Infrared Thermography in the Tokamak à Configuration Variable
Martim Zurita, H. Reimerdes, C. Colandrea, H. Elaian, M. Pedrini · 2026
Read Infrared Thermography in the Tokamak à Configuration Variable by Martim Zurita, H. Reimerdes, C. Colandrea, H. Elaian, M. Pedrini (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-11912.txt.
Introduction
Zurita and co-authors describe infrared thermography at the TCV tokamak, which watches the surface heat of graphite tiles. Three camera systems cover the lower inner wall, the floor, and the lower outer wall. Tile temperatures feed the THEODOR code to infer heat flux profiles after each discharge. A reduced frame mode pushes capture past 10 kHz for fast events. Recent upgrades add heated tiles, new rooftop tiles, and a long-pass filter that dims stray plasma light. Stray light and tile property estimates remain the main error sources.
Catalog record
| CREATOR | Martim Zurita, H. Reimerdes, C. Colandrea, H. Elaian, M. Pedrini |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 57291 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.11912
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Infrared Thermography in the Tokamak à Configuration Variable Thanks: Author list at the end of the paper. Martim Zurita, H. Reimerdes, C. Colandrea, H. Elaian, M. Pedrini Abstract In the Tokamak à Configuration Variable (TCV), infrared thermography (IR) is currently composed of the horizontal, vertical, and tangential infrared systems (HIR, VIR, TIR), which all use Equus 81k M cameras. The IR diagnostics obtain the surface temperature of TCV’s graphite tiles for post-discharge analysis. Target heat flux profiles are inferred from the tile temperature with the THEODOR ( Th ermal E nergy O nto D ivert or ) code. Fast transient analysis is possible in reduced frame mode, with acquisition frequencies above 10kHz. The main views are the lower inner wall for HIR, the floor for VIR, and the lower outer wall for TIR. The HIR camera can also be moved to view the midplane inner wall, while TIR can be moved to see the midplane inner wall and the upper outer wall, mainly to measure synchrotron radiation and heat deposition due to runaway electrons. Recent developments in TCV’s IR systems include (i) tile diffusivity and conductivity measurements to assure the precision of heat flux estimates; (ii) the addition of one new VIR heated valley tile and two rooftop TIR tiles, for measurements of fast heat flux transients; (iii) the implementation of long-pass wavelength filter of 4095 nm, to diminish the measurement of plasma parasitic infrared light, mainly from deuterium 5 → 4 5\rightarrow 4 emission at 4051 nm. Despite these developments, the main sources of uncertainty for IR in TCV are still parasitic infrared light and the determination of the surface layer heat transmission factor, both of which mainly affect the VIR system. École Polytechnique Fédérale de Lausanne (EPFL), Swiss Plasma Center (SPC), CH-1015 Lausanne, Switzerland 1 Introduction Infrared thermography (IR) is an imaging technique that translates the infrared radiation emitted by an object into its surface temperature. This is done by employing the total photon radiance from blackbody theory, R_{BB} [ 1 ] : R_{BB}(T)=\int_{\lambda_{1}}^{\lambda_{2}}\frac{2c}{\lambda^{4}}\frac{d\lambda}{\exp\left(\frac{hc}{\lambda k_{B}T}\right)-1}\quad\left[\frac{\text{1}}{\text{sr}\;\text{s}\;\text{m}^{2}}\right] (1.1) where c=2.99792\times 10^{8}\;\text{m/s} is the light speed, h=6.62607\times 10^{-34}\;\text{J s} is Planck’s constant, k_{B}=1.38065\times 10^{-23}\;\text{J/K} is Boltzmann’s constant, T is the object surface temperature, and \lambda_{1} and \lambda_{2} are the minimum and maximum wavelength measured by the camera detector. R_{BB}(T) measures the number of photons per solid angle, time and area emitted by the object. Inverting R_{BB}(T) gives the surface temperature as a function of the infrared radiation. In tokamaks and stellarators, infrared thermography is used to evaluate the temperature of the device walls in contact with the plasma. By employing heat diffusion equations, the heat flux due to the plasma can then be estimated. This paper details the IR systems of the Tokamak à Configuration Variable (TCV) [ 2 ] . Sec. 2 describes TCV’s infrared cameras; Sec. 3 , the calibration to temperature with thermocouples; Sec. 4 , the target heat flux estimation with the THEODOR code; Sec. 5 , the windows used for IR in TCV; Sec. 6 , the tiles — with emphasis for their thermal properties and new tiles developed for fast transient analysis; Sec. 7 discusses the infrared parasitic plasma radiation measured by the cameras; Sec. 8 shows a comparison between the VIR and TIR heat flux profiles and Sec. 9 summarizes the results. 2 TCV’s infrared cameras 2.1 General information All infrared cameras in TCV are currently IRCAM Equus 81k M (specifications in Table 2.1 ), adapted by the manufacturer to withstand ‘high’ magnetic fields (0.1-0.2 T). The detector is a focal plane array (FPA) made of mercury cadmium telluride (MCT). A Stirling cooler refrigerates the detector, minimizing its own infrared emission and reducing noise due to thermally excited electron holes. The sensor is a complementary metal oxide semiconductor (CMOS), with each pixel having its own amplifier. The length of each pixel (i.e. the pixel pitch) is of 30\;\mu\text{m} . The integration mode of the data is integrate-then-read. The cameras’ spectral response goes from 3.7\;\mu\text{m} to 4.8\;\mu\text{m} , falling in the mid-wavelength infrared (MWIR) range. Nevertheless, as of TCV’s discharge 79993, all three cameras are filtered with a 4095\;\text{nm} long-pass wavelength filter to decrease the amount of infrared light measured from the plasma. The maximum and minimum frame sizes are respectively 320x256 (approximately 81 thousand pixels, hence the camera name Equus 81k M) and 64x2 pixels. Decreasing the frame size increases the acquisition frequency, as described in Eq. ( 2.1 ). The spatial resolution ranges from about 2.0mm/pixel (VIR and TIR) to 1.3mm/pixel (HIR). The minimum integration time (i.e. the interval the camera pixels are exposed to light) is \Delta t_{int}=1\;\text{$\mu$s} , which can be used for cases with very high tile temperatures (e.g. HIR of discharge 89000, not shown). Nevertheless, the standard value is \Delta t_{int}=1\;\text{ms} for L-mode, ohmically heated plasmas. TCV’s infrared cameras are controlled by a C++ program with MDS events (Sec. 3.1 of [ 3 ] ). As of 2025, the quality of the data is monitored shot by shot in TCV’s DdJ panel [ 4 ] . Table 2.1: General information about TCV’s IR cameras Equus 81k M by IRCAM. Detector MCT FPA Sensor CMOS Integration mode Integrate-then-read Wavelength sensitivity 3.7\;\mu\text{m} (or 4.1\;\mu\text{m} filtered) to 4.8\;\mu\text{m} Max window size 320x256 pixels Min window size 64x2 pixels Pixel pitch 30\;\mu\text{m} Spatial resolution 1.3mm/pixel (HIR) and 2mm/pixel (VIR & TIR) Integration time 1\;\mu\text{s} to 10\;\text{ms} Acquisition frequency 200 Hz (standard) to 20kHz (max) 2.2 Interval between frames As described in the cameras user manual and in Sec. 3.1.2 of [ 3 ] , the minimum interval between frames of TCV’s 81k M IR cameras is \displaystyle\Delta t_{frame}^{(min)} \displaystyle=\Delta t_{int}+\Delta t_{read}+\Delta t_{reset} (2.1) \displaystyle=\Delta t_{int}+N_{px}/(4f_{clk})+20/f_{clk} where \Delta t_{int} is the integration time, i.e. the interval in which the camera receives light for each frame. \Delta t_{read}=N_{px}/(4_{clk})=N_{px}/27 is the readout time of the data, N_{px} is the number of pixels used by the camera (max 320\times 260=81920 , min 64\times 2=128 , as in Table 2.1 ) and f_{clk}=6.75\;\text{MHz} is the camera mother-clock frequency. \Delta t_{reset}=20/f_{clk}=20/(6.75\;\text{MHz})=3\;\mu\text{s} is estimated as the time to reset the pixels for the next measurements. It is also possible to change f_{clk} to 10\;\text{MHz} , decreasing \Delta t_{read} by 32%. However, this leads to pixels interfering to each other (pixel cross-talk), which can distort the data. Substituting f_{clk}=6.75\;\text{MHz} in Eq. ( 2.1 ) gives the general formula for the minimum interval between frames \Delta t_{frame}^{(min)}=\Delta t_{int}+\frac{1}{27}N_{px}\;\mu\text{s}+3\;\mu\text{s} (2.2) In full frame mode, N_{px}=320\times 260=81920 , yielding \Delta t_{read}=3.0\;\text{ms} and hence \Delta t_{full\;frame}^{(min)}=\Delta t_{int}+3.0\;\text{ms} (2.3) Since in general \Delta t_{int}<3\;\text{ms} , the full frame interval is dominated by the readout time and the maximum acquisition frequency is then obtained for \Delta t_{int}\rightarrow 0 , f_{acq}=1/\Delta t_{full\;frame}^{(min)}=333\;\text{Hz} . In practice, to assure reliable measurements, the used interval between frames is increased by a factor of 10% and an adjustment \Delta t_{adjust} is added, \displaystyle\Delta t_{frame} \displaystyle=(\Delta t_{int}+\Delta t_{read}+\Delta t_{reset}+\Delta t_{adjust})\cdot 1.1 (2.4) \displaystyle=(\Delta t_{int}+N_{px}/27\;\mu\text{m}+3\;\mu\text{s}+15\;\mu\text{s})\cdot 1.1 \Delta t_{adjust}=15\;\mu\text{s} is important for low number of pixels ( N_{px}<2400 ), specially at extremely low integration times ( 1\sim 5\;\mu\text{s} ) to assure the identification of the frame instant. Decreasing the readout time too much can also lead to frames not being recorded and thus a minimum