Contact us
My IOPscience
Thomson scattering on COMPASS — commissioning and first data
This content has been downloaded from IOPscience. Please scroll down to see the full text.
2012 JINST 7 C01074
View the table of contents for this issue, or go to the journal homepage for more
Download details:
IP Address:
This content was downloaded on 31/03/2015 at 04:05
Please note that terms and conditions apply.
R ECEIVED: November 17, 2011
ACCEPTED: December 27, 2011
P UBLISHED: January 19, 2012
O CTOBER 13–19, 2011
a,c R. Scannell,d M. Tripsky,c V. Weinzettl,a M. Hron,a
M. Aftanas,a,b,1 P. Bohm,
a M. Walshe and P. Bilkovaa
R. Panek,a J. Stockel,
a Institute
of Plasma Physics, Academy of Sciences of the Czech Republic, v.v.i.,
Za Slovankou 3, 182 00 Prague 8, Czech Republic
b Faculty of Mathematics and Physics, Charles University,
Ke Karlovu 3, 121 16 Prague, Czech Republic
c Faculty of Nuclear Sciences and Physical Engineering, Czech Technical University,
Brehova 7, 11519 Prague 1, Czech Republic
d Culham Centre for Fusion Energy, Culham Science Centre
Abingdon, Oxfordshire OX14 3DB, United Kingdom
e ITER Organization
CS 90 046, St Paul lez Durance Cedex 13067, France
E-mail: [email protected]
A BSTRACT: High-resolution edge and core Thomson scattering systems have been designed with
the main aim to investigate electron density and temperature profiles on the COMPASS tokamak
(R = 0.56 m, a = 0.18 m, BT max = 2.1 T). The spatial resolution is optimized namely for the pedestal
studies (radial spatial resolution ∼ a/100). Currently, the core Thomson scattering system is commissioned, calibrated and tested in tokamak discharges. This contribution describes particular steps
of optical alignment and calibrations. Moreover, we show control tools developed for electronic
settings, calibrating procedures and laser beam position measurement. Both the calibration data,
i.e. Raman scattering obtained in the N2 at pressures up to 200 mbar, and the first Thomson scattering data measured by the system in the H discharges are presented.
K EYWORDS : Plasma diagnostics - charged-particle spectroscopy; Plasma diagnostics - interferometry, spectroscopy and imaging
1 Corresponding
c 2012 IOP Publishing Ltd and SISSA
2012 JINST 7 C01074
Thomson scattering on COMPASS — commissioning
and first data
Outline of diagnostic
Calibration and settings
3.1 Settings and calibration of the detection part of the diagnostic system
3.1.1 Electronic set-up of the polychromators
3.1.2 Spectral calibration
3.2 Alignment of the system and laser beam properties
3.3 Absolute calibration of the whole diagnostic by means of Raman scattering
First data of Thomson scattering diagnostic in a plasma on COMPASS
The COMPASS tokamak (R = 0.56 m, a = 0.18 m, BT = 0.8–2.1 T, κ = 1.6) originally from Culham Centre for Fusion Energy (CCFE) Culham, United Kingdom, has been reinstalled in Institute
of Plasma Physics (IPP) Prague, Czech Republic [1]. The high resolution multi-point Thomson
scattering (TS) diagnostic has been designed to provide the electron temperature and density profiles on the COMPASS tokamak [2]. The diagnostic consists of two sub-systems, observing core
and edge plasma region. In this paper, commissioning of both core and edge TS systems and the
first measurements performed by the core TS system are presented.
Outline of diagnostic
Thomson scattering (TS) diagnostic on COMPASS is based on Nd:YAG (yttrium aluminium garnet) and APD (Avalanche photodiodes) technology [2]. The diagnostic consists of two parts, one
focused on the core plasma region, the other on the edge of the plasma (table 1, for details see [2]).
Two identical and independent Nd:YAG lasers have been incorporated and thus better flexibility
achieved [3]. The lasers can be fired either simultaneously to double the probing energy (3 J), or
independently when electron density is high enough to obtain a sufficient amount of scattered light.
Then the lasers can be operated in regime doubling the repetition rate (60 Hz), or in “burst mode”
with arbitrary pulse separation from 1 µs to 16.6 ms. Laser beams propagate from the top of the
tokamak downwards and the scattered radiation is collected by two collection objectives (for core
and for edge region). Collection optics has been designed in IPP AS CR, v.v.i. [4]. Optical fibres
have been designed to match well with parameters of collection lenses as well as with parameters
2012 JINST 7 C01074
Table 1. Overview of parameters of the TS diagnostic on Compass.
Field of view
Number of points
Number of polychromators
Vertical spatial resolution
Temporal resolution
of the laser beam [2]. Both the core TS and edge TS diagnostics are equipped with a split fibre for
a purpose of alignment.
Scattered light is analysed in polychromators with spectral filters, where it is detected by
Avalanche photodiodes (APD) cascade. The fourth generation of polychromators designed on
MAST, CCFE, U.K. (original design by WSL — M.J. Walsh) is used for COMPASS TS diagnostics [5–7]. A new set of spectral filters had been adapted for needs of COMPASS, taking into
account a lower temperature range expected there in comparison with the MAST tokamak [2, 8].
The COMPASS TS data acquisition has been built as a modular system providing easy expansion and flexibility. An embedded computer allows future real-time data processing. The polychromator outputs (scattered signals) are registered on both sixty synchronized 2-channel 8-bit
R and two 961 GSample/s analog-to-digital convertor (ADC) cards from National Instruments
channel 16-bit 500 kSamples/s ADC from D-tAcq. The ADCs have variable input ranges from
50 mV to 5 V.
Calibration and settings
Settings and calibration of the detection part of the diagnostic system
Electronic set-up of the polychromators
The signal from each APD is electronically processed by amplifiers integrated on board of each
polychromator. For each of the signals two outputs are used - a fast channel (high-pass filter from
200 kHz) and slow channel (low-pass filter up to 200 kHz). The fast channel is used to measure
scattered signal waveform of duration of approximately 10 ns. There, slow fluctuations, caused
by plasma events, are filtered out, because it would limit the ADC dynamic range. This has very
little effect on scattered signal since the bandwidth of the laser pulse is 40 MHz. The slow channel
passes low frequency signal corresponding to plasma background light.
The scattered radiation intensity is proportional to electron density. The signal level in different
spectral channels (channel ratios) varies with electron temperature and can be different by a factor
of 3 at constant electron density.
2012 JINST 7 C01074
Electron temperature range
Electron density resolution
Core TS
Edge TS
−38 to 195 mm above mid-plane 200–300 mm above mid-plane
8.1–12.4 mm
3–4.6 mm
16.6 ms, 33.3 ms,
burst regime (two samples distant by 1 µs or more @ 30 Hz)
30 eV–5 keV with error below 10% for electron density 1019 m−3
Error below 10% for electron density 1019 m−3
two independent Nd:YAGs, 1064 nm, 1.5 J @ 30 Hz each
1 GSample/s
The APD gain increases with the APD reverse bias voltage and decreases with room temperature. The APD gain is set to about 250 kV/W at 25◦ C in a reverse bias voltage specified by
the manufacturer and it is set to be constant over all operation temperatures by using of a thermal
compensation circuit with a temperature sensitive diode inside the APD package.
R environment has
To preset the bias voltage and setup temperature compensation, LabVIEW
been used to make monitoring, setup and testing of parameters more efficient.
Spectral calibration
Spectral calibration, necessary for determining electron temperature from Thomson scattering measurement, is a process of obtaining the spectral response of the individual channels in polychromators (and optical fibre bundles) using a white-light source with measured spectral density. Polychromators used on the TS diagnostic on COMPASS are equipped with five spectral filters.
Schematic layout of the spectral calibration is in figure 1. During measurement, a halogen
lamp (Oriel LSB116/5) coupled to monochromator (Newport Cornerstone 260, grating model
77990) is used. Light with wavelengths shorter than 610 nm is blocked by a longpass filter to
assure that higher order of diffraction originating on the monochromator grating does not affect the
calibration process. Chopper is incorporated to the calibration scheme to provide additional precision by subtracting background signal. Monochromator scannes the spectrum over the wavelength
range of 750–1064 nm. Its output is coupled to the polychromator and the response is recorded.
The white light source spectra is measured with a calibrated detector (Oriel 71651 and ThorLabs
PDA200C) — figure 2.
The control program written in LabVIEW makes the entire process of spectral calibration
routine. The program sets the monochromator, registers APD signals and records signal from the
calibrated detector. Entire wavelength range is scanned and these values are routed to the MatLab
code for calculating signal from each channel as a function of electron temperature (figure 3).
The final calibration data are stored in text files and are used during the TS measurement by data
processing routine to calculate electron temperature Te and density ne .
2012 JINST 7 C01074
Figure 1. Scheme of spectral calibration. The whole process is driven and controlled automatically by a
Figure 3. Final spectral calibration data: ratios of channel signals as a function of electron temperature.
Alignment of the system and laser beam properties
Laser beam propagates through plasma. Scattered light is focused by collection optics on the optical fibre bundles and transferred to the polychromators. If laser is misaligned, then only portion
of the scattered light is collected by the optical fibre and that might lead to uncertainty of electron
density determination. Therefore, the design of detection part of the TS system on COMPASS has
been done to avoid the effect of misalignment on the data [2]. A special split fibre bundle has been
designed and incorporated into the core fibre array in the middle of the core field view. Laser beam
width is calculated from a scan of left and right fibre half ratios over different positions of split fibre
bundle (figure 4). Consequently, optimal alignment is found looking for the left and right fibre half
ratio close to 1:1. A real setup of the system alignment is realized with help of Raman scattering process in the tokamak vessel filled by high pressure Nitrogen (pressure 240 mbar) allowing
measurements of left-right intensity ratios without performing a plasma discharge.
2012 JINST 7 C01074
Figure 2. Raw data of spectral calibration: channel (specified by filter parameters — center wavelength
and bandwidth) sensitivity (left y-axis); sensitivity of calibrated detector (right y-axis), power spectrum of
white-light source (right y-axis).
The laser beam width was computed iteratively by comparing the measured and simulated left
and right fibre half ratios (figure 4 right). The simulations were performed for different width of
the laser beam (in the figure 4 the simulated profile is for 930 µm laser beam width) of Gaussian
and flat-top profiles.
Laser beam width has been found to be approximately 1 mm, what is in good agreement with
the value determined by optical considerations [3], which was about 1.2 mm.
Absolute calibration of the whole diagnostic by means of Raman scattering
The absolute calibration of the TS diagnostic can be done either by Rayleigh or by Raman scattering. An advantage of Raman scattering is that the scattered radiation is at different but close
wavelengths to the incident radiation [9]. Therefore, the same arrangement as for TS measurement
can be used. Raman scattering in nitrogen has been used to perform absolute calibration of TS on
COMPASS. The polychromators possess spectral channels close to the laser wavelength and, at the
same time, very good rejection of the laser wavelength at these channels (laser light rejected by
order of 4.8), that enables this kind of calibration.
Raman calibration for core TS diagnostic has been performed in the vessel filled in with nitrogen at five different values of pressure between 27 mbar and 199 mbar. During Raman calibration, signals in two spectral channels with band closest to the laser wavelength (center 1057.7 nm,
bandwidth 5.5 nm and center 1049.3 nm, bandwidth 12 nm) are detected. Since the polychromator
transmits blue-shifted wavelengths, only anti-Stokes lines are considered for calibration. A detailed
description of calculation of the integrated Raman lines can be found in [9–11].
Raman cross-section (RCS) has been calculated by multiplying of integrated Raman line intensities and transmission of spectral filters measured during spectral calibration (figure 5).
The linear dependence of the detected signal on gas pressure is shown at figure 6. Linearity
is very good (correlation factor is 0,998–0,999). Slope of the fitted line is used for calculation of
2012 JINST 7 C01074
Figure 4. left: Principle of the laser beam width calculation: Calculation was iteratively performed as a
convolution (blue and red solid line) of the laser beam profile (dashed line) with transmission function of the
left and the right half of the split fibre bundle (with cut-out of the fibre bundle halves on background). right:
measured (dots) and simulated (line) split fibre signal ratio as a function of laser image position. In figure,
the signal ratio is taken as a ratio of the smaller value to higher one.
Raman factor R (figure 7):
R(j) = Temp ∗ kB ∗ (re2 /RCS(j)) ∗ slope(j),
where Temp is room temperature [K], kB is Boltzmann constant, re is electron radius [cm], RCS(j)
is Raman cross-section [cm2 ] for polychromator j and slope(j) is slope of linear fit of integrated
signal of Raman scattered light over pressure for polychromator j.
The Raman factor is directly used for calculation of electron density from the Thomson scattered signal.
First data of Thomson scattering diagnostic in a plasma on COMPASS
One of the first profiles of electron temperature and electron density measured on the COMPASS
tokamak is shown in figure 8. The plasma parameters of the shot are shown in figure 9. The
scattered data were of good quality, free of any stray-light. For our configuration of Thomson
scattering diagnostic, estimated number of photons on photodiode is 2000, with 250 kV/W gain
factor of photodiode the expected integrated signal is 0.1 nVs. This rough estimate is in a good
agreement with measured data, the signals in first spectral channels were around 0.2 nVs.
The Thomson scattering system on COMPASS has been successfully used for the first time in
February 2011, performing an absolute calibration by means of Raman scattering in nitrogen. The
first Thomson scattering signal has been measured in June 2011, while in September 2011 a full
2012 JINST 7 C01074
Figure 5. Detected Raman cross-section for each of 29 polychromators and two spectral channels close to
the laser wavelength (1057.7 nm and 1049.3 nm)
Figure 7. Raman factor calculated for 13 polychromators (core diagnostic part). The values have been
averaged over both 1st and 2nd spectral channels.
2012 JINST 7 C01074
Figure 6. Integrated signal of Raman scattered light for several pressures detected by the 1st and 2nd spectral
channels of one of polychromator. Full line indicates a linear fit to measured data.
Figure 9. Plasma parameters of the shot No. 2165: triggers for lasers, plasma current (upper picture) and
both radial and vertical positions of a plasma (bottom picture).
core profile has been obtained. Quality of Thomson scattered data is good, no straylight was
The edge TS diagnostic is going to be tested after the edge collection optics assembly that is
under way in these days.
2012 JINST 7 C01074
Figure 8. Profiles of electron temperature and electron density along the z axis measured at time 982.52 ms
(left column) and 1015.85 ms (right column). See the plasma parameters in figure 9.
The work was performed and supported from the grant GA CR No. 202/09/1467, under the contract
of association between EURATOM and IPP.CR and within the framework of the European Fusion
Development Agreement. The views and opinions expressed herein do not necessarily reflect those
of the European Commission. Authors would like to acknowledge the whole COMPASS team.
Authors wish to thank for kind help of MAST TS team from CCFE in United Kingdom.
[2] P. Bilkova et al., Design of new Thomson scattering diagnostic system on COMPASS tokamak, Nucl.
Instrum. Meth. A 623 (2010) 656.
[3] P. B¨ohm et al., Laser system for high resolution Thomson scattering diagnostics on the COMPASS
tokamak, Rev. Sci. Instrum. 81 (2010) 10D511.
[4] P. Bilkova et al., Progress of development of Thomson scattering diagnostic system on COMPASS,
Rev. Sci. Instrum. 81 (2010) 10D531.
[5] M. Walsh et al., Combined visible and infrared Thomson scattering on the MAST experiment, Rev.
Sci. Instrum. 74 (2003) 1663.
[6] R. Scannell et al., Design of a new Nd:YAG Thomson scattering system for MAST, Rev. Sci. Instrum.
79 (2008) 10E730.
[7] R. Scannell et al., A 130 point Nd:YAG Thomson scattering diagnostic on MAST, Rev. Sci. Instrum.
81 (2010) 10D520.
[8] M. Aftanas, R. Scannell, P. B´ılkov´a, P. B¨ohm, V. Weinzettl and M. Walsh, Design of Filters for
COMPASS Thomson Scattering Diagnostics, in WDS’09 Proceedings of Contributed Papers: Part II Physics of Plasmas and Ionized Media, Prague (2009) p. 144 [ISBN 978-80-7378-102-6].
[9] D.H. Froula, S.H. Glenzer, N.C. Luhmann Jr. and J. Sheffield, Plasma Scattering of Electromagnetic
Radiation: Theory and Measurement Techniques, Elsevier (2011) [ISBN: 978-0-12-374877-5].
[10] C.M. Penney et al., Absolute rotational Raman cross sections for N2,O2, and CO2, J. Opt. Soc. Am.
64 (1974) 712.
[11] R. Scannell, Investigation of H-mode Edge Profile Behaviour on MAST using Thomson Scattering,
Ph.D. Thesis (2007).
2012 JINST 7 C01074
[1] R. Panek, Reinstallation of the COMPASS-D tokamak in IPP ASCR, Czech. J. Phys. 56 (2006) B125.