National Institute for Fusion Science

In the magnetic confinement devices that are represented by the Large Helical Device (LHD), we confine high-temperature plasma in the magnetic field container, which is invisible to the human eye. Because plasma steadily diffuses to the edge region of the container, part of the plasma is stopped by the divertor. In the future fusion reactor the divertor will be exposed to high heat load.

Thus, at the National Institute for Fusion Science, we are advancing with research on divertors with high efficiency of heat removal. The divertors will use tungsten, which has the highest melting point among the metals, at the places that face plasma. The absorbed heat is removed by water flowing through pipes made from a copper alloy that has good thermal conductivity that touches the back side of the tungsten plate. In order to heighten heat-removal efficiency it is necessary to “strongly bond” the tungsten and the copper alloy. Regarding the bonding of metals, it may be thought that it suffices to use welding. However, although both being the metal, tungsten and copper alloys do not meld because their qualities differ. Thus, connection by welding is extremely difficult. By inserting thin material like paper which performs the role of a bonding agent such as brazing material between the two metals, the “braze-jointing method,” which melts the brazing material at more than 900 degrees Celsius and makes the metals attach together, is commonly used. Because the high-temperature thermal expansion coefficients (the ratio at which volume changes due to temperature change) of tungsten and copper alloys greatly differ, in the past not only the brazing material but also soft materials that served as cushions (called intermediate strata) were necessary. However, due to the insertion of the intermediate strata, because the number of surfaces that must be brazed (bounded interfaces) increases, the strength of the divertor structure will weaken. And, further, the heat removal performance will fall. Moreover, there is the problem of production costs increasing. “Strongly bonding” tungsten and copper alloys is more difficult than thought.

Here, at the National Institute for Fusion Science we have developed a new brazing method that does not use intermediate strata in affixing tungsten and copper alloys. That is, even if we do not use intermediate strata, where the brazing material (brazing connection layer) itself performs the role of a cushion, we have succeeded in producing a divertor test device with powerful and tough bonding and with excellent cooling performance.

There are two factors in this success. The first factor is that we used together the copper alloy called oxide dispersed strengthening copper (ODS-Cu) and the brazing material BNi-6 (composed of 89% nickel and 11% phosphorous). The second factor is that we optimized the thickness of the brazing material, the heat treatment temperature at the time of melting and affixing the brazing material, and the cooling period. The results from experiments that added mechanical load from outside to the joints of tungsten and the copper alloys produced by this method showed thin brazing material similar to paper (brazing material) absorbed shocks by changing the shape, and we realized a powerful bond. Moreover, in results from heat load experiments using an electron beam on the divertor test device, even for heat loads predicted for the future fusion reactor, we confirmed that the temperature of tungsten will be limited to approximately 350 degrees Celsius and the high heat removal performance. A temperature of 350 degrees Celsius is a sufficiently low temperature compared to the temperature of the BNi-6 brazing material melting point (875 degrees Celsius) and the temperature at which tungsten becomes brittle (1,500 degrees Celsius). In this newly established connection method, it is thought that because there are no intermediary layers the heat received by tungsten meets little resistance when it flows to the copper alloy, and that this contributes to the high heat removal performance.

The brazing connection method and the divertor test device utilizing the method that have been newly established in this research realized not only the high performance of the divertor to be used in the fusion reactor but also a great reduction in production costs that are likely to be an issue when constructing the fusion reactor. In the future, using the above-mentioned methods, we will aim at designing and manufacturing a divertor that is stable and operational for a long period by producing a large-scale divertor test device whose structure will be similar to that of the divertor to be utilized in the fusion reactor.