EUROfusion Consortium, JET, Culham Science Centre, Abingdon, OX14 3DB, UK
1MIT PSFC, 175 Albany Street, Cambridge, Massachusetts 02139, USA2Ecole Polytechnique Fédérale de Lausanne (EPFL), SPC, CH-1015 Lausanne, Switzerland3CCFE, Culham Science Centre, Abingdon, OX14 3DB, UK4Department of Physics and Astronomy, UCI, California 92697, USA5CEA, IRFM, F-13108 Saint-Paul-lez-Durance, France6PPPL, Princeton University, Princeton, New Jersey 08543, USA7See the author list of X. Litaudon et al., Nucl. Fusion 57, 102001 (2017).
N. Fil1, M. Porkolab1, V. Aslanyan1, P. Puglia2, S. E. Sharapov3, S. Dowson3, H. K.
Sheikh3, S. Taimourzadeh4, L. Shi4, Z. Lin4, P. Blanchard2, A. Fasoli2, D. Testa2, J.
Mailloux3, M. Tsalas3, M. Maslov3, A. Whitehead3, R. Scannell3, S. Gerasimov3, S.
Dorling3, G. Jones3, A. Goodyear3, K. K. Kirov3, R. Dumont5, G. Dong6 and JET
Contributors7
Nicolas Fil, APS DPP 2018 – Poster | 05 November 2018
Abstract
The resonant detection and measurement of the damping rates of Alfvén Eigenmodes (AEs) is of
critical importance to the design of experiments and development of models of AE stability [1].
We performed experimental measurements on JET with the AE Active Diagnostic (AEAD), and
theoretical modelling using state of the art MHD and gyrokinetic codes.
The AEAD has undergone a major upgrade [2]. It can provide a state of the art excitation and
real-time detection system thanks to its new amplifiers, filters, digital control system and to the
newly installed magnetic probes. Weakly-damped AEs have been resonantly probed with
external antennas. With GTC [3] we have simulated both stable and unstable AEs by using
equilibria and diagnostic data from JET pulses dedicated to TAEs studies. Good agreement was
obtained between simulations and experiments which adds confidence to further predictions for
next-step burning plasma experiments, including JET and ITER.
[1] W. W. Heidbrink, Phys. Plasmas 15, 055501 (2008)
[2] P. Puglia et al., Nucl. Fusion 56, 112020 (2016)
[3] Z. Wang et al., Phys. Rev. Lett. 111, 145003 (2013)
*This work has been carried out within the framework of the EUROfusion Consortium and has received funding
from the Euratom research and training programme 2014-2018 under grant agreement No 633053. The views
and opinions expressed herein do not necessarily reflect those of the European Commission.
This work has been part-funded by the RCUK Energy Programme [grant number EP/P012450/1] Support for
MIT was provided by the US DOE / DE-FG02-99ER54563, for the Brazilian group the FAPESP Project
2011/50773-0, and for the Swiss group in part by the Swiss NSF.
| PAGE 2
Nicolas Fil, APS DPP 2018 – Poster | 05 November 2018
Physics issues
| PAGE 3
For typical fusion plasma parameters, the phase velocity of the Alfvén wave is just lower than fusion-
born alpha particle speed.
⇒ Possibility of Alfvén wave-particle resonance: effective exchange of energy and momentum with the modes. The exchange time scale is much shorter than alpha particle relaxation time, then:
⇒ self-heating process can be compromised ⇒ damages to the first wall by the ejection of highly energetic alphas
Dispersion relation with (solid) and without
(dashed) toroidal coupling of the waves
Excite AEs with an external antenna of variable frequency
[W.W. Heidbrink, Phys. Plasmas 15, 055501 (2008)]
Frequency
gap
In toroidally confined plasmas:
⇒ the index of refraction is periodic
⇒ gaps appear in the continuous spectrum
Possibility of weakly-damped Alfvén Eigenmodes (AEs)
in these gaps.
Frequency (left) and mode structure (right).
Nicolas Fil, APS DPP 2018 – Poster | 05 November 2018
Leadership of project
Analysis software and
physics
Engineering support
Part-time PostDoc
National Instrument units for
control/digitization
Engineering support
PostDoc time for operation, analysis
and modelling work
A collaborative international project at JET
| PAGE 4
Modelling activities
Use of Gyrokinetic Toroidal Code
(GTC)
Project management
Engineering and system
integration
Redesign of the whole system
Responsible for ongoing
maintenance
Engineering of amplifiers
Control system and
commissioning
Engineering support
Filters and test modules
Nicolas Fil, APS DPP 2018 – Poster | 05 November 2018
Aims of the project
| PAGE 5
Wide range of experiments and theoretical studies of Alfvén Eigenmodes (AEs): structure,
frequencies and stability.
The AEAD can probe stable AEs of the plasma and measure its net damping rate.
Upgraded AEAD system covers the toroidal mode number (n) range that is anticipated to be
the most unstable in ITER, 4 ≤ n ≤ 15
The AEAD will be continuously operated in the full range of isotope experiments preceding a
full DT campaign.
Modelling work, gyrokinetic simulations of AEs in collaboration with UC Irvine to supplement
the ongoing use of ideal MHD codes, such as MISHKA.
Experimental and theoretical studies of Geodesic Acoustic Modes (GAMs), Beta (Acoustic)
AEs (BAE/BAAEs) and Reverse Shear AEs (RSAEs) which are predicted to be of
importance by drift-kinetic and gyrokinetic theory.
Precise validation of the quantitative model of each of the individual damping mechanisms,
and their significance in different ITER and DEMO relevant scenarios.
Nicolas Fil, APS DPP 2018 – Poster | 05 November 2018
Aims of the AEAD upgrade
The AEAD has gone through a major upgrade to allow for the probing of AEs in plasma
configurations which were not possible in the past due to insufficient power. Use of
independently controlled amplifiers to allow arbitrary phasing which is crucial to couple to
modes with high toroidal number n.
Individual 4kW RF amplifiers
Separate excitation & real time control of relative phase between antenna currents
Increased RF current (15A limit).
Increased frequency range of operation, 10kHz – 1000 kHz
Selection of the antennas’ toroidal spectrum for Alfvén Eigenmodes of interest with
toroidal mode number 4 ≤ n ≤ 15
New Digital Control System (named Master Driver)
To control the system and drive the currents in the individual antenna coils with the
desired frequency, amplitude and relative phase.
New set of filters for lower frequencies have been procured and commissioned last year
Study of GAMs, BAE/BAAEs and RSAEs
| PAGE 6
Nicolas Fil, APS DPP 2018 – Poster | 05 November 2018
JET upgraded Alfvén Eigenmode Active
Diagnostic (AEAD)
| PAGE 7
8 in-vessel antennas – 18 turns
4 on each side of the torus (180o)
(2 antennas on Octant 8 unavailable)
Photo: Antennas 1 – 4 Octant 4
Matching units:
Filters available:
125 – 250 kHz
75 – 150 kHz
25 – 50 kHz
19th low pass filters
-70 dB of the 3rd harmonic
200 kHz/s frequency sweeps
Nicolas Fil, APS DPP 2018 – Poster | 05 November 2018
Matching Unit – Low Pass Filters
19th order low pass Chebyshev filters
70 dB attenuation for third harmonic
200 kHz/s frequency sweeps
Limited at 15A and 1.1kV at feedthrough (protected
from overvoltage and overcurrent in real time)
| PAGE 8
Frequency (kHz)
Goal: -70dB
250 kHz filter50 kHz filter
Cut-off frequency
Nicolas Fil, APS DPP 2018 – Poster | 05 November 2018
Master Driver (MD)
To control the system and drive the currents in the individual antenna
coils with the desired frequency, amplitude and relative phase.
Essentially a National Instruments (NI) PXI express chassis.
LabView Real Time (RT) and Field Programmable Gate Array (FPGA)
software performing the various amplifier control functions
Creates reference signal for up to 8 antennas
Real time control < 1𝑚𝑠Frequency control ∆𝑓 < 0.1%Phase control ∅ ± 3°
| PAGE 9
Nicolas Fil, APS DPP 2018 – Poster | 05 November 2018
AEAD capabilities, improvement ongoing
6 switching amplifiers: power of 4kW with pulse times of 15 seconds
Operation close to feed-through limits of 1.1kV and current limit of 15A
Phase errors depends on the frequency range, target of few degrees
| PAGE 10
Operation with ASYNC shots (vacuum):
50 100 150 200 250
Freq (kHz)
I (A
)
12
10
8
6
4
50 100 150 200 250
Freq (kHz)
Z(O
hm
s)
Nicolas Fil, APS DPP 2018 – Poster | 05 November 2018
JET restart discharges
SYNC shots (plasmas) with the 125 – 250 kHz filters:
Operation of the AEAD at the end (63s to 66s) of repeated 1.85MA/2.2T shorter He4
pulses with BreakDown in He.
Reproductive observation on these pulses, JPN ∈ 93066, 93072 (on 24/10/2018)
| PAGE 11
Similar
observations
with:
T001
T002
T007
T008
T009
H302
H303
H304
H305
Phase control disable
Nicolas Fil, APS DPP 2018 – Poster | 05 November 2018
Antennas
Damping rate measurement
Modes appear as resonances on magnetic sensors at the
mode frequency
Damping rate proportional to “quality factor” (q) – width of
the resonance
Equivalent to damped harmonic oscillator
Transfer function:
| PAGE 12
Nicolas Fil, APS DPP 2018 – Poster | 05 November 2018
TAEs damping rate measurements in M15-24
experiments
Damping rates of
marginally stable TAEs
have been measured in
M15-24 experiments
(Real) Frequencies
consistent with nearby
unstable modes (at
51.05s compared to
50.6s)
Too few coils to perform
reasonable mode
number analysis.
For next campaigns:
25 magnetic coils
installed during last
shutdown
| PAGE 13
Nicolas Fil, APS DPP 2018 – Poster | 05 November 2018
Modelling activities, GTC
New collaboration with UC Irvine to work with the Gyrokinetic Toroidal Code (GTC)
Gyrokinetic simulations of low frequency AEs.
To supplement the ongoing use of ideal MHD codes, such as MISHKA.
Used to determine the structure, frequencies and stability of AEs in JET plasmas.
GTC electromagnetic simulations have bulk and “fast” ions, treated gyrokinetically or with a
reduced MHD model
Electrons are treated with a fluid-kinetic model [1] or with a fluid (adiabatic) response
| PAGE 14
[1] Z. Wang, Z. Lin et al. Phys. Plasmas 22, 022509 (2015).
Gyrokinetic ions/fast ions Fluid-kinetic electrons
- Electrostatic potential
Nicolas Fil, APS DPP 2018 – Poster | 05 November 2018
Mode structure and growth in GTC
Equilibrium chosen when many unstable
modes are observed; n=5 chosen
Fully gyrokinetic ions, kinetic electrons and
fast ions at ~747 keV (500 keV at mode
location)
Net growth rate and frequency deduced from
the oscillations of highest amplitude poloidal
harmonic (m=11)
Antenna structure is composed of two
poloidal harmonics to resemble fast ion
mode structure (m=11, m=12)
Radial profile of each is Gaussian
Amplitude response has beat pattern which
peaks (time indicated by red cross)
| PAGE 15
Nicolas Fil, APS DPP 2018 – Poster | 05 November 2018
Synthetic antenna (I)
Maximum amplitudes of beat pattern
(as denoted by red cross) similar to a
driven damped harmonic oscillator
Spectral response is similarly given
by the following:
Damping rates of marginally stable
modes can therefore be determined
| PAGE 16
Nicolas Fil, APS DPP 2018 – Poster | 05 November 2018
Synthetic antenna (II)
Changing the physics mode allows the effect of different damping mechanisms to be
determined
Electron Landau damping: kinetic/adiabatic electrons
Ion Landau damping: gyrokinetic/MHD-like ions
| PAGE 17
Mechanism Drive/damping
Continuum 0%
Electron Landau -0.09%
Ion Landau -1.64%
Radiative -1.18%
Energetic
particle
+4.29%
Nicolas Fil, APS DPP 2018 – Poster | 05 November 2018
Modes relative to Alfvén Continuum
ALCON routine calculates
continuum structure
including acoustic couplings
Modes plotted relative to
observed plasma-frame
frequency
Error bars equivalent to
FWHM
| PAGE 18
Nicolas Fil, APS DPP 2018 – Poster | 05 November 2018
Three measurements of marginally stable TAEs made in two discharges by the Alfvén
Eigenmode Active Diagnostic (AEAD)
Frequency and damping rate deduced from transfer function:
Small number of remaining magnetic probes leads to large uncertainty in damping rate
and mode number (the latter cannot be consistently identified)
GTC comparison to AEAD, antenna measurements
| PAGE 19
Nicolas Fil, APS DPP 2018 – Poster | 05 November 2018
GTC comparison to AEAD (I)
Multiple candidate modes must be
simulated
Labels refer to the dominant poloidal
harmonics driven by the synthetic
antenna
This signal is peaked between the two
harmonics’ rational surfaces
Closest match to the observed modes
(red lines) in plasma-frame frequency
and damping rate (n=6, m=5,6) is the
best candidate to describe the
observed mode
| PAGE 20
Nicolas Fil, APS DPP 2018 – Poster | 05 November 2018
GTC comparison to AEAD (II)
Closest candidate modes from
GTC have been matched to the
AEAD measurements
Uncertainty in damping rate
arises from variability of the
different coils
Uncertainty in the frequency
arises from uncertainty in plasma
rotation
Error bars denote rotation
uncertainty based on
spectroscopic plasma rotation
measurements
| PAGE 21
[V. Aslanyan et al. submitted
to Nuclear Fusion]
Nicolas Fil, APS DPP 2018 – Poster | 05 November 2018
Ongoing work on the AEAD
Improve the control of the antennas currents/voltages
Improve the phase control between the antennas
Enable the real-time tracking mode of the AEs
Target specific antennas’ toroidal spectrum for AEs of interest.
Measurements of damping rates of TAEs in plasmas with a variety of q-profiles and with
the effects of fast ions from neutral beams and ICRH.
Detection and measurements of low frequency modes, BAEs, BAAEs, GAMs and RSAEs
| PAGE 22
Magnetic signal from JPN 54895
showing GAMs.
Use of the 25 – 50 kHz filters
to observe these modes on JET
AELM can track and differentiate
modes in real time
Nicolas Fil, APS DPP 2018 – Poster | 05 November 2018
Future work, Modelling
Extensive use will be made of the “antenna” version of GTC to determine the stability of
predicted modes.
Damping mechanisms will be identified by GTC simulations
Extensive comparisons between GTC simulations and the existing codes CSCAS, MISHKA
and CASTOR-K will be made.
Simulations will be extended to lower frequency modes, such as BAEs BAAEs, and GAMs to
form an important synergy with AEAD measurements.
| PAGE 23
Purely theoretical, not observed
experimentally: EP-driven mode
observed by changing the FI
temperature
Characteristic of BAE:
• Single m harmonic and high
growth rate, 𝜸/𝝎~𝟐𝟎%• f~40 kHz