating magnetic perturbations. The algorithm assumes that perturbations of known shape are rotating rigidly, but dynamically derives and updates the rotation frequency to improve phase and gain accuracy of the control signals. Experiments were set up to control four rotating n = 1 perturbations at different poloidal angles. The perturbations are treated as coupled in frequency but independent of amplitude and phase, so that the system effectively controls a helical n = 1 perturbation with unknown poloidal spectrum. The control system suppresses the amplitude of the dominant 8 kHz mode by up to 60%. Deviation from the optimal feedback phase combines suppression with a speed up or slow down of the mode rotation frequency. The feedback performance is found to exceed previous results obtained with an FPGA- and Kalman-filter based control system without requiring any tuning of system model parameters.

We report on the design of a new plasma control system for the HBT-EP tokamak that utilizes a graphical processing unit (GPU) to magnetically control the 3D perturbed equilibrium state [1] of the plasma. The control system achieves cycle times of 5μs and I/O latencies below 10μs for up to 96 inputs and 64 outputs. The number of state variables is in the same order. To handle the resulting computational complexity under the given time constraints, the control algorithms are designed for massively parallel processing. The necessary hardware resources are provided by an NVIDIA Tesla M2050 GPU, offering a total of 448 computing cores running at 1.3GHz each. A new control architecture allows control input from magnetic diagnostics to be pushed directly into GPU memory by a D-TACQ ACQ196 digitizer, and control output to be pulled directly from GPU memory by two D-TACQ AO32 analog output modules. By using peer-to-peer PCI express connections, this technique completely eliminates the use of host RAM and central processing unit (CPU) from the control cycle, permitting single-digit microsecond latencies on a standard Linux host system without any real-time extensions.

We report high-resolution detection of the time-evolving, three-dimensional (3D) plasma magnetic structure of wall-stabilized tokamak discharges in the High Beta Tokamak – Extended Pulse (HBT-EP) device. Measurements and control experiments are conducted using a newly-installed adjustable conducting wall made up of 20 independent, movable, wall segments that have been precision located and equipped with 120 modular control coils and 216 poloidal and radial magnetic sensors. The control coils are energized with high-power solid-state amplifiers, and massively-parallel, high-throughput feedback control experiments can be performed using low-latency connections between input and output CompactPCI modules and a graphics processing unit (GPU). The time evolution of unstable and saturated wall-stabilized external kink modes are studied in detail with and without applying magnetic perturbations with the control coils. The 3D dynamic structure of the magnetic field surrounding the plasma is defined with biorthogonal decomposition using the full set of magnetic sensors without the need to fit either a Fourier or a model-based basis. Naturally occurring external kinks are composed of independent helical modes, frequently having m/n = 3/1 and 6/2 helicity, that are seen to modulate each other in time. When magnetic perturbations are applied by energizing the control coils, the resonant magnetic response can be either linear, saturated, or disruptive depending upon the amplitude of the applied perturbation and on plasma’s current profile and rotation rate. Active feedback experiments have been conducted using 40 magnetic sensors and 40 control coils, and initial results show mode amplification and suppression as a function of feedback phase angle.

Multimode Structure of Resistive Wall Modes near the Ideal Wall Stability Limit Jeffrey Peter Levesque This thesis presents the first systematic study of multimode external kink structure and dynamics in a tokamak using a high-resolution magnetic sensor set. Multimode effects are directly measured, rather than inferred from anomalies in single-mode behavior. In order to accomplish this, an extensive set of 216 poloidal and radial magnetic field sensors has been installed in the High Beta Tokamak – Extended Pulse (HBT-EP) device for high-resolution measurements of three-dimensional mode activity. An analysis technique known as biorthogonal decomposition (BD) is described, and simulations are presented to justify its use for studying kink mode dynamics in HBT-EP data. Coherent activity of multiple simultaneous modes is observed using the BD without needing to define a mode structure basis beforehand. Poloidal mode numbers up to m = 8 are observed via sensor arrays with full 360◦ coverage. Higher poloidal mode numbers are suggested by the data, but cannot be well-resolved with the available diagnostics. Toroidal mode numbers up to n = 4 are observed. Non-rigid, multimode activity is observed for coexisting external kinks having m/n = 3/1 and 6/2 structures. Despite sharing the same helicity and same resonant surface, rotation of 6/2 modes is independent of 3/1 mode rotation – the n = 2 mode does not simply rotate with double the frequency of the n = 1 mode. During periods of 3/1-dominated activity, the 6/2 mode is observed to modulate the 3/1 amplitude, and in brief instances can overpower the 3/1. Statistical analysis over many shots reveals the multimode nature of the 3/1 kink to be more significant when the resonant q = 3 surface begins internal, then is ejected from the plasma. This inference is based on the relative amplitudes of secondary modes during 3/1-dominated activity, as well as spectral content of the modes. Conformal conducting wall segments were also retracted away from the plasma surface using low-order poloidal and toroidal asymmetries to excite measurable differences in low mand n-number modes. Kink mode amplitudes increase as the wall segments are withdrawn, and non-symmetric wall configurations modulate the amplitude and frequency of the rotating modes depending upon their toroidal orientation with respect to the non-symmetric wall. Modulations of mode amplitude and rotation are larger for the toroidal wall asymmetry than for the poloidal wall asymmetry.