Status of the Commissioning of the ILA

1
Status of theCommissioning of the ILA

• Presented by F. Durodié
• 4-Jul-2008

2
Remarks

• Contribution of many colleagues in and outside
  the ILA (core) team
• Most of the (higher performance) data was
  obtained on the lower half
• technical difficulties on the control of C2
• lost control definitively after 23/7
• uncertainties in ZT (upper as well as lower
  half) appear to affect the upper half more
• run away when closing in on match of one of the
  2 matching solutions

3
Highest power coupled in L-mode

• 1.85 MW (APTL) - 0.35 MW (losses) 1.50 MW
  coupled
• L-mode ROG 4 cm

4
Highest Voltages (L-mode)

• 35 kV _at_ 1.5 MW (APTL) - 0.35 MW (losses)
  1.15 MW coupled
• pulses in preparation of L? H mode transitions
• using ROG sweeps 4cm ? 6cm and activiate
  algorithm only after sweep

5
Highest Voltages (H-mode)

• Lower half H-mode ROG 4cm
• 36 kV (- 38 kV max) _at_ 1.3 MW (APTL) - 0.35 MW
  (losses) 1.05 MW coupled

6
H-mode ROG scans

• Higher sensitivity to tripping at larger ROG
• Preliminary need to asses size of ELMs in
  various pulses
• (possibly slightly ROG dependent ?)

ROG
ROG
7
Trips during ELMs (73603)
max(VSWR34,VSWR78)
Da V

• On the smaller ELMs the VSWR decreases
• cfr offset matched 2nd stage
• Trips only at (somewhat) larger ELMs
• Time resolution on RF data not sufficient to
  determine if it is lack of load tolerance or a
  genuine arc

8
Control difficulties of C2

• before 23/7
• substantially rescaled signal
• scaling more or less stable over the previous
  weeks
• Noise on remaining sensor did not affect RF
  signals
• Bosh controller appeared to be able to filter
  the noise and keep C2 in a more or less steady
  position
• from 23/7
• extreme difficulties to find the correct
  possition for C2
• scaling modified
• RF signal fluctuations synchronized with noise
  on position sensor
• Capacitor position is no longer sufficiently
  steady

Match control issue
9
Probable ZT estimation issues

• Despite several attempts to approach the
  matching solution of the upper half from
  different starting points this solution appears
  unstable
• The sign of the reactive part to ZT , XT, that
  must be used to balance the voltages depends on
  the matching solution however on the lower half
  the sign of XT is independent of the chosen
  matching solution
• from there we estimate that there is an offset
  of about 2 Ohm on XT (XT reads too high)
• the current thinking on the origin of this
  discrepancy is the observed cross-talk on the
  digital databus signals between APDM and PXI.

10
Simulation
Measurement
Solution 1 ZT61j Solution 2 ZT6-1.7j
Solution 1 ZT6-1j Solution 2 ZT6-3j
RDL balanced Toroidally unbalanced (v3/v71.2)
Fully balanced

• For solution 1, Im(Zt) must be positive
• The simulation give a positive imaginary part for
  solution 1 even if the interstrap coupling
  increase
• Poloidal interstrap increase involes the absolute
  (Im(Zt)) raising to reach a perfect match

Capacitor positions are closed to the measurment
ones and global behavior is fine exept for the
unbalanced strap voltage Something wrong on the
electrical behavior or on the calibration (APTL)
APTL phasing and module ?
A. Argouarch
11
APDM cross-talk
13 deg

• One possible suspect cross-talk between
  APTL forward and reflected voltages in the APDMs
• has been tested yesterday (first impression)
  effect minimal

12
ILA Session 02-Jun-2008 / 73007

• Matching
• ROG
• ILA long pulse (56 ? 61s)
• 4cm (ILA profile)

13
ILA Session 02-Jun-2008 / 73016

• Matching
• ROG
• ILA long pulse (56 ? 61s)
• 4cm (ILA profile)

14
SMAD / SHAD

• SHAD
• So far the SHAD has not yet detected any arc
  (PJ)
• (despite there have been arcs)
• SMAD
• Principles verified on slow (JPF) data
• Appear OK near match
• tendency to drift on some pulses (DVE)
• SMAD coefficients have been computed
• SMAD operates on the raw APDM data calibration
  of signals is folded in SMAD coeffs (EL)
• No data taken until now with SMAD hardware
• issues with APDM SMAD databus handshake
• however expect to be able to take data in the
  coming days

15
SMAD principle on JPF data
72871
16
Conclusions

• ILA Encouraging initial results on H-mode,
  however
• still uncertainties on ZT which could affect the
  performance
• Load tolerance to be optimized
• fast DAQ (CODAS) and SMAD will be available in
  the coming days
• Arcs vs. Mismatch
• ? Sensitivity to ROG
• Substantial difficulties
• Control of C2
• installation of ex-vessel monitoring planned
• caveat will require a learning curve to operate
• Control of match on RDL12
• ? Uncertainties in ZT or somehow linked to
  control of C2
• Matching the whole ILA
• very sensitive to relative phasings of RDLs
• Not much progress yet on dynamic simulation
• ? Origin of difficulty of matching RDL12
• Substantial worries