Abstract
The Ion Cyclotron Range of Frequencies (ICRF) is an important auxiliary plasma heating method in the Chinese Fusion Engineering DEMO Reactor (CFEDR). A coupled ICRF power $P_c$ of 10 MW during the ramp-up phase and 20 MW during the flat-top phase is required, launched from an ICRF antenna located at an equatorial port, to obtain or sustain conventional H-mode in CFEDR. Using the CFETR-ICRF antenna model, we use RAPLICASOL code to evaluate ICRF wave coupling across various plasma conditions and toroidal phasings. The analysis of the antenna coupling resistance $R_c$, power spectrum, parallel wavenumber $k_\parallel$, surface current distribution on the straps $J_s$, and parallel electric field $E_\parallel$ under different edge electron density $n_e$ profiles and toroidal phasings for frequencies between 40 MHz and 90 MHz are presented.
Introduction
In fusion reactors like the CFEDR, heating plasma to the extreme temperatures required for fusion is a critical challenge. ICRF heating serves as a primary auxiliary heating method, acting similarly to a high-power microwave for ions. However, its effectiveness is limited by two main factors: coupling efficiency and impurity production. Energy must "tunnel" through a cut-off layer at the plasma edge; if this gap is too wide or the antenna phasing is suboptimal, power transfer is inefficient. Furthermore, high-power RF waves can generate strong parallel electric fields ($E_\parallel$) at the antenna surface, leading to sputtering of impurities (like tungsten) into the plasma, which can degrade performance. This study aims to identify antenna configurations that optimize power coupling while minimizing these parasitic fields.
Methods and Modeling
The research utilizes the RAPLICASOL code, a Finite Element Method (FEM) based tool within the COMSOL environment, to model antenna coupling. The antenna design is based on an ITER-like configuration, consisting of 4 poloidal sub-straps and 6 toroidal columns.
The plasma is represented using a "cold plasma approximation" with the Stix dielectric tensor. Simulations were conducted across a frequency range of 40 to 90 MHz, testing three toroidal phasings: $(0, \pi, 0, \pi, 0, \pi)$, $(0, \pi, \pi, 0, 0, \pi)$, and $(0, 0, \pi, \pi, 0, 0)$. Realistic boundary conditions were derived from integrated modeling of CFEDR conventional H-mode operation, including core electron temperatures of 30 keV and densities of $1.5 \times 10^{20} \text{ m}^{-3}$. To evaluate coupling performance, three distinct edge density profiles (high, mid, and low) were analyzed.
Key Insights and Results
Coupling Resistance and Phasing
The study found a significant trade-off involving the parallel wavenumber ($k_\parallel$). While lower $k_\parallel$ values generally lead to higher coupling resistance ($R_c$), extremely low values can result in poor core absorption. The $(0, 0, \pi, \pi, 0, 0)$ phasing demonstrated the highest coupling resistance but presented the most risk regarding absorption profiles.
Impurity Mitigation
A key finding was the performance of the $(0, \pi, \pi, 0, 0, \pi)$ phasing. Compared to the standard $(0, \pi, 0, \pi, 0, \pi)$ dipole phasing, it significantly reduced the parallel electric field at the antenna edges. At 60 MHz in high-density plasma, the mean parallel electric field $\langle |E_\parallel| \rangle$ was reduced from 8.764 kV/m to 3.359 kV/m, suggesting a substantial reduction in potential impurity generation.
Operating Bands
The antenna system exhibits "effective operating bands" near 60 MHz and 90 MHz, corresponding to Helium-3 and Hydrogen minority heating schemes, respectively. These peaks in coupling resistance are enhanced by the impedance transformation properties of the antenna's internal components, such as the water tank and transmission line sections.
Discussion and Future Work
While the cold plasma approximation provides a robust baseline for coupling analysis, it ignores kinetic effects at the plasma edge that could influence wave behavior. Future work will involve integrating these coupling results with core-absorption codes to ensure that high-coupling phasings also provide efficient central heating. Additionally, more detailed studies on the segmented strap design are needed to fully quantify its advantages over traditional long-strap configurations, particularly regarding current distribution uniformity.
Conclusion
This modeling effort identifies optimal antenna phasings for CFEDR that balance high power coupling with low impurity production. The $(0, \pi, \pi, 0, 0, \pi)$ phasing emerges as a promising candidate for sustaining high-performance H-mode plasmas by effectively managing the edge electric fields while maintaining sufficient coupling resistance.
Summary updated