Transcript Initial results from the new helical Mirnov array for the H

Initial results from the new helical
Mirnov array for the H-1NF heliac
S.R. Haskey, B.D. Blackwell, M.J. Hole, D.G. Pretty, J. Howard
Plasma Research Laboratory, Research School of Physics and Engineering, Australian National University, Canberra,
Australia
Abstract
Results - Overview
plasmas from two poloidal Mirnov arrays [1], we have installed a
new helical Mirnov array with a higher frequency response and
improved toroidal resolution. The array is placed very close to the
plasma as it twists around H-1, in areas of strong mode activity,
with varying degrees of “favourable curvature” in the MHD sense
but almost constant magnetic angle which makes phase variation
with coil position near linear. The low shear heliac geometry [2]
provides a unique opportunity for detailed study of Alfvén
eigenmodes, which could be a serious issue for future fusion
reactors.
ramping experiments has been obtained. Several spectrograms
Tens of thousands of magnetic fluctuations were observed over
showing the variety of observed mode activity are shown in fig. 3.
four κh scans (κh =helical current/ring current). Preliminary results
from data mining on the phases of these fluctuations has been
performed using the SWM11 program [3]. The largest clusters are
shown in fig. 6a. One possible type of wave causing the
fluctuations is the global Alfvén eigenmode (GAE). GAE
frequency predictions in cylindrical geometry are shown in fig. 6b
along with the observed datapoints. Qualitative scaling agreement
with κh is observed; however a scale factor of 0.27 has been
applied to the theoretical values to obtain quantitative agreement.
Fig 3. Several
Results – Magnetic configuration scan, data
Motivated by observations of magnetic fluctuation activity in H-1 A large dataset from several rotational transform scans and density mining and mode analysis
spectrograms showing
a variety of mode
activity including
single and multiple
modes, mode
transitions and higher
frequencies.
Array description and location
The helical array is composed of sixteen three-axis Mirnov coils
(fig. 1c) which follow the helical winding of H-1 through one
revolution (fig. 1a,1b,2). This places the array close to the plasma
in areas of strong mode activity. The entire array is encased in a
thin stainless steel bellows to minimise attenuation whilst
providing vacuum integrity. A single former is placed inside a
copper sputtered Pyrex tube to provide higher frequency
measurements (fig. 1d). Additionally amplifiers with digitally
switchable filters and gain were developed for the array (fig. 1f).
This allows amplifier settings for each shot to be recorded along
with the data in an MDSplus database.
(b)
(a)
(c)
(a)
Results – Data analysis
The array output has been analysed using several techniques including
singular value decomposition (SVD), data mining, analytic phase
calculation and an averaged coherence picking process. SVD analysis
of certain modes shows clear separation into rotating cosine and sinelike components (fig. 4a) and the helical variation of phase is near
linear (fig. 4b,d). Mode identification using the phase change across
the array provides mode numbers which in many cases are consistent
with the rotational transform for the configuration under
consideration. Other modes which have completely different phase
behaviour with helical angle are under investigation (fig. 4e).
(b)
(c)
Array
(a)
Chronos
Singular
Values
(d)
(e)
(b)
(d)
(e)
Time (ms)
(f)
PSD of chronos
Topos
(g)
Fig 1. (a), (b) Rendering and photo of the array
twisting around the central ring conductor. (c)
Three orthogonal coils are wound onto each
former. (d) One former is placed inside a test tube
sputtered with copper to provide higher frequency
response. (e) Positive coil directions for the three
orthogonal coils at each former location. (f) PCB
panel with six of the pre-amplifiers developed for
the array. (g) Argon discharge in H-1.
(d)
(c)
(e)
Array
(a)
Fig 4. (a) SVD of the array data is dominated by a pair of singular values with highly
coherent (orthogonal) sinusoidal singular vectors indicating a travelling wave. (b,c)
Phase difference across the array shows near linear behaviour. (d) Raw (red) and
filtered (black) coil signals showing the phase relationship. (e) Not all coil-phase
relationships are as linear – suspected coupled modes.
Results – ICRH modulation and density ramping
(b)
Experiments have been conducted where a step in the ICRH power is
applied to determine the relationship between density fluctuations and
frequency (fig. 5). Qualitative agreement with the predicted frequency
for the global Alfvén eigenmode (GAE) in cylindrical geometry is
obtained; however, a scale factor of 0.27 is required for quantitative
Antenna currents (red and black)
agreement.
(c)
(d)
Helical
Poloidal
Density
(blue)
Fig 2. (a) Poincaré plots at different toroidal locations with black dots
representing coil location. The coil location relative to the bean shaped
magnetic surfaces is near constant. (b) Coil locations for one of the existing
poloidal arrays (numbered). (c) Poloidal (red) and toroidal (blue) directions
(d) The coil locations are plotted on a poloidal-toroidal lab frame plane with
a simplified travelling wave. The array traces out a near “helical line” as it
crosses both the toroidal and poloidal directions.
Fig 6. Results from data mining (a,b), cluster phase analysis (d,e) and a plot of rotational
transform in radius v kh space (c)
The mean and standard deviation of the cumulative phases for
clusters 3 and 7 are shown along with predicted cumulative phase
for various mode numbers in fig. 6d and fig. 6e respectively.
Analysis of the phases of shows closest agreement with (n,m) =
(4,3) and (5,4) respectively. This agrees with the zero in the
rotational transform shear being close to 5/4 and 4/3 for the κh
values that these clusters exist (fig. 6c). These findings agree with
previous findings using the poloidal arrays [1]. Analysis of the
polarization of these fluctuations is currently under investigation
along with a comparison with optical emission observations.
Conclusions and future work
A 48 coil helical Mirnov array and associated systems have been
installed in the H-1 heliac. Some results from the initial analysis
have been presented showing a variety of interesting behaviour.
We plan to further investigate the various frequency scaling
relations to plasma parameters as well as relating the identified
modes to theory in a more realistic geometry. We are in the process
of comparing the data to the output of several other diagnostics
including new synchronous imaging technique results. Finally, the
H-1 heliac has recently received a major upgrade to the ICRH
heating system and an upgrade to the interferometer system is
underway. These new systems as well as plans to use an active
excitation antenna should not only provide a wider range of plasma
conditions but also provide substantially more density information
and variety of fluctuations to use in our analysis.
Acknowledgements: John Wach and the H-1 team for machine operations. This work
was supported by AINSE and ANSTO.
References:
[1]
D. G. Pretty, PhD Thesis, Australian National University, 2007.
[2]
B. D. Blackwell et al., 21st IAEA Fus. Energy Conf., Chengdu, China (2006).
[3]
M. Shinder, A. Wong, A. Meyerson, NIPS 2011.
Fig 5. Black overlay in the spectrogram is the predicted frequency for the GAE in
cylindrical geometry with a short time delay showing qualitative agreement; however, for
quantitative agreement a scale factor of λ=0.27 is required
For additional information contact: Shaun Haskey
Plasma Research Laboratory, RSPE, Australian National University
[email protected]