THE 14th INTERNATIONAL STELLARATOR WORKSHOP

Electron Cyclotron Heating in Various Magnetic Configurations of HSX*

J.N. Talmadge, K.M. Likin, A. Abdou, A. Almagri, D.T. Anderson, F.S.B. Anderson ,
J. Canik, C. Deng, S.P. Gerhardt, K. Zhai

The HSX Plasma Laboratory

University of Wisconsin-Madison, USA

Abstract
Thomson scattering and diamagnetic loop measurements indicate that the central electron
temperature and stored energy increase linearly with power. Experimentally it is found that the central
electron temperature is roughly independent of density. These findings are consistent with a thermal
conductivity that scales inversely with the density. Typically in good confinement configurations, the
stored energy shows a peak at low density and is constant at the higher densities, in contradiction to
the model. On the other hand, in configurations that poorly confine trapped particles, the stored
energy increases linearly with density, as expected. From measurements of X-ray emission and
absorbed power, as well as calculations of the absorption efficiency from ray tracing, it is concluded
that at low densities a nonthermal electron population accounts for a significant fraction of the stored
energy for the good confinement configurations.

Introduction
Plasma heating in HSX is done at a magnetic field of 0.5 T using the extraordinary
wave at the second harmonic of the electron cyclotron frequency. In order to investigate the
improved confinement properties of the quasihelically symmetric configuration, a 28 GHz
gyrotron is used to heat electrons to the low collisionality regime. So far, up to 100 kW has
been launched from the low field side of the device in the form of a Gaussian beam with a
spot size of 4 cm. In HSX, neoclassical transport can be greatly increased with a set of
auxiliary coils that adds a toroidal mirror term to the magnetic field spectrum. In the Mirror
configuration, the field is increased at the location of the ECH antenna, while in the anti-
Mirror configuration the field is decreased. Trapped particles are well confined for the QHS
configuration, while they are lost from the outboard side of the torus at the location of the
antenna in the Mirror configuration. However, trapped particles launched near the antenna for
the anti-Mirror configuration are lost from both the inboard and outboard sides of the torus.

Neoclassical Transport
Transport in HSX is analyzed using the ASTRA code.1 The particle and heat fluxes in
ASTRA, including the off-diagonal terms, are calculated by taking moments of the
monoenergetic diffusion coefficient. These diffusion coefficients are obtained for a broad
range of test particle energies, densities and electric fields using a Monte Carlo code and then
fit to a six-parameter analytic expression originally developed by Shaing2 and later modified
by Painter and Gardner3. The expression can be written in the following manner:

π ν ω 2 = C1ν~ 2 + C2 (ω E + ω B ) 2 + C3ω B2 + C4 ω B ν~
D= ε t C6 Vd2
2 ω2
(1)
ν E , ε =r/R
ω B = −C5Vd , Vd = K , ν~ = , ωE = t
eBr C6 rB
 This simple form of the monoenergetic
diffusion coefficient fits the numerical
data over a broad range of electric fields,
ASTRA:QHS magnetic fields, collisionality, particle
energy and particle mass. One particular
advantage of using this expression is that
it is a fast way of calculating the
ambipolar radial electric field by setting
the ion and electron fluxes equal to each
other, Γi(r,Er) = Γe(r,Er), where r is a flux
ASTRA: Mirror
surface label.

In the ASTRA code, the power

Fig. 1: Central electron temperature as a
function of line average density. Also shown
are the ASTRA calculations for the QHS and
Mirror configurations (Er = 0, dashed line;
ambipolar Er, solid line).
deposition profile is determined from a
ray tracing calculation.4 Typically the
single pass absorption coefficient is
about 0.4 and rises with density. The
deposition profile tends to be very
peaked with central resonance.

Electron Temperature Scaling

Since electron transport in
stellarators is typically not solely
neoclassical, the electron thermal
ASTRA: QHS
conductivity is specified as the sum of a
neoclassical term plus an anomalous
component. Previously we used an
ASTRA: Mirror anomalous thermal conductivity given by
ASDEX L-mode scaling5 that is
proportional to Te3/2. This yields a
dependence of the temperature on the
density and power, T ~ (P/n)0.4 and a
stored energy that goes as W ~ n0.6P0.4.
Fig. 2: Central Te versus absorbed power. Recently the central channel of a 10-
channel Thomson system with a
Nd:YAG laser has been made operational. As seen in Figure 1, the central temperature is
roughly independent of density, except perhaps at the lowest densities for the QHS
configuration. This data is with a constant input power of 40 kW. Figure 2 shows the central
electron temperature for a fixed average density of 1.5 × 1018 m-3, as the gyrotron power is
varied from 20 kW to 100 kW. The absorbed power is estimated from the difference in the
slopes of the stored energy after the gyrotron turn-off. From the figure, it can be seen that for
the QHS configuration, the central Te appears to scale linearly with power. The two scaling
results indicate that the ASDEX L-mode model for the anomalous thermal conductivity does
not follow the experimental scaling.

Instead we find better agreement with the experimental data for an anomalous thermal
conductivity that scales as χe,an = 10.35/ne where the conductivity is in units of m2/s and the
density is in units of 1018 m-3. This gives a temperature dependence that is independent of
density and linear with the power. Modeling of the particle diffusion coefficient based on
 calibrated Hα measurements and
DEGAS calculations gives a similar
dependence on the plasma density.6
This sort of Alcator-like dependence
ISS95 for the thermal conductivity was found
to be a good fit to the data for the early
Heliotron-E ECH results with a 28
ASTRA: Mirror: GHz gyrotron.7 In Figures 1 and 2 are
also plotted the ASTRA calculation of
the central electron temperature for the
ASTRA: QHS QHS and Mirror configurations. For
the QHS configuration, the transport is
solely determined by the anomalous
transport since the neoclassical fluxes
are so low. For the mirror
Fig. 3: Stored Energy versus absorbed power configuration however two curves are
plotted, corresponding to the case
when the radial electric field is set to
zero, as well as when the ambipolar
electric field is determined. For these
calculations, the radial electric field is
solely given by the electron root when
the ion flux is set equal to the electron
flux. Such a high positive electric field
lowers the electron fluxes considerably
for the Mirror and makes the transport
properties of this configuration similar
to the QHS mode of operation.
However, there is sufficient reason to
believe that the prediction of an
electron root plasma may be too
optimistic.8 Thus we expect the
Fig. 4: Stored Energy versus line average experimental data for the Mirror
density configuration to fall between the two
limits of the ASTRA calculation. At
this time, there is insufficient data to distinguish which of the two predictions for the Mirror
configuration might best approximate the experimental results.

Stored Energy Scaling

Figure 3 shows that the stored energy increases linearly as a function of the absorbed
power for the QHS and Mirror configurations at a fixed density of 1.5 × 1018 m-3. This is in
agreement with the Alcator-like model for anomalous transport, rather than the ASDEX L-
mode model. The difference between the QHS and Mirror stored energy reflects the 15%
lower plasma volume for the Mirror case. From the figure, it can be seen that only at the
higher powers does the stored energy agree with ISS95 scaling, although the functional
dependence on the power is very different. We would expect from the model that the stored
energy should increase linearly with the density. Figure 4 shows, however that for a constant
input power of 40 kW, this is not so for the QHS configuration and for central resonance in
the Mirror mode. In contrast, Figure 5 shows that for Mirror outboard resonance where
 Fig. 5: Stored energy versus density for Fig. 6: Hard X-ray emission and stored
Mirror outboard resonance. energy as a function of density for
Mirror central resonance.
1 confinement of trapped electrons is poor,
the dependence of the stored energy on
0.8 Te = 0.4 keV the density does agree with the model.
Absorption

0.6
To understand why the
0.4 dependence of the stored energy on the
density does not agree with the model,
0.2 Te from exp. we measured the hard X-ray emission
0 from the plasma using a collimated
CdZnTl detector. Figure 6 shows that the
0 1 2 3 4 emission increases with density to a
3
Line Average Density, 1018 m- maximum at 0.5 × 1018 m-3 and then falls
off sharply at higher densities. Similarly
Fig. 7: Ray tracing calculation of absorption the stored energy also shows a peak at
efficiency assuming constant peak temperature
0.5 × 1018 m-3, but remains roughly
of 0.4 keV. Also shown is calculation based on
constant at higher densities. The
experimental measurements.
conclusion from this comparison of the
X-ray emission to the stored energy is that a nonthermal component in the plasma makes a
significant contribution to the stored energy at the lower densities.

Ray Tracing and Absorbed Power

This conclusion is further supported by a comparison of the heating efficiency from a
ray tracing code compared to measurements of the absorbed power.4 Figure 7 shows the
results of the ray tracing calculation. The single-pass absorption is calculated assuming a
parabolic density profile and and exponential electron temperature profile. With a central
temperature assumed to be 0.4 keV, the single-pass absorption coefficient is 0.4 at a line
average density of 1.5 × 1018 m-3. Using the Thomson scattering data for the central
temperature there is slightly more absorption at the lower densities, but both calculations
show that the absorption efficiency should increase with density up to the cut-off at a line
average density of 3 × 1018 m-3. Calculations of multi-pass absorption indicate that the total
absorption efficiency can be as high as 70% for two passes.

Figure 8 shows the measured absorption efficiency versus plasma density for the QHS
and Mirror configurations. The absorption efficiency is measured using a set of microwave
 QHS Mirror

1 1

Absorption
MD #1 MD #1
Absorption 0.8 MD #2 0.8 MD #2
MD #3 MD #3
0.6 0.6
MD #4 MD #4
0.4 0.4
0.2 0.2
0 0
0 1 2 3 4 0 1 2 3 4

Line Average Density, 1018 m-3 Line Average Density, 1018 m-3

Fig. 8: Multi-pass absorption coefficients for QHS and Mirror configurations,

measured by microwave antennas.

antennas, four of which are plotted in the figure. For both configurations the efficiency is in
the range of 0.8-0.98. It can be seen that for the Mirror configuration the efficiency drops at
densities less than 0.5 × 1018 m-3, while for the QHS configuration it remains high.
Interestingly, for the anti-Mirror configuration in which the confinement of trapped particles
near the ECH antenna is the worst of the three configurations, the measured absorption
efficiency is never higher than 0.6, while the stored energy remains low at only 5-7 J. For this
configuration, there are no hard X-rays even at the low densities. Hence, the discrepancy
between the ray tracing calculations and measured absorption efficiency at low density is
most likely due to additional absorption on a nonthermal electron population. Furthermore,
the differences in absorption measured for the QHS, Mirror and anti-Mirror configurations
can be explained by differences in trapped particle confinement.

Conclusion
In summary, anomalous transport in HSX is best described by an Alcator-like thermal
conductivity in which the thermal conductivity is inversely proportional to the density.
Departures of the experimental results from this model, as with the experimental stored
energy as a function of plasma density, can best be explained by a nonthermal component of
the electron distribution that is accounting for a significant fraction of the stored energy.

*This work is supported by DOE Grant DE-FG02-93ER54222

References:
1
N. Karulin, “Transport Modeling of Stellarators with ASTRA”, IPP 2/328, Dec. 1994.
2
K.C. Shaing, Phys. Fluids 27 (1984) 1567.
3
S.L. Painter and H.J. Gardner, Nucl. Fusion 33 (1993) 1107.
4
K. Likin et al., 30th EPS Conference on Controlled Fusion and Plasma Physics, to be
published.
5
N. Karulin, “Transport Modeling of Stellarators with ASTRA”, IPP 2/328, Dec., 1994
6
J. Canik et al., “Particle Neutral Density Modeling and Measurements in HSX”, this
conference.
7
H. Zushi et al., Nucl. Fusion 24 (1984) 305.
8
H. Maassberg et al., Phys. Fluids B 5 (1993) 3627.