Armoring of the Wendelstein 7-X divertor-observation immersion-tubes based on NBI fast-ion simulations

https://doi.org/10.1016/j.fusengdes.2019.01.099Get rights and content

Highlights

  • Simulations predicted NBI fast ion losses to sensitive immersion tubes.

  • Protective stainless steel collars were installed on short notice.

  • First NBI experiments confirmed the need for armoring.

Abstract

The first neutral beam injector (NBI) experiments of the Wendelstein 7-X stellarator took place in summer 2018. The modelling of the fast ion production and slowing down processes predicts losses of the NBI fast ions to the first wall on the order of 15%. One location receiving a high load (possibly peaking at several M W/m2) is the immersion tube for optical and infrared monitoring of the divertor targets. The stainless steel face of the tube has three vacuum windows, which are sensitive to temperature gradients and overheating. To protect the windows from damage caused by the fast ions, different heat load mitigation techniques were investigated. Given the available time and resources until the first NBI experiments, a protective stainless steel collar mounted at the front of the immersion tubes was regarded the most realistic solution. This contribution describes the fast ion modelling of the loads, the new design, thermal modelling of the design, and finally experimental experience with the protective collar showing heat loads in excess of 1.5 M W/m2. The fast ion heat loads have been assessed computationally with the ASCOT code and experimentally with thermal imaging.

Introduction

The Wendelstein 7-X (W7-X) stellarator has modular superconducting coils and the field is optimized i.a. for improved fast ion confinement. It has been in operation since 2015, and the first neutral beam injection (NBI) experiments took place in summer 2018. The machine has quite radial NBI injection geometry [1] with initial NBI particle pitch-angles only 17° or 27° from perpendicular. This leads to significant NBI fast ion losses [2] to the plasma facing components (PFCs) [3]. The most critical of the loaded components was found to be a set of immersion tubes holding cameras for divertor observations [3], [4].

The immersion tubes are part of the main vacuum barrier, as the cameras are in air and look at the plasma through three thin vacuum windows. The immersion tubes are in outboard side ports, that have identifiers consisting of the letters “AEF”, followed by two digits depicting the toroidal section. Thus, the tubes are located in a local magnetic well (“ripple well”) between the main coils. Furthermore, the tubes slightly protrude towards the plasma from the surrounding PFCs. Therefore, certain class of ripple trapped fast ions drift to the immersion tube and produce local heat loads of several M W/m2.

The stainless steel immersion tubes are cooled by compressed air and the system is designed to remove the heating by plasma radiation (typical design value would be 150 k W/m2) as well as the heating from the cameras inside the tube. The immersion tubes were designed to only be used in the first campaigns of W7-X. A future steady state compatible endoscope will use water cooling and will not have fragile vacuum windows.

A number of methods to alleviate the fast ion heating problem were explored. In 2017, the solution was expected to be optimization of magnetic field of W7-X by utilizing the flexible set of 7 different shapes of main field coils. Certain coil current combinations (magnetic configurations) seemed to remove the fast ion loading from the immersion tubes. However, in December 2017 it was discovered that such currents would induce intolerable electromechanical loads to the coils [5].

Thereafter, different alternative solutions were quickly explored: retracting the immersion tube away from the plasma, welding the most endangered windows shut, or adding a protective collar around the immersion tube face. Only the last solution was realistic in the very short time before the vacuum vessel was closed at the end of March 2018 for the next campaign.

This contribution describes the final design of the protective collar and first experimental results with fast ion heating. The paper starts with Section 2 describing the simulation results of the problem, then the implemented collar is described in section 3. Sections 4 and 5 describe the simulated fast ion and thermal plasma heat loads with the collar, respectively. Section 6 introduces the simple thermal model used to calculate heat fluxes from the experimental surface temperatures reported in Section 7. A summary concludes the contribution.

Section snippets

Fast ion loading without armoring

This section describes the original issue of fast ion heating of the immersion tube. The problem was first predicted in detailed modelling: the immersion tube model was included in the configuration scan simulations performed in [3] with the ASCOT suite of codes [6], [7]. The codes model the NBI injection geometry, neutral beam ionization, orbit following, fast ion diffusion, slowing down and wall collisions.

The results indicate that there is strong heating of the upper corners of the upper

The implemented armor

Fig. 2 shows the final design for the protective collar. This design was reached after first using simple shapes to find the optimal dimensions for the component. The collar extends ∼4 cm above the face of the immersion tube, which could not be extended without limiting the field of view of the camera. The gap in the collar is in an unloaded part which would obscure key areas of interest for the thermal camera next to it.

The material was chosen to be the 1.4429 stainless steel. A carbon collar

Fast ion loading with the armoring

The final collar geometry was imported to ASCOT and the fast ion load was calculated. It was found that the collar stops the fast ions from reaching the immersion tube face, as illustrated in Fig. 3. The collar fast ion load can reach up to several M W/m2 in the highest loading configuration expected for 2018 experiments (the so called “high iota” configuration).

The simulations were performed with magnetic fields with various plasma pressure to magnetic pressure ratios (βs). When the average

Thermal plasma loading

The original immersion tube protrudes slightly from the surrounding PFCs, so extending the tube by further 4 cm may bring it into contact with thermal plasma. This was studied by moving the immersion tube model with a 4 cm thick collar closer to the plasma and performing field line diffusion calculations [9] to assess the heat load in different magnetic configurations and different plasma 〈β〉s. In these calculations the plasma energy is modelled with markers, that flow out along the magnetic

Thermal modelling

In the case of fast ion heating of the collar, the dominating heat source is the fast ion heat flux arriving to the surface of the collar. The temperatures are too low for radiation losses to be significant. The temperature behaviour is dominated by heat conduction into the collar.

The change of surface temperature ΔT of a semi-infinite solid due to a constant heat flux q (W/m2) as a function of time t is given by the equation ([10])ΔT(t)=q2πλρcpt,where the constants for the collar steel are

Experimental results

In August 2018, the first injections of NBI power to W7-X plasma took place. The length of NBI injection was gradually increased while continuously monitoring the plasma and plasma facing components for signs of NBI fast ion loading. The main and most successful tool for this were the thermal cameras in the AEF-port immersion tubes. The thermal imaging confirmed the predicted fast ion hot spot locations. (Quantitative comparison is beyond the scope of the current contribution.)

Fig. 5 shows a

Summary

This paper summarizes the outcome of armoring the AEF-port immersion tubes against fast ion impacts in the Wendelstein 7-X stellarator. The problem of NBI heatloads was identified in fast ion simulations. The fragile vacuum windows near a fast ion hot-spot were protected by installing a protective stainless steel collar in front of each immersion tube. The design was evaluated with further fast ion and field line diffusion calculations. First experimental measurements of NBI fast ion loads show

Acknowledgements

The authors would like to thank all the participants of the CCB246 meetings on the fast ion heating issue. The work was partly funded by Walter Ahlström foundation. This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014–2018 and 2019–2020 under Grant Agreement No. 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission.

References (10)

There are more references available in the full text version of this article.

Cited by (14)

View all citing articles on Scopus
View full text