Research and development of copper joint for ITER In-Vessel Coils

The ITER In-Vessel Coils (IVCs) are crucial components of the Tokamak fusion reactor. They are located within the vacuum vessel (VV) behind the blanket shield modules and are responsible for compensating for fast perturbations of the plasma. There are a total of 176 joints used to connect the IVCs, feeders and feedthrough. The joint structure consists of a central copper conductor, a mineral (MgO) insulation layer, and an outer stainless-steel jacket. Given the inaccessibility of the IVCs once the blanket modules are installed, the highest level of quality is required for the joints. One key item is the welding of the copper conductor which has to be done inside the VV. To address the high thermal conductivity of copper conductor and the challenging constraints working inside the VV, a preheating method by the arc of the welding head has been developed. Macro- and micro-structure examination were performed on the welding cross sections, and tensile tests were carried out to check the strength of the copper joint, which demonstrated that arc preheating does not have a significant impact on the heat-affected zone. After more than 40 welding attempts and several design updates, the copper joint has been successfully developed and can be repeatedly and successfully welded.


Introduction
The International Thermonuclear Experimental Reactor (ITER) facility currently under construction at Cadarache, France, is set to become the world's largest tokamak fusion reactor [1]. The In-Vessel Coils (IVCs) including edgelocalized mode and vertical stabilization coils are mainly used * Authors to whom any correspondence should be addressed.
Original content from this work may be used under the terms of the Creative Commons Attribution 4.0 licence. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
to control the position of plasma, which are located inside the vacuum vessel (VV) behind the blanket shield module. An overview of the IVC system is given in figure 1. The main components are described in [2].
To withstand the severe conditions, including mechanical loads due to steady state and transient magnetic fields, high radiation flux (around 3.8 W cm −3 ) and high temperature (up to 420 • C), and provide the required functionality, the IVCs are made of stainless steel mineral insulated conductor as shown in figure 2 [3].
The system comprises a total of 176 joints for interconnecting the IVCs, feeders, and feedthroughs. The joint structure is the same as the structure of the conductor of the IVCs [4]. It consists into a central conductor which  is a water-cooled copper tube (150 • C during operations), a mineral (MgO) insulation layer and an outer stainlesssteel jacket. The joints provide containment of the cooling water in the copper tube, the electrical continuity of the copper conductor, insulation continuity of the MgO insulation layer and vacuum tightness and structural continuity of the steel jacket.
All the joints have to be manufactured inside the ITER VV after the coils and feeders have been installed. One of the main constraints in making joints is the limited available space of 60 mm from the copper conductor surface to the others as shown in figure 3. All tools for copper welding must fit and operate within the available space. Another constraint is that the heat input to the neighboring components has to be limited such that the temperature of those components is kept below 200 • C.
Due to the IVCs' inaccessibility after the installation of the Blanket Modules, the highest level of joint quality is required. As the first step, the welding quality of the copper conductor will directly affect the subsequent process of the joint manufacturing. It shall be qualified in accordance with ASME Sec. VIII Div. 2 and ASME Sec. IX. Considering the strict constraints in the VV and the need for high-quality welding, successfully welding the copper conductor in one attempt is a significant challenge.

Materials and preparation
In order to ensure that the parameters of the welding experiment can be used in the joint fabrication, the copper tube in the experiment is the same as the final IVC conductor. The base material for the copper is a high purity Cu-OFE CW009A. The detailed chemical composition of the Copper tube is respectively presented in table 1.
To achieve control and reproducibility, the orbital welding equipment is used for the welding experiment to produce high quality welds and develop the weld procedure. Figure 4 shows the orbital welding head which is developed by the Institute of Plasma Physics Chinese Academy of Sciences (ASIPP) and POLYSOUDE (Shanghai) company. Welding used for joining Cu tube is done by Tungsten-inert-gas (TIG) welding technique with the φ1.0 mm ERCu as the filler material. TIG welding is one of the most important welding methods, which uses a single, binary or multivariate mixture gas as shielding gas [5]. Table 2 shows the chemical composition of the filler material.

Design of the copper tube joint
As a key part of the complete joint, copper joint plays a very important role as mentioned in the introduction. It requires    that the copper joint shall have equivalent mechanical and electrical performance as the copper tube itself. According to the ASME Sec. VIII Div. 2, the copper joint is defined as a butt joint. The thickness of the copper tube is 6 mm and the outer diameter is 46 mm, the groove is an U-shape as category B type 3 as shown in figure 5. Figure 6 shows the detailed parameters of the copper joint. The nose measures 2 mm in length and 2 mm in thickness. To reduce the penetration on the inner face, the angle of the groove is defined from 105 • to 135 • and have been optimized through welding trails.

Developed of the preheating method
The melting point of the copper is about 1000 • C. The melting temperature of the base material is difficult to reach locally because of the high thermal conductivity and thickness of the copper tube. The molten pool cannot be formed during the welding process. Therefore, the copper tube needs to be preheated before welding. In addition, preheating can reduce the temperature difference and slow down the cooling rate to reduce the welding stress. Three preheating methods were developed during the experiment.
To preheat the copper tube, resistance wire heating belts were employed. They are prepared to fit the outer diameter of the tube so that a proper heat conductivity is assured as shown in the figure 7. Since the length of the bare copper tube is only 80 mm, the heating belt cannot be tightly wrapped up, which leads to low heating power. As a consequence, it takes more than one hour to preheat at 300 • C-400 • C. Moreover, removing the heating belt and installing the welding machine head causes the temperature of the copper tube to decrease to the point that it cannot meet the welding requirements.
An additional preheating method is the induction heating as shown in the figure 8. Induction heating is much faster than the heating belts. The temperature of the copper tube can reach over 600 • C quickly, but it is difficult to accurately control because the length of the bare copper tube is relatively short. After the induction heating, the induction coil needs to be removed and the welding machine installed. During this process, the copper tube is also cooled by air, resulting in a decrease in temperature.
Preheating by the arc of the welding head is shown in figure 9. The temperature of the copper tube is relatively uniform after preheating, and the welding procedure can be directly started after the temperature is reached. Comparing the three preheating methods, arc preheating is obviously more advantageous than the others.

Welding procedure
In order to optimize the specific angle of the groove, analyze the influence of the arc preheating on the heat affected zones (HAZs) and define the final welding parameters, more than 40 different welding trials were carried out. The detailed welding process as follows: • Sample preparation: prior to experimentation, the copper tube's inner and outer surfaces were meticulously prepared through grinding and cleansing with acetone to ensure optimal cleanliness. • Sample assembly: the copper tubes are affixed to the structural framework which emulates the joint's position, and subsequently, the backing gas is introduced into the tubes. The temperature sensors are fixed on the surface of each copper tube 30 mm away from the groove to collect the temperature during the welding process.
• Position adjustment: align the copper tube and identify the welding location. • Preheating: preheat the copper tube using an arc to a temperature range of 300 • C-400 • C. • Welding: one pass by one pass, the temperature of the inter-pass beyond 350 • C. • Recording: accurately record all temperatures and parameters during the welding process for subsequent analysis.

Study of the groove angle on the internal protrusion
Given the crucial functional significance of the joint, it is paramount to minimize the degree of internal protrusion, as illustrated in figure 10. Adhering to the most exacting standards of acceptance (Level B of ISO 5817), the internal protrusion must satisfy the following prescribed criteria: h ⩽ 1 + 0.2b, but max. 3 mm.
Where b is the width of weld reinforcement (being 2 mm here) and h is the height of the protrusion. Therefore, the requirement of the protrusion is: h ⩽ 1.4 mm. Figure 11 displays the correlation between the groove angle and the internal penetration. Evidently, the protrusion is contingent on the magnitude of the breach angle. Greater angles of breach result in significantly larger protrusions compared to smaller angles. Based on this outcome, the optimal angle of breach is determined to be 105 • . The final welding parameters are presented in table 3.

Study of the arc preheating on the HAZ
The utilization of arc preheating in welding copper tubes may have deleterious effects on their structural integrity. To assess the likelihood of hot cracking or micro-structural alterations, it is recommended to undertake a metallographic analysis. Given the critical role of joint welding in ensuring the longevity of the conductor and the stringent reliability demands placed on it, it is of utmost importance to conduct a thorough metallographic examination of welded joint specimens, encompassing both the weld and the heat-affected zone.
Two OFE copper welded tube samples were produced and provided to CERN's Materials and Metrology laboratory for macro and micrographic examination of the weld. Two cross sections per sample were analyzed, investigating the hardness profiles of the welds and the bulk material [7]. One sample was welded with a 45 • angle (sample 45 • ), and the second one with a 90 • orientation with respect the vertical axis (sample 90 • ) as shown in figure 12.
The macro-structure and micro-structure are examined using stitched images (up to 50× magnification per individual image for the macroscopic examination) obtained with a Keyence VHX 6000 optical microscope. Two cross-sections per sample were mechanically ground to a final 1200 mesh size and then polished to a mirror-like finishing. Afterwards, they were etched by immersion with etchant #31 of ASTM   E407 (ammonium persulfate 10% in volume) to reveal the macro-structure. Figures 13 and 14 show the macro-structure of the weld and its surroundings for the two orientations. Additionally, a sample far enough from the weld and HAZ was prepared with the same procedure ( figure 15) to confirm that the macro-structure observed at the extremities of figures 13 and 14 is consistent with the macro-structure of the parent metal shown in figure 15. HAZ was assessed using the macro-structural observation and based on the grain size variation. Indeed, the arc preheating process can induce a grain growth in the vicinity of the weld location, due to the increase of temperature operated. HAZ size is rather homogeneous for all the analyzed cross sections. The measurements are summarized as below (see table 4):     Additionally, the affected zone is further investigated by hardness measurements to see if the arc preheating has a wider impact in a broader zone. Micro-hardness tests of the samples was performed following ISO 6507-1, with a load  The measurements are coherent with an annealed OFE copper material. The hardness in the weld is slightly higher than both the HAZ and the parent material. This is consistent with the higher rate of solidification in the weld bead during TIG welding process and lesser purity of the filler.

Mechanical performance test on the welds
Two copper joints were manufactured according the finalized weld parameters and preheated by the arc of welding head. Nondestructive test (NDT), such as leak test, penetration test and x-ray test were carried out on these samples and no macro defects were detected. After NDT, the joints were put in a vacuum furnace and heated treatment at 240 • C for 24 h. 8 subsize specimens were sectioned following the ISO 4136 standard by electrical discharge machining from the joints after heat treatment to be carried out tensile test at room temperature (RT) and 130 • C. Configuration and dimensions of the subsize specimens are shown in figure 17.
From figure 18 can be observed that the failure during tensile loading took place at the base material for all the specimens with significant necking. This implies that plastic deformation in the base material occurs during the tensile

Conclusion
The copper conductor joint for ITER IVCs has been developed. To achieve control and reproducibility, full penetration TIG welding is done with an orbital welding tool. During the operation and cooling processes, inert gas flows continuously on the inner and outer sides of the joint. The angle of the groove is optimized at 105 • to reduce the protrusion on the inner face. To address the high thermal conductivity of copper and space constraints in the welding area, preheating methodology by the arc of the welding head is also developed. Macro-structure and micro-structure examinations were performed on four cross sections. HAZ assessment was carried out based on a grain size variation on the crosssections. Micro-hardness test were made to assess the influence of the arc preheating step on the HAZ. No significant difference was found between the hardness in the HAZ and in the base metal. Based on these observations, arc preheating does not have a significant impact on the HAZ. Tensile tests were carried out to check the strength of the welded joint. The results showed that the failure during tensile loading took place at the base metal for all the specimens. Therefore, the copper joint has been successfully developed and its reproducibility fully achieved.