Dust Measurements With The DIII-D Thomson system The DIII-D Thomson scattering system, consisting of eight ND:YAG lasers and 44 polychromator detection boxes, has recently been used to observe the existence of dust in the SOL and divertor regions during normal plasma operation. In order to maintain absolute sensitivity calibrations of the Thomson system, each polychromator box contains detector channel with filter at the laser wavelength that is sensitive to Rayleigh/Mie scattered light. Dust particles residing the region of the laser beam viewed by a detector produce a very large scattered light signal in this Rayleigh channel compared to the usual Thomson scattering from the plasma electrons. Background from stray light scattered in the machine and neutrons can be measured and average dust density profiles for various plasma configurations are determined.
Motivation Multiple sources of dust exist at DIII-D Flakes from redeposited CxDy films High Heat Flux Generation Particles from vent activities Degradation of grafoil compliant layers No Current Dust Measurements During Run Campaigns Dust of interest for safety and plasma contamination
Thomson System Measures Electron Temperature and Density Throughout DIII-D 3 paths through the vessel to cover from the center to the edge and the divertor Lasers follow a common path below the vessel and split into 3 groups before entering vessel Scattered light from the laser beams is images on arrays of fiber optics and transported to 44 polychromators which measure the temperature and density
Polychromators are a very flexible method of measuring Thomson scattered light 5 to 7 Bandpass filters are used in each polychromators to divide the scattered light into wavelength ranges The light which passes each bandpass filter is focused on an APD for detection Electronics have background subtracted and background outputs to obseved the scattered light and monitor the background
Polychromators filters are chosen for sensitivity to a wide range of temperatures Polychromators are sensitive to a wide range of temperatures from 1 ev to 10keV One bandpass filter measures light at the laser wavelength for an absolute density calibration by Rayleigh scattering in Argon Narrow 1062nm channel is used for low temperature plasmas in the divertor and edge of the core
Cameras monitor the alignment of lasers before entering and after the machine Lasers beams path through focusing lenses before entering the machine 3mm wide laser beam in the vessel Imaging fiber bundle views 10mm high by 5mm wide region along the beam Beam profile measurement is important for calculating dust particle sizes
Rayleigh Scattering Is Used for Absolute Density Calibrations Rayleigh scattering: = 10 /3k 4 a 6 Typical Laser Energy 0.5J Typical sensitivity for the DIII- D Thomson system 400 counts/torr Ar/J Stray light levels (signal at 0 Torr) are generally low except for a few divertor and horizontal channels
Laser channels are not used for Thomson scattering but are sensitive to dust Consider a single large particle producing a signal equivalent to the Rayleigh scattering from 1Torr of Argon (3.3 10 16 cm -3 at 22 o C) Laser Flat power profile Radius 0.15cm Scattering region 1cm long 200cnts/Torr = 8.66 10-14 /Ar atom Ar -> a = 1.1A radius Size = 1.1 (200/8.66 10-14 ) 1/6 Particle = 400A = 40nm radius
Sample Dust Event Events with large signals compared to the typical stray light are observed during plasma operations by the Rayleigh channels of the Thomson system Neighboring channels do not see any increase in scattered light in the Rayleigh channel during the event Neighboring channels seeing increasing signals from light coming from the inner wall
Background Light is monitored for every channel Background light is monitored by a DC coupled output for each channel before every laser pulse Neighboring channels view overlapping regions of the inner wall Events due to light from the wall show on multiple channels Narrow filters transmit little background light High background light events must be excluded due to amplifier saturation at high signal levels
Selection Criteria Check candidate events where plasma current > 500kA Look for large changes in signal compared to events before and after the candidate event Require large changes in signal for nearby data with the same laser (stray light can vary significantly with laser parameters) Ignore events which show simultaneously in multiple channels (mis-aligned laser pulses can produce high stray light levels on multiple channels simultaneously) Require low background light levels to eliminate saturation and high noise levels (detector readout can oscillate between different signal levels with the background level approaches opamp saturation
Many Shots Have No Events and Less Than 1% of Slices Contain A Candidate Event Resulting events mostly occur near the edge of the plasma Dust density = 4000 m -3 in SOL Large number of shots are scanned to produce statistics
Additional Detector Added to the System to Detect Gamma And Neutron Induced Events Noise from sources other than scattering from dust needs to be measured to exclude events from other sources Similar detector and electronics are mounted behind an opaque plate to monitor background Events are believed to be due to and n fluxes in the Thomson room Event rate on the neutron detector is similar to event rates in the center of plasma and on shifted wavelength channels
Event Rates in Wavelength Shifted Channels Is Proportional to the Neutron Rate Signal from scintilators used to monitor the neutron rate is correlated with the event rate in the non-yag channels Primary source of noise events seen on the monitor detector is neutrons from the plasma Additional statistics are required to compare the neutron detector to the neutron rate directly
The Pulse Height of the Observed Signals Often Saturate The YAG Channel Amplifier Intensity of the signals provides information about the size of the scattering particle Strong (a 6 ) dependence on particle size produces tail of large signals as particle size increases Majority of observed events saturate the detector on the laser line channel Largest signal events are from dust particles too large for a Rayleigh approximation
1062 nm Filters Are Sensitive to YAG Light with a lower gain than the YAG channels Divertor and Edge core polychromators contain 1062nm filters for resolving low temperatures in the plasma A small fraction of the ND:YAG light can be observed in these filters Leakage can be used to extend the range of resolved signals in the Thomson system Suggests some dust events produce signals up to 20 times the saturation current of the laser channel detector
Divertor analysis is complicated by large stray light levels, stray light fluctuations and low temperature plasmas Light scattered from the exit window and entrance baffles can produce strong YAG light backgrounds and is very alignment sensitive 3000 Stray light can saturate divertor YAG channels Pulsed Signal Shot 118237 Lowest channels which view the floor closest to the entrance cannot be used if high stray light levels Only 1 laser so the data sample of the core 2000 1000 Lowest channel Second channel Third High density, low temperature plasmas can produce significant signals on 1062nm and YAG filters 0 2000 0 2000 4000 6000 Time (ms)
Estimates of the Particle Size Can Be Made From The Pulse Height Justin Burkart (GP1.0037 Tuesday Morning) Has a Detailed Presentation on Particle Size estimates from the spectra data Size distribution is convoluted with a laser power profile to produce the observed spectra Deconvolution with a model of the laser profile can produce a size distribution for the observed particles
Dust events are observed during the pre-fill before shots Thomson trigger modified to take data during the gas prefill for 300ms before the shot Events can be observed before the shot but only a limited amount of data was taken with the Thomson system before the shot After shot analysis is not possible for most shots due to very high stray light levels due to vibrations Events before the shot most likely on first shot of the day
Dust Densities Can be Correlated With Many Plasma Parameters Beginning studies of dust events correlations with plasma parameters Comparison of dust events/slice for various parameters averaged over the time period searched for dust events USN operations during 2005 were significantly more likely to produce dust events than LSN Higher injected neutral beam power and confinement are also correlated with higher dust densities
The Gap Between the Inboard Wall and the Plasma Influences the Observed Event Rate The inner wall may be a source of the dust observed The observed event rate increases for shots with a low inboard gap The outboard gap does not significantly influence the event rate More work will be required to understand the factors which influence the dust production rate
Summary Dust can be observed in the DIII-D vessel with the Thomson scattering system. Dust events can produce very large signals relative to the signals normally seen from Rayleigh and Thomson scattering Background of large signal events is primarily due to neutrons and gammas striking the APDs in the Thomson room. A reference detector has been added to the system to monitor these events Dust analysis can be performed on data taken during plasma operations by the Thomson system Dust is rarely observed before plasma discharges and after discharge analysis is limited by vibrations of the Thomson system during disruptions Spectral information can be used to calculate size distributions of the observed particles
Upgrades for next run campaign Better monitoring of beam profiles for size distribution calculations Switching 1062nm filters to optimize spectral measurements for channels with high dust rates More data with the neutron detector to better understand background levels Extend studies to look at dust density under different plasma conditions and new lower divertor configuration Completion of YAG alignment system to limit stray light variations in the divertor