Reconstruction of plasma shape and plasma energy in Spherical Tokamak Globus-M.

1
Reconstruction of plasma shape and plasma energy
in Spherical Tokamak Globus-M.

• S. Bender1, V.Gusev, A.Detch, Yu.Kostsov1, R.
  Levin, K. Lobanov1, N.Sakharov.
• A.F.Ioffe Physico-Technical Institute,
  St.Petersburg, Russia
• 1 D.V. Efremov Institute of Electrophysical
  Apparatus, St. Petersburg, Russia

2
Spherical tokamak Globus-M

• Tokamak Parameters
• Major radius R 0.36m
• Minor radius a 0.24m
• Aspect ratio A gt 1.5
• Achived Plasma Current Ip lt 370kA
• Toroidal magnetic field BT 0.07 - 0.55??
• Achieved plasma parameters
• Plasma density ltngt lt 1.2?1020 m-3
• Electron Temperature ?? lt 1000 ??
• Ion Temperature ?i lt 600 ??
• Elongation k 1.1 2.2
• Triangularity ? 0.1 0.45
• Safety factor q95 ? 2.1

3
Magnetic diagnostic analyses
Magnetic field coils Magnetic flux loops
Local measurements in toroidal direction Averaging of toroidal heterogeneity
Direction of coil magnetic axis Toroidal magnetic field error Scalar measurements
Magnetic field measurements are indirect measurement for equilibrium code gt Need a high accuracy of measurements Cannot be used without magnetic flux loops Direct measurements for equilibtium code. Can be used without magnetic coils.
Need a high accuracy of calibration Measurements of vacuum vessel current density
4
Magnetic loops quantity and coordinate selection
requirements

• 1. Quantity of magnetic loops must be minimum
• 2. Magnetic diagnostic must work If 1-2 loops are
  damaged
• 3. Magnetic loops cannot cross the aperture of
  ports
• 4. The reason for selection of magnetic loops
  coordinate is minimization of magnetic flux
  extrapolation error to closed circuit.
• for measurements near ports and other
  diagnostics
• for poloidal field reconstruction before plasma
  discharge

5
Upgrade of magnetic diagnostic

• Globus-M magnetic diagnostics
• 9 Rogowsky coils for measurements current in
  tokamak coils
• 21 magnetic flux loops
• 64 poloidal magnetic field detectors
• Rogowsky coil positioned outside the vacuum
  vessel for vacuum vessel current measurement
• Rogowsky coil positioned inside the vacuum vessel
  for plasma current measurement
• 2 diamagnetic flux loops

Magnetic flux loops
In winter of 2004-2005 14 additional magnetic
flux loops closed toroidally were installed
inside the vacuum vessel
6
Globus - M model
PF coils and magnetic flux loops In the
model we use real dimensions, coordinates and the
number of turns for each PF coils and magnetic
flux loops. Besides the signals of the magnetic
flux loops the EFIT input data include currents
in the PF coils and the induced toroidal current
in the vacuum vessel Limiter In the
limiter magnetic configurations the plasma
outmost closed magnetic surface is determined by
the graphite limiter on the vessel central
cylinder.
7
Globus - M model

• Vacuum vessel
• The vacuum vessel is an all-welded
  stainless steel construction with the
  characteristic wall thickness 2-3 mm except the
  outer ring of 14 mm thickness. Due to the small
  wall thickness the vessel electric resistance to
  the toroidal current is about 0.1 mOhm. For this
  reason the current flowing through the vessel in
  the toroidal direction does not exceed 40-50 kA
  in plasma current ramp-up phase and 15-20 kA
  during the plasma current plateau. The last
  values are small in comparison with the plasma
  current values. However, the spatial distribution
  of the vacuum vessel current is taken into
  account. For this purpose the vessel is
  extrapolated by the set of 21 rings corresponding
  to the number of loops. Current in each ring is
  determined according to the induced voltage
  measured by the closest loop.

8
Globus - M model

• Central Solenoid
• The central solenoid is extrapolated with
  two-layer coil wound by the condactor of 2020
  mmmm cross section with a uniform current
  density distribution. In the experiments the
  central solenoid operates in current swing
  regime. The plasma breakdown is initiated in
  phase of positive current ramp down. The starting
  value of the current is 50-55 kA per turn. The
  plasma current plateau is sustained during the
  solenoid negative current half-wave, where
  current reaches the values of 40-45 kA at the end
  of plasma shot. In this phase the central
  solenoid stray magnetic field plays a significant
  role in the formation of plasma magnetic
  configuration despite the energizing of
  compensation coils for the stray field
  correction. Especially important the stray field
  asymmetry relatively the tokamak midplane caused
  by some inaccuracy in the solenoid manufacture
  and the tokamak assembly. The adequate model
  describing the errors in the solenoid
  constraction should be developed on the base of
  magnetic measurements. At present the central
  solenoid asymmetry is described by the coil
  vertical shift of 3.5 mm above the midplane.

9
Formation of detached magnetic configuration

• A systematic observation of reconstructed plasma
  magnetic configurations revealed that both
  X-points are usually located inside the vacuum
  vessel volume during a most part of the plasma
  current plateau phase. At the same time the
  plasma is usually limited by the vessel wall and
  the outmost closed magnetic surface is determined
  by the plasma contact with the graphite limiter
  on the vacuum vessel inner cylinder near the
  tokamak midplane. The formation of a fully
  detached plasma was demonstrated in the
  experiments.
• The transition from limiter to X-point
  configuration was accompanied by some increase of
  the plasma vertical elongation. Another evidence
  of the plasma detachment is a significant
  decrease in the intensity of impurities emission
  observed by collimated detectors in the midplane.
  Figure illustrates at least by a factor of 3-4
  drop in the intensity of the OIII and CIII lines
  during a few milliseconds interval.
• Figure shows also the variation of plasma
  thermal energy obtained from EFIT analysis.
  However, for further study of plasma energy
  balance the EFIT energy must be verified by
  kinetic measurements of the plasma temperature
  and density spatial distributions.

10
Monitoring of the plasma vertical position

• Before the installation of 14 additional magnetic
  flux loop the EFIT analysis was based on the
  experimentally measured values of the plasma
  current, currents in PF coils and the plasma
  radial position. The reconstruction was performed
  for magnetic configurations symmetrical
  relatively the midplane. First measurements using
  new magnetic loops have demonstrated a tendency
  to the plasma vertical shift towards the vessel
  lower dome and a formation of a single null
  configuration as the amplitude central solenoid
  negative current increased. Most plasma shots
  were terminated by the internal reconnection
  event (IRE) accompanied by vertical disruption
  even at large safety factor values.

gt
11
Monitoring of the plasma vertical position

• The input signal for the plasma vertical position
  control was a radial magnetic flux measured by
  two loops positioned on the top and lower vessel
  domes. The plasma vertical position was monitored
  according to the preprogrammed radial magnetic
  flux waveform. Usually the radial flux value
  closed to zero was chosen for the steady state
  phase of the plasma discharge. However, the EFIT
  analysis revealed a strong, up to 10-15 cm,
  vertical displacement of the plasma geometrical
  center when the automatic system sustained the
  radial magnetic flux according to the
  preprogrammed zero value. This effect is caused
  by the superposition of the asymmetric relatively
  the midplane central solenoid stray field and the
  radial symmetric magnetic field produced by the
  horizontal field coils in the feedback control
  contour. The absolute value of the plasma
  vertical shift was larger for larger values of
  the plasma vertical elongation. Note, that IRE
  occurs in plasma with higher vertical elongation
  and therefore with higher value of the safety
  factor near the plasma boundary, but at larger
  vertical shift.

12
Monitoring of the plasma vertical position

• The plasma column can be shifted back towards the
  center of the vacuum vessel by a correction of
  the preprogrammed radial magnetic flux wave form.
  The time evolution of the plasma current and
  plasma vertical position in the shots with
  different programs of radial magnetic flux. In
  most shots the correction of the plasma vertical
  position led to an increase of the plasma shot
  length.

13
Measurements of plasma total stored energy
Elongated plasma cross-section (kgt1.3) lets to
reconstruct total plasma energy.
14
Conclusions

• A successive upgrade of magnetic diagnostics have
  been performed in Globus-M experiments. A new
  diagnostic together with improve model of the
  tokamak magnets and the vacuum vessel made
  possible EFIT reconstruction equilibrium magnetic
  configurations of different types by using a
  small number of magnetic loops. The magnetic
  loops and the specially developed fast data
  acquisition system will be employed further for a
  real time digital control of the plasma shape.
• Detached plasmas limited by a single null
  separatrix were achieved in tight plasma vacuum
  vessel configuration.
• The influence of the central solenoid asymmetric
  stray field on the plasma shape was studied. The
  application of additional radial magnetic field
  in the plasma current plateau phase led to the
  plasma stability enhancement and as a result to
  an increase of the plasma shot length.
• This work was supported by Rosatom and RFBR
  grants (03-02-17659, 05-02-17773)