Status of the High-k Scattering System

D.R. Smith, E. Mazzucato, T. Munsat, H. Park, and D. JohnsonPrinceton Plasma Physics Lab

L. Lin, C.W. Domier, M. Johnson,

and N.C. Luhmann, Jr.University of California at Davis

46th APS-DPP Meeting

November 15-19, 2004

Savannah, GA

Motivation, Goal, and Method

Motivation: Electron thermal transport in all magnetic confinement devices remains

greatly enhanced beyond neoclassical levels In NSTX ion thermal transport is at or near neoclassical levels and

electron heat conduction is the dominant thermal loss channel Gyrokinetic simulations of NSTX show ITG/TEM modes suppressed due to

E×B flow shear, but ETG modes with k┴e≈1 remain robust

Goal: Demonstrate the presence or absence of ETG turbulence in NSTX If present, characterize the effect on electron thermal transport

Method: Direct experimental observation of e-scale turbulent fluctuations

by the coherent scattering of electromagnetic waves

National Spherical Torus Experiment

Major Radius, R 0.85 m

Minor Radius, a 0.67 m

Aspect Radio, A ≥ 1.27

Elongation, ≤ 2.6

Triangularity, ≤ 0.8

Plasma Current, IP 1.5 A

Toroidal Field at R0, BT.3-.45 T

NB Power, PNB 7 MW

HHFW Power, PHHFW 6 MW

Toroidal Beta, T ≤ 40%

Normalized Beta, N ≤ 9

Pulse Length 1 s

ETG Turbulence and Electron Thermal Transport

• NBI Heated Discharges– Most NBI power expected to deposit on

electrons, but Ti > Te observed

– Ion thermal transport at or near neoclassical levels

• Believed due to E×B flow shear rate greater than ITG/TEM growth rate

– Electron thermal transport remains poor• Flat Te profile in core

• ETG turbulence suspected

• HHFW Heated Discharges– Strong core electron heating as expected

– Ion thermal transport above neoclassical levels

• Large ITG/TEM growth rates when Te>Ti

– Electron thermal transport still the dominant loss mechanism

• Evidence for electron ITB

• ETG turbulence suppressed when Te>Ti

LeBlanc et al., NF 44, 513 (2004)Synakowski et al., PPCF 44, A165 (2002)

Growth ratesfor NBI heateddischarges.

Scattering of Electromagnetic Waves

Collective fluctuations with scale lengths much larger than the Debye length can scatter electromagnetic waves

11

Dk

Incident Beam:

Scattered Beam:

Turbulent Fluctuation:

Conservation of Energy:

Conservation of Momentum:

ii k

,

ss k

,

k

,

is

kkk is

A high frequency probe beam gives

sisi kk and

The wave vectors describe an isosceles triangle to givethe Bragg relation where is the scattering angle

2sin2 ikk

sk

ik

k

Spectral Density and Scattered Power

The spectral density is related to the Fourier transform and auto-correlation of the electron density ne where V is the scattering volume and T is the observation time

The density fluctuation magnitude is related to the spectral density

,

2

1~ 34

2

kSdkdnVT

e

The scattered power is experimentally measured. It is related to the spectral density.

where is the solid angle of observation, r0 is the classical electron radius, and L is the longest length of the scattering volume.

trne , trki

e

TV

FTe etrndtrdkn

, , 3

kiACe

FTe

endd

knVT

kS

,

, 1

,

3

2

trntrndtrdVT

n ee

TV

ACe , ,

1, 3

, 2d

20

2

kSLPrdd

Pi

s

The total power scattered from a coherent fluctuation with a single Fourier component is

This expression is used to estimate the minimum detectable density fluctuation for a receiver system.

~4

1 2222

0 iies PLnrP

High-k System Overview

Bay K

Bay H

• Probe beam– High power BWO millimeter wave source– Launched toroidally– Collimated with waist at scattering region

• Low loss waveguide transmission– HE11 mode in corrugated waveguide– Couples to free space Gaussian beam

• Low loss water-free quartz windows– Etalon optimized– 85% transmission

• Inboard / outboard launch configurations– Inboard: kr = 0 - 20 cm-1

– Outboard: kr = −20 - 0 cm-1

• Superheterodyne receiver system(UC Davis)

– Five channels– Low noise mixers and amplifiers

Inboard Launch

Smith et al., RSI 75, 3840 (2004)

Probe Beam

Frequency fi 280 GHz

Wavelength i 1.07 mm

Incident Power Pi 200 mW

1/e2 Intensity Half-Width w0 4 cm

Wave Number ki = 2⁄ i 58.7 cm-1

Beam Divergence = 2i⁄ w0

0.98°

Rayleigh Length zR = w02⁄ i 470 cm

• High frequency probe beam– Low frequency drift-wave turbulence withf ≤ 1 MHz, so the Bragg relation applies

• Toroidal launch configuration– Beam launched ~5 in. above toroidal

midplane and directed ~5° downward– Measure primarily kr

• Good wave number resolution– k ~ 2⁄ w0 = 0.5 cm-1

• Diffraction will be significant• Circularly polarized

• Linear GK theory predicts the turbulent spectrum peaks at

• Scattering angles = 0°-20° will cover |k| = 0-20 cm-1

2. ek

Scattering Volume and Spatial Resolution

• For isotropic fluctuations, the scattering volume is the overlap of the probe beam and receiving beam

With k = 10 cm-1 → r = 94 cm This is not satisfactory.

However, the situation is different for anisotropic fluctuations…

2/sin

24 00

w

k

wkr i

• Drift wave turbulence is anisotropic with the fluctuation wave vector nearly perpendicular to the local magnetic field.

This imposes an additional constraint on the scattering process. Some regions common to the launch and receiving beams are, in essence, “detuned” because the magnetic field changes direction. Thus, the scattering volume is constricted and spatial resolution improves.

0 ||kkBk

• Toroidal curvature of the magnetic field detunes the detector from some scattering regions within its field of view.

• This beneficial effect is most pronounced at small scattering angles and low aspect ratio.

• As a low aspect ratio device, NSTX is inherently well suited for spatially resolved scattering measurements.

Anisotropic Fluctuations and Scattering Volume Constriction

• Consider a ray-like probe beam launched on the toroidal midplane with the detector also located on the midplane. Fluctuation wave vectors are necessarily radial.

• If fluctuations were isotropic, the detector could be aimed anywhere along the probe beam and still detect scattered signal.

• However, with anisotropic fluctuations, the detector must be aimed at the particular region of the probe beam that satisfies the condition above. If the detector is aimed at other regions of the probe beam, scattered signal will not be observed.

Mazzucato, PoP 10, 753 (2003)

Ray Tracing Simulations and Instrument Selectivity

•System design based upon ray tracing simulations that include diffraction and refraction effects.

•The instrument selectivity function (plotted below) illustrates scattering volume constriction due to toroidal curvature.

–For k=10 cm-1 → r ≈ 13 cm

Gaussian Probe Beam

• The probe beam is transmitted to Bay H via corrugated waveguide. Upon launch from waveguide, the beam expands to the proper size.

• A focusing mirror creates a highly collimated beam with the beam waist positioned at the scattering location.

• The beam enters the vessel through a water-free quartz window and gate valve.

• An in-vessel steering mirror directs the beam for either inboard or outboard launch.

• An in-vessel collection mirror focuses scattered signal for exit through five water-free quartz windows.

• The scattered beams converge for coupling into waveguide.

• Beam parameter

• Transformation for free space propagation

• Transformation at optical element

zwi

zRzq 2

11

dzqdzq

fzqzq beforeafter

111

ikzzw

rz

z

zR

ri

eeezw

wEzrE R

2

22

arctan0

0,

Corrugated Waveguide and Beam Splitter

• Circular Corrugated Waveguide– Low loss propagation in HE11 mode

• Matches to gaussian beam in free space• 2×10-7 np/m attenuation

– Geometry• 2.5 in. inner diameter• .01 or .02 in. corrugations

– 90° Miter bends for turns

• Wire Grid Beam Splitter– Needed to tap off BWO

LO for receiver system– 10 m wires spaced 25

m apart– Located near Bay H– BWO LO transported via

waveguide to receiver system below Bay K

Cavallo et al., RSI 61, 2395 (1990)

Etalon Optimized Quartz Windows

Adjacent transmitted beams have a phase difference

Single pass reflection, absorption, and transmission factors are

The transmitted power including all internal reflections is

The transmitted power has a local maximum when

cos

4dn

ARTeAn

nR d

1 1 1

12

2sin41 22

2

0 RR

T

P

PT

,3,2,1 with cos2

2 Nn

NdN

• Windows are a significant source of power loss for millimeter waves

• Windows must be thick enough to withstand vacuum forces, but as thin as possible to minimize power loss

• Besides diamond, water-free quartz is the best material for high strength and low power loss

– Index of refract: n = 1.96– Absorption coef: = .08 Np/cm– Rupture stress: P = 90 MPa

• Etalon analysis gives the thickness for a transmission local maximum

– 84.6% transmission at 34th transmission peak

Launch Configurations

OutboardLaunch

InboardLaunch

AutoCAD Design

BWO LO

Bay HLaunch

Bay KReceive

SphericalFocusing Mirror

f = 1.5 m

Steering Mirror forInboard/Outboard

Launch

BeamSplitter

Collection Mirror

f = 1.5 m

Bay H Launch

Bay K Receive

FiveScatteredSignals

Launch Port

Receiving Port

BWO Millimeter Wave Source and Power Supply

• Thomson CSF BWOModel CO 10-1

– O-type backward wave oscillator (BWO)

– High power• ~200 mW

– Frequency tunable• 275-290 GHz• ~15 MHz/V

– Lifetime: ~2000 hrs

• Siemel Power Supply– High voltage: 12 kV– Low ripple: < 15 mV– Anode current controls

BWO power output– Cathode voltage controls

BWO frequency– Configured to lower

BWO filament current between shots to conserve BWO lifetime

F-bandWaveguide

Output

CoolingLines

High Voltage LinesThomson CSF BWO

Siemel Power Supply

Superheterodyne Receiver System (UC Davis)

Scattered Power Calculation

~4

1 2222

0 iies PLnrP

The receiver system NEP is 7×10−13 W.

This is the minimum detectable scattered power.

Ps 7×10−13 W

Pi 0.1 W

r0 2.82×10−15 m

L 5 cm

i 1.07 mm

ne 5×10−19 m-3

313 m 105.3~ en

Minimum detectabledensity fluctuation

6107~

e

e

n

n

3101~

Tee

e

Lkn

n

For ETG fluctuations, theory predicts

Fukui University

Possible Upgrade: 20 W Gyrotron at 354 GHz

•November 2004: Installation of in-vessel hardware and port covers

•Additional Ex-Vessel Tasks– Motorize actuators– Fabricate beamsplitter from wire

grid and miter bends– Purchase focusing mirror– Finalize design and construct

optical boxes for Bays H and K– Control computer (control room)– DAQ computer (test cell)

Current Status and Remaining Tasks

•Major Hardware Acquired– BWO/power supply– Bay H port cover– Bay K port cover– Receiver system– Entrance/exit windows– Steering mirror– Collection mirror– Heavy-duty actuators– Waveguide and miter bends

Summary

• Electron thermal transport remains poor in all magnetic confinement devices

• ETG modes with k┴e≈1 may drive electron thermal transport in NSTX

• Scattering measurements can provide direct experimental observation

of ETG modes

• The high-k scattering system is currently being installed on NSTX

– BWO millimeter wave source

– Toroidal launch of probe beam to observe kr spectra

– 200 mW probe beam at 280 GHz and 8 cm waist

– Low aspect ratio geometry improves spatial resolution

– Five channel, low noise receiver system to observe fluctuations with

|kr|=0-20 cm-1

– Low loss waveguide transmission and water-free quartz windows

• Possible upgrade to 20 W gyrotron at 354 GHz