# Real-Time Detection of Overloads on the Plasma-Facing Components of Wendelstein 7-X

^{1}

^{2}

^{3}

^{4}

^{*}

^{†}

## Abstract

**:**

## 1. Introduction

^{2}. The divertor tiles are made of a Carbon Fibre Composite (CFC) layer joined to a CuCrZr cooling structure. The divertor’s maximum operational temperature is limited by a Cu interlayer, which should not exceed a sustained temperature of 475 ${}^{\circ}$C. This temperature is reached at 10 MW/m

^{2}when the surface temperature is 1200 ${}^{\circ}$C [4,5]. The other water-cooled PFCs are the baffles, the heat shields, the wall and pumping gap panels, and the poloidal closures. The baffles and the heat shields are made of fine-grain graphite tiles, which are limited to 400 ${}^{\circ}$C because of the brazed joint between the cooling pipe and the heat sink. The wall and pumping gap panels and the poloidal closure are made of water-cooled stainless steel, which cannot exceed 200 ${}^{\circ}$C. See Figure 3 for an overview of the PFCs and their maximum operational temperatures.

## 2. Real-Time Imaging System

## 3. Image Analysis

#### 3.1. The Scene Model

#### 3.2. Image Processing Pipeline

#### 3.3. Thermal Overload Detection Algorithm

^{2}] and a time interval $\Delta t$ [s]. ${C}_{m}$ is a material-dependent constant (${C}_{m}=\pi \xb7\lambda \xb7\rho \xb7{C}_{p}/4\times {10}^{-6}$ [s(kW)

^{2}/(m

^{4}K

^{2})], where $\lambda $ is the diffusion coefficient, $\rho $ the material density and ${C}_{p}$ the specific heat capacity).

- (1)
- After the acquisition and calibration of the thermal image T, the image is filtered to avoid false alarms due to remaining uncorrected bad pixels and noise coming from neutrons. It is clear that hot pixels can trigger false alarms, but also cold pixels must be removed as they can suddenly become hot, creating an apparent (false) peak of heat flux. This may lower the threshold in excess, triggering a false alarm. We remove hot pixels with a spatial filter, a morphological opening with a structuring element of 3 × 3 pixels and 8-connectivity. Cold pixels are removed by a complementary morphological closing followed by a reconstruction [14]. The uncertainty of the temperature calibration is then added to the thermal image resulting in the corrected image ${T}^{\prime}$.
- (2)
- Following, the corrected thermal image is averaged over time $\overline{{T}^{\prime}}$ with a moving average to reduce its noise. The average window size n has to be kept small (2 to 5 frames) to avoid adding too much delay to the system.
- (3)
- The temperature threshold ${T}_{xy}^{th}$ [K] is computed with Equation (3) for each pixel $(x,y)$ in real-time using the model in Equation (2), the reaction time ${t}_{r}=0.16$ s, the temperature limit ${T}_{xy}^{limit}$ [K] of the PFC, the current heat flux ${q}_{xy}$ [kW/m
^{2}], and the material constant ${C}_{xy}^{m}$ [s(kW)^{2}/(m^{4}K^{2})]. The scene model provides the temperature limit and material constant for each pixel.$${T}_{xy}^{th}\left(t\right)={T}_{xy}^{limit}-{q}_{xy}\left(t\right){\left(\frac{{t}_{r}}{{C}_{xy}^{m}}\right)}^{l}$$ - (4)
- Note that Equation (3) requires the computation of the heat-flux in real-time. Currently, 2-dimensional thermal calculations in real-time are not possible at W7-X, since we cannot assume toroidal symmetry [15]. The heat-flux can be estimated roughly according to Equation (1) using the increase of the averaged corrected temperature between two frames $\Delta \overline{{T}_{xy}^{\prime}}$ and the frame time interval ${t}_{f}=0.01$ s, resulting in Equation (4).$${q}_{xy}\left(t\right)=\frac{\Delta \overline{{T}_{xy}^{\prime}}\left(t\right)}{{\left(\frac{{t}_{f}}{{C}_{xy}^{m}}\right)}^{l}}$$
- (5)
- $\Delta \overline{{T}_{xy}^{\prime}}$ is then time-filtered in Equation (5) with a learning rate $\alpha $ to avoid triggering false alarms due to fast transients.$$\Delta \overline{{T}_{xy}^{\prime}}\left(t\right)=(1-\alpha )\Delta \overline{{T}_{xy}^{\prime}}(t-{t}_{f})+\alpha \left(\right)open="("\; close=")">\overline{{T}_{xy}^{\prime}}\left(t\right)-\overline{{T}_{xy}^{\prime}}(t-{t}_{f})$$Fast transients refer to fast physical events with high heat flux that last for a short period of time, not enough to actually heat and damage the PFCs. However, the temperature prediction is based on the measured heat flux, and it reacts assuming that the current heat flux is sustained for 0.16 s. A short high increase of heat flux can then trigger a false alarm, even lasting just few ms. To prevent this, the heat flux has to be averaged over time to react only to sustained heat flux changes.
- (6)
- In case the heat-flux calculation under-estimates the real heat flux, we follow a conservative approach to stay on the safe side. The temperature threshold finally used (Equation (6)) is the minimum between the computed according to Equation (3) and an upper-bound ${T}_{xy}^{th\_max}$:$${T}_{xy}^{th}\left(t\right)=min\left(\right)open="\{"\; close="\}">{T}_{xy}^{limit}-{q}_{xy}\left(t\right){\left(\frac{{t}_{r}}{{C}_{xy}^{m}}\right)}^{l},{T}_{xy}^{th\_max}$$The upper-bound values are:
- Divertor targets: 1034 ${}^{\circ}$C
- Baffles and heat shields: 387 ${}^{\circ}$C
- Pumping gap: 172 ${}^{\circ}$C

These values are computed according to Equation (2) considering a ${t}_{r}={t}_{max\_delay}=0.11$ s (without safety margin) and assuming a heat-flux of 10 MW/m^{2}for the divertors, 0.5 MW/m^{2}for the baffle and heat shields and 0.2 MW/m^{2}for the pumping gap. Note that we do not require here the safety margin of the reaction time because the calculation of the threshold is deterministic. The safety margin was added to account for the uncertainty of the heat-flux estimation. Also note that setting a temperature threshold upper-bound is equivalent to setting a heat flux lower-bound or a minimum value of the heat flux (see Equation (2)). - (7)
- An estimated risk ${\widehat{r}}_{xy}$ is then computed in Equation (7) for each pixel with the averaged corrected temperature $\overline{{T}_{xy}^{\prime}}$, the temperature threshold ${T}_{xy}^{th}$ and a correction factor to consider the presence of surface layers ${S}_{xy}$ [13].$${\widehat{r}}_{xy}\left(t\right)=\frac{\overline{{T}_{xy}^{\prime}}\left(t\right)}{{T}_{xy}^{th}\left(t\right)}(1-{S}_{xy})$$Note that the estimated risk image is normalized for the different temperature limits and heat fluxes on the different PFCs, allowing to detect overloaded regions in the entire field of view naturally, avoiding the use of regions of interest.
- (8)
- The overloaded regions are detected by clustering all pixels that have an estimated risk equal or greater than 1. If the cluster’s area (in physical units) is bigger than a minimum area ${A}_{min}$ an alarm is triggered. The scene model also provides the physical area covered by each pixel in the image.

## 4. Results from OP1.2 Campaign

## 5. Towards Feedback Control of Thermal Loads

## 6. Conclusions

## Author Contributions

## Funding

## Data Availability Statement

## Acknowledgments

## Conflicts of Interest

## References

- Endler, M.; Baldzuhn, J.; Beidler, C.D.; Bosch, H.S.; Bozhenkov, S.; Buttenschön, B.; Dinklage, A.; Fellinger, J.; Feng, Y.; Fuchert, G.; et al. Wendelstein 7-X on the path to long-pulse high-performance operation. Fusion Eng. Des.
**2021**, 167, 112381. [Google Scholar] [CrossRef] - Klinger, T.; Andreeva, T.; Bozhenkov, S.; Brandt, C.; Burhenn, R.; Buttenschön, B.; Fuchert, G.; Geiger, B.; Grulke, O.; Laqua, H.P.; et al. Overview of first Wendelstein 7-X high- performance operation. Nucl. Fusion
**2019**, 59, 112004. [Google Scholar] [CrossRef] - Jakubowski, M.; Endler, M.; Feng, Y.; Gao, Y.; Killer, C.; König, R.; Krychowiak, M.; Perseo, V.; Reimold, F.; Schmitz, O.; et al. Overview of the results from divertor experiments with attached and detached plasmas at Wendelstein 7-X and their implications for steady-state operation. Nucl. Fusion
**2021**, 61, 106003. [Google Scholar] [CrossRef] - Boscary, J.; Peacock, A.; Stadler, R.; Mendelevitch, B.; Tittes, H.; Tretter, J.; Smirnow, M.; Li, C. Actively Water-Cooled Plasma Facing Components of the Wendelstein 7-X Stellarator. Fusion Sci. Technol.
**2017**, 64, 263–268. [Google Scholar] [CrossRef] - Stadler, R.; Vorköper, A.; Boscary, J.; Cardella, A.; Hurd, F.; Li, C.; Mendelevitch, B.; Peacock, A.; Pirsch, H. The in-vessel components of the experiment W7-X. Fusion Eng. Des.
**2009**, 84, 305–308. [Google Scholar] [CrossRef][Green Version] - Huber, A.; Kinna, D.; Huber, V.; Arnoux, G.; Balboa, I.; Balorin, C.; Carman, P.; Carvalho, P.; Collins, S.; Conway, N.; et al. The near infrared imaging system for the real-time protection of the JET ITER-like wall. Phys. Scr.
**2017**, T170, 014027. [Google Scholar] [CrossRef] - Huber, V.; Huber, A.; Kinna, D.; Matthews, G.F.; Sergienko, G.; Balboa, I.; Brezinsek, S.; Lomas, P.J.; Mailloux, J.; McCullen, P.; et al. The Software and Hardware Architecture of the Real-Time Protection of In-Vessel Components in JET-ILW. Nucl. Fusion
**2019**, 59, 076016. [Google Scholar] [CrossRef] - Herrmann, A.; Drube, R.; Lunt, T.; De Marné, P. Real-time protection of in-vessel components in ASDEX Upgrade. Fusion Eng. Des.
**2011**, 86, 530–534. [Google Scholar] [CrossRef][Green Version] - Sieglin, B.; Faitsch, M.; Herrmann, A.; Brucker, B.; Eich, T.; Kammerloher, L.; Martinov, S. Real-Time capable Infrared Thermography at ASDEX Upgrade. Rev. Sci. Instrum.
**2015**, 86, 113502. [Google Scholar] [CrossRef] [PubMed] - Jakubowski, M.; Drewelow, P.; Fellinger, J.; Puig Sitjes, A.; Wurden, G.; Ali, A.; Biedermann, C.; Cannas, B.; Chauvin, D.; Gamradt, M.; et al. Infrared imaging systems for wall protection in the W7-X stellarator. Rev. Sci. Instrum.
**2018**, 89, 10E116. [Google Scholar] [CrossRef] - Pisano, F.; Cannas, B.; Fanni, A.; Sias, G.; Jakubowski, M.W.; Drewelow, P.; Niemann, H.; Puig Sitjes, A.; Gao, Y.; Moncada, V.; et al. Tools for Image Analysis and First Wall Protection at W7-X. Fusion Sci. Technol.
**2020**, 76, 933–941. [Google Scholar] [CrossRef] - Herrmann, A.; Sieglin, B.; Faitsch, M.; Team, A.U. Surface Temperature Measurement of In-Vessel Components and Its Real-Time Validation. Fusion Sci. Technol.
**2016**, 69, 569–579. [Google Scholar] [CrossRef] - Ali, A.; Niemann, H.; Jakubowski, M.; Pedersen, T.S.; Neu, R.; Corre, Y.; Drewelow, P.; Sitjes, A.P.; Wurden, G.; Pisano, F.; et al. Initial results from the hotspot detection scheme for protection of plasma facing components in Wendelstein 7-X. Nucl. Mater. Energy
**2019**, 19, 335–339. [Google Scholar] [CrossRef] - Vincent, L. Morphological Grayscale Reconstruction in Image Analysis: Applications and Efficient Algorithms. IEEE Trans. Image Process.
**1993**, 2, 176–201. [Google Scholar] [CrossRef][Green Version] - Sieglin, B.; Faitsch, M.; Herrmann, A.; Martinov, S.; Eich, T.; Team, A.U. Real-Time infrared thermography at ASDEX Upgrade. Fusion Sci. Technol.
**2016**, 69, 580–585. [Google Scholar] [CrossRef] - Sitjes, A.P.; Gao, Y.; Jakubowski, M.; Drewelow, P.; Niemann, H.; Ali, A.; Moncada, V.; Pisano, F.; Ngo, T.T.; Cannas, B.; et al. Observation of thermal events on the plasma facing components of Wendelstein 7-X. J. Instrum.
**2019**, 14, C11002. [Google Scholar] [CrossRef] - Mitteau, R.; Belafdil, C.; Balorin, C.; Courtois, X.; Moncada, V.; Nouailletas, R.; Santraine, B. WEST operation with real time feed back control based on wall component temperature toward machine protection in a steady state tungsten environment. Fusion Eng. Des.
**2021**, 165, 112223. [Google Scholar] [CrossRef] - Martin, V.; Moncada, V.; Travere, J.M.; Loarer, T.; Brémond, F.; Charpiat, G.; Thonnat, M. A Cognitive Vision System for Nuclear Fusion Device Monitoring. In International Conference on Computer Vision Systems; Springer: Berlin/Heidelberg, Germany, 2011; pp. 163–172. [Google Scholar]
- Gao, Y.; Geiger, J.; Jakubowski, M.W.; Drewelow, P.; Endler, M.; Rahbarnia, K.; Bozhenkov, S.; Otte, M.; Suzuki, Y.; Feng, Y.; et al. Effects of toroidal plasma current on divertor power depositions on Wendelstein 7-X. Nucl. Fusion
**2019**, 59, 106015. [Google Scholar] [CrossRef] - Pisano, F.; Cannas, B.; Fanni, A.; Sias, G.; Gao, Y.; Jakubowski, M.; Niemann, H.; Sitjes, A.P. Learning control coil currents from heat-flux images using convolutional neural networks at Wendelstein 7-X. Plasma Phys. Control. Fusion
**2020**, 63, 025009. [Google Scholar] [CrossRef] - Böckenhoff, D.; Blatzheim, M. Application of improved analysis of convective heat loads on plasma facing components to Wendelstein 7-X. Nucl. Fusion
**2019**, 59, 086031. [Google Scholar] [CrossRef] - Böckenhoff, D.; Blatzheim, M.; Hölbe, H.; Niemann, H.; Pisano, F.; Labahn, R.; Pedersen, T.S. Reconstruction of Magnetic Configurations in W7-X using Artifical Neural Networks. Nucl. Fusion
**2018**, 58, 056009. [Google Scholar] [CrossRef][Green Version]

**Figure 1.**W7-X five-fold modules with its 10 divertor units, 5 upper and 5 lower divertor units. The divertor targets intersect the magnetic islands at the edge for power and particle exhaust.

**Figure 2.**The divertor units are 5 m long and 1 m wide with several target modules: vertical targets (TM1V-TM3V), low-iota targets (TM1H-TM4H), low-load target modules (TM5H-TM6H) and high-iota targets (TM7H-TM9H). The divertor water-cooled tiles are made of CFC (Carbon Fibre Composite) except the two low-load central target modules, which are made of fine-grain graphite.

**Figure 3.**Infrared image overlaid on the CAD of the different plasma-facing components with their maximum operational temperatures.

**Figure 4.**The immersion tubes provide an optical resolution of the order of 5 to 20 mm/pixel on the divertor, depending on the distance and the viewing angle of the divertor. In the high-iota region, the lower resolution may lead to measuring lower temperatures from hot spots smaller than the resolution limit. The immersion tubes are equipped with microbolometer cameras IRCam Caleo 768k L covering the spectral range from 8 to 10 $\mathsf{\mu}$m at a frame rate of 100 Hz.

**Figure 5.**The steady-state endoscopes consists of an off-axis Cassegrain optical system with a pinhole aperture of 6 mm. A set of mirrors transmits the light and a dichroic splitter divides it into visible and infrared beams, which are detected by the cameras after being corrected by a group of lenses. They have an optical resolution of 8 mm/pixel and they are equipped with an Indium Antimonide (InSb) infrared camera covering the spectral range from 2 to 5.7 $\mathsf{\mu}$m (filtered at 4.1 $\mathsf{\mu}$m) at a frame rate of 100 Hz.

**Figure 6.**An MTCA-based (Micro Telecommunication Computing Architecture) frame grabber acquire the infrared images and synchronizes them with the W7-X universal time. The raw images are transferred to the Image Analysis System, a set of real-time computers, one for each camera, that calibrates and analyses the infrared data. The video streams are analyzed through computer vision techniques to detect overloads. When the integrity of the PFCs is compromised, the system triggers the Interlock. We will implement some of these algorithms on a GPU to speed up computation. The detected overloads and alarms are sent to the Thermal Event Monitoring system in the control room and registered into a database, the Archive, and the Logbook for future reference.

**Figure 7.**The scene model is a multi-channel image matching the camera field of view, each channel providing different PFCs properties for each pixel: identifier, material, emissivity, heat flux limit, temperature limit and pixel resolution.

**Figure 8.**The system calibrates and analyses each frame of a pulse. The calibrated images are analysed by the Thermal Overload Detection algorithm. If there is risk of overload of the PFCs, the system sends an alarm to the interlock to stop operation. The total delay of the pipeline must be less than 110 ms, including the stopping of the heating systems. The scene model and the emissivity correction map provide the thermal properties of the imaged components and their temperature limits. A surface layer map is used to adjust the risk in presence of surface layers. The non-uniformity correction (NUC), emissivity correction map, and surface layer map are updated periodically with dedicated procedures.

**Figure 9.**On the divertor targets, if the algorithm triggers the interlock when it detects that the temperature breaches the limit (${T}_{limit}=1200$ ${}^{\circ}$C), the temperature will exceeded this limit before the heating is interrupted due to delay of the system. To prevent this, it is required to stop at a lower threshold (${T}_{th}$). At 1000 ${}^{\circ}$C it takes 0.16 s to reach the temperature limit when the heat flux is 10 MW/m

^{2}(the maximum heat flux allowed). This provides us with a reference reaction time ${t}_{r}=0.16$ s. The system, however, must stop at a lower threshold if the detected heat flux is higher than 10 MW/m

^{2}or it can stop at a higher threshold if the detected heat flux is lower.

**Figure 10.**Distribution of the anticipation times on the validation dataset relative to the time when the temperature reaches the limit ${t}_{T={T}_{limit}}$. All alarms are triggered in time (with ${t}_{anticipation}>{t}_{max\_delay}$). Only one false alarm is triggered, considering a false alarm when ${t}_{anticipation}>1$ s.

**Figure 11.**A baffle overload example. (

**a**) Maximum of the temperature over time. (

**b**) Maximum of the estimated risk over time; the strike-lines which are within the temperature limits of the divertor are barely visible; the overload of the baffle stands out. (

**c**) Maximum temperature for each PFCs over time; the inner baffle reaches 400 ${}^{\circ}$C at 5 s. (

**d**) Time evolution of the maximum temperature normalized by the temperature limit over the image, time evolution of the maximum averaged temperature normalized by the temperature limit over the image, and Fast Interlock Signal (0.5 corresponding to a warning and 1.0 to an alarm); We can observe that an alarm is triggered 0.39 s before the maximum temperature reaches the limit (when the baffle temperature reaches 400 ${}^{\circ}$C) well in advance to compensate for the 0.11 s of the system delay.

**Figure 12.**Divertor overload example. (

**a**) Maximum of the temperature over time. (

**b**) Maximum of the estimated risk over time. (

**c**) Maximum temperature for each PFCs over time; the low-iota divertor target reaches 1200 ${}^{\circ}$C at 0.4 s. (

**d**) Time evolution of the maximum temperature normalized by the temperature limit over the image, time evolution of the maximum averaged temperature normalized by the temperature limit over the image, and Fast Interlock Signal (0.5 corresponding to a warning and 1.0 to an alarm); We can observe that an alarm is triggered 0.12 s before the maximum temperature reaches the limit (when the divertor temperature reaches 1200 ${}^{\circ}$C) just in time to compensate the 0.11 s of the system delay.

**Figure 13.**The machine is protected with a dual detection scheme. The Thermal Overload Detection algorithm detects when a PFC has risk of overloading and triggers the interlock to stop operation. The Thermal Event Detection algorithm promptly detects the thermal events, tracks them over time, classifies them and computes their risk. It sends all this information to the feedback control system. When an event reaches a certain risk, the feedback control has to take action to prevent the overloading of the PFCs and the plasma interruption.

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |

© 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Puig Sitjes, A.; Jakubowski, M.; Naujoks, D.; Gao, Y.; Drewelow, P.; Niemann, H.; Fellinger, J.; Moncada, V.; Pisano, F.; Belafdil, C.;
et al. Real-Time Detection of Overloads on the Plasma-Facing Components of Wendelstein 7-X. *Appl. Sci.* **2021**, *11*, 11969.
https://doi.org/10.3390/app112411969

**AMA Style**

Puig Sitjes A, Jakubowski M, Naujoks D, Gao Y, Drewelow P, Niemann H, Fellinger J, Moncada V, Pisano F, Belafdil C,
et al. Real-Time Detection of Overloads on the Plasma-Facing Components of Wendelstein 7-X. *Applied Sciences*. 2021; 11(24):11969.
https://doi.org/10.3390/app112411969

**Chicago/Turabian Style**

Puig Sitjes, Aleix, Marcin Jakubowski, Dirk Naujoks, Yu Gao, Peter Drewelow, Holger Niemann, Joris Fellinger, Victor Moncada, Fabio Pisano, Chakib Belafdil,
and et al. 2021. "Real-Time Detection of Overloads on the Plasma-Facing Components of Wendelstein 7-X" *Applied Sciences* 11, no. 24: 11969.
https://doi.org/10.3390/app112411969