NSTX-U Supported by Physical Characteristics of Neoclassical Toroidal Viscosity in Tokamaks for Rotation Control and the Evaluation of Plasma Response Coll of Wm & Mary S. A. Sabbagh 1 , R.E. Bell 2 , T.E. Evans 3 , N. Columbia U Culham Sci Ctr CompX York U Ferraro 3 , I.R. Goumiri 4 , Y.M. Jeon 5 , W.H. Ko 5 , Y.S. General Atomics Chubu U FIU Fukui U INL Park 1 , K.C. Shaing 6 , Y. Sun 7 , J.W. Berkery 1 , D.A. Hiroshima U Johns Hopkins U Hyogo U LANL Gates 2 , S.P. Gerhardt 2 , S.H. Hahn 5 , C.W. Rowley 4 Kyoto U LLNL Kyushu U Lodestar Kyushu Tokai U MIT 1 Department of Applied Physics, Columbia University, New York, NY NIFS Lehigh U 2 Princeton Plasma Physics Laboratory, Princeton, NJ Niigata U Nova Photonics U Tokyo ORNL 3 General Atomics, San Diego, CA JAEA PPPL 4 Princeton University, Princeton, NJ Inst for Nucl Res, Kiev Princeton U Ioffe Inst Purdue U 5 National Fusion Research Institute, Daejeon, Republic of Korea TRINITI SNL 6 National Cheng Kung University, Tainan, Taiwan Chonbuk Natl U Think Tank, Inc. UC Davis NFRI 7 ASIPP, Hefei Anhui, China UC Irvine KAIST UCLA POSTECH 25th IAEA Fusion Energy Conference UCSD Seoul Natl U U Colorado ASIPP U Illinois October 14th, 2014 CIEMAT U Maryland FOM Inst DIFFER U Rochester ENEA, Frascati St. Petersburg, U Tennessee CEA, Cadarache U Tulsa IPP, Jülich U Washington Russian Federation IPP, Garching U Wisconsin ASCR, Czech Rep X Science LLC V1.9m
The physical characteristics of NTV investigated in tokamaks for rotation control and the evaluation of plasma response Motivation Low magnitude ( d B/B 0 ~ O (10 -3 )) 3D magnetic fields are used favorably used in tokamaks (e.g. ELM suppression, MHD mode control) 3D fields of this magnitude can produce neoclassical toroidal viscosity (NTV), which can: K.C. Shaing, et al., Nucl. Fusion 54 (2014) 033012 • Alter plasma rotation K.C. Shaing, et al., IAEA FEC 2014 Paper TH/P1-11 • Significantly reduce fusion gain, Q, by increased alpha particle transport ( d B/B 0 ~ O (10 -4 )) Therefore, it is important to understand NTV in tokamaks, backed by accurate (~ O (1)) quantitative modeling Outline NTV physical characteristics NTV comparison of theory to experiment NTV experiments and assessment of plasma response Application of NTV to plasma rotation control for NSTX-U NSTX-U NSTX 25 th IAEA Fusion Energy Conference: Characteristics of NTV for Rotation Control / Plasma Response (S.A. Sabbagh, et al.) October 14 th , 2014 2
Neoclassical Toroidal Viscosity (NTV) can be studied through the application of 3D fields in tokamaks NSTX 3D coils Theory: NTV strength varies with plasma collisionality n , d B 2 , rotation K.C. Shaing, M.S. Chu, C.T. Hsu, et al., KSTAR 3D coils PPCF 54 (2012) 124033 NTV force in “ 1/ n ” collisionality regime p 1 1 3 1 i i e B R 2 I ( ) n t NC 2 3 / 2 t B R n t i ( 1 / ) K.C. Shaing, et al., 5/2 T i plasma rotation PPCF 51 (2009) 035004 NSTX-U NSTX 25 th IAEA Fusion Energy Conference: Characteristics of NTV for Rotation Control / Plasma Response (S.A. Sabbagh, et al.) October 14 th , 2014 3
NTV physical characteristics are generally favorable for rotation control alteration by n = 2 Non-resonant NTV characteristics (e.g. in applied field configuration NSTX and KSTAR) in NSTX 3D field configurations with dominant toroidal 30 mode number n > 1 can alter the plasma t(s) rotation profile, , without mode locking outer region 0.575 braking saturates Experimentally, NTV torque is radially 0.585 at this time extended, with a relatively smooth profile 0.595 20 NTV changes continuously as the applied 3D 0.605 /2 (kHz) field is increased 0.615 0.625 T NTV is not simply an integrated torque applied at the plasma boundary, but a radial profile – e.g. shear can be changed 10 These aspects are generally favorable for rotation control; give potential mode control Questions remain 0 e.g. Is there hysteresis when is altered by 0.9 1.0 1.1 1.2 1.3 1.4 1.5 NTV? R(m) NSTX-U NSTX 25 th IAEA Fusion Energy Conference: Characteristics of NTV for Rotation Control / Plasma Response (S.A. Sabbagh, et al.) October 14 th , 2014 4
KSTAR experiments show essentially no hysteresis in steady-state profile vs. applied 3D field strength KSTAR non- resonant (“n = 2 ”) NTV experiments Experiment run to produce 3D field current (kA/t) various steady-state with different 3D field evolution The steady-state rotation profile reached is generally 3D field current stepped up / down independent of the starting point of depends just on the applied Plasma rotation profiles 3D field current level 0 kA/t important for rotation control (krad/s) 2.7 kA/t Absence of hysteresis further confirmed in very recent 3.3 kA/t experiments with 6 steps in 3D field current 3.8 kA/t NSTX-U NSTX 25 th IAEA Fusion Energy Conference: Characteristics of NTV for Rotation Control / Plasma Response (S.A. Sabbagh, et al.) October 14 th , 2014 5
Neoclassical Toroidal Viscosity varies as d B 2 , and T i 2.27 in KSTAR experiments, expected by theory n = 2 field current vs. time Best fit: d B 2 Plasma rotation profile vs. time Best fit: T i 2.27 steady-state reached each d B step NTV torque T NTV expected to scale as d B 2 and T i 2.5 in the “ 1/ n regime” Y.S Park, et al., IAEA FEC 2014: EX/P8-05 (Fri. PM) NSTX-U NSTX 25 th IAEA Fusion Energy Conference: Characteristics of NTV for Rotation Control / Plasma Response (S.A. Sabbagh, et al.) October 14 th , 2014 6
3D field perturbation experiments conducted to measure the T NTV profile in NSTX High normalized beta plasma targets typically chosen Typically near or above n = 1 no-wall limit (for higher T i ) Apply or otherwise change 3D field on a timescale significantly faster than the momentum diffusion time, t m Analysis before/after 3D field application isolates T NTV in the momentum diffusion equation; -dL/dt = T NTV dL/dt measured experimentally and compared to theoretically computed T NTV on this timescale dL/dt profile can change significantly on timescales > t m , (diffuses radially, broadens, leads to significant error compared to T NTV ) Focus on non-resonant applied 3D field configurations To avoid driving MHD modes Resonant fields (e.g. n = 1) are more strongly screened by plasma NSTX NSTX-U 25 th IAEA Fusion Energy Conference: Characteristics of NTV for Rotation Control / Plasma Response (S.A. Sabbagh, et al.) October 14 th , 2014 7
Theoretical NTV torque density profiles, T NTV are computed for NSTX using theory applicable to all collisionality regimes 3D field definition Non-axisymmetric coils fully modelled in 3D d B b B / B B 0.4 0.2 z(m) plasma displacement 0 -0.2 -0.4 -1.5 General considerations -1 1.5 1 -0.5 0.5 0 In tokamaks, not typically y(m) 0 0.5 x(m) -0.5 1 -1 1.5 -1.5 measured, can lead to large error NTV analysis of NSTX – data interfaced “Fully - penetrated field constraint” used to define to NTVTOK B b 2 D • (Y. Sun, Liang, Shaing, et al., NF 51 (2011) 053015) Singularities avoided by standard finite island width assumption Use Shaing’s “connected NTV model”, For NSTX, | | ~ 0.3 cm << 0.5 r i , covers all n , superbanana plateau regimes therefore, ion banana width- (K.C. Shaing, Sabbagh, Chu, NF 50 (2010) 025022) averaging is used for ion channel • Full 3D coil specification and d B spectrum, Can explain why strong resonant ion and electron components computed, peaks in NTV profile are not observed in experiment no aspect ratio assumptions NSTX-U NSTX 25 th IAEA Fusion Energy Conference: Characteristics of NTV for Rotation Control / Plasma Response (S.A. Sabbagh, et al.) October 14 th , 2014 8
Measured NTV torque density profiles quantitatively compare well to computed T NTV using fully-penetrated 3D field n = 2 coil configuration n = 3 coil configuration NSTX NSTX Experimental Experimental -dL/dt -dL/dt NTVTOK NTVTOK y N y N T NTV (theory) scaled to match peak value of measured -dL/dt Scale factor ( (dL/dt)/T NTV ) = 1.7 and 0.6 (for cases shown above) – O(1) agreement O(1 ) agreement using “fully -penetrated 3 D field” indicates that plasma response is not strongly amplified from this “vacuum field assumption” ( T NTV ~ d B 2 ) NSTX-U NSTX 25 th IAEA Fusion Energy Conference: Characteristics of NTV for Rotation Control / Plasma Response (S.A. Sabbagh, et al.) October 14 th , 2014 9
Recommend
More recommend