MPTS Phase 4

1
MPTS Phase 4

• Benoit P. LeBlanc
• PPPL
• 19 June 2009

2
Outline

• Scope of work for Phase 4
• Technical issues
• Seven unused fiber bundles
• Gradient location database
• Previous inputs
• A partial suggestion

3
Scope of Phase 4 and More

• Install 10 possibly 12 -- new radial channels
• Use 6-filter polychromator as in phases 1 2
• Install two new alignment channels
• Rebuild phase-3 polychromators
• Add 1064-nm filter for Rayleigh noise rejection
  (unlikely)
• Repair MPTS hardware three items
• Calibrate 40-channel system
• Spectral, fast/slow, gain before Rayleigh/Raman
• Preparation work for high-Te channels
• To be implemented in NSTX-U 10keV (on the back
  burner)

4
Technical Issues

• Plan is to complete more of existing hardware
• 7 unused fiber one bundles sees RF NB
• 16 unused sample-and-hold slots
• Changes will occur in the polychromator room
• Develop a technique to unglue output end of fiber
  bundles tricky but feasible
• Reroute a large number of fiber bundles
• Split bundles need to be physical close order
  of 6
• Long sequence of activities, which cannot be
  stopped before MPTS is back online 5 months
• Upgrade includes the construction of a fourth
  polychromator tower
• All new polychromators and ancillary hardware
  need to be ready at the end of FY10 experimental
  run
• Massive calibration work following successful
  hardware installation

5
Polychromator Stacked in Towers

• Current 3 tower each house 10 polychromators
• Each polychromator is attached to a movable
  shelf
• Split bundles need to be next to each other
  vertically
• The fourth tower will have 12 shelves

6
Seven fiber bundles remain unused
Seven unused bundles indicated in green R 31.8,
86.4, 96.3, 104.9, 112.5, 154.6, 157.7 cm
7
Gradient location database

• 1446 shots from 2008
• Use spline Te and ne profiles to compute the
  location of maximum gradients, average over times
  0.15s 0.9max(ts_time)
• Database includes 29 shots suggested by Rajesh
  Maingi
• Use this data to identify new channel locations

8
Location of ?Te vs. location of ?neDistribution
plot
9
Location of ?Te vs. location of
?neDistribution plot redux
Shots suggested by R. Maingi
10
Location of ?Te vs. location of ?ne2D histogram
plot
11
Histogram of ?Te location
1
2
unused
split bundle
3
Phase 1 Phase 2 Phase 3 Unused
12
Histogram of ?ne location
1
2
split bundle
unused
3
Phase 1 Phase 2 Phase 3 Unused
13
Comments for Team

• Rajesh Maingi
• To better resolve the H-mode pedestal gradients,
  I suggested the spatial resolution be optimized
  from R141-147cm. Practically this means adding
  4-5 channels.
• Shell works best with smaller gap
• Jon Menard
• I am concerned about LLD being successful in
  increasing D pumping
• and forcing us to run larger outer gaps to reduce
  metal influx
• What range of outer gaps would the proposals
  below support?, i.e.
• have good gradient resolution for H-mode for our
  typical profiles.
• Some of our transport barrier shots are clearly
  under-resolved in
• the core, so this might warrant some additional
  coverage there.
• Crudely, I tend to think resolving the edge is
  2-3x more important
• than the core programmatically
• Joel Hosea
• Good documentation of pedestal foot, re ELMS
• Ben LeBlanc
• The outermost bundle sees the BN limiter
  antenna was moved inward in 2004

14
Use Existing Data to Estimate Error

• Found among 2008 shots six pairs of successive
  plasmas (S217 open S218 open), followed by
  (S217 open S218 Shut)
• Check shots for similar time evolution etc.
• Took data between 0.05s and 0.5s -- avoid
  saturation
• Compare data with both shutters open Tef0,
  dTef0, nef0 with data with one shutter closed
  Tef1, dTef1, nef1
• Select radial positions with raw signal (S218
  shot/open) of 0.3 or 0.5.

15
Fiber Bundle divided in Two
16
Fiber Bundle divided in Three
17
A SuggestionRevision June 2009

• Propose 10 new channels, but will aim at 12
• Cannot fit more channels in the schedule
• One new radial channels at R154.6 cm (unused FB)
• One new radial channels at R31.8 cm (unused FB)
• Four to six(?) new radial channels in pedestal
  region
• Done by splitting instrumented bundles in two
• No unused bundles available
• Location of higher-resolution channels to be
  chosen by team/experts recommendation
• 2008 data suggest centering around 141 cm
• How far in before splitting bundles is no longer
  useful
• Four to six in the core and ITB region
• Four unused bundles in the core region

18
Edge Pedestal Density
Phase 1 Phase 2 Phase 3 Unused
19
Te ITB Profile
Phase 1 Phase 2 Phase 3 Unused