Translatable in-vessel mirrors have enabled the DIII-D Thomson scattering system to diagnose the divertor plasma in high triangularity shaped plasmas. Previous divertor Thomson scattering measurements in DIII-D were restricted to spatial locations along a Nd:YAG laser beam that was directed through a vertical port. This only allowed measurements to be made in low triangularity shaped plasmas. The new mirrors re-route the laser underneath floor tiles to a position of smaller major radius as necessary for high triangularity plasmas. New in-vessel collection optics transmit scattered light from regions inaccessible to external lenses. Damage to mirrors and high stray light levels are challenges that were overcome to successfully make these measurements. Through the careful use of baffles and light shields, stray light leakage into polychromator detector channels was reduced to negligible levels, allowing temperature measurements below 1 eV. The system is described and the initial results presented.

Thomson scattering measurements of electron temperature and density in the divertor region of the DIII-D tokamak have been done for the past 20 years.1,2 Recently, divertor Thomson scattering (DTS) systems have been designed for NSTX,3 MAST,4 and ITER5,6 to diagnose this important region of the plasma.

Previously, the location of the divertor Thomson measurements on DIII-D were restricted to the area above a vertical port near the center of the machine where the laser beam could be directly injected into the vacuum vessel. Magnetically sweeping the divertor region across the laser beam enabled 2D measurements of this complex region. This system worked well for low triangularity shaped plasmas.

As the understanding of the advantages of highly shaped plasmas on stability and confinement improved,7 operation in high triangularity shaped plasmas became desirable. In 2006, the DIII-D divertor region was modified to accommodate high triangularity shaped plasmas by the introduction of a cryogenically pumped divertor shelf.8 However, operation at high triangularity moved the divertor plasma away from the laser beam and precluded the Thomson scattering measurements of this important region.

This paper, together with a previous paper,9 describes the design and initial results of a Thomson scattering system in which the laser beam is introduced into the vacuum vessel at one location and then moved to another location through a series of mirrors located underneath the protective wall tiles. Figure 1 shows the old location of the laser beam on the divertor shelf and the new location on the divertor floor. Also shown are the locations of two other laser beams for the core10 and tangential11 Thomson scattering systems.

FIG. 1.

Location of the laser beams for Thomson scattering measurements in DIII-D. Also shown are the standard shelf Divertor Thomson Scattering (DTS) measurement locations and the new floor DTS locations.

FIG. 1.

Location of the laser beams for Thomson scattering measurements in DIII-D. Also shown are the standard shelf Divertor Thomson Scattering (DTS) measurement locations and the new floor DTS locations.

Close modal

It had been thought that stray light from in-vessel mirrors would be so strong as to preclude the measurements, and laser beam intensities on the mirror so great that the mirrors would be immediately damaged. Through the use of advanced high-quality interference filters in the polychromators, high damage threshold laser mirrors, and careful design, these challenges have been overcome and successful Thomson scattering measurements have been made in high triangularly shaped, diverted plasmas in DIII-D.

Thomson scattering measurements are available in both the shelf and floor configurations, as shown Fig. 2. Either a moveable mirror located below the shelf is pneumatically retracted to allow the laser beam to continue to the shelf location at R = 1.485 m, or it is inserted to redirect the laser beam inboard to a second mirror which directs it vertically up through a 1 cm hole in the floor tiles at R = 1.335 m. External collection optics1 are placed to look down from a port above the divertor region and image the laser beam on an array of fiber optic bundles which carries the light to 8 remotely located polychromators.12 When the laser beam is moved to the floor configuration, the fiber bundles are moved along a precision track to keep the laser beam in focus. Details of the design can be found in Ref. 9.

FIG. 2.

Schematic of the lower divertor showing the re-directed laser beam to the floor region, in-vessel relay optical elements, and the existing optics of the current system with adjustable focus. The separatrix of a high-triangularity plasma discharge is overlain.

FIG. 2.

Schematic of the lower divertor showing the re-directed laser beam to the floor region, in-vessel relay optical elements, and the existing optics of the current system with adjustable focus. The separatrix of a high-triangularity plasma discharge is overlain.

Close modal

The lower portion of the floor region is not visible to the external collection optics, so an in-vessel lens system located under the shelf relays the image of the laser beam in this region through a window of another port where it is subsequently imaged on another fiber bundle array.

In the shelf configuration, the laser beam exits the vessel through a symmetric port at the top of the machine and is absorbed by an external beam dump. In the floor configuration, the laser beam passes through a 2 cm hole in a ceiling tile and is absorbed in a glass absorbing beam dump located behind the tile.9 

Both systems use the same polychromators optimized for low Te measurements in the divertor region (0.5-500 eV) and the same 50 Hz Nd:YAG laser, nominally 1 J and 10 ns pulses.

The introduction of mirrors in the vessel so close to the scattering volume presented a number of challenges. Scattering of laser light from the mirrors, even at very low values, would create significant stray light that could easily overwhelm the Thomson signal. To minimize the collection of stray light, the laser path was completely baffled from the in-vessel optics. In addition, the hole through the shelf was sealed with a shutter that closes as the mirror is inserted. This prevents stray light from entering the vessel, except for the 1 cm hole in the tile where the laser enters. The in-vessel lenses are shrouded and baffled to limit their exposure to light from unwanted sources. Advanced hard-coated interference filters13 were used in the polychromators, and these have an optical density (OD) greater than 5 (except for the 1061.9 nm filter which has an OD of 3) at the laser wavelength to further reject any collected stray light. The laser beam dump uses absorbing glass at Brewster’s angle to minimize any reflected light. These techniques were proved effective at eliminating stray light from the Thomson signals even at wavelength channels near the laser line (needed for low Te measurements).

Polychromators for the divertor region have an additional long wavelength side (1067 nm) filter channel to extend and improve their low Te measurement capability. The change in the central wavelength of this filter from laser line (Δλ) is placed between Δλs of the first two filters on short wavelength side of the laser line (Fig. 3). This improves the measurement of narrow, very low Te (<0.5 eV) scattered spectra (Fig. 4). When not used for divertor measurements, these polychromators are also used for core and tangential measurements by simply substituting a core fiber optic for a divertor fiber optic. In order to accommodate the higher Te values in the core, an additional 800 nm (200 nm wide) filter was also added to extend the high Te range to over 8 keV.

FIG. 3.

Interference filter transmission of filters on both the long and short wavelength sides of the laser line (vertical dashed line).

FIG. 3.

Interference filter transmission of filters on both the long and short wavelength sides of the laser line (vertical dashed line).

Close modal
FIG. 4.

Calculated expected fractional Te measurement error from an 8 channel polychromator (black) compared to a previous 6 channel polychromator (red). The addition of the 1067 nm filter extends the range to less than 1 eV, and an 800 nm filter increases the range to over 8 keV.

FIG. 4.

Calculated expected fractional Te measurement error from an 8 channel polychromator (black) compared to a previous 6 channel polychromator (red). The addition of the 1067 nm filter extends the range to less than 1 eV, and an 800 nm filter increases the range to over 8 keV.

Close modal

Typically, in Thomson scattering systems, the laser is focused at the scattering volume. Due to the presence of mirrors so close to the scattering volume, the laser beam was focused at the center of the vessel to allow the beam to expand somewhat at the entrance and exit where laser mirrors are located. This reduced the laser intensity from 11 J/cm2 at the focus (3 mm diameter) to 3.3 J/cm2 at the mirrors (5.5 mm diameter). High damage threshold, dielectric, ion beam sputtering coated mirrors (40 J/cm2) that can tolerate 400 °C bake out temperatures were used.14 In addition, the laser divergence is controlled by warming up the laser sufficiently before permitting it to enter into the system. An improperly warmed up laser could focus near one of the mirrors and damage it. To further minimize the fluence on the laser mirrors, the Q-switch timing is adjusted to throttle the laser energy to the minimum required to obtain a useful signal. Typically, 0.2 J are sufficient for high density divertor operation. Finally, the input plasma facing mirror has a shutter linked to the moveable mirror which opens only when the system is in use. This protects the mirror from glow discharge cleaning, boronization operations, and general plasma exposure during experiments for which the system is not in use.

Initial alignment of the system is done by personnel in the vessel during vents. The moveable mirror is pushed against a stop where its front surface encounters three precision aligned small raised surfaces near the edge of the mirror that define a reproducible position. The second mirror is aligned in a kinematic mount and locked in place. The overall laser alignment is done at fiducials at the entrance and exit windows. Alignment is checked during operations using an in situ alignment monitor consisting of an array of 5 fiber optics that span the laser beam and measure its position relative to the collection fibers.2 

Raw data in digitizer counts of the scattered light for the lowest spatial channel are shown in Fig. 5 for a shot, in which the divertor is swept past the floor divertor system as shown in Fig. 6. The scattered signal is the output of a 30 ns gated integrator centered on the laser pulse after a 30 ns delay line subtraction of the background light.15,16 This channel, whose wavelength is centered at 1062 nm, shows no signal until plasma is present at about 2.5 s. This demonstrates the effectiveness of the filter in suppressing stray light and permits the measurement of low Te plasmas.

FIG. 5.

Raw data in digitizer counts of the Thomson scattered signal from the lowest floor divertor channel for shot 173031. The signal is from the 1062 nm wavelength channel and light produces negative going signals. The dashed line represents the no-light signal level. The laser starts firing at −1.5 s and the plasma discharge starts at 0.0 s.

FIG. 5.

Raw data in digitizer counts of the Thomson scattered signal from the lowest floor divertor channel for shot 173031. The signal is from the 1062 nm wavelength channel and light produces negative going signals. The dashed line represents the no-light signal level. The laser starts firing at −1.5 s and the plasma discharge starts at 0.0 s.

Close modal
FIG. 6.

Reconstructed separatrix locations relative to the floor divertor Thomson scattering (DTS) chord locations at various times during the x-point sweep of shot 173031.

FIG. 6.

Reconstructed separatrix locations relative to the floor divertor Thomson scattering (DTS) chord locations at various times during the x-point sweep of shot 173031.

Close modal

Unfortunately, the 1064 nm channel, used for Rayleigh scattering calibration of the system, is saturated and cannot be used for absolute density calibration. For the results reported here, the density calibration for the shelf system was extended to the floor system, taking into account the change in the optical magnification and collection solid angle between the two systems.

Temporal evolution of Te and ne for three of the floor divertor channels are shown in Fig. 7. Temperatures near 5 eV and densities of 1 × 1020 m−3 are measured near the floor as the outer strike point moves across the lowest spatial location. The large spikes in Te are due to the effects of ELMs which periodically exhaust heat and particles from the main plasma to the divertor. The spatial profile of Te and ne measured at 2536 ms is shown in Fig. 8. The error bars represent the 1-sigma uncertainty of the fit to the raw data.

FIG. 7.

Time evolution of (a) the variation of the outer strike point (ROSP) and the floor DTS (RFLOOR_DTS) location, and (b) Te and (c) ne for three of the floor divertor channels.

FIG. 7.

Time evolution of (a) the variation of the outer strike point (ROSP) and the floor DTS (RFLOOR_DTS) location, and (b) Te and (c) ne for three of the floor divertor channels.

Close modal
FIG. 8.

Spatial profile of (a) Te and (b) ne at 2536 ms of shot 173031.

FIG. 8.

Spatial profile of (a) Te and (b) ne at 2536 ms of shot 173031.

Close modal

A composite 2D map of Te, ne, and pe based on measurements as the divertor leg is swept past the laser beam for several plasma discharges is shown in Fig. 9. The data are mapped to magnetic flux surfaces and displayed at a fixed x-point location. Low Te and pe near the floor may indicate partial plasma detachment in this case.

FIG. 9.

Composite 2D constructions of (a) Te, (b) ne, and (c) pe based on measurements as the divertor leg is swept past the measurement locations shown as red circles at R = 1.335 m. The data are then mapped to magnetic flux surfaces and displayed at a fixed x-point location.

FIG. 9.

Composite 2D constructions of (a) Te, (b) ne, and (c) pe based on measurements as the divertor leg is swept past the measurement locations shown as red circles at R = 1.335 m. The data are then mapped to magnetic flux surfaces and displayed at a fixed x-point location.

Close modal

Because high stray light levels prevent calibration using Rayleigh scattering in argon, Raman scattering in nitrogen will be used to absolutely calibrate the density. Raman scattering produces signals in wavelength channels near the laser line which do not suffer from stray light pollution. This signal is much weaker than the Rayleigh signal and therefore requires higher pressures to obtain good signal-to-noise ratios. Typically, argon pressures of 4 Torr are used for Rayleigh scattering calibration. Nitrogen pressures of 100 Torr are expected for Raman calibration.

There are also plans to expand the system to include a scanning laser and optics system that can provide many vertical beams and produce 2D measurements of Te and ne in the entire divertor region without having to scan the plasma across the laser beam. It is envisaged to use an external fast-steering mirror to move the laser beam across a stationary multi-element mirror to direct the beam to different vertical chords on a pulse by pulse basis.

We gratefully acknowledge the skillful technical assistance of J. Kulchar, D. Ayala, and C. Liu. This work is supported by the U.S. DOE under No. DE-FC02-04ER54698 and by LLNL under No. DE-AC52-07NA27344.

DIII-D data shown in this paper can be obtained in digital format by following the links at https://fusion.gat.com/global/D3D_DMP.

This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise, does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.

1.
T. N.
Carlstrom
 et al.,
Rev. Sci. Instrum.
66
,
493
(
1995
).
2.
T. N.
Carlstrom
 et al.,
Rev. Sci. Instrum.
68
,
1195
(
1997
).
3.
A. G.
McLean
 et al.,
Rev. Sci. Instrum.
85
,
11E825
(
2014
).
4.
J.
Hawke
 et al.,
J. Instrum.
8
,
C11010
(
2013
).
5.
E. E.
Mukhin
 et al.,
Nucl. Fusion
54
,
043007
(
2014
).
6.
E. E.
Mukhin
 et al.,
J. Instrum.
7
,
C02063
(
2012
).
7.
C.
Holcomb
,
J.
Ferron
,
T.
Luce
,
T.
Petrie
,
P.
Politzer
,
C.
Challis
,
J.
DeBoo
,
E.
Doyle
,
C.
Greenfield
,
R.
Groebner
 et al.,
Phys. Plasmas
16
,
056116
(
2009
).
8.
P.
Anderson
,
Q.
Hu
,
C.
Murphy
,
E.
Reis
,
Y.
Song
, and
D.
Yao
,
Fusion Eng. Des.
82
,
1756
(
2007
).
9.
F.
Glass
 et al.,
Rev. Sci. Instrum.
87
,
11E508
(
2016
).
10.
T. N.
Carlstrom
 et al.,
Rev. Sci. Instrum.
63
,
4901
(
1992
).
11.
D. G.
Nilson
 et al., General Atomics Report GA-A23198 presented at LAPD,
1999
.
12.
T. N.
Carlstrom
 et al.,
Rev. Sci. Instrum.
61
,
2858
(
1990
).
13.
Alluxa, 3660 North Laughlin Road, Santa Rosa, CA 95403; www.alluxa.com.
14.
Advanced Thin Films, 5733 Central Avenue, Boulder, CO 80301; www.advancedthinfilms.com.
15.
T. M.
Deterly
,
B. D.
Bray
,
C.-L.
Hsieh
,
J. A.
Kulchar
,
C.
Liu
, and
D. M.
Ponce
,
IEEE Trans. Plasma Sci.
38
,
1699
(
2010
).
16.
D. M.
Ponce-Marquez
,
B. D.
Bray
,
T. M.
Deterly
,
C.
Liu
, and
D.
Eldon
,
Rev. Sci. Instrum.
81
,
10D525
(
2010
).