Published by Fusion Energy Division, Oak Ridge National Laboratory Building 5700 P.O. Box 2008 Oak Ridge, TN 37831-6169, USA
Editor: James A. Rome Issue 120 June 2009
E-Mail: jar@ornl.gov Phone (865) 482-5643
n the Web at http www r g v sci fed ste ews
with a rotational transform = /2 = 0.1, …, 1 for a typi-
Photogrammetric survey
of magnetic field lines in the stellarator WEGA
Introduction
The confinement quality of the magnetic field in a stellara- tor is generally very sensitive to field errors, e.g., devia- tions in the position and orientation as well as the shape of the magnet coils. It is therefore necessary to characterize the magnetic configuration and prove the existence of closed and nested flux surfaces. Since the vacuum field in a stellarator—produced only by the external coils—is already confining, this check for field integrity can be done experimentally without plasma. A common method [1, 2] of surveying the field structure utilizes an electron beam, which follows the field line due to reduced perpen- dicular mobility. A fluorescent mesh or a moving fluores- cent bar is necessary to detect the intersection points of the beam in a poloidal cross section. So far, only the detection principle limits the measured data to a two-dimensional cut view. The electron beam trajectory follows the whole field line on its flux surface, so that an alternative beam measurement technique can offer spatial information.
Using a background gas, it is possible to visualize the beam via electron-impact excitation of the gas particles [3]. This luminescent trace is visible throughout the stel- larator and can be observed from different viewpoints and measured by means of photogrammetry. The trace is lim- ited in range due to a finite initial energy of the electrons and a steady energy loss in the excitation processes. This less invasive technique, which provides a three-dimen- sional (3D) measurement of the magnetic field lines, was applied and optimized at the WEGA stellarator at Max- Planck-Institut f.ir Plasmaphysik, Greifswald.
The experimental setup
The experiment WEGA is a classical stellarator with a major radius R0 = 72 cm and a minor radius r0 = 19 cm for the toroidal plasma vessel. Forty toroidal and two helical copper coils produce a magnetic field B0 = 50–600 mT
cal flat-top time of tM = 30 s. A more detailed description of the stellarator WEGA can be found in [4].
The free electrons for the beam were provided by a heated tungsten wire and accelerated to an initial energy E0 = e•UA with the acceleration voltage UA applied between the wire (cathode) and metal cup (anode) with a hole. The ves- sel was at the same potential as the anode to screen the electric field outside the electron gun. Figure 1 shows the measurement setup at the stellarator using four viewpoints with overlapping fields of view. Inside the vessel, which was evacuated to a pressure p between 103 and 106
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.
ak Ridge National Laboratory is managed by UT-Battelle, LLC, for the U.S. Department of Energy.
Fig. 1. Experimental setup with toroidal field coils (dark blue), vacuum vessel (grey), luminescent trace of the electron beam (red), four observation cameras (black) and six reference marks mounted on the high field side (white dots with rings).
mbar, a set of reference marks was installed to determine the exact camera positions. The pictures were taken by a Starlight ExpressTM HX916 monochrome camera using a Peltier-cooled CCD chip with a resolution of 1300 1030 pixels, allowing detailed images and long exposures of up to tE = 5 s. To retrieve the 3D coordinates of the visible trace, the commercial photogrammetry software AICON 3D-Studio [5] was used. The program calculates the light ray going from the visualized field line through the optical system to the pixel on the image sensor with a ray tracing algorithm. By taking two different pictures from the same section of the field line transition, the intersection points of the respective calculated light rays provide the coordi- nates of the measured trace. The precise optical and geo- metric properties of the lens system and image sensor were checked in a prior calibration. In addition, the exact loca- tions and orientation of the cameras were determined using a calibrated set of reference points visible in the background of every image.
Results
The electron beam was optimized for high luminescence, long range, and small divergence. The main dependences occurred for different gas pressures p, gas types, field strengths B0, and acceleration voltages UA. The effect of these parameters on the range of the visible trace was addi- tionally compared with a collision model taking into account inelastic and elastic scattering of electrons, atoms, and ions. Although qualitative agreement was found, charge accumulation, space charge effects, and increases in pitch angle seemed to be non-negligible but unconsid- ered factors.
Regarding the electron gun, the shape and design of the anode had a strong influence on both the emission current
and the distribution of electrons with different pitch angles in the beam.
For application at WEGA, the range of the visible trace was not maximized but limited to ensure easy distinction of the different revolutions. As a result, photos like that in Fig. 2 were taken for a magnetic configuration with 0.2 at B0 589 mT. Argon served as background gas at p 4
104 mbar. The emitted electrons with a current of IE 2
mA had an initial energy of E0 300 eV. Thus, the beam created a visible trace of 5 revolutions or s 24 m in length and a beam cross-section diameter d 5 mm.
As shown in Fig. 2, photos of the luminescent traces and the reference points have been taken sequentially and merged via photo editing to handle the high contrast between reference points, traces and background. These perspective pictures from four angles of view were pro- cessed by the photogrammetry software, and a measured point cloud of the trace was obtained.
It was possible to measure up to the fourth revolution of one magnetic field line in the stellarator with an accuracy of about ±5 mm. By measuring part of the fourth revolu- tion, this is equivalent to a sensitivity in the rotational transform of = 0.004. The sketches in Fig. 3 show a comparison between the measured points and a numerical simulation [6] of a drift path of equivalent electrons. As can be seen, the trajectories are not exactly the same; rather, they are shifted and tilted. This produces an addi- tional deviation up to r = 7 mm at the fourth revolution. This deviation results from the determination of the cam- era positions using inaccurate positions of the reference marks and problems with merging the sequential photos.
Fig. 2. Photo of one visualized magnetic field line with five transitions in the field of view and six reference marks inside the torus.
Outlook
The proof of principle for this 3D measurement technique for magnetic field lines was given, although some work on improving the accuracy remains to be done. In particular, better camera position determination by using a more accurate mount for the reference marks would be benefi- cial. In order to prevent unintended interaction with the plasma in later experiments, it would also be preferable to use distinct features of the plasma vessel as reference points rather than additional marks mounted on a carrier.
Nevertheless, this technique presents an alternative way of characterizing vacuum stellarator fields, and its use for the first magnetic field tests at Wendelstein W7-X is under discussion. Additional fields of application are the align- ment and adjustment of toroidally separated plasma diag- nostics aimed at specific parts of the magnetic field, such as the magnetic axis or the last closed flux surface. Fur- thermore, this principle could be used for an experimental 3D investigation of trapped particle orbits, with an appro- priate electron gun that emits particles with a high fraction of the impulse perpendicular to the magnetic field line.
Fig. 3. Comparison of measured data (red) with numeri- cally calculated drift trajectories (orange) viewed from above (top) and inside (bottom) on the plasma vessel (dark blue contours). The black figure represents the six con- nected reference marks.
References
M. Otte and J. Lingertat, "Initial Results of Magnetic Surface Mapping in the WEGA Stellarator,” Proc. 29th EPS Conf. Plasma Phys. Contr. Fusion, Montreux, 17– 21 June 2002, ECA, Vol. 26B, p. 5.036 (2002).
M. Otte et al., "Magnetic Flux Surface Measurements at Wendelstein W7-AS,” Stellarator News 100, p. 2–5 (2005).
O. Neubauer, F. H. Bohn et al, "Measurement of the Vertical Magnetic Field in TEXTOR using the Electron Beam Technique,” Fusion Sci. Technol. 31, 154–158 (1997).
M. Otte et al., "The WEGA Stellarator: Results and Prospects,” AIP Conf. Proc. 993, 3–10 (2008.
AICON 3D Studio, AICON 3D Systems GmbH, www.aicon.de.
A. Werner, Max-Planck-Institut f.ir Plasmaphysik, Greifswald, private communication.
Electron cyclotron emission from the WEGA stellarator
After installation of the new 28-GHz electron cyclotron resonance heating (ECRH) system (gyrotron) at the stel- larator WEGA (Max-Planck-Institut fir Plasmaphysik, Greifswald) electron temperatures clearly exceeding
15 eV are expected [1]. In this paper we describe the recently installed electron cyclotron emission (ECE) diag- nostic. Because ECE is the standard temperature diagnos- tic in fusion research, it should be tested at WEGA. The first tests presented here have been carried out using a 2.45-GHz ECRH system with electron temperatures of about 5 eV. Under these conditions the emission is domi- nated by bremsstrahlung; however, the existence of ECE has been demonstrated.
WEGA is a medium-sized classical = 2, m = 5 stellarator with major radius R 72 cm and maximum plasma radius a 11 cm. For central power deposition in the plasma using the 10-kW, 28-GHz ECRH system, the magnetic flux density on the plasma axis must be B0 = 0.5 T to be resonant at the second harmonic of the electron cyclotron frequency. In addition to this resonant case, the plasma can be heated non-resonantly at B0 = 0.5 T by the 2.45-GHz system (26 kW). Nearly the same electron densities up to
ne = 5 x 1018 m3 with typical electron temperatures Te =
5 eV can be reached in both cases.
To calculate the total expected microwave emission of the plasma, an integration of the emission coefficient along the sight line must also include absorption. In the case of high optical depth ( » 1 ), line integration of the local absorption coefficient along the propagation direction of the electromagnetic wave leads to a direct proportionality of observed intensity to Te. The second harmonic of the perpendicularly polarized wave (X2-mode) has the highest optical depth and is generally used for temperature diag- nostic [2]. The optically thin plasmas at WEGA have <
0.1 in the expected X2 frequency range from 25 GHz to 35 GHz, so that wall reflections at the metallic walls of the torus have to be considered. This leads to an observed radiation temperature Ts =
Te(1 e)/(1 e) with a typical value of Ts = 0.1Te for
a reflection coefficient > 0.9.
To detect the microwave radiation emitted by the plasma, a multichannel radiometer system was installed (Fig. 1).
Fig. 1. Block diagram of the ECE diagnostic.
The radiation is received by a horn antenna and preampli- fied by a low-noise-amplifier (LNA) with a noise tempera- ture of 230 K and 22 dB gain. Conversion to a lower frequency band by a heterodyne technique is not necessary due to the availability of amplifiers in the frequency range up to 40 GHz [2]. The signal is split by means of power dividers and bandpass filters into 12 frequency bands. The power of the incident radiation is measured using incoher- ent detectors. The radiometer filter bank could be taken from the earlier ECE system of Wendelstein 7-AS (W7- AS) [3]. Of the existing 12 channels, 7 are within the expected electron cyclotron frequency range on WEGA. To protect the sensitive preamplifier against the strong 28- GHz radiation of the gyrotron, a notch filter based upon lateral coupling of cavity resonators at a waveguide was built. The H011 mode was selected as the induced wave field at 28 GHz. Figure 2 shows the transmission charac- teristic of the notch filter using four serial connected reso- nators.
Fig. 2. Notch filter loss I versus frequency f. Grey shaded frequency ranges show bandpass filters of the radiometer channels. The H011 main resonance at 28 GHz induces
suppression up to 45 dB.
The vertical green lines indicate the region of expected ECE at second harmonic according to the plasma area within the last closed flux surface (LCFS), see Fig. 3 for orientation.
has the ability to detect radiation temperature values downward to 0.1 eV. For better comparability Fig. 5 depicts the radiation temperature vs time and the mid-fre- quency of the channels.
Fig. 3. Pointier plot of a WEGA configuration with a = 1/5 and the magnitude of magnetic flux density color coded in the background.
An overlap of different harmonics does not exist. The fil- ter was designed in such a way that the sensitivity of chan- nels covering the ECE frequency range is only weakly affected by parasitic resonances. The sensitivity of the channels was determined with the aid of a hot-cold tech- nique [2] using blackbody radiation at room temperature and liquid nitrogen. The measured values are in the region of
30 V/K.
The first measurements with the new microwave diagnos- tic investigated plasmas using nonresonant heating at
2.45 GHz and 0.5 T and demonstrated the performance of the whole ECE system. Evidence for cyclotron emission was found in a specially designed discharge scenario shown in Fig. 4 where two time windows with different magnetic field are compared.
The magnetic flux density was chosen such that the ECE X2 emission is once outside the detection range of the radiometer (B0 = 0.35 T, fc,X2 = 19.6 GHz) whereas the ECE shifts towards the central frequency channels for the higher field (B0 = 0.49 T, fc,X2 = 27.5 GHz). To guarantee a constant radiation level, the heating power Pout was adjusted at 0.49 T such that the line-integrated density ndl was kept constant. The agreement of electron density ne and electron temperature Te profiles was checked with Langmuir probes. The bottom time trace shows as an example the corresponding radiation temperature of the channel with a mid-frequency of 31.5 GHz. The measured value correlates with the line-integrated density and the magnetic flux density. It also makes clear that the system
Fig. 4. Time traces of magnetic flux density B0 on the axis, heating power Pout, line-integrated density II d1, and radia- tion temperature TS.
Fig. 5. Radiation temperature TS vs time t and mid-fre-
quency f of the radiometer channels.
A monotonic decrease of the radiation temperature with higher frequency is obvious. With step-up of the magnetic flux density from 0.35 T to 0.49 T, the radiation tempera- ture in the channels with additionally expected ECE increases, too. Before this rise can be identified as cyclo- tron emission, the source of the emission for magnetic field values up to 0.35 T must be clarified. A third har- monic in this amplitude would demand an increase of emission in the two channels above 37 GHz at the mag- netic field of 0.49 T. Taking into account that low-temper- ature plasmas are dominated by collisions, bremsstrahlung
also has to be considered. The integrated value of the emission coefficient of bremsstrahlung and cyclotron radi- ation is in the same order of magnitude but the typical exponential decay of bremsstrahlung with frequency is not expected in the microwave range. However, applying a simple one-dimensional radiative transfer model results in a similar frequency dependency of their expected radiation temperature [4]. Figure 6 shows the radiation temperature vs the frequency at a magnetic flux density of 0.35 T.
Fig. 7. Difference of radiation temperatures TS at magnetic flux density of 0.49 T and chosen reference field versus mid-frequency f of channels.
Fig. 6. Radiation temperature TS vs mid-frequency f of the radiometer channels at a magnetic flux density of 0.35 T.
The vertical black lines indicate the frequency range with expected cyclotron emission at a magnetic flux density of
0.49 T. ECE is expected only in the channels outlined in red. The red curve shows the calculated radiation tempera- ture of bremsstrahlung, taking wall reflections based on this simple model into account, confirming the observed magnitude as well as the basic frequency dependence. However, to make a unique identification of the observed jump in the radiation temperature with ECE, a reference field larger than 0.35 T is necessary so that the reference case channels with expected ECE fall in the frequency range of the diagnostic. This is used for Fig. 7, which plots the increase of the radiation temperature at 0.49 T with respect to two different reference fields.
The overlapping channels are outlined in red. At both ref- erence fields (0.40 T and 0.44 T), these channels are sepa- rated by an edge from the other ones. This observation shows ECE as an additional radiation process. Hence a mixing of two different radiation processes is reflected in this temperature range. At electron temperatures above
15 eV the possibility of their quantitative determination via ECE is expected. However, considering the wave- length and plasma dimensions a full-wave calculation is additionally necessary for this purpose.
Acknowledgement
The authors are grateful to Thomas Geist, Friedrich Wag- ner, and Heinrich Laqua for their contributions to the experiment and thank Dieter ABmus and Ralf Gerhardt for their technical support.
References
H. P. Laqua, D. Andruczyk, E. Holzhauer, S. Marsen,
M. Otte, Y. Y. Podoba, J. Preinhealter, J. Urban, and
G. B. Warr, "Electron Cyclotron Wave Experiments at the WEGA Stellarator,” Proc 34. EPS Conf. Plasma Phys., Warsaw 2007, European Conf. Abstracts 31F P- 1.154 (2007).
H. J. HartfuB, T. Geist, and M. Hirsch, "Heterodyne methods in millimetre wave plasma diagnostics with applications to ECE, interferometry and reflectometry,” Plasma Phys. Control. Fusion 39, 1693–1769 (1997).
Ch. Fuchs and H. J. HartfuB, "Technology of the new W7-AS broadband radiometer system,” Fusion Eng. Des. 53, 451–456 (2001).
G. Bekefi, Radiation Processes in Plasmas, John Wiley and Sons, Inc., New York, London, Sydney (1966).
Electrostatic potential mea- surement using a 6-MeV heavy ion beam probe in LHD
Introduction
A heavy ion beam probe (HIBP) is a unique diagnostic tool to measure the electrostatic potential, together with the potential and density fluctuations, directly and simulta- neously in high-temperature magnetically confined plas- mas. In nonaxisymmetric plasmas, the radial electric field is a key parameter for both neoclassical transport and anomalous transport.
In the Large Helical Device (LHD) at Toki, Japan, an HIBP has been in development since the beginning of the project. Recently, potential profiles and fluctuations have been observed using the HIBP [1, 2]. Here the HIBP sys- tem is described and some observations using the HIBP are reported.
HIBP system on LHD
In an HIBP, singly charged positive ions are injected into a plasma as a primary beam, and doubly charged ions obtained by ionization due to collision with the plasma on the path of this beam are detected (referred to as the sec- ondary beam). In order to extract the probe beam from the plasma, the Larmor radius of the probe beam must be comparable with the size of the plasma device. Since the magnetic field in LHD is up to 3 T and the device size is a few meters, singly charged heavy ions with energies of
several million electron volts are required for the primary beam. To date, the highest energy achieved has been 2 MeV for the HIBP in the TEXT-U tokamak, in which some technical difficulties were experienced in high-volt- age operation. In order to reduce difficulties, a tandem accelerator and a new type of energy analyzer have been adopted in the LHD HIBP.
Unlike previous HIBPs, a tandem accelerator is used in the LHD HIBP to reduce the acceleration voltage. A tandem accelerator has an ionizer, which is a gas cell for the LHD HIBP, at the center, and high positive voltage is applied there. Singly charged negative ions, which are Au- for the LHD HIBP, are initially injected into the accelerator and accelerated towards the gas cell. They are ionized posi- tively (Au+, Au2+, …) by collisions with neutral particles in the gas cell. Then, they are accelerated again toward the other side port, which is grounded. Thus, the tandem accelerator can accelerate the beam particles twice, and only half of the acceleration voltage corresponding to the required energy of a singly charged positive ion beam is required. On the other hand, the tandem accelerator requires negative ions for the initial beam. From the point of view of ion mass and reliability as a negative ion source, gold ions (Au-) were selected and have been devel- oped for the LHD HIBP. The singly charged positive ions (Au+) which are accelerated and extracted from the accel- erator are injected into plasmas. A beam energy of 6 MeV is needed to measure the potential using the Au+ beam at the magnetic axis with a magnetic field strength of 3 T, so the maximum acceleration voltage of the tandem accelera- tor is 3 MV.
Fig. 1. Schematic view of the LHD HIBP.
Ste arator News -7- June 2009
The probe beam is injected through the bottom port of the LHD vacuum vessel, and the secondary beam obtained by ionization due to collision with the plasma is selected and detected at the horizontal port. Octupole deflectors are installed at the injection and detection ports, and both are used to control the beam trajectory in three dimensions. This "active trajectory control” method was developed for an HIBP in the Compact Helical System (CHS). The geometry of the LHD HIBP is shown in Fig. 1.
The variation in beam energy, which corresponds to the electrostatic potential at the ionization position of the probe beam, must be analyzed. Parallel-plate electrostatic analyzers are usually used for HIBPs because they have the desirable property of second-order focusing for the incident angle. However, this type of analyzer requires an impractical voltage of several hundred kilovolts for the MeV-range beam of the LHD HIBP. Hence, we have devised a new energy analyzer with two sets of electrodes: a tandem parallel-plate energy analyzer. The incident angles of the tandem energy analyzer are designed as 6 degrees for the first electrode and 10 degrees for the sec- ond electrode, in contrast to the 30 degrees of a conven- tional parallel-plate analyzer. Consequently, the required voltages can be reduced to 56.5 kV on the first electrode and 113.6 kV on the second electrode for a 6-MeV beam, and this new analyzer has been verified to work success- fully in the LHD HIBP system.
The observable area is shown in Fig. 2. The potential pro- file can now be measured in the core region by the HIBP. Since the radial electric field Er in the edge region can be measured using charge-exchange spectroscopy (CXS) in LHD, the Er profile in the whole region can be measured with the HIBP and CXS. Complementary measurements using both the HIBP and CXS will advance the study of the physics of radial electric field formation in helical plasmas.
Fig. 2. Observable region. (a) Rax = 3.6 m, BQ = 100%, =
1.254. (b) Rax = 3.75 m, BQ = 100%, = 1.254. The beam
energy is shown in the legend in each figure. The sample
volume traces the marked curve during a sweep of the probe beam. The sample volume of the probe beam with the same Larmor radius ( 1b Eb Bt ) traces the same
curve, where Eb and Bt are the beam energy and the toroi- dal magnetic field strength, respectively.
Potential profile and fluctuation measurement
In this section, as an example of potential measurement by the HIBP, the change in the potential profile during elec- tron heating and the characteristic potential fluctuations in plasmas with weak magnetic shear are shown.
In LHD, the rotational transform is controlled by use of electron cyclotron current drive (ECCD) 3 . Weak or neg- ative magnetic shear profiles are formed during ECCD in the co-direction, which increases the rotational transform in the core region, and then characteristic magnetic and electrostatic fluctuations are observed.
The plasma is produced and sustained by balanced tangen- tial neutral beam injection (NBI), and the line-averaged electron density is 0.1 1019 m3 and constant. ECCD is superposed for 0.6 s, and weak magnetic shear profiles are formed by ECCD as shown in Ref. 3 . The electron tem- perature and electrostatic potential profiles during and after ECCD are shown in Fig. 3. A positive electric field is formed in the core region during ECCD because of the electron heating, and the transition to the electron root pre- dicted by neoclassical theory qualitatively accounts for this behavior of the potential profiles.
Fig. 3. (a) Electron temperature profiles and (b) electro- static potential profiles. The plasma is produced and sus- tained by balanced tangential NBI, and ECCD is superposed from 1.3 to 1.9 s. The reference electrostatic potential is on the LHD vacuum vessel.
During ECCD, several magnetic fluctuations are observed at frequencies 40 kHz, and their frequencies change gradually with a time constant of a few hundred millisec- onds as shown in Fig. 4(a), though the electron tempera- ture and density are constant. The frequency shift seems to reflect a change in the rotational transform profile. Similar fluctuations are observed in plasmas with weak or reversed magnetic shear produced by neutral beam current drive (NBCD) [4], and the fluctuations are the reversed- shear-induced Alfven eigenmodes (RSAEs).
tion of the HIBP is reciprocated from ~ 0.1 to ~ 0.4 at a frequency of 10 Hz (Fig. 4(b)) and the spectrogram includes not only the temporal evolution but also the spa- tial structure. As shown in Fig. 4(c), several modes whose frequencies vary in the frequency range of 40 kHz and more in the potential fluctuation are observed, as in the magnetic fluctuation. In addition to that, a potential fluctu- ation with a constant frequency of about 35 kHz during ECCD and 20 kHz just after ECCD is observed. The fre- quency depends on the square root of the electron temper- ature, and its absolute value almost agrees with the geodesic acoustic mode (GAM) frequency. The mode is not observed in plasmas without tangential NBI. Thus, the mode with the constant frequency will be the energetic particle-induced GAM.
As mentioned above, since the spectrogram shown in Fig. 4(c) includes information about the spatial structure as well as the temporal evolution, the spatial distribution of the potential fluctuations can be estimated. Figure 5 shows the distribution of the power of each mode [5]. The lower frequency mode exists in inner region, and the mode with the GAM frequency is excited in the innermost region.
At present, the mode structure is not identified. A multi- channel detector, which is under development, will help us clarify more precise structure of the potential profile and fluctuation.
From the point of view of turbulence measurement, the signal-to-noise ratio is not sufficient to detect the turbulent potential fluctuation. The development of a high-power negative ion source and more efficient detector is neces- sary, and it is under way.
Fig. 4. (a) Spectrogram of the magnetic fluctuation. (b) Measurement position of the HIBP. (c) Spectrogram of the electrostatic potential fluctuation.
Fluctuations are also observed in the potential spectrogram measured using the HIBP, where the measurement posi-
Fig. 5. Spatial distribution of coherent potential fluctuations in each frequency range. The data from 1.75 to 1.80 s are analyzed in Fig. 4.
Summary
An HIBP using a 3-MV tandem accelerator has been installed in LHD and used to measure the electrostatic potential in core plasmas.
The measured potential profile shows the transition from weak negative Er to strong positive Er, and the behavior is qualitatively accounted for by neoclassical theory.
From the point of view of the fluctuation measurement, the electrostatic potential fluctuations relating to the RSAE and the energetic particle-induced GAM are mea- sured locally and directly, and their spatial distributions are clarified.
Direct measurement of the potential and its fluctuation by the LHD HIBP will contribute to better understandings of plasma confinement.
References
T. Ido et al., Plasma Fusion Res. 3, 031 (2008).
A. Shimizu et al., Plasma Fusion Res. (submitted, 2009).
S. Kubo et al., Paper EX/P6-14 presented at the 22nd IAEA Fusion Energy Conference, Geneve, 2008 (http:/
/www-pub.iaea.org/MTCD/Meetings/FEC2008/
ex_p6-14.pdf)
K. Toi et al., Paper EX/P8-4 presented at the 22nd IAEA Fusion Energy Conference, Geneve, 2008 (http:/
/www-pub.iaea.org/MTCD/Meetings/FEC2008/
ex_p8-4.pdf)
T. Ido et al., Rev. Sci. Instrum. 79, 10F318 (2008).
Ste arator News -10- June 2009