5 October 2020 in Plasma Physics and Controlled Fusion
We demonstrate 2D reconstructions of the fast-ion velocity distribution from two-view fast-ion D-alpha (FIDA) measurements at the EAST tokamak. By expressing the distribution in a basis relying on the fast-ion slowing-down process in fusion plasmas, the smoothness and velocity-space resolution of reconstructions are improved. We reconstruct distributions of fast ions born from simultaneous co- and counter-current neutral beam injection and detect the expected distinct change in fast-ion birth pitch when comparing discharges utilizing different neutral beam injectors. For purely co-current injection, we find a good agreement between TRANSP-predicted and reconstructed fast-ion densities, pressures and current densities for energies above 20 keV. We furthermore illustrate the improvement of the reconstructed high-energy range (> 40 keV) of the distribution by combining FIDA with neutron emission spectroscopy measurements with the compact single-plate EJ301 scintillator.
1 April 2021 in Review of Scientific Instruments
Fast ions in fusion plasmas often leave characteristic signatures in the plasma neutron emission. Measurements of this emission are subject to the phase-space sensitivity of the diagnostic, which can be mapped using weight functions. In this paper, we present orbit weight functions for the TOFOR and NE213 neutron diagnostics at the Joint European Torus, mapping their phase-space sensitivity in 3D orbit space. Both diagnostics are highly sensitive to fast ions that spend a relatively large fraction of their orbit transit times inside the viewing cone of the diagnostic. For most neutron energies, TOFOR is found to be relatively sensitive to potato orbits and heavily localized counter-passing orbits, as well as trapped orbits whose “banana tips” are inside the viewing cone of TOFOR. For the NE213-scintillator, the sensitivity is found to be relatively high for stagnation orbits.
29 September 2022 in Nuclear Fusion
Fast ions in the MeV-range can be diagnosed by neutron emission spectroscopy (NES) and gamma-ray spectroscopy (GRS). In this work, we present orbit weight functions for one-step fusion reactions, using NES and GRS diagnostics on perpendicular and oblique lines-of-sight (LOS) at Joint European Torus (JET) as examples. The orbit weight functions allow us to express the sensitivities of the diagnostics in terms of fast-ion (FI) orbits and can be used to swiftly reproduce synthetic signals that have been computed by established codes. For diagnostically relevant neutron energies for the D(D, n)3He reaction, the orbit sensitivities of the NES diagnostics follow a predictable pattern. As the neutron energy of interest increases, the pattern shifts upwards in FI energy. For the GRS diagnostic and the T(p,γ)4He reaction, the orbit sensitivity is shown to be qualitatively different for red-shifted, blue-shifted and nominal gamma birth energies. Finally, we demonstrate how orbit weight functions can be used to decompose diagnostic signals into the contributions from different orbit types. For a TRANSP simulation of the JET discharge (a three-ion ICRF scenario) considered in this work, the NES signals for both the perpendicular and oblique LOS are shown to originate mostly from co-passing orbits. In addition, a significant fraction of the NES signal for the oblique LOS is shown to originate from stagnation orbits.
30 May 2023 in Nuclear Fusion
We compute reconstructions of 4D and 5D fast-ion phase-space distribution functions in fusion plasmas from synthetic projections of these functions. The fast-ion phase-space distribution functions originating from neutral beam injection (NBI) at TCV and Wendelstein 7-X (W7-X) at full, half, and one-third injection energies can be distinguished and particle densities of each component inferred based on 20 synthetic spectra of projected velocities at TCV and 680 at W7-X. Further, we demonstrate that an expansion into a basis of slowing-down distribution functions is equivalent to regularization using slowing-down physics as prior information. Using this technique in a Tikhonov formulation, we infer the particle density fractions for each NBI energy for each NBI beam from synthetic measurements, resulting in six unknowns at TCV and 24 unknowns at W7-X. Additionally, we show that installing 40 LOS in each of 17 ports at W7-X, providing full beam coverage and almost full angle coverage, produces the highest quality reconstructions.
1 May 2021 in Review of Scientific Instruments
The relationship between simulated ion cyclotron emission (ICE) signals s and the corresponding 1D velocity distribution function of the fast ions triggering the ICE is modeled using a two-layer deep neural network. The network architecture (number of layers and number of computational nodes in each layer) and hyperparameters (learning rate and number of learning iterations) are fine-tuned using a bottom-up approach based on cross-validation. Thus, the optimal mapping of the neural network in terms of the number of nodes, the number of layers, and the values of the hyperparameters, where θ is the learned model parameters, is determined by comparing many different configurations of the network on the same training and test set and choosing the best one based on its average test error. The training and test sets are generated by computing random ICE velocity distribution functions f and their corresponding ICE signals s by modeling the relationship as the linear matrix equation Wf = s. The simulated ICE signals are modeled as edge ICE signals at LHD. The network predictions for f based on ICE signals s are on many simulated ICE signal examples closer to the true velocity distribution function than that obtained by 0th-order Tikhonov regularization, although there might be qualitative differences in which features one technique is better at predicting than the other. Additionally, the network computations are much faster. Adapted versions of the network can be applied to future experimental ICE data to infer fast-ion velocity distribution functions.
15 February 2024 in Plasma Physics and Controlled Fusion
We present a new version of the FILDSIM code (Galdon-Quíroga et al 2018 Plasma Phys. Control. Fusion 60 105005), which significantly refines the modelling of the fast-ion loss detector (FILD) signal. We demonstrate that the FILD weight functions computed using this new version of FILDSIM are more accurate relative to synthetic benchmarks than those computed using the previous version. Thus, the new version enables higher-quality velocity-space sensitivity modelling and reconstructions. We validate the improvements on experimental data from discharge #75620 at TCV. Additionally, we present a novel approach for characterizing FILDs through a gross FILD measurement and a gross weight function based on the calculations from the new version of FILDSIM. We use them to characterize the TCV FILD.