1 FIP/P4-8

Design Development of the ITER Divertor Diagnostic Systems in Japan 1K. Itami, 1T. Sugie, 2M. Takeuchi, 1 H. Ogawa, 1S. Kitazawa, 1T. Maruyama, 1T. Ono, 1T. Shimada, 1M. Ishikawa, 1Y. Kawano, 3E. Veshchev, 3P. Andrew, 3M. Walsh 1 National Institutes for Quantum and Radiological Science and Technology, Naka, Ibaraki, 311-0193 Japan 2 Sanoh Industrial Co., Ltd., Koga, Ibaraki, 306-0041 Japan 3 ITER Organization, St. Paul-lez-Durance, France

E-mail contact of main author: itami.kiyoshi@qst.go.jp Abstract. Advanced design for Divertor Impurity Monitor (DIM) and IR Thermography (IRTh) is being developed in Japan to contribute to the divertor plasma control and research in ITER through the spectroscopic measurement with high spatial and spectral resolutions and accurate measurement of the target temperature. Through the detailed design activity, necessary design improvement for the front end optics was highlighted.

1. Divertor Impurity Monitor (DIM) system

Divertor Impurity Monitor (DIM) spectroscopically measures two dimensional profiles with spatial resolution of 20 - 40 mm of metallic impurities (such as W, Be), injected impurity gas (Ar, Ne, Kr), fuel ratio, helium exhaust, ion temperature and ionization fronts by measuring light in the wavelength range of λ = 200nm – 1000 nm from the divertor plasmas [1-2]. Therefore, its measurement contributes the plasma operation and the control of the divertor plasmas, as well as the divertor plasma research with help of the plasma modelling.

There are three port systems in DIM. The upper port #1 (UP1) system (71ch) is viewing the divertor plasmas in the same port section. The equatorial port #1 (EQ1) system (71ch) is viewing tangentially downward the divertor plasma in the port #4 section. The lower port #2 system (LO2) has two optical systems viewing through a gap between the divertor cassettes and one optical viewing both the inner and outer divertor plasma from the front end optics under the dome. The dome optics is expected to resolve the ionization and recombination of fuel ions to characterize the plasma detachment at the strike points with 20 mm resolution. In the optical design, spot diagrams of each port optics have shown that the spot size is less than 10 mm and good enough for the spatial resolutions of 20 - 40 mm in the wavelength range of λ = 200 nm – 1000 nm. [3].

Fig.1 Viewing fan of UP1 and EQ1 port system

2 FIP/P4-8

Front end optics in each port systems collects light in the field of view and the relay optics relays images of the divertor with a spatial resolution of 20 – 40 mm and the Cassegrain optics focuses those images on the multi-channel fibres in the port cell. The ultra violet light (λ = 200nm – 450 nm) is transferred to the UV spectrometers in the port cells via the 1 m short bundle fibres. The visible and near infrared light (λ = 400 nm – 1000 nm) is transferred via optical fibres to the spectrometers in the diagnostic room with spectral resolutions of 0.01 nm - 1 nm and time resolutions of 1 ms – 10 ms. Since DIM ports are located at the opposite side of the Tokamak building, the optical fibre bundles as long as 200 m are laid along the gallery in B1 level (LO2 system), L1 level (EQ1 system) and L2 level (UP1 system). Among the three port systems, detailed design of the UP1 and EQ1 systems has almost completed toward the preliminary design review (PDR). In the PDR, the design consistency, soundness of the diagnostic components against the possible load conditions, reliability, safety and redundancy are reviewed, in addition to the compliance to the measurement requirement, as a nuclear system to be integrated in ITER as the experimental nuclear fusion reactor. The optical component design of UP1 system is shown in Fig. 2. The optical component design EQ1 system is very similar to that of UP1 system, while configuration of the front end mirrors are different. The front end optics consists of four metallic mirrors. The first (M1) and the second (M2) mirrors are flat mirrors, while M3 is an axisymmetric aspheric mirror and M4 is a toroidal mirror. M1 is inclined 45 degrees to enable the mirror cleaning discharge in the stationary toroidal field. A gas-pressure driven shutter is allocated near the entrance pupil to protect M1. The relay optics outside of vacuum windows has a function of optical axis alignment. Adjustment mechanism of four mirrors (M5, M6, M9, M10) and the chromatic correction lenses (L3+L4) are utilized, so that the light flux is focused to multi-channel fibres by Cassegrain imaging optics. All metallic mirrors are aluminium coated for the highest reflectivity except the first mirror (M1) made of molybdenum, which is resilient to sputtering. Throughput of the collection optics is estimated to be 0.05 – 0.16 in the wavelength range of λ = 200nm – 1000 nm. Since very accurate alignment of mirrors including non-spherical mirrors is required, the front end mirrors and pneumatic shutter for protection of the first mirror are integrated into a single assembly for both the UP1 and EQ1 system. Integration of mirror cleaning system inside the mirror box is under discussion. Structural integrity of the assemblies of the UP1 and EQ1 front end optics was accessed by finite element (FE) analysis, based on the load specification for the generic upper port and equatorial port. The earth gravity (9.8 m/s2), seismic load,

Fig. 2 UP1 Optics of the front end optics, relay optics and imaging optics.

3 FIP/P4-8

thermal load during normal and baking (250 ºC) operation and electromagnetic (EM) load were used for the assessment of the stress to the structural material (Stainless Steel SS316L). The thermal load during the normal operation (THO) for the UP1 optical assembly, as shown in Fig. 3 (a), is nuclear heating of 0.2 MW/m3 and the plasma radiation of 0.935 kW/m to M1 mirror and the shutter blade. Temperature profile of the mirror box in Fig. 3 (b) shows large gradient of temperature at the bottom side. The maximum Von Mises (VM) stress is 123 MPa at lower plate, as shown in Fig. 3 (c). This value is lower than the allowable stress of SS316L (381 MPa at 150 ºC, 3Sm). Fig. 3 (d) shows temperature profile of M1 for the THO thermal load. The maximum VM stress is 76 MPa at the edge of the mirror holder, as shown in Fig. 3 (e). In both of the mirror box and M1, VM stress is lower than the allowable stress of SS316L (381 MPa at 150 ºC, 3Sm).

The most significant electromagnetic (EM) load for the UP1 optical assembly occurs at the plasma disruptions during the upward displacement in 36ms (VDE-III). The transient EM force vector profile of M4 is shown in Fig. 3 (f). The maximum displacement is 45 µm at Mirror. The maximum stress is 27 MPa on the back plate, as shown in Fig. 3 (g). This value is lower than allowable stress of SS316L (127 MPa at 150 ºC, Sm).

M1 Mirror

M2 Mirror

M3 Mirror

M4 Mirror

Shutter blade

(to Divertor)

(to Relay Optics)

(a) (b)

(c)

(d)

(e)

(f)

(g)

(h)

(i)

Fig. 3 (a) Inside of the UP1 optical assembly. (b)-(e): THO load case; (b) Temperature profile and (c) VM stress profile of Mirror Box. (d) Temperature profile and (e) VM stress profile of M1. (e)-(f) VDE-III load case; (f) EM force vector profile and (g) VM stress profile of M4. (h)-(i) SL-2 load case; (h) Displacement Profile and (i) VM stress profile of shutter: Red tubs in figures indicate positions of the maximum.

4 FIP/P4-8

While thermal load during normal operation and EM load during the plasma major disruptions are design drivers for all components of the UP1 optical assembly, stress due to SL-2 seismic load as a single load, as large as 4 G, was verified. Displacement profile of Shutter for SL-2 at baking temperature is shown in Fig. 3 (h). The maximum displacement is 0.6 mm at Shutter plate. The maximum VM stress is 43 MPa at base plate of shutter, as shown in Fig. 3 (i). This value is lower than allowable stress of SS316L (107 MPa at 300 ºC, Sm).

In summary, structural integrity to all of the single loads was guaranteed. The maximum value of the scalar sum of primary stresses and secondary stresses (THO+EM (VDE-III) +SL-1 +DW) is 209 MPa, which is lower than allowable stress of 381 MPa. Here DW is the dead weight and SL-1 is 1/3 of SL-2. Hence the structural integrity of the UP1 optical assembly is guaranteed. In the similar way, the structural integrity of the EQ1 optical assembly is guaranteed.

2. IR Thermography (IRTh) System

Among the ITER diagnostic systems which monitor the first walls, IR Thermography (IRTh) is specialized to measure the target temperature with fast time resolution and high spatial resolution and to investigate divertor plasma physics, such as transient heat fluxes in events of ELMs and the plasma disruptions, as well as heat flux in steady state. It is required to measure temperature profiles of 200 – 1000 °C with the time resolution of 2 ms and that of 1000 – 3600 °C with the time resolution of 20 µs. The IR Thermography (IRTh) system is located in the equatorial port 17 (EQ17). The inner divertor optics is viewing the inner divertor target in the port 16 section and the outer divertor optics is viewing the outer divertor target in the port 3 section. (Fig. 4)

The optical design is strongly restricted by the allocated space for the IRTh system in EQ17 port plug, where the front end optics for viewing the inner and outer divertor is integrated with other five diagnostic and a GDC systems. Details are described with Fig, 5. The numerical aperture (NA) to the divertor must be as low as possible from the allowable size for front end mirrors, while twice of the diffraction limit s, determined by the formula, s = 0.61 λ /(NA), is not larger than the required spatial resolution of 3 mm. Here = 5 µm was chosen and NA must be smaller than 0.002. The front end optics transfers image of the target area to

Fig. 4 Layout of IRTh optical components

5 FIP/P4-8

the intermediate image just outside of the vacuum mirror, enabling adjustment of the optical axis and focusing for mintenance. The numerical aperture of the intermediate image, NAIM, is determined by the envelop of the space usable for the IRTh optics in the central DSM. Therefore, diameter of the intermediate image is 130 mm and the vacuum window size must be as large as 120 mm, which is the maximum size of the vacuum window in ITER. The intermediate image is relayed to the port cell by the relay optics and focused by Cassegrain optics to the intermediate images for the two wave measurement and spectrometer. The intermediate images for the two wave measurement is branched to 3 µm band pass image measurement and 5 µm band image measurement by IR camera with detector size of 12 mm x 16 mm. One hundred positions are selected on the intermediate image for spectrometer and those images are transferred via IR fibres to the prism spectrometer and IR spectrum at the selected positions is measured by IR camera.

The important design issue for the front end mirrors is how to control deformation due to volumetric heating due to the nuclear heating and surface heating by the ITER plasma operation. Molybdenum (Mo) coating on the polished stainless steel is expected by considering the manufacturability of mirrors and reflectivity of light of IR wavelength. Deformation of the front end mirrors by heating by the ITER plasma was investigated by using ANSYS code. In the simulation of IM1 (the first mirror of the inner optics), the nuclear heating power of 0.128 MW/m3 and the plasma radiation power of 5.1 kW/m2 was applied. Cooling water pressure of 4 MPa, cooling water temperature of 343 Kelvin, cooling water heat transfer coefficient of 7436 W/m2/K were assumed. A water cooling channel of the mirror was redesigned from that shown in Fig. 6 (a) to that shown in Fig. 6 (b), since profiles of the surface temperature and deformation were asymmetric due to the water pressure effect. Profiles of the surface temperature and deformation are shown in Fig. 6 (c) and (d), respectively. Peak to Peak (P-P) deformation of µm order is too large, since less than 1/10 of IR wavelength is required. Further improvement was not obtained by increasing length of the water channel, while the temperature profile was more flatten. Therefore, in order to cope with the surface heating, the cooling surface level, which is defined by the depth of the water channel under the surface, was moved upward to the level where total heating power of 74 W

Fig. 5 Optical design of the inner optics of the IRTh system. Focal planes for IR cameras are indicated by red circles.

6 FIP/P4-8

is dividend into two of 37 W, as shown in Fig. 6 (e). Deformation is significantly improved as shown in Fig. 6 (g), The mirror and cooling channel made of Mo could significantly flatten and P-P temperature of 0.281 °C and P-P deformation of 0.016 µm was obtained, because of the higher heat conductivity and lower coefficient of thermal expansion than those of SS316L, while manufacturability must be assessed.

Most of the front end mirror is non spherical mirror except IM1 and deformation analysis was carried out for non spherical mirrors. While the curvature is slightly changed by the isothermal expansion, it is possible to compensate by adjustment of the focusing in the backend. Therefore, deformation of a non spherical mirror is defined by deviation from the surface with the isothermal expansion of the mirror. Suppression of the deformation in the non spherical mirrors was more difficult than that in the flat mirror, when the surface heating was applied. In the simulation of OM1 (the first mirror of the outer optics), the nuclear heating power of 0.152 MW/m3 and the plasma radiation power of 3.3 kW/m2 was applied, the P-P deformation could not be smaller than 0.96 µm. It is recently pointed out that power deposition to the first mirrors of IRTh due to the stray ECH wave power is an order of the nuclear heating power. Therefore, the more sophisticated optical and mechanical design of front end mirrors are required before the finalizing those designs, as well as the accurate estimation of the ECH power deposition to the optical components.

For accurate measurement of the tungsten (W) target temperature, it is required to monitor changes in the emissivity due to changes in the W surface property during the plasma operation. Therefore, the in-situ calibration was investigated in a laboratory by using IR laser with λIR = 3.22 µm, as shown Fig. 7 (a). The results were; (i) Emissivity of the W sample increased from 0.2 to 0.6 when the surface roughness changed from 0.3 µm to 5.9 µm from the measurement by IR camera and thermocouples. (ii) Back scattering light of IR laser would be observable in the viewing lines of the IRTh sight, if the surface roughness was 1 µm or larger than 1 µm, as shown in Fig. 7(b). On the other hand, reflection of the IR laser light was strongly peaked at the specular angle on the W sample with 0.3 µm roughness and very small scattered light was observed at the other angles. (iii) It was experimentally validated the formula, “ε (λIR) = 1 – f (λIR)” for the W target from the measurement using IR laser, IR camera and thermocouples. Here ε (λIR) and f (λIR) are the emissivity and reflectivity,

Fig. 6 (a) The first design of cooling channel, (b) Redesigned cooling channel, (c) Surface temperature profile and (d) Deformation profile of IM1 to the volumetric and surface heating, (e) Upward move of the cooling surface, (c) Surface temperature profile and (d) Deformation profile of IM1 after upward move of the cooling surface.

7 FIP/P4-8

respectively. In conclusion, feasibility of in-situ calibration of by using IR laser light was demonstrated for the first time [4].

3. Update of neutron transport analysis

Recently, the neutron transport calculation using C-lite model ver.2 become applicable to the ITER diagnostic systems. Therefore, nuclear analysis to the DIM and IRTh systems have been updated. MCNP ver.5 (Monte Carlo N-Particle Transport Code) code and FENDL-2.1 (Fusion Evaluated Nuclear Data Library) was used. Toroidally symmetrical source with 14 MeV mono energy neutron source was used. For nuclear heating calculation, ITER inductive operation scenario with 500 MW fusion power was used. Although the guideline requests the minimum NPS of 109, NPS of our analysis is 108 order, statistical errors of less than 10% were satisfied. Revised nuclear heating profile is shown for the front end of DIM UP1 optics in Fig .8 (a) and the front end of IRTh inner and outer optics. Nuclear heating power in the front end mirrors and shutters is larger than that used for the detailed design by a factor of two to three. These results suggest that more efficient heat removal from the front end mirrors and shutters must be investigated for the DIM and IRTh systems.

Fig.7 (a) IR laser spot imaged by IR camera (b) Scattering angle dependence

Fig. 8 Nuclear heating profile in front end of (a) DIM UP1 optics and (b) IRTh optics

8 FIP/P4-8

References

[1] Kawano, Y., et al. Proc. 24th IAEA Fusion Energy Conference, ITR/P5-35 (2012). [2] Kajita, S., et al., J. Nuclear Materials 463 (2015) 936. [3] Kitazawa, S., et al. Fusion Eng. and Des. 101 (2015) 209-217. [4] Takeuchi, M., et al., to be published in Fusion Sci. Tech. 69, 655 (2016).