- Email: [email protected]
Measurement of peripheral plasma turbulence using a fast camera in Heliotron J
Journal of Nuclear Materials 390–391 (2009) 432–435
Contents lists available at ScienceDirect
Journal of Nuclear Materials journal homepage: www.elsevier.com/locate/jnucmat
Measurement of peripheral plasma turbulence using a fast camera in Heliotron J N. Nishino a,*, T. Mizuuchi b, S. Kobayashi b, K. Nagasaki b, H. Okada b, F. Sano b, S. Yamamoto b, K. Kondo c a b c
Graduate School of Engineering, Hiroshima University, 1-4-1 Kagamiyama, Higashi-Hiroshima 739-8527, Japan Institute of Advanced Energy, Kyoto University, Japan Graduate School of Energy Science, Kyoto University, Japan
a r t i c l e PACS: 52.55.Hc 52.25.Fi 52.25.Gj
i n f o
a b s t r a c t Using a fast camera with a tangential view, filamentary structure in a peripheral region was observed during L and H-mode, and the apparent motion of filamentary structure across the magnetic field was also observed. In the L-mode the cross-field scale of the filamentary structure was smaller than that in the H-mode. In the L-mode the apparent motion was counterclockwise in a camera view. During L–H transition the apparent motion was stopped, and in the H-mode the apparent motion changed direction, and the speed of the apparent motion nearly doubled. If this apparent motion was due to Er B drift, Er should be positive in the L-mode and Er should be negative in the H-mode. To get the density threshold for the H-mode a directional gas puff was used, and it was found that the dithering phase appeared after low frequency MHD activity (1–2 kHz), even though the density was higher than that of the H-mode threshold. Ó 2009 Elsevier B.V. All rights reserved.
1. Introduction It is very important to study peripheral turbulence due to the relationship between the energy confinement and H-mode physics. Fast cameras had been installed in Heliotron J [1] to get information on the peripheral plasma behavior several years ago. In our previous work, using a combination method of a fast camera (Ultima-SE, Photron) and a small movable carbon target a structure of a low frequency (5–6 kHz) edge plasma oscillation in high electron density ECH (70 GHz, 0.3 MW) discharges was observed just after the L–H transition [2]. Also a combination of a fast camera (FXK4, NAC image technology), movable probe and directional gas puff technique gave us important information that the spatial profile of turbulent burst in the edge plasma were observed as filamentary structure in the L-mode [3–5]. Each burst of the ion saturation current of a peripheral probe corresponded to the filamentary structure hitting the probe. Recently, the turbulent structures during the L-mode and Hmode were observed with/without a directional gas puff [6]. It was clear that the spatial structure of peripheral turbulence during the L-mode was different from that of the H-mode. Larger crossfield scale of filamentary structure was observed in the H-mode, however, smaller cross-field scale of that was observed in the Lmode. Moreover, these filamentary structures sometimes became dark in the H-mode, and they were not detected by the fast camera.
* Corresponding author. E-mail address: [email protected] (N. Nishino). 0022-3115/$ - see front matter Ó 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.jnucmat.2009.01.119
Therefore, the apparent motion of the filamentary structure seemed to be related to the confinement mode. In this paper, the latest results of Heliotron J plasma measurement by the fast cameras are reported. 2. Experimental setup The tangential port was provided with the fast camera to observe peripheral plasma behavior easily. Fig. 1 shows the locations of the tangential port (this experiment, green) and horizontal port (past experiments, blue) for the fast camera and the locations of other system (ECH, NBI, ICRF, etc.). Fig. 2 shows typical image of the fast camera from the tangential port. The image is the initial discharge phase of Heliotron J plasma. The bright string in the image became thicker and darker, and the equilibrium state of Heliotron J plasma was obtained within a few ms after this image. Two top ports of Heliotron J vacuum vessel and ICRF antenna are seen. 3. Results and discussion In high density ECH (70 GHz, 0.3 MW) discharges the abrupt drops of Da signal were observed and the H-mode plasmas were obtained. In these shots the reproducibility was very good, and typical waveforms of H-mode are shown in Fig. 3. H-mode began at 252.8 ms and the diamagnetic signal increased rapidly from the L–H transition. Typical period of the H-mode in these shots was about 5 ms. During H-mode the electron density also increased rapidly, and this increment was not controllable. There-
433
N. Nishino et al. / Journal of Nuclear Materials 390–391 (2009) 432–435
fore, the hot plasma was extinguished with the radiation collapse after H-mode in several shots. Fig. 4(a) and (b) show the tangential camera images of the Lmode plasma and the H-mode plasma, respectively. The emission from the peripheral plasma region was relatively clear in the Hmode (inside the white dashed ellipses), and the size of the emission region of the H-mode plasma was somewhat larger than that of the L-mode in the images. This implied that the electron temperature was higher in H-mode and it pushed Ha emission region outward. From this tangential view the filamentary structures in the peripheral region were observed in both modes (white arrows). These filamentary structures are widely seen in the various machines such as tokamaks/ST [7–9], and it is believed that they are parallel to the magnetic field. In these H-mode experiments, motion of the filamentary structures perpendicular to the magnetic
Wb (A.U.) I (A.U.) P (A.U.)
Fig. 1. Location of tangential view for the fast camera in Heliotron J.
field was recognized in the camera images in the L- and H-modes and motion was stopped during the L–H transition. The direction of motion during L-mode is opposite to that of H-mode. If the motion would be poloidal rotation, the rotation direction was counterclockwise during the L-mode, and the rotation direction during the H-mode was clockwise. The directions of the magnetic field and the camera view were counterclockwise at the horizontal plane (see Fig. 1) in these shots. Therefore, if the filamentary structures would rotate by Er B drift, Er should be positive in the L-mode and Er should be negative in the H-mode. These results were consistent with the past results on the H-mode review in Heliotron J [10]. To see this motion easily the phase image of selected frequency component are used for the author group [2,5,6]. Using time-dependent FFT analysis for each pixel data spectra and its phase are calculated for each pixel. Therefore, the images of phase for every frequency component are obtained. This phase image means 2D k–x (x is corresponding to the selected frequency) diagram. In other words the phase image shows 2D the wave propagation map. If the motions of phase image for many frequency components are almost similar, this motion will be plasma motion. The low frequency of 8.75 kHz signal was relatively strong in the power spectra of the camera pixel data
0.15 0.1 0.05 0 0.4 0.3 0.2 0.1 0.2
PECH
Dalpha
Diamag
0.1 0
200
300
time (ms) Fig. 3. Typical waveform of H-mode in ECH discharge. Upper: ECH power, middle: Da signal, lower: diamagnetic signal.
Fig. 2. Tangential view from the fast camera. This is a part of the original image (448 368 pixels, 10,000 FPS), and it was rotated to meet the major radius to right direction.
Fig. 4. Tangential views of the L-mode and H-mode plasmas: (a) L-mode and (b) Hmode (both views are not rotated).
N. Nishino et al. / Journal of Nuclear Materials 390–391 (2009) 432–435
#28407
Fast Langmuir Probe H-mode
1.5 1 0.5 0 200
220
240 time (ms)
260
280
Fig. 6. Ion saturation current of the fast Langmuir probe (ECH plasma).
this time (see Fig. 3). From 240 ms to transition time the filamentary structures were seen to rotate mainly counterclockwise but sometime those structures stop to rotate and/or those structures rotate clockwise for very short period (a few frames, <1 ms). Therefore, in this period it seems the L-mode plasma dither to shift to the H-mode plasma. This phenomenon was named as ‘dithering phase’ in Ref. [10]. In the dithering phase the diamagnetic signal increases with time (see Fig. 3) and the electron density also increased (not shown due to the signal jump after the H-mode). The increments of both signals were about the same in an average of these H-mode shots. Therefore, the gross temperature seems to be constant, and the increment of plasma pressure is due to mainly the increment of the plasma density. Usually, the gas puff (normal piezo valve) was used to get the threshold electron density for H-mode plasma; however, it is not so easy to adjust the gas quantity correctly due to recycling. To raise the electron density the directional gas puff (DGP) was tried to use in the serial experiment with ECH (70 GHz, 0.35 MW) and NBI (23 kV, 0.25 MW [absorbed power]). Fig. 7 shows typical waveform of plasma parameters. The trigger of DGP is 200 ms, and after a few ms delay the gas was injected to the main plasma. The camera view shows that DGP provided working gas at the local position (not shown in the figure). The light emission was affected by DGP
-3 -2
#30126 Line Averaged Electron Density
-1 5 4 3 2 1 0 0.02
Power (a.u.)
0 0.2
NBI ECRH
0.1
I (a.u.)
0.020 OV
0.01
0.01
I (a.u.)
Power (a.u.)DENSITY (E19m-2)
during the L-mode and no strong peak was obtained in the power spectra during the H-mode (not shown in the figure). To evaluate the rotation effect with two-dimension this low fluctuation frequency was chosen to see how the filamentary structures in Heliotron J plasma behaves during the L- and H-modes. Fig. 5 shows one of two-dimensional phase images during L-mode, H-mode and L–H transition. The frequency of the fluctuation signal depicted on each frame was 8.75 kHz and the fast camera was operated with 40 000 frames per second (FPS). We used 32 frames for time-dependent FFT. The color is corresponding to the phase p to p. The same color region has the same phase. These images show the propagation phenomena at the selected frequency. The left picture shows phase image during the L-mode, the center one shows the phase image during the L–H transition. The right one shows phase image during the H-mode. The dashed ellipse shows the top region of vacuum vessel. As already mentioned, the emission from the peripheral plasma was easily measured in this region, and the filamentary structures in the H-mode were seen to be somewhat wider than that of the L-mode. A rotation speed of 1–2 km/s was roughly estimated in the images during the H-mode and it was about half speed during the L-mode. The toroidal magnetic field is about 1.5 T. Therefore, according to this rough estimation from the images Er is 2 kV/m during the H-mode, and Er is +1 kV/m during the L-mode. These results were consistent with Ref. [10]. The period of the L–H transition was also deduced in the images. The number of the camera frames when plasma rotation stopped is only one at 40 000 frames per second. That means the transition period in Heliotron J plasma was below 50 ls (two frames), because the shutter was open in these experiments. Fig. 6 shows the ion saturation current of the movable probe of the same shot. The location of this probe is shown in Fig. 1. The burst signal of the ion saturation current is the maximum around 230 ms. From the power spectra of this ion saturation current (not shown in the figure) the fluctuation in all frequency components was suppressed dramatically during the H-mode, however, the fluctuation was already suppressed from 240 ms, especially in the low frequency range. Da signal also decreased slightly from
I. S. (a.u.)
434
CIII 0 2
0 Visible monitor
I (a.u.)
1
Fig. 5. 2D-Phase images during the L-mode, the L–H transition, and the H-mode. Phase image in dashed circle region moves counterclockwise during the L-mode. During the L–H transition the motion of the phase image in dashed circle region become slow, and at the transition the rotation of the phase image stops. After the transition phase image begins to rotate but the direction is clockwise [6].
0 -0.4 Diamag signal -0.3 -0.2 -0.1 0 180 190 200
210 time (ms)
220
230
240
Fig. 7. Typical waveform of the directional gas puff experiment (ECH + NBI).
N. Nishino et al. / Journal of Nuclear Materials 390–391 (2009) 432–435
0.4 0.3 0.2 0.1 0 2 1 0 -1 -2 200
Fast Langmuir Probe
435
The feature of the dithering phase looks like that of the elmy Hmode in tokamaks. It is needed to know what is the difference between the dithering phase in Heliotron J and the elmy H-mode in tokamaks. To compare these phenomena carefully the additional experiments are planned again.
Magnetic Probe 4. Conclusion
210
220
230
time (ms) Fig. 8. Ion saturation current and magnetic probe signal in DGP experiment (ECH + NBI).
for 10 ms from camera images, especially almost plasma region was bright during the initial few ms (also see visible monitor signal in Fig. 7). Fig. 8 shows the ion saturation current and magnetic probe signals in this shot. From the power spectra of both signals the fluctuation was almost suppressed just after DGP was injected, and its state lasted for a few ms. It is thought that the improved mode appeared, however, Da emission did not drop. Actually there were several shots that Da emission dropped. It is easy to recognize the latter as H-mode, but the former is difficult to distinguish the H-mode and the other. In either case after the improved mode ends, then the low frequency MHD activity (LFMHD) was found at 1–2 kHz frequency region. The confinement energy sometimes was lowered by this LFMHD. Typically, LFMHD continued for 10 ms. 2D phase images from camera data show that in many shots plasma recovered the dithering phase after LFMHD ends, even though the electron density is higher than that of H-mode threshold. This experiment was a first attempt to raise the density using DGP. Therefore, if DGP will be used more sophisticatedly for gas quantity, it is hoped that LFMHD is avoided. As mentioned above, if the H-mode was once gotten, the electron density increased rapidly in Heliotron J plasma and it was uncontrollable. On the contrary, in the dithering phase the electron density seemed to be controllable and this phase was longer than the period of the Hmode plasmas. Therefore, to get high-pressure (or high b) plasmas this dithering phase might be preferable in Heliotron J. In these experiments the motion of the filamentary structures during the L- and H-modes and the dithering phase was identified.
Using the fast camera, peripheral turbulence of Heliotron J plasma was observed successfully. In particular, the apparent motion of the filamentary structure near the plasma edge was seen. If this motion would be poloidal rotation, the direction of the motion suggests that Er was negative in the H-mode and Er was positive in the L-mode. Moreover, the deduced Er value was consistent with the results of the past H-mode experiments. The dithering phase was also identified and the transition period can be estimated by the fast camera images. Using the directional gas puff (GDP) it is found the improved mode continued for a few ms just after DGP. Also, the low frequency MHD activity is found after the improved mode. This MHD activity was extinguished as the dithering phase appeared, even though the density was higher than that of H-mode threshold. Acknowledgements This work is performed with the support and under the auspices of the Bi-directional Collaborative Research Program (NIFS04KUHL006). The author thanks to everyone who are concerned with this experiment, especially Photron inc., for borrowing new fast camera. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10]
T. Obiki, T. Mizuuchi, K. Nagasaki, et al., Nucl. Fusion 41 (2001) 883. N. Nishino, K. Takahashi, H. Kawazome, et al., J. Nucl. Mater. 337 (2005) 1073. N. Nishino, T. Mizuuchi, K. Kondo, et al., J. Nucl. Mater. 363–365 (2007) 628. T. Mizuuchi et al., J. Nucl. Mater. 337–339 (2005) 332. N. Nishino et al., J. Plasma Fusion Res. 1 (2006) 035. N. Nishino, T. Mizuuchi, et al., Plasma Fusion Res. 3 (2008) S1023. S.J. Zweben, D. Manos, R.V. Bundy, et al., J. Nucl. Mater. 145 (1987) 250. R.J. Maqueda, G.A. Wurden, S.J. Zweben, et al., Rev. Sci. Instrum. 72 (2001) 931. S.J. Zweben, R.J. Maqueda, G.P. Stotler, et al., Nucl. Fusion 44 (2004) 134. F. Sano, T. Mizuuchi, K. Kondo, et al., Nucl. Fusion 45 (2005) 1557.
Contents lists available at ScienceDirect
Journal of Nuclear Materials journal homepage: www.elsevier.com/locate/jnucmat
Measurement of peripheral plasma turbulence using a fast camera in Heliotron J N. Nishino a,*, T. Mizuuchi b, S. Kobayashi b, K. Nagasaki b, H. Okada b, F. Sano b, S. Yamamoto b, K. Kondo c a b c
Graduate School of Engineering, Hiroshima University, 1-4-1 Kagamiyama, Higashi-Hiroshima 739-8527, Japan Institute of Advanced Energy, Kyoto University, Japan Graduate School of Energy Science, Kyoto University, Japan
a r t i c l e PACS: 52.55.Hc 52.25.Fi 52.25.Gj
i n f o
a b s t r a c t Using a fast camera with a tangential view, filamentary structure in a peripheral region was observed during L and H-mode, and the apparent motion of filamentary structure across the magnetic field was also observed. In the L-mode the cross-field scale of the filamentary structure was smaller than that in the H-mode. In the L-mode the apparent motion was counterclockwise in a camera view. During L–H transition the apparent motion was stopped, and in the H-mode the apparent motion changed direction, and the speed of the apparent motion nearly doubled. If this apparent motion was due to Er B drift, Er should be positive in the L-mode and Er should be negative in the H-mode. To get the density threshold for the H-mode a directional gas puff was used, and it was found that the dithering phase appeared after low frequency MHD activity (1–2 kHz), even though the density was higher than that of the H-mode threshold. Ó 2009 Elsevier B.V. All rights reserved.
1. Introduction It is very important to study peripheral turbulence due to the relationship between the energy confinement and H-mode physics. Fast cameras had been installed in Heliotron J [1] to get information on the peripheral plasma behavior several years ago. In our previous work, using a combination method of a fast camera (Ultima-SE, Photron) and a small movable carbon target a structure of a low frequency (5–6 kHz) edge plasma oscillation in high electron density ECH (70 GHz, 0.3 MW) discharges was observed just after the L–H transition [2]. Also a combination of a fast camera (FXK4, NAC image technology), movable probe and directional gas puff technique gave us important information that the spatial profile of turbulent burst in the edge plasma were observed as filamentary structure in the L-mode [3–5]. Each burst of the ion saturation current of a peripheral probe corresponded to the filamentary structure hitting the probe. Recently, the turbulent structures during the L-mode and Hmode were observed with/without a directional gas puff [6]. It was clear that the spatial structure of peripheral turbulence during the L-mode was different from that of the H-mode. Larger crossfield scale of filamentary structure was observed in the H-mode, however, smaller cross-field scale of that was observed in the Lmode. Moreover, these filamentary structures sometimes became dark in the H-mode, and they were not detected by the fast camera.
* Corresponding author. E-mail address: [email protected] (N. Nishino). 0022-3115/$ - see front matter Ó 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.jnucmat.2009.01.119
Therefore, the apparent motion of the filamentary structure seemed to be related to the confinement mode. In this paper, the latest results of Heliotron J plasma measurement by the fast cameras are reported. 2. Experimental setup The tangential port was provided with the fast camera to observe peripheral plasma behavior easily. Fig. 1 shows the locations of the tangential port (this experiment, green) and horizontal port (past experiments, blue) for the fast camera and the locations of other system (ECH, NBI, ICRF, etc.). Fig. 2 shows typical image of the fast camera from the tangential port. The image is the initial discharge phase of Heliotron J plasma. The bright string in the image became thicker and darker, and the equilibrium state of Heliotron J plasma was obtained within a few ms after this image. Two top ports of Heliotron J vacuum vessel and ICRF antenna are seen. 3. Results and discussion In high density ECH (70 GHz, 0.3 MW) discharges the abrupt drops of Da signal were observed and the H-mode plasmas were obtained. In these shots the reproducibility was very good, and typical waveforms of H-mode are shown in Fig. 3. H-mode began at 252.8 ms and the diamagnetic signal increased rapidly from the L–H transition. Typical period of the H-mode in these shots was about 5 ms. During H-mode the electron density also increased rapidly, and this increment was not controllable. There-
433
N. Nishino et al. / Journal of Nuclear Materials 390–391 (2009) 432–435
fore, the hot plasma was extinguished with the radiation collapse after H-mode in several shots. Fig. 4(a) and (b) show the tangential camera images of the Lmode plasma and the H-mode plasma, respectively. The emission from the peripheral plasma region was relatively clear in the Hmode (inside the white dashed ellipses), and the size of the emission region of the H-mode plasma was somewhat larger than that of the L-mode in the images. This implied that the electron temperature was higher in H-mode and it pushed Ha emission region outward. From this tangential view the filamentary structures in the peripheral region were observed in both modes (white arrows). These filamentary structures are widely seen in the various machines such as tokamaks/ST [7–9], and it is believed that they are parallel to the magnetic field. In these H-mode experiments, motion of the filamentary structures perpendicular to the magnetic
Wb (A.U.) I (A.U.) P (A.U.)
Fig. 1. Location of tangential view for the fast camera in Heliotron J.
field was recognized in the camera images in the L- and H-modes and motion was stopped during the L–H transition. The direction of motion during L-mode is opposite to that of H-mode. If the motion would be poloidal rotation, the rotation direction was counterclockwise during the L-mode, and the rotation direction during the H-mode was clockwise. The directions of the magnetic field and the camera view were counterclockwise at the horizontal plane (see Fig. 1) in these shots. Therefore, if the filamentary structures would rotate by Er B drift, Er should be positive in the L-mode and Er should be negative in the H-mode. These results were consistent with the past results on the H-mode review in Heliotron J [10]. To see this motion easily the phase image of selected frequency component are used for the author group [2,5,6]. Using time-dependent FFT analysis for each pixel data spectra and its phase are calculated for each pixel. Therefore, the images of phase for every frequency component are obtained. This phase image means 2D k–x (x is corresponding to the selected frequency) diagram. In other words the phase image shows 2D the wave propagation map. If the motions of phase image for many frequency components are almost similar, this motion will be plasma motion. The low frequency of 8.75 kHz signal was relatively strong in the power spectra of the camera pixel data
0.15 0.1 0.05 0 0.4 0.3 0.2 0.1 0.2
PECH
Dalpha
Diamag
0.1 0
200
300
time (ms) Fig. 3. Typical waveform of H-mode in ECH discharge. Upper: ECH power, middle: Da signal, lower: diamagnetic signal.
Fig. 2. Tangential view from the fast camera. This is a part of the original image (448 368 pixels, 10,000 FPS), and it was rotated to meet the major radius to right direction.
Fig. 4. Tangential views of the L-mode and H-mode plasmas: (a) L-mode and (b) Hmode (both views are not rotated).
N. Nishino et al. / Journal of Nuclear Materials 390–391 (2009) 432–435
#28407
Fast Langmuir Probe H-mode
1.5 1 0.5 0 200
220
240 time (ms)
260
280
Fig. 6. Ion saturation current of the fast Langmuir probe (ECH plasma).
this time (see Fig. 3). From 240 ms to transition time the filamentary structures were seen to rotate mainly counterclockwise but sometime those structures stop to rotate and/or those structures rotate clockwise for very short period (a few frames, <1 ms). Therefore, in this period it seems the L-mode plasma dither to shift to the H-mode plasma. This phenomenon was named as ‘dithering phase’ in Ref. [10]. In the dithering phase the diamagnetic signal increases with time (see Fig. 3) and the electron density also increased (not shown due to the signal jump after the H-mode). The increments of both signals were about the same in an average of these H-mode shots. Therefore, the gross temperature seems to be constant, and the increment of plasma pressure is due to mainly the increment of the plasma density. Usually, the gas puff (normal piezo valve) was used to get the threshold electron density for H-mode plasma; however, it is not so easy to adjust the gas quantity correctly due to recycling. To raise the electron density the directional gas puff (DGP) was tried to use in the serial experiment with ECH (70 GHz, 0.35 MW) and NBI (23 kV, 0.25 MW [absorbed power]). Fig. 7 shows typical waveform of plasma parameters. The trigger of DGP is 200 ms, and after a few ms delay the gas was injected to the main plasma. The camera view shows that DGP provided working gas at the local position (not shown in the figure). The light emission was affected by DGP
-3 -2
#30126 Line Averaged Electron Density
-1 5 4 3 2 1 0 0.02
Power (a.u.)
0 0.2
NBI ECRH
0.1
I (a.u.)
0.020 OV
0.01
0.01
I (a.u.)
Power (a.u.)DENSITY (E19m-2)
during the L-mode and no strong peak was obtained in the power spectra during the H-mode (not shown in the figure). To evaluate the rotation effect with two-dimension this low fluctuation frequency was chosen to see how the filamentary structures in Heliotron J plasma behaves during the L- and H-modes. Fig. 5 shows one of two-dimensional phase images during L-mode, H-mode and L–H transition. The frequency of the fluctuation signal depicted on each frame was 8.75 kHz and the fast camera was operated with 40 000 frames per second (FPS). We used 32 frames for time-dependent FFT. The color is corresponding to the phase p to p. The same color region has the same phase. These images show the propagation phenomena at the selected frequency. The left picture shows phase image during the L-mode, the center one shows the phase image during the L–H transition. The right one shows phase image during the H-mode. The dashed ellipse shows the top region of vacuum vessel. As already mentioned, the emission from the peripheral plasma was easily measured in this region, and the filamentary structures in the H-mode were seen to be somewhat wider than that of the L-mode. A rotation speed of 1–2 km/s was roughly estimated in the images during the H-mode and it was about half speed during the L-mode. The toroidal magnetic field is about 1.5 T. Therefore, according to this rough estimation from the images Er is 2 kV/m during the H-mode, and Er is +1 kV/m during the L-mode. These results were consistent with Ref. [10]. The period of the L–H transition was also deduced in the images. The number of the camera frames when plasma rotation stopped is only one at 40 000 frames per second. That means the transition period in Heliotron J plasma was below 50 ls (two frames), because the shutter was open in these experiments. Fig. 6 shows the ion saturation current of the movable probe of the same shot. The location of this probe is shown in Fig. 1. The burst signal of the ion saturation current is the maximum around 230 ms. From the power spectra of this ion saturation current (not shown in the figure) the fluctuation in all frequency components was suppressed dramatically during the H-mode, however, the fluctuation was already suppressed from 240 ms, especially in the low frequency range. Da signal also decreased slightly from
I. S. (a.u.)
434
CIII 0 2
0 Visible monitor
I (a.u.)
1
Fig. 5. 2D-Phase images during the L-mode, the L–H transition, and the H-mode. Phase image in dashed circle region moves counterclockwise during the L-mode. During the L–H transition the motion of the phase image in dashed circle region become slow, and at the transition the rotation of the phase image stops. After the transition phase image begins to rotate but the direction is clockwise [6].
0 -0.4 Diamag signal -0.3 -0.2 -0.1 0 180 190 200
210 time (ms)
220
230
240
Fig. 7. Typical waveform of the directional gas puff experiment (ECH + NBI).
N. Nishino et al. / Journal of Nuclear Materials 390–391 (2009) 432–435
0.4 0.3 0.2 0.1 0 2 1 0 -1 -2 200
Fast Langmuir Probe
435
The feature of the dithering phase looks like that of the elmy Hmode in tokamaks. It is needed to know what is the difference between the dithering phase in Heliotron J and the elmy H-mode in tokamaks. To compare these phenomena carefully the additional experiments are planned again.
Magnetic Probe 4. Conclusion
210
220
230
time (ms) Fig. 8. Ion saturation current and magnetic probe signal in DGP experiment (ECH + NBI).
for 10 ms from camera images, especially almost plasma region was bright during the initial few ms (also see visible monitor signal in Fig. 7). Fig. 8 shows the ion saturation current and magnetic probe signals in this shot. From the power spectra of both signals the fluctuation was almost suppressed just after DGP was injected, and its state lasted for a few ms. It is thought that the improved mode appeared, however, Da emission did not drop. Actually there were several shots that Da emission dropped. It is easy to recognize the latter as H-mode, but the former is difficult to distinguish the H-mode and the other. In either case after the improved mode ends, then the low frequency MHD activity (LFMHD) was found at 1–2 kHz frequency region. The confinement energy sometimes was lowered by this LFMHD. Typically, LFMHD continued for 10 ms. 2D phase images from camera data show that in many shots plasma recovered the dithering phase after LFMHD ends, even though the electron density is higher than that of H-mode threshold. This experiment was a first attempt to raise the density using DGP. Therefore, if DGP will be used more sophisticatedly for gas quantity, it is hoped that LFMHD is avoided. As mentioned above, if the H-mode was once gotten, the electron density increased rapidly in Heliotron J plasma and it was uncontrollable. On the contrary, in the dithering phase the electron density seemed to be controllable and this phase was longer than the period of the Hmode plasmas. Therefore, to get high-pressure (or high b) plasmas this dithering phase might be preferable in Heliotron J. In these experiments the motion of the filamentary structures during the L- and H-modes and the dithering phase was identified.
Using the fast camera, peripheral turbulence of Heliotron J plasma was observed successfully. In particular, the apparent motion of the filamentary structure near the plasma edge was seen. If this motion would be poloidal rotation, the direction of the motion suggests that Er was negative in the H-mode and Er was positive in the L-mode. Moreover, the deduced Er value was consistent with the results of the past H-mode experiments. The dithering phase was also identified and the transition period can be estimated by the fast camera images. Using the directional gas puff (GDP) it is found the improved mode continued for a few ms just after DGP. Also, the low frequency MHD activity is found after the improved mode. This MHD activity was extinguished as the dithering phase appeared, even though the density was higher than that of H-mode threshold. Acknowledgements This work is performed with the support and under the auspices of the Bi-directional Collaborative Research Program (NIFS04KUHL006). The author thanks to everyone who are concerned with this experiment, especially Photron inc., for borrowing new fast camera. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10]
T. Obiki, T. Mizuuchi, K. Nagasaki, et al., Nucl. Fusion 41 (2001) 883. N. Nishino, K. Takahashi, H. Kawazome, et al., J. Nucl. Mater. 337 (2005) 1073. N. Nishino, T. Mizuuchi, K. Kondo, et al., J. Nucl. Mater. 363–365 (2007) 628. T. Mizuuchi et al., J. Nucl. Mater. 337–339 (2005) 332. N. Nishino et al., J. Plasma Fusion Res. 1 (2006) 035. N. Nishino, T. Mizuuchi, et al., Plasma Fusion Res. 3 (2008) S1023. S.J. Zweben, D. Manos, R.V. Bundy, et al., J. Nucl. Mater. 145 (1987) 250. R.J. Maqueda, G.A. Wurden, S.J. Zweben, et al., Rev. Sci. Instrum. 72 (2001) 931. S.J. Zweben, R.J. Maqueda, G.P. Stotler, et al., Nucl. Fusion 44 (2004) 134. F. Sano, T. Mizuuchi, K. Kondo, et al., Nucl. Fusion 45 (2005) 1557.