Consideration of signal to noise ratio for an imaging bolometer for ITER.

An infrared imaging video bolometer (IRVB) is proposed for ITER having a tangential view of the entire ITER cross section. For the initial estimate of the signal level, a 840 m3 plasma is assumed to uniformly radiate 67.3 MW. A more detailed estimate of the signal strength is provided by synthetic images based on radiation data from SOLPS and SANCO models for the edge and core plasma, respectively. The Pt foil used as the radiation absorber would have the dimensions of 7 × 9 cm2 and a thickness of 16 µm that will stop 95% of the radiated power. Two different InSb based IR cameras having a sensitivity of 15 mK are considered for measuring the temperature rise of the foil due to the radiation. The first has 1280 × 1024 pixel2 and a frame rate of 105 fps. The second has 640 × 512 pixel2 and a frame rate of 1000 fps. The resulting IRVBs have 40 × 30 pixel2, 10 ms time resolution, and a signal to noise ratio (SNR) of 17 and 20 × 15 pixel2, 3 ms time resolution, and a SNR of 35, respectively. The synthetic image data give SNRs of 30 and 59, respectively.

Signal to noise ratio of upgraded imaging bolometer for KSTAR.

An InfraRed imaging Video Bolometer (IRVB) was installed on KSTAR in 2012 having a ∼2 µm × 7 cm × 9 cm Pt foil blackened with graphite and a 5 mm × 5 mm aperture located 7.65 cm from the foil with 16 × 12 channels and a time resolution of 10 ms. The IR camera was an Indigo Phoenix (InSb, 320 × 256 pixels, 435 fps, <25 mK). In 2017, the IRVB was upgraded by replacing the IR camera with a FLIR SC7600 (InSb, 640 × 512 pixels, 105 fps, <25 mK). The aperture area was reduced by approximately half to 3.5 mm × 3.5 mm, and the number of channels was quadrupled to 32 × 24. A synthetic image derived using the projection matrix for the upgraded IRVB from a Scrape Off Layer Plasma Simulator (SOLPS) model with 146 kW of total radiated power had a maximum signal of 7.6 W/m2 and a signal to noise ratio (SNR) of 11. Experimental data for a plasma with parameters similar to the SOLPS model (total radiated power of 158 kW) had a maximum signal of 12.6 W/m2 and noise equivalent power density (SNR) of 0.9 W/m2 (14).

Forward projection matrix derivation through Monte-Carlo ray-tracing of KSTAR infra-red imaging video bolometer (IRVB).

The infrared imaging video bolometer (IRVB) as a foil bolometry technique can be an alternative solution to the conventional resistive bolometer due to its electro-magnetic immunity and 2D plasma profiles. The plasma profile of the IRVB cannot be directly derived from the foil images due to the difference between the foil pixel number and the plasma pixel number and the line integrated nature of the incident rays on the foil. So, it needs tomography such as the Phillips-Thikhonov algorithm. The projection matrix constructing the foil image from the plasma very directly influences the tomography. So, the projection matrix needs to be constructed precisely. For the technique calculating the precise projection matrix, a forward Monte-Carlo ray-tracing method is proposed here, and this technique can provide the detailed descriptions of the foil image. And it can give enhanced performance in the reconstructions of the plasmas with spatially localized power.

Reconstruction of radiation profiles near the plasma boundary using an infrared imaging video bolometer in KSTAR.

An infrared imaging video bolometer using tomographic inversion can provide the total radiated power and 2-D radiation profiles, which are crucial information for impurity seeding experiments. Because large amounts of impurities exist at the plasma edge, accurate reconstruction of the radiation profiles near the material boundary is an important issue. In this study, two methods of boundary condition treatment are compared. One involves the exclusion of plasma pixels outside the boundary before reconstruction, whereas the other excludes these pixels after reconstruction. Phantom reconstruction tests are performed with D-shaped and divertor radiation profiles, and the second method shows an improvement in the boundary-reconstruction results compared with the first method. Using the second method, the radiation profiles of krypton (Kr) seeded H-mode plasmas in KSTAR are reconstructed. A significant amount of input power is dissipated through the Kr radiation. The 2-D reconstructed radiation profiles show that Kr mostly accumulates at the plasma core rather than at the edge.

Development of bolometric tomography technique for high-contrast radiation distribution in toroidal devices.

An extended Computed Tomography (CT) technique with a priori information based on Phillips-Tikhonov regularization has been developed to handle a high-contrast radiation distribution, which can result in large reconstruction errors in the region where radiation intensity is low, with few line of sight (LOS) data. Reference profiles generated from LOS data for every time slice are employed as the a priori information. In the extended technique, the weighting parameter for the reference profile is automatically determined from the LOS data, to avoid an inappropriate reference. The extended technique has been examined with a reconstruction test with the previously designed CT system using an infrared imaging video bolometer in JT-60SA. In the reconstructed profile of the high-contrast radiation distribution, the extended technique shows improvement of the weak radiation region, such as the scrape-off layer, without the deterioration of the strong radiation region, such as around the divertor. The results indicate that the extended CT technique can help to handle a high-contrast radiation distribution with few LOS data.

Conceptual design of imaging bolometer for use of computed tomography in JT-60SA.

An InfraRed imaging Video Bolometer (IRVB) system in JT-60SA has been designed for the two-dimensional (2D) Computed Tomography (CT) analysis of radiation. To achieve complete viewing of the whole poloidal cross section for plasmas having a low aspect ratio, a new IRVB concept, which has two sets of pin-hole camera systems viewing two different directions, has been introduced. Fields of view of the IRVB have been successfully designed by moving apertures to be sufficient for CT use without installation in tangential ports. The size of the aperture, which determines the width of the sightline of the IRVB channels, has also been optimized with respect to the reconstruction error with well-established solvers of tomography. The optimization indicates that the 7 mm square aperture is the best for this system, and the CT measurement without this optimization can become degraded. A synthetic image using an estimated radiation profile shows that the signal to noise (S/N) ratio of the designed IRVB is large enough to identify the incident radiation power. The result indicates that the designed IRVB can be used for the 2D CT measurement of radiation in JT-60SA.

Using a priori knowledge for developing bolometric tomography in toroidal devices.

In tomographic imaging of magnetically confined toroidal plasmas, a countermeasure against missing observation has been studied in terms of the adoption of prior information based on modelled plasma profiles. The Tikhonov regularization for image reconstruction is extended by the use of the Euclidean distance. A procedure of model fitting is designed in order to adaptively generate the reference image. The new method is tested on a typical example of ill-conditioned tomography, that is, the three-dimensional imaging-bolometer tomography in the large helical device. It has been found that the new method is useful for diminishing artifacts and thus for better recognizing the radiation structure of plasma.

Three-dimensional tomographic imaging for dynamic radiation behavior study using infrared imaging video bolometers in large helical device plasma.

A three-dimensional (3D) tomography system using four InfraRed imaging Video Bolometers (IRVBs) has been designed with a helical periodicity assumption for the purpose of plasma radiation measurement in the large helical device. For the spatial inversion of large sized arrays, the system has been numerically and experimentally examined using the Tikhonov regularization with the criterion of minimum generalized cross validation, which is the standard solver of inverse problems. The 3D transport code EMC3-EIRENE for impurity behavior and related radiation has been used to produce phantoms for numerical tests, and the relative calibration of the IRVB images has been carried out with a simple function model of the decaying plasma in a radiation collapse. The tomography system can respond to temporal changes in the plasma profile and identify the 3D dynamic behavior of radiation, such as the radiation enhancement that starts from the inboard side of the torus, during the radiation collapse. The reconstruction results are also consistent with the output signals of a resistive bolometer. These results indicate that the designed 3D tomography system is available for the 3D imaging of radiation. The first 3D direct tomographic measurement of a magnetically confined plasma has been achieved.

Improved signal to noise ratio and sensitivity of an infrared imaging video bolometer on large helical device by using an infrared periscope.

An Infrared imaging Video Bolometer (IRVB) diagnostic is currently being used in the Large Helical Device (LHD) for studying the localization of radiation structures near the magnetic island and helical divertor X-points during plasma detachment and for 3D tomography. This research demands high signal to noise ratio (SNR) and sensitivity to improve the temporal resolution for studying the evolution of radiation structures during plasma detachment and a wide IRVB field of view (FoV) for tomography. Introduction of an infrared periscope allows achievement of a higher SNR and higher sensitivity, which in turn, permits a twofold improvement in the temporal resolution of the diagnostic. Higher SNR along with wide FoV is achieved simultaneously by reducing the separation of the IRVB detector (metal foil) from the bolometer's aperture and the LHD plasma. Altering the distances to meet the aforesaid requirements results in an increased separation between the foil and the IR camera. This leads to a degradation of the diagnostic performance in terms of its sensitivity by 1.5-fold. Using an infrared periscope to image the IRVB foil results in a 7.5-fold increase in the number of IR camera pixels imaging the foil. This improves the IRVB sensitivity which depends on the square root of the number of IR camera pixels being averaged per bolometer channel. Despite the slower f-number (f/# = 1.35) and reduced transmission (τ0 = 89%, due to an increased number of lens elements) for the periscope, the diagnostic with an infrared periscope operational on LHD has improved in terms of sensitivity and SNR by a factor of 1.4 and 4.5, respectively, as compared to the original diagnostic without a periscope (i.e., IRVB foil being directly imaged by the IR camera through conventional optics). The bolometer's field of view has also increased by two times. The paper discusses these improvements in apt details.

Calibration of a thin metal foil for infrared imaging video bolometer to estimate the spatial variation of thermal diffusivity using a photo-thermal technique.

A thin metal foil is used as a broad band radiation absorber for the InfraRed imaging Video Bolometer (IRVB), which is a vital diagnostic for studying three-dimensional radiation structures from high temperature plasmas in the Large Helical Device. The two-dimensional (2D) heat diffusion equation of the foil needs to be solved numerically to estimate the radiation falling on the foil through a pinhole geometry. The thermal, physical, and optical properties of the metal foil are among the inputs to the code besides the spatiotemporal variation of temperature, for reliable estimation of the exhaust power from the plasma illuminating the foil. The foil being very thin and of considerable size, non-uniformities in these properties need to be determined by suitable calibration procedures. The graphite spray used for increasing the surface emissivity also contributes to a change in the thermal properties. This paper discusses the application of the thermographic technique for determining the spatial variation of the effective in-plane thermal diffusivity of the thin metal foil and graphite composite. The paper also discusses the advantages of this technique in the light of limitations and drawbacks presented by other calibration techniques being practiced currently. The technique is initially applied to a material of known thickness and thermal properties for validation and finally to thin foils of gold and platinum both with two different thicknesses. It is observed that the effect of the graphite layer on the estimation of the thermal diffusivity becomes more pronounced for thinner foils and the measured values are approximately 2.5-3 times lower than the literature values. It is also observed that the percentage reduction in thermal diffusivity due to the coating is lower for high thermal diffusivity materials such as gold. This fact may also explain, albeit partially, the higher sensitivity of the platinum foil as compared to gold.

A reconsideration of the noise equivalent power and the data analysis procedure for the infrared imaging video bolometers.

The infrared imaging video bolometer (IRVB) used for measurement of the two-dimensional (2D) radiation profiles from the Large Helical Device has been significantly upgraded recently to improve its signal to noise ratio, sensitivity, and calibration, which ultimately provides quantitative measurements of the radiation from the plasma. The reliability of the quantified data needs to be established by various checks. The noise estimates also need to be revised and more realistic values need to be established. It is shown that the 2D heat diffusion equation can be used for estimating the power falling on the IRVB foil, even with a significant amount of spatial variation in the thermal diffusivity across the area of the platinum foil found experimentally during foil calibration. The equation for the noise equivalent power density (NEPD) is re-derived to include the errors in the measurement of the thermophysical and the optical properties of the IRVB foil. The theoretical value estimated using this newly derived equation matches closely, within 5.5%, with the mean experimental value. The change in the contribution of each error term of the NEPD equation with rising foil temperature is also studied and the blackbody term is found to dominate the other terms at elevated operating temperatures. The IRVB foil is also sensitive to the charge exchange (CX) neutrals escaping from the plasma. The CX neutral contribution is estimated to be marginally higher than the noise equivalent power (NEP) of the IRVB. It is also established that the radiation measured by the IRVB originates from the impurity line radiation from the plasma and not from the heated divertor tiles. The change in the power density due to noise reduction measures such as data smoothing and averaging is found to be comparable to the IRVB NEPD. The precautions that need to be considered during background subtraction are also discussed with experimental illustrations. Finally, the analysis algorithm with all the improvements is validated and found to reproduce the input power well within 10% accuracy. This article answers many fundamental questions relevant to the IRVB and illustrates the care to be exercised while processing the IRVB data.

Development of imaging bolometers for magnetic fusion reactors (invited).

Imaging bolometers utilize an infrared (IR) video camera to measure the change in temperature of a thin foil exposed to the plasma radiation, thereby avoiding the risks of conventional resistive bolometers related to electric cabling and vacuum feedthroughs in a reactor environment. A prototype of the IR imaging video bolometer (IRVB) has been installed and operated on the JT-60U tokamak demonstrating its applicability to a reactor environment and its ability to provide two-dimensional measurements of the radiation emissivity in a poloidal cross section. In this paper we review this development and present the first results of an upgraded version of this IRVB on JT-60U. This upgrade utilizes a state-of-the-art IR camera (FLIR/Indigo Phoenix-InSb) (3-5 microm, 256 x 360 pixels, 345 Hz, 11 mK) mounted in a neutron/gamma/magnetic shield behind a 3.6 m IR periscope consisting of CaF(2) optics and an aluminum mirror. The IRVB foil is 7 cm x 9 cm x 5 microm tantalum. A noise equivalent power density of 300 microW/cm(2) is achieved with 40 x 24 channels and a time response of 10 ms or 23 microW/cm(2) for 16 x 12 channels and a time response of 33 ms, which is 30 times better than the previous version of the IRVB on JT-60U.