Feasibility study for velocity space tomography using fast-ion D-α spectroscopy in KSTAR

Abstract

Since the fast ions with energies several times greater than the main ions significantly influence the plasma heating and self-sustaining high temperature of the heated plasma, understanding the fast-ion behavior and transport is essential. Especially, the fast-ion distribution function is an important parameter since the fast-ion behavior and transport are largely affected by their phase space region. However, measuring the fast ion distribution by direct measurements is a challenging task. Instead, velocity space tomography techniques have been proposed to estimate the distribution by solving inverse problems using diagnostic data. Although KSTAR is equipped with fast ion diagnostic instruments, velocity space tomography has not been attempted. Using diagnostic data, we aim to evaluate the fast ion velocity distribution in KSTAR experimental conditions. As a first step, we explored the feasibility of applying fast-ion Dα spectroscopy (FIDA) diagnostics to velocity space tomography. Before applying real experimental data, phantom tests were performed to assess the tomographic reconstruction quality of the currently installed FIDA systems. We calculated the phase space weight function based on FIDASIM using KSTAR experimental data for the phantom test. We developed the code for tomography using Tikhonov regularization. Two kinds of phantoms were used in the test. The first phantom was the fast-ion distribution phantom obtained based on the calculation of the neoclassical transport, and the other one was uniformly distributed multiple Gaussian peaks in the phase space. Also, we introduced several evaluation metrics to estimate the reconstruction quality. Then, we conducted the phantom test based on the developed code and calculated the phase space weight function using existing FIDA systems in KSTAR. The results revealed that the currently available system in KSTAR can only access the limited pitch angles of high-energy regions. Therefore, reconstruction performance is insufficient with the existing system. Especially, we verified that the spatially-biased weight function of the system is vulnerable to Tikhonov regularization and the spatially less-biased weight function gives better reconstruction quality. Therefore, the spatial bias of the weight function in the region of interest was identified as a crucial factor for the determination of the reconstruction performance. Based on this, we optimized the position of the newly installed system to obtain a wider coverage in phase space and spatially less-biased weight function. Adding a new vertical view FIDA system improved the reconstruction performance. By adding a new FIDA system, we confirmed that KSTAR FIDA systems can reasonably reconstruct the velocity space distribution at high-energy regions containing the most pitch ranges and it is expected to be utilized in further fast-ion research in KSTAR.