Effect of neoclassical poloidal viscosity and resonant magnetic perturbation on the response of the m/n=1/1 magnetic island in LHD B. Huang 1, S. Satake 2, R. Kanno 2, Y. Narushima 2, S. Sakakibara 2, S. Ohdachi 2, and LHD experiment group 1 The Graduate University for Advanced Studies[SOKENDAI] 2 National Institute for Fusion Science Plasma conference 214, Nov, 181, 214 B. Huang et. al. (SOKENDAI) 1 / 17
Outline 1 Abstract 2 The Motivation of this work 3 Simulation with FORTEC-3D 4 Simulation Results Radial Profile of NPV and E r 5 Simulation Results Radial Profile of NPV and E r NPV density dependence on R-axis ( @ ι 1 surface) Depandence of NPV on E r and Collisionality 6 Simulation Results with RMP NPV density on R-axis Depandence of NPV (each (m,n) mode) 7 Conclusion and Future work B. Huang et. al. (SOKENDAI) 2 / 17
Abstract We employ FORTEC-3D to investigate the dependence of neoclassical poloidal viscosity(npv) on magnetic configuration of LHD, and the effect of resonant magnetic perturbation (RMP) on NPV. For the m/n = 1/1 island formation in LHD, the threshold of RMP amplitude depends on the magnetic axis position in LHD. On the other hand, neoclassical transport theory predicts that the NPV also correlates to the magnetic axis position. By δf simulation, we investigate NPV variation in LHD plasmas with plasma profiles and ambipolar electric field E r. Thus, we study m/n=1/1 RMP effect on NPV. B. Huang et. al. (SOKENDAI) 3 / 17
RMP pentrates plasma as its amplitude exceeds the threshold. m/n = 1/1 RMP in LHD 1 RMP penetrate plasma as perturbation amplitude exceeds threshold. 2 The threshold correlates to magnetic axis position. 3 The Neoclassical Poloidal viscosity (NPV) also depends on magnetic axis position. With LHD experiment data (temperature, density, R ax ), we investigate the basic dependence of NPV in LHD. B. Figure Huang et. 1: al. Rax-RMP (SOKENDAI) amplitude 4 / 17
Simulate NPV with FORTEC-3D We employ drift-kinetic equation for distribution function δf(x,v) f(x,v) f m (x,v) δf t +(v +v D ) δf + v f v C T(δf) = (v D + v v )f m +Pf m. Taking the moment of (1) (1) t mnu e θ = e θ P +e nu Φ (2) NPV is difined as e θ P δp B. (3) B θ FORTEC-3D simulates drift kinetic Eq. and evaluate the δp δp = δp +δp δp = d 3 v m 2 v 2 δf, δp = d 3 v mv 2 δf B. Huang et. al. (SOKENDAI) 5 / 17
Simulation Result We scan the magnetic axis from 3.55[m] to 3.8[m] for ramped-up/down RMP experiments respectively as we show in the fig.(1). In the fig.(2) We investigate NPV amplitude and E r radial profiles without RMP. We simulate the ambipolar radial electric field E r on flux surface by using Γ i (from FORTEC-3D) and Γ e (from GSRAKE[4]). For the investigation of NPV without RMP effect, we collect the data when RMP was shielded. B. Huang et. al. (SOKENDAI) 6 / 17
Simulation Results Scan the magnet axis from 3.55[m] to 3.8[m] by FORTEC-3D NPV[N/m 2 ] NPV [N/m 2 ].5 -.5-1 -1.5 LHD RMP-ramp-up Er 116765, R=3.55-6 116765, R=3.55.5 116761, R=3.6 118891, R=3.6 118891, R=3.6-8 116761, R=3.6-3 116738, R=3.65 116738, R=3.65-3.5 116735, R=3.7-1 116735, R=3.7 121762, R=3.75 121762, R=3.75 116727, R=3.8 116727, R=3.8-4 -12.3.4.5.6.7.8.9 1.3.4.5.6.7.8.9 1 fig(2a) Radial profile of NPV in ramp-up cases fig.(2b) Radial profile of E r in ramp-up cases -.5-1 -1.5 LHD RMP-ramp-down 2-4 LHD RMP-ramp-up -4.5-3 -6 R365 (116785) R365 (116785) -3.5 R365 (116788) -8 R365 (116788) -4 R37 (116791) R37 (116791) R375 (116795) -1 R375 (116795) -4.5 R38 (11684) R38 (11684) R36 (118888) R36 (118888) -5-12.3.4.5.6.7.8.9 1.3.4.5.6.7.8.9 1 fig(2c) Radial profile of NPV in ramp-down cases fig.(2d) Radial profile of E r in ramp-down cases E r [kv/m] 2 LHD RMP-ramp-down B. Huang et. al. (SOKENDAI) Figure 2: Radial Profile of NPV and E. 7 / 17
Simulation Results NPV density depends on E r ( @ ι 1 surface) NPV[norm. by r * dp/dr] -.1 -.8 -.6 -.4 -.2 116795 118888 116788 121781 116791 11684 116738 116785 116735 116761 116727 118891 116765 up down 121762 NPV[norm. by r * dp/dr] -.1 -.8 -.6 -.4 -.2 116727 118891 121762 116765 up down 118888 116795 121781 116788 11684 116791 116738 116785 116735 116761.4.5.6.7.8.9 1 1.1 1.2 1.3 1.4 nu* -5-4.5-4 -3.5-3.5-1.5-1 -.5 Er Figure 3: ν -NPV. Figure 4: Dependence of NPV on E r B. Huang et. al. (SOKENDAI) 8 / 17
Simulation Results NPV depends on R ax but Collisionality NPV[norm. by r * dp/dr] -.1 -.8 -.6 -.4 118888 116788 116761 -.2 118891 116727 116765 3.5 3.55 3.6 3.65 3.7 3.75 3.8 3.85 Rax up down 116795 121781 116791 11684 116738 116785 116735 121762 Figure 5: Dependence of NPV on R ax We expected that NPV 1 nu,but in fig.(3), the collisionality is not the major factor dominating the amplitude of NPV. As neoclassical theory expected, the fig.(4) shows that the NPV amplitude is proportional to 1/ E r. In the fig.(5), NPV increases as R ax increases but it is still determined by plasma profile, too. For example, density and ion temperature may be the potential factor. For shot#116727, NPV amplitude decreases due to large E r amplitude. For shot#11888, NPV is increased but E r amplitude is small. B. Huang et. al. (SOKENDAI) 9 / 17
Simulation Results with RMP NPV density on R-axis NPV [N/m 2 ] -.7 -.6 -.5 -.4 -.3 LHD 116765t5178 (R ax =3.55) m/n=1/1 RMP -.2 w/o RMP db=.1% -.1 db=.3% db=1.% db=3.%.5.6.7.8.9 1 Figure 6: Radial profile of NPV with RMP in a case R ax = 3.55 NPV [N/m 2 ] -1 -.8 -.6 -.4 -.2 LHD 118891t4848 (R ax =3.6) m/n=1/1 RMP (db=.1%) (db=.3%) (db=1.%) (db=3.%) iota=1.5.6.7.8.9 1 Figure 7: Radial profile of NPV with RMP in a case R ax = 3.6 B. Huang et. al. (SOKENDAI) 1 / 17
Simulation Results with RMP NPV density on R-axis NPV [N/m 2 ] -3.5-3.5 LHD 116735t5233 (R ax =3.7) m/n=1/1 RMP -1.5 w/o RMP db=.1% -1 db=.3% db=1.% db=3.% -.5.5.6.7.8.9 1 Figure 8: Radial profile of NPV with RMP in a case R ax = 3.7 In figs.(6), (7) and (8), we investigate the NPV with different m/n= 1/1 RMP amplitude using a simple model δb () 2. The RMP effect is obvious near m/n= 1/1 surface. We found that below 1% δb/b amplitude, NPV in LHD is almost unaffected by the RMP field. B. Huang et. al. (SOKENDAI) 11 / 17
Simulation Results Depandence of NPV (each (m,n) mode) (1,) NPV[N/m^2] -.8 -.9-1 -1.1-1.2-1.3-1.4-1.5-1.6 =.69 =.79 =.89, iota=1.1.1-1.2-1.3-1.4-1.5-1.6-1.7-1.8-1.9.1.2.1.1 -.8-1.2-1 -1.4-1.6-1.8.2.4.6.1.1 (1,1) NPV[N/m^2] -.1 -.2 -.3 -.4 -.5 -.6 -.7 -.8.1.1.6.4.2 -.2 -.4 -.6 -.8 -.1 -.12 -.14.1.1 -.1 -.2 -.3 -.4 -.5 -.6 -.7 3.55 3.6 3.7.1.1 (2,1) NPV[N/m^2] 1.4 1.3 1.2 1.1.9 1.8.7.6.5.4.3.1.1 1.3 1.25 1.2 1.15 1.1 1.5 1.1.1 2 1.8 1.6 1.4 1.2 1.8.6.4.1.1 delta B delta B delta B Figure 9: Depandence of NPV at B(1,), B(1,1) and B(2,1) mode B. Huang et. al. (SOKENDAI) 12 / 17
Simulation Results Depandence of NPV (each (m,n) mode) B amplitude.4.3.2.1 -.1 -.2 R=3.55 (1,) (2,1) (1,1) (3,1) R=3.6 (1,) (2,1) (1,1) (3,1) R=3.7 (1,) (2,1) (1,1) (3,1).1.2.3.4.5.6.7.8.9 1 Figure 1: B m,n amplitudes in R ax = 3.7, R ax = 3.65 and R ax = 3.55. In fig.1, B 1, and B 2,1 are main components.the tow components are changed a little due to the shift in Rax. By contrast, B 1,1 and B 3,1 are minor but changed a lot due to the shift in Rax. In fig.11, NPV 1, and NPV 2,1 sum is small to contribute to total NVP too much. In Fig.12, NPV 1,1 and NPV 3,1 sum is main contribution for NPV. B. Huang et. al. (SOKENDAI) 13 / 17
Simulation Results Depandence of NPV (each (m,n) mode) NPV[N/m 2 ] 2 1-1 NPV 1,,R=3.55 R=3.6 R=3.7 NPV 2,1,R=3.55-3 R=3.6 R=3.7-4 NPV 1, +NPV 2,1 R=3.55 R=3.6 R=3.7-5.1.2.3.4.5.6.7.8.9 1 Figure 11: NPV contribution of B(1,) and B(2,1) without RMP in R ax = 3.7, R ax = 3.65 and R ax = 3.55. NPV[N/m 2 ] 1.5 -.5-1 -1.5 NPV tot, R=3.55 R=3.6 R=3.7 NPV.5 (1,1) +NPV (3,1) R=3.55 R=3.6 R=3.7-3.1.2.3.4.5.6.7.8.9 1 Figure 12: NPV contribution of B(1,1) and B(3,1) without RMP in R ax = 3.7, R ax = 3.65 and R ax = 3.55. B. Huang et. al. (SOKENDAI) 14 / 17
Depandence of NPV (each (m,n) mode) We construct the NPV by e θ P = e θ P m,n B nδ m,n Q m,n. (4) m,n m,n 1 As expected, the RMP, mode B(1,1), affects surface ι 1 and the effect is proportional to RMP amplitude. 2 In the other surface (=.69 and.79), it is small and negligible to the RMP effect on mode B(1,1). The strong RMP influences the (1, )- and (2, 1)- modes and changes their amplitudes. This suggests that the strong RMP δb m,n couple with the pressure perturbation of the different modes, Q m,n. 3 Practically δb/b 1 4 in LHD but we employ δb/b 1 2 in our test. It is anticipated that the RMP field does not change NPV amplitude in the practical LHD experiment. B. Huang et. al. (SOKENDAI) 15 / 17
Conclusion and Future work Conclusion NPV amplitude becomes larger as R ax moves outward. Following the results, we are able to see the influence near the resonant surface (ι 1) if the RMP amplitude δb/b >.1. Practically RMP amplitude is δb/b 1 4 in LHD. We found that m/n = 1/1 RMP less than 1% cannot affects NPV. The strong RMP influences the mode B(1,) and B(2,1) and changes their amplitudes. Future work Study the relation between the NPV torque with the threshold RMP amplitude from the LHD experiments. Compare the E r profile obtained from the ambipolar condition with the measured. B. Huang et. al. (SOKENDAI) 16 / 17
Referece and Acknowledgment Referece [1] Sakakibara,et al.,(213). In Nuclear Fusion 53, p. 431. [2] Narushima, Y.,et al.,(211). In Nuclear Fusion 51, p. 833. [3] Satake, S.,et al.(211). In Plasma Physics and Controlled Fusion 53, p. 5418. [4] C D Beidler and W D D haeseleer (1995) Plasma Phys. Control. Fusion 37 463 Acknowledgment This work is pported by NIFS collaborative Research Programs NIFS13KNST51, and JSPS Grant-in-Aid for Young Scientists (B), No. 237681. Part of calculations was carried out using the HELIOS supercomputer at IFERC-CSC, under the ITER-BA collaboration implemented by Fusion for Energy and JAEA. This presentation received the financial support program for international conference presentations provided by the Course-by-Course Education Program of SOKENDAI. B. Huang et. al. (SOKENDAI) 17 / 17