All opinions expressed herein are those of the authors and should not be reproduced, quoted in publications, or
used as a reference without the author’s consent.
Oak Ridge National Laboratory is managed by UT-Battelle, LLC, for the U.S. Department of Energy.
Published by Fusion Energy Division, Oak Ridge National Laboratory
Building CR 5600 P.O. Box 2008 Oak Ridge, TN 37831-6169, USA
Editor: James A. Rome Issue 137 June 2012
E-Mail: jamesrome@gmail.com Phone (865) 482-5643
On the Web at http://www.ornl.gov/sci/fed/stelnews
Bulk and fast ion diagnostic in
the Large Helical Device using
collective Thomson scattering
Understanding the behavior of bulk and fast ions is a
major concern in fusion devices. A collective Thomson
scattering (CTS) diagnostic is one of candidates to measure
those charged particles and their fusion products
especially in high energy region. In the Large Helical
Device (LHD), a CTS diagnostic system has been developed
utilizing a high-power 77-GHz gyrotron as a probing
beam. A CTS diagnostic system has been developed to
measure the velocity distribution function in JET, TEXTOR,
and ASDEX-UG, and has been designed for ITER
[1–3]. These systems use high-power gyrotrons with frequencies
of 140, 110, 105, and 60 GHz and powers of a
few hundred to ~1000 kW to generate a probing beam.
The fast ion velocity distribution and dynamics are
obtained from the frequency spectrum of scattered radiation
only when the collective condition, 1/kD>1, is satisfied,
where D is the Debye length, and the fluctuation
vector is given by the incident and scattered
wave vectors, ki and ks, respectively (see Fig. 1).
The scattered radiation is very weak, so it is essential to
utilize a high-power coherent source. Gyrotrons with 60–a
few hundred GHz satisfy the collective condition and are
suitable for CTS diagnostics. The CTS diagnostic requires
a sensitive receiver system to detect the scattered radiation
from the plasma. In 2008 we designed the LHD CTS
receiver system [4], and after that we installed it and
obtained initial results for scattered spectra measured by
the CTS diagnostic [5].
The scattered radiation is resolved into 8 channels at the
receiver system. For more accurate velocity distribution
function measurements, the number of channels is
increased from 8 to 32 channels. These results have
already been reported in Refs. [6, 7]. Figure 2 shows a
broadband heterodyne receiver system, which mainly consists
of a notch filter, a mixer with a local oscillator, some
low-noise RF amplifiers, a filter bank, diodes, and video
amplifiers, which are connected to the data acquisition
Fig. 1. Vector diagram of incident and scattered waves.
Probing beam
Receiving beam
ki
ks
k􀁢
k = ks – ki
In this issue . . .
Bulk and fast ion diagnostic in the Large Helical
Device using collective Thomson scattering
Collective Thomson scattering (CTS) experiments
have started and made progress in the Large Helical
Device (LHD) at the National Institute for Fusion Science
(NIFS) in Toki, Japan. .................................... 1
CoordinatedWorking Group Meeting (CWGM9)
for Stellarator-Heliotron Research
The ninth Coordinated Working Group Meeting
(CWGM9) was held on 28 January 2012 at Australian
National University, with 15 experts participating. .. 4
Stellarator News -2- June 2012
system. The scattered radiation is reconstructed to obtain
the ion velocity distribution. The 32 channels were calibrated
using measurements of liquid nitrogen radiation or
electron cyclotron emission (ECE) during plasma discharges.
An amplitude-modulated probing beam with a
frequency of 50 Hz is injected into the plasma for the subtraction
of the background ECE. The probing and receiving
beams are controlled by the final steering mirrors (Fig.
3). The scattered radiation from the overlap between the
probing and the receiving beams is transmitted to the front
end of the CTS receiver system by the waveguides for
electron cyclotron heating (ECH).
Fig. 2. The CTS diagnostic system with a megawatt gyrotron probing beam and the CTS broadband receiver system for
the detection of scattered radiation.
Notch
filter
Local oscillator
(74GHz)
Mixer <6GHz
amplifier
32 channel
filters
32 channel
video amplifier
Data acquisition
Receiving
beam
Spectrum ± 3 GHz
Probing
beam from the gyrotron
77GHz
LHD
plasma
32 channel
diodes
77GHz gyrotron Filter bank of the CTS receiv er
Fig. 3. Final steering mirrors inside the LHD vacuum vessel for probing and receiving beams.
Stellarator News -3- June 2012
The time evolution of the measured CTS spectrum is
shown in Fig. 4. The plasma is sustained by two auxiliary
neutral beams: a perpendicular (NB4) and a parallel (NB3)
beam. The k vector, where the beam overlap exists, is
nearly perpendicular to the magnetic field. Therefore, the
intensity of the CTS spectrum in this case is considered to
be sensitive to the perpendicular NB4 beam. From the
above viewpoint, NB4 is injected, followed by an increase
of intensity and frequency spread (proportional to particle
velocity), as shown by the arrows in Fig. 4.
We have started comparison of both experiment and simulation
to understand both the CTS spectrum and the use of
predictive simulation for fast ion confinement. Figure 5 is
an example, which is calculated by an orbit-following
Monte Carlo code (MORH) [8]. The fast-ion velocityspace
distribution is calculated when one perpendicular
and two parallel beams are injected. As the observation
direction is at an angle of 100 degrees (the k vector in the
figure), all particles are projected onto the k vector. After
the treatment of this geometrical effect, we can obtain the
one-dimensional velocity distribution in the bottom graph
of Fig. 5. The next step is the development of a scattering
spectrum analysis code that can handle any kind of velocity
distribution, to compare with simulated and experimental
results.
Work is in progress to improve the signal to noise ratio for
the CTS receiver, high-purity probing beam, and fast scanning
mirror system in preparation for the forthcoming
LHD experimental campaign in summer 2012.
Masaki Nishiura
National Institute for Fusion Science
Toki, Japan
E-mail: nishiura@nifs.ac.jp
Fig. 4. Spectrogram of scattered radiation as measured by
the CTS diagnostic in LHD shot 101719. The frequency is
proportional to the velocity ( ). The 1-GHz frequency
shift corresponds to the energy of the40-keV NB4.
-1.0
-0.5
0.0
0.5
1.0
Frequency (GHz)
7.5 7.6 7.7 7.8 7.9 8.0
Time (s)
Notch filter region
NB#3
NB#4
NB#3: 175keV
NB#4: 32-35keV
v  k  
1D velocity distribution function (simulated)
Fig. 5. Top: Simulation of fast ion distribution during NB2,
NB3, and NB4 injection using the MORH code. Bottom: 1D
velocity distribution calculated from the simulated velocityspace
distribution with a k vector of 100 degrees.
6
4
2
0
v􀂌 (106m/s)
-6 -4 -2 0 2 4 6
6
4
2
0
v􀂌 (106m/s)
-6 -4 -2 0 2 4 6
6
4
2
0
v􀂌 (106m/s)
-6 -4 -2 0 2 4 6
v|| (106m/s)
NB2
NB3
NB4
kδ
􃻌
kδ
kδ
5
4
3
2
1
0
g(u) (1012m/s4)
-4 -2 0 2 4
u (106m/s)
NB3
NB2
Rax=3.6m, Bt=-2.4T, t=6.05s, r/a=0
NB4
Bulk
Sum
Stellarator News -4- June 2012
References
[1] H. Bindslev, J. A. Hoekzema, J. Egedal, J. A. Fessey,
T. P. Hughes, and J. S. Machuzak, Phys. Rev. Lett. 83,
3206 (1999).
[2] S. B. Korsholm, H. Bindslev, F. Meo, F. Leipold, P. K.
Michelsen, S. Michelsen, P. Woskov, E. Westerhof,
FOM ECRH Team, J. W. Oosterbeek, J. Hoekzema, F.
Leuterer, D. Wagner, and ASDEX Upgrade ECRH
Team, Rev. Sci. Instrum. 77, 10E514 (2006).F. Meo, H.
Bindslev, S. B. Korsholm, E. L. Tsakadze, C. I. Walker,
P. Woskov, and G. Vayakis, Rev. Sci. Instrum. 75, 3585
(2004).
[3] M. Nishiura, K. Tanaka, S. Kubo, T. Saito, Y. Tatematsu,
T. Notake, K. Kawahata, T. Shimozuma, and T. Mutoh,
Rev. Sci. Instrum. 79, 10E731 (2008).
[4] M. Nishiura, S. Kubo, K. Tanaka, N. Tamura, T. Shimozuma,
T. Mutoh, K. Kawahata, T. Watari, T. Saito,
Y. Tatematsu, T. Notake and LHD experiment group,
Journal of Physics: Conference Series, 227, 012013
(2010).
[5] S. Kubo, M. Nishiura, K. Tanaka, T. Shimozuma, Y.
Yoshimura, H. Igami, H. Takahashi, T. Mutoh, N.
Tamura, Y. Tatematsu, T. Saito, T. Notake, S. B. Korsholm,
F. Meo, S. K. Nielsen, M. Salewski, and M. Stejner,
Rev. Sci. Instrum. 81, 10D535 (2010).
[6] Masaki Nishiura, Shin Kubo, Kenji Tanaka, Namiko
Tamura, Takashi Shimozuma, Takashi Mutoh, Kazuo
Kawahata, Tetsuo Watari, Teruo Saito, Yoshinori
Tatematsu, and LHD Experiment Group, Plasma Fusion
Res. 6, 2402068 (2011).
[7] Ryosuke Seki, Yutaka Matsumoto, Yasuhiro Suzuki,
Kiyomasa Watanabe, Kiyotaka Hamamatsu, and Masafumi
Itagaki, Plasma Fusion Res. i5, 014 (2010).
CoordinatedWorking Group
Meeting (CWGM9) for
Stellarator-Heliotron Research
The 9th Coordinated Working Group Meeting (CWGM9)
was held on 28 January 2012 at Australian National University,
Canberra Australia, with15 experts participating.
The meeting was composed of six sessions including the
Opening, which discussed the evolution of CWGM and its
central role for international collaboration in the International
Energy Agency (IEA) Implementing Agreement for
Cooperation in Development of the Stellarator-Heliotron
(S-H) Concept (http://iea-shc.nifs.ac.jp/). The list of joint
papers originated from CWGM was introduced to stimulate
more joint activities to systematize the physics understandings
in S-H plasmas. The achievements of CWGM8
(Stellarator News, Issue 131, April 2011, http://
www.ornl.gov/sci/fed/stelnews/issue131.pdf) were also
briefly mentioned.
Following the Opening, in the Energetic Particles session,
D. A. Spong (ORNL) reported on a new code benchmark
study of linear growth rates of energetic particle driven
Alfvén eigenmodes (AE modes) in S-H plasmas. Currently,
five different models (MEGA-R, AE3D-K, local
analytic, CAS3D-K, and CKA-EUTERPE) are participating
in the benchmark, and two toroidal AE (TAE) modes
that were observed in LHD were selected for analysis. At
this stage areas of both similarity and difference have been
found among the codes, depending on how the particle
weights are evolved and which of the two modes is examined.
As the study progresses, it is expected to sort out the
differences and to be extended to observations from other
devices. It was also pointed out that for ITER, the impact
of a large alpha particle population on AE modes is an
important issue, and contributions from the S-H community
are anticipated for discussions in the ITPA.
F. Castejon (CIEMAT) discussed the dynamics of fast ions
coming from neutral beam injection (NBI) heating in
three-dimensional (3D) systems using Monte Carlo codes,
without any assumptions on the diffusive nature of transport,
on the size of orbits, or on the conservation of kinetic
energy. The codes FAFNER2 and HFREYA are linked to
ISDEP (Integrator of Stochastic Differential Equations for
Plasmas) in order to study such dynamics in TJ-II and
LHD plasmas. The steady-state distribution function is
obtained at several radial positions and from it, several
interesting quantities, such as poloidal and toroidal rotation
velocities, can be estimated. It has been found that the
CNPA (Compact Neutral Particle Analyzer) spectra in TJII
are in good agreement with the simulations. The slowing-
down time is obtained by NBI blip experiments and
simulations for TJ-II and LHD, with good agreement in
both devices, showing also a good similarity with the
Spitzer slowing-down formula in LHD and a strong difference
in TJ-II, attributed to the larger radial extent of ion
orbits.
The importance of the validation of numerical codes
against experimental results and the resulting increased
accuracy of predictions for energetic particle issues (especially
for ITER) were pointed out.
A new Equilibrium in Experiment session was launched at
CWGM9, reflecting increased interest in and activities on
equilibrium reconstruction based on the progress of profile
measurement, numerical code applications for experimental
analysis, and the extension of the International Stellarator-
Heliotron Profile Database (ISH-PDB), to increase its
Stellarator News -5- June 2012
usability and scope. J. Geiger (Max-Planck, Greifswald)
used function parameterization to try to shortcut the process
for the interpretation of experiments (equilibrium
reconstruction) by utilizing pre-calculated VMEC equilibria.
In practice, reconstruction of flux surface shape is represented
by Fourier components, Rmn and Zmn, and then
quantities such as reff (radial coordinate), coil currents,
plasma current and plasma pressure, etc., can be represented
using them, in W7-AS. This approach has been
implemented for W7-X as a Web service. It will be utilized
for mapping of diagnostics and simulations of diagnostic
measurements.
C.Suzuki (NIFS) applied real-time magnetic coordinate
mapping (so-called TSMAP) in LHD, which is based on
searching for the “best-fit” coordinate mapping from a
wide range of precalculated VMEC equilibria to make the
measured electron temperature (by Thomson scattering)
symmetric.
The progress of ion temperature (Ti) measurement by
XICS (X-ray Imaging Crystal Spectrometer) in LHD and
related equilibrium reconstruction research was reported
by N. A. Pablant (PPPL). This diagnostic allows Ti measurements
to be made under plasma conditions where
existing diagnostics (e.g., CXRS) cannot be operated. The
system is now fully commissioned and can provide lineintegrated
measurements of Ti and Te. Local profiles of Ti
and Te are found through Doppler tomography utilizing
the known plasma equilibrium. Initial comparisons against
Thomson and CXRS show good agreement, demonstrating
the applicability of this technique to helical geometries.
There was discussion on how this diagnostic could
be integrated into transport codes to provide Ti profiles to
study heat transport in LHD. N. A. Pablant also presented
a report on the S-H equilibrium reconstruction activity
being conducted at PPPL.
The STELLOPT code, has been developed for S-H equilibrium
reconstruction. It optimizes the VMEC input
parameters to obtain a best match to diagnostic data using
a modified Levenberg-Marquardt algorithm. Reconstructions
were shown for several LHD discharges where optimization
was done to match the following parameters: coil
currents, stored energy, net toroidal current, and electron
pressure. Work is in progress to add pitch angle measurements
from the motional Stark effect diagnostic. Current
development of STELLOPT is aimed at improving performance
and user interaction to facilitate its routine use. The
codes PIES and SPEC were introduced; they allow investigation
of equilibria in the absence of good flux surfaces
and complement work done using HINT2. Discussion
addressed the integration of STELLOPT results into diagnostic
interpretation and transport studies. It was proposed
that comparisons of STELLOPT equilibria with those
determined though functional parameterization and
VMEC database approaches, as presented in this session,
could be used to examine the accuracy of these fast lookup
techniques.
The progress of the equilibrium registration attached to the
registered experimental profiles in the ISH-PDB was
reported by M.Yokoyama (NIFS), who presented as an
example the examination of an LHD high-Ti discharge.
The TSMAP-defined VMEC equilibrium (re-calculated by
inputting parameters corresponding to the “best fit”) was
registered as a test case. The process for utilizing the equilibrium
was also explained. Those interested in analyzing
a registered discharge using their numerical codes can start
the calculations either from a registered VMEC input or
from the output files. This should speed up the transition
of ISH-PDB from the “storage” phase to the “utilization”
phase.
It was noticed that various approaches for equilibrium
reconstruction/specification have been in progress (there
was also a session in the following 18th International Stellarator-
Heliotron Workshop), and it was pointed out that
comparisons on these approaches would be beneficial to
increase their validity and accuracy.
In the Transport Analysis session, M.Yokoyama (NIFS)
presented an analysis of steady-state power balance in
LHD, using TR-SNAP. This code has now been linked to
TSMAP by establishing the interface to describe measured
temperature and density profiles as a function of reff and to
acquire NBI energy/port-through power. This suite was
demonstrated and is now available to collaborators. It can
replace the TSMAP-defined equilibrium by equilibria constructed
in different ways, as discussed in the Equilibrium
in Experiment session. This will provide an opportunity to
test the impact of equilibrium properties on steady-state
power balance analysis in an easy way.
A. Wakasa (Kyoto U.) reported on the status of the
TASK3D development for the predictive simulation of
reachable temperatures (and the resulting profiles) in LHD
including turbulent transport modeling in addition to the
established neoclassical diffusion coefficient database,
DGN/LHD. It was pointed out that validation experiments
in LHD should increase the accuracy of the prediction, and
that a comparative study on a module-to-module basis,
with, for example, the predictive transport code developed
at IPP, would be valuable.
Rotation and momentum transport issues have gathered a
lot of attention recently. In the Rotation session, recent
activity at TJ-II was presented by J. A. Alonso (CIEMAT).
Mass flows in (nonsymmetric) S-H plasma are expected to
be dictated by NC ambipolarity and the parallel force balance
equation. These devices are, therefore, best suited for
testing neoclassical predictions on flow damping. External
biasing provides a controlled perpendicular force that is
Stellarator News -6- June 2012
easy to quantify. TJ-II has recently pursued different
experimental approaches to test the pre-eminent role of
neoclassical mechanisms in flow pattern regulation. At
CWGM9, there was only one presentation on this topic.
However, similar studies have been or are being performed
in other devices, and thus a coordinated action
seems appropriate, which could also help to expand our
understanding of flows in symmetric configurations. J. A.
Alonso has visited Heliotron J and LHD in March 2012,
and moved to launch such a coordinated action.
We also had a session on Pellet Fueling. G. Motojima
(NIFS) is starting to deal with fueling issues in a S-H
power plant. The significance of pellet injection as the
main fueling in S-H plasmas was discussed. Capability for
higher-density plasma confinement in S-Hs provides an
attractive scenario for a power plant. The key to make this
realized is a reliable pellet injection system, such as the
one operating on LHD. Research on pellet fueling was
introduced. The conceptual design for Heliotron J was also
presented. The pellet injection system for TJ-II, constructed
with U.S. collaboration, is now nearly finished.
Such installations will certainly enhance the collaborative
research among several devices on physics and technological
issues related to pellet injection and facilitate future
collaborative development.
Opportunities for collaborative research on H-1NF were
discussed by D. Pretty (ANU). In order to facilitate H-1NF
data access for collaborators, a new data access system
(H1DS) has been developed to provide a data interface
that is simple and intuitive for new users and collaborators.
The H1DS design utilizes the hypertext transfer protocol
(HTTP) to provide an extensible web-service based
application programming interface (API), which can interface
with standard data analysis languages (Python, IDL,
Matlab, Labview, etc.). A web-based MDSplus data
viewer is now online at http://h1ds.anu.edu.au/mdsplus,
providing simple navigation and basic processing of data.
The new data system also includes a summary database,
configurations database, and centralized documentation. A
point of discussion was the possibility of deploying the
H1DS software with datasets from other (non-MDSplus)
S-H devices. Currently the Web service only supports
MDSplus, but it is not very tightly coupled to it, so it
would not take too long to implement on other systems
(~ few weeks development time).
Progress towards an MHD Documentation Database
(MDDB), occasionally discussed at previous CWGMs,
was also introduced. This project aims to provide a reference
documentation database of MHD modes using data
mining techniques. Fluctuations are found by scanning
Mirnov signals and “fingerprinting” coherent fluctuations
by the phase difference between adjacent Mirnov channels;
clustering algorithms are then used for unsupervised
identification of the same fluctuation across many shots.
These techniques have been used to identify Alfvén eigenmodes
and other modes in H-1, TJ-II, and Heliotron J for
thousands of shots. In order to generate the MDDB database,
processing has started on a larger dataset, with
~50,000 shots and including LHD and W7AS; improved
clustering algorithms capable of scaling to more than 100
data points are being explored. The discussion focused on
what properties of fluctuations should be recorded when
scanning Mirnov signals: for example, frequency spectral
width may reveal growth rates.
Finally, it was pointed out that, in addition to widening the
range of topics (the extent of the mountain), “flagship”
topics (a peak of the mountain) should be intensively dealt
with, so that CWGM becomes more visible and relevant in
world-wide fusion research, through more outreach activities.
Strategic discussions on how to make CWGM evolve
will be continuously made in preparation for the next
(10th) CWGM. The 10th CWGM, now being planned, is
to be held 6–8 June 2012, in Greifswald, Germany. Your
interest and participation are anticipated. The CWGM10
Web site is at
http://www.ipp.mpg.de/~dinklage/CWGM10/.
The materials presented at the 9th CWGM are available at
http://ishcdb.nifs.ac.jp/cwgm9.html.
Acknowledgments
Dr. Boyd Blackwell and Dr. David Pretty (Australian
National University) are greatly appreciated for their
excellent hosting of CWGM9, particularly on the occasion
just before the 18th International Stellarator-Heliotron
Workshop, for which they also made extensive efforts as
its local organizers. The 9th CWGM is partly supported by
NIFS (National Institute for Fusion Science)/NINS
(National Institutes of Natural Sciences) under the project
“Promotion of the International Collaborative Research
Network Formation.”
M. Yokoyama (NIFS), on behalf of all of the participants in
CWGM9.
E-mail: yokoyama@LHD.nifs.ac.jp