1. Study on Electron Cyclotron Heating (ECH)
Pre-ionization in Versatile Experiment Spherical Torus (VEST)
JongGab Jo, H. Y. Lee, Y. H. An, K. J. Chung and Y. S. Hwang
Department of Nuclear Engineering, Seoul National University, Seoul , Korea

2. Motivation & Objectives Experimental Setup
Contents
Introduction
Motivation & Objectives
Experimental Setup
ECH system and diagnostics in VEST
Experimental condition
Experimental Result
The effect of ECH power on pre-ionization
The effect of TF strength on pre-ionization
Comparison between O-mode and X-mode injection
Discussion
Budden analysis
Higher ECH power density operation
Summary & Conclusion
Reference

3. Introduction Motivation & Objectives
ECH pre-ionization is widely used method in tokamak to reduce the required loop voltage for breakdown and save the magnetic flux consumption.
Especially pre-ionization is essential for Spherical Torus (ST) which has difficulty in start-up due to lack of space for the center stack.
In ST devices which have characteristics of high beta operation and consequent low cutoff density, alternative method using the Electron Bernstein Wave (EBW) has been studied.
Preliminary experiment for VEST pre-ionization has been conducted in linear device.

4. Introduction Motivation & Objectives
H.Y. Lee(WP58)
Bt center
2.45GHz microwave
The results of preliminary experiment show that LFS X-mode injection produces the largest electron density.
Production of overdense plasma by XB mode conversion.
ECH launching system of VEST has been designed in a low field side injection configuration by accounting the preliminary experimental results in linear device.

5. Introduction Motivation & Objectives
In most previous works, radial profile of the electron density and temperature in pre-ionization phase has not been studied in detail. However, propagation, absorption and reflection of EC wave are significantly associated with profile of the plasma parameters.
Investigate the heating mechanism of ECH / EBW with direct measurement of the electron density and temperature profile when only the toroidal magnetic field is applied and no loop voltage are provided to find effective pre-ionization condition.

6. Experimental Setup ECH System and diagnostics in VEST
94GHz Interferometry
2.45GHz, 6kW, CW
2.45GHz, 3kW, pulse
Triple Probe
2.45GHz, 6kW microwave generator and 3kW magnetron is installed in main chamber of VEST.
Low field side X-mode injection configuration.
WR284 / WR340 rectangular waveguide for TE10 mode propagation.
Directional coupler and rf power meter for microwave power monitoring.
A triple probe is fabricated and installed to diagnose the time varying plasma density and temperature during discharges and compared with interferometer to confirm the validity of data.

7. Experimental Setup Experimental Condition
Operation Parameters
Value
Gas Species
Hydrogen
Base pressure
1.5e-6 Torr
Operating Pressure
2e-5 Torr
TF Current
3.8 / 5.4 / 6.7 / 8.2 kA
ECH Power
2 / 3 / 4 / 6 kW
Driving Frequency
2.45GHz
Wave Polarization
X-mode

8. Experimental Result The effect of ECH power on pre-ionization
UHR
Power absorption in UHR(ne) and ECR(Te).
Initial breakdown occurs in ECR, and then UHR move outward with electron density build-up.
Doppler shift and relativistic effect in wave-particle resonance condition.

9. Experimental Result The effect of TF strength on pre-ionization (ne)

10. Experimental Result The effect of TF strength on pre-ionization (Te)
2nd
1st
2nd
1st
1st
2nd

11. Experimental Result Second harmonic heating
Te [eV]
TF Current: 3.8kA
1st ECR
2nd ECR
Electron temperature peak is located in the 1st ECR at the beginning of breakdown, and then another peak near the 2nd ECR layer appears at the ECH power ramp-up phase. (threshold of ECH power)
Second harmonic heating is observed when both 1st and 2nd ECR layer exist in chamber but X2 mode breakdown without 1st ECR layer is fail.
Pre-heated plasma will be needed for second harmonic heating (FLR effect)

12. Experimental Result Comparison between O-mode and X-mode injection
X wave ~ X wave
O wave X wave
RF power meter with directional coupler to collect the chosen wave polarization
X-mode injection is slightly better than O-mode.
Power meter data shows that many of injected O-wave is converted into X-mode in the chamber unlike X-mode injection.
X-mode has a high rate of single pass absorption while O-mode experiences multiple reflection and then converted X-mode is absorbed in the fundamental ECR and UHR layer.

13. Discussion 1D Inhomogeneous cold plasma model
UHR, R-cutoff doublet
(Budden model) [1, 2]
L-cutoff, UHR, R-cutoff triplet [3, 4]
R-cutoff
UHR
Low Field Side X-wave Injection
L-cutoff a a
R-cutoff
UHR
Low Field Side X-wave Injection
Budden Parameter
This problem has no analytic solution
distance between R-cutoff and UHR layer.
Transmission(T), Reflection(R) and
Conversion(C) Coefficients
Maximum mode conversion coefficient can even be 1.
L-cutoff plays a role of increasing conversion efficiency.
XB mode conversion

14. Discussion 1D Inhomogeneous cold plasma model
If the change of plasma parameters and magnetic field is large over a wavelength, inhomogeneous plasma model must be applied for describing the behavior of the wave in resonance and cutoff layer.
If thickness of the evanescent region is sufficiently smaller than wavelength, the incident wave can propagate beyond the R-cutoff and reach to the UHR layer. (XB mode conversion)
Maximum electron density of VEST pre-ionization plasma is lower than L-cutoff density.
Budden model can be applied for describing the X-wave in the UHR, R-cutoff pair region.

15. Discussion Time evolution of electron density and coefficients
(a): breakdown & ECH power ramp-up
(b): ECH power flat top ~3kW
(c): ECH power ramp-up
(d): ECH power flat top ~6kW
Estimation of tunneling and conversion effect by Budden analysis.
Time evolution of reflection coefficient calculated from experimental data show good agreement with measured reflected microwave power.

16. Discussion Electron density scale length and magnetic field effect
Distance between the UHR and R-cutoff can be expressed by density scale length and magnetic field within the limit of
(k, Ln: evaluated at the R-cutoff)
density scale length is the dominant factor and magnetic field strength is a second factor in determining the Budden parameter.
TF Current T R C
8.2kA
0.2754
0.5251
0.1995
6.7kA, 5.4kA
0.05
0.9
3.8kA
0.123
0.7691
0.1079
When the TF current is 8.2kA, the tunneling and mode conversion can occur due to the very steep density gradient in edge region supported by the chamber wall and high density plasma is produced.
As TF current decrease from 8.2kA to 5.4kA , the resonance layer and the electron density peak move away from the outer chamber wall increasing the density scale length. The transmission and conversion coefficients are decreased maintaining low plasma density in intermediate TF strength.
When the TF current is 3.8kA, the transmission and conversion coefficient are increased again because of low TF strength.

17. Discussion High density plasma production in low TF strength
High density plasma comparable to that of the highest TF strength is produced near the center stack unlike the result of Budden analysis.
1. Reflected wave from inner wall of the chamber makes situation similar to triplet case increasing mode conversion efficiency. The inner wall plays a role of high density cutoff layer like L-cutoff that reflects incident X-wave back into the resonance layer.
2. Power density absorbed in resonance layer is increased with decreased radius of resonance layer. As TF current decrease, volume of the energy absorption zone is decreased increasing the power density in this region.

18. Discussion Higher ECH power density operation
VEST
Linear Device
Volume [m-3]
3.68
7.07e-3
Heating Power [kW] 6
0.9
Heating Power Density [kW/m-3]
1.63
127
Maximum Electron Density [1017 m-3] ~1 ~7
Two magnetrons (~6kW) will be installed.
Line density measured in VEST is not saturated until current maximum ECH power (~6kW) is injected.
Heating power density in linear device is almost 78 times than VEST.
Higher electron density and mode conversion efficiency are expected with increasing ECH power.

19. Summary & Conclusion
X-wave injected from low field side is absorbed in UHR (ne) and fundamental ECR (Te) layer.
The second harmonic heating effect is observed but only when the pre-heated plasma exists and sufficient ECH power is transferred to the plasma.
O-wave injected from low field side is converted into X-mode in the chamber and then absorbed in UHR and fundamental ECR layer.
Tunneling and mode conversion phenomena of X-wave in VEST are well described and XB mode conversion efficiency is calculated by Budden analysis.
High density plasma is produced when the peak of density profile is near the inner wall or outer wall with the aid of high X-B mode conversion efficiency.
More ECH power is needed to produce overdense plasma in VEST by accounting the results of the linear device.
Cutoff-resonance-cutoff triplet that can be made by inner wall or overdense plasma will be an important factor for effective pre-ionization.

20. Reference
[1] Budden, The propagation of radio waves. (Cambridge University Press, Cambridge, 1985).
[2] D Gary Swanson, Plasma Waves, 2nd Edition. (Institute of Physics Publishing, Bristol and Philadelphia, 2003).
[3] A. K. Ram, A. Bers, S. D. Schultz, and V. Fuchs, ‘‘Mode conversion of fast Alfven waves at the ion-ion hybrid resonance’’, Phys. Plasmas 3, 1976 (1996).
[4] A. K. Ram and S. D. Schultz, ‘‘Excitation, propagation, and damping of electron Bernstein waves in tokamaks’’, Phys. Plasmas 7, 4084 (2000).