Device and use of the device for reflecting microwave radiation during electron-cyclotron resonance heating

Abstract

The present invention relates to a device 100 for reflecting microwave radiation during electron- cyclotron resonance heating (ECRH). The device 100 comprises a bulk portion 200 and at least one reflective layer 300a, 300b, 300c, wherein the bulk portion 200 includes dielectric material and the at least one reflective layer 300a, 300b, 300c is made of an electrically conductive material which has an electrical conductivity in a range of 0.1 - 70 MS / m, and wherein the thermal expansion coefficient of the bulk portion 200 differs from the thermal expansion coefficient of the at least one reflective layer 300a, 300b, 300c by an amount in a range of 0 - 20x10-6 / K, and wherein a cooling means 400a, 400b, 400c is provided at or at least partially within the bulk portion 200.

Description

[0001] Device and use of the device for reflecting microwave radiation during electron-cyclotron resonance heating

[0002] Technical field

[0003] The present invention relates to a device and use of the same for reflecting microwave radiation during electron-cyclotron resonance heating (ECRH).

[0004] Background

[0005] In thermonuclear fusion devices, a plasma consisting mainly of hydrogen isotopes is heated up to temperatures in the order of 108K to allow fusion reactions to take place. The plasma is suspended (i.e. confined) in a vacuum chamber by extremely strong magnetic fields (up to about 10 T). Plasma heating is the main purpose of ECRH systems, whose operating principle is based on the injection of high-power microwave beams (up to about 1 MW per beam) into the plasma, which then absorbs the beam power and heats up. Microwaves are usually generated away from the machine and are transmitted to the vacuum chamber by means of a series of mirrors, typically made of metallic materials. The last of these mirrors is called a launching mirror. It is often placed inside the vacuum chamber, and reflects the incident microwave beam into the plasma. To fulfill the purpose of the ECRH system and to efficiently inject the power into the plasma, the launching mirror is equipped with a horizontally and vertically movable support system which allows the injection angle to be changed.

[0006] Conventional launching mirrors as well as other mirrors along the transmission lines of the ECRH system are typically made of electrically conductive material. The main reason is that good electrical conductors efficiently reflect microwaves, thus minimizing power loss which would otherwise be lost by mirror absorption. Another reason is that good electric conductors are in general also good thermal conductors. The thermal conductivity property is particularly important in mirrors that are exposed to microwave beams for more than a few seconds, which results in excessive mirror temperatures and thus possibly damage and misalignment of the mirror. Hence, the mirrors in ECRH systems which operate for several tens of seconds (long- pulse operation) need to be actively cooled. Conventional launching mirrors are therefore made entirely of copper, stainless steel, or they exhibit a multilayer configuration with a copper surface and a stainless-steel bulk body.

[0007] The powerful magnetic fields used to ensure the confinement of the plasma may be subject to intense variations. During normal machine operations, magnetic fields vary with relatively small-time gradients to ensure the stability of the plasma and preserve its fundamental parameters. However, there are events (disruptions), which are undesired, but yet difficult to avoid, which result in a sudden loss of plasma stability and thus in a sudden interruption of the experiment due to cool down of the plasma within time intervals on the order of a few milliseconds. In such circumstances, the strength of the magnetic field inside the machine varies rapidly and significantly, with time gradients up to 100 or 1000 times higher than those during normal machine operation.

[0008] The variation of the magnetic fields inside the machine induces eddy currents in all electrically conductive components, including the launching mirrors of the ECRH system, which are particularly vulnerable to mechanical loads, as they are movably connected to the support system. These mechanical loads are absorbed by the various components of support system, such as the mechanical transmissions and the actuators, which must therefore be sized to not only meet dynamic requirements (positioning speed and accuracy, acceleration capacity), but also have to withstand electromagnetic loads, particularly those resulting from disruptive events.

[0009] In addition, eddy currents yield a net moment of the Lorentz force produced by the magnetic field on the moving charges in the electrically conductive components, which is often referred as magnetic torque. Magnetic torques have several undesirable effects on the launching system. First, they can cause misalignment of the ECRH system components, such as mirrors, and reduce plasma launching efficiency. For mirrors reflecting high-power beams, alignment and preservation of beam shape is of great importance. Failure in doing so may direct the reflected beam in unwanted directions or change its power distribution (propagation modes), thus possibly damaging surrounding objects as well as having detrimental effects on mirror functionality. Furthermore, magnetic torques are also responsible for structural stress exerted on ECRH system components, for example launching mirrors. Long-term exposure to strong magnetic torques could lead to wear or failure in the structural components of the ECRH system.

[0010] In metals, a reduction of electrical conductivity, which is advantageous for reducing electromagnetic effects due to the strong variation of the magnetic field during disruptive events, is accompanied by a disadvantageous decrease in thermal conductivity. State-of-the- art mirrors, made of different metals, experience high thermal stress in large-scale machines where the magnetic field undergoes significant changes during plasma heating. One possible solution for reducing thermal stress on components is to increase their size, so that the absorbed thermal energy distributes in a larger volume, resulting in lower average temperature.

[0011] However, this solution has a substantial disadvantage. As the size of the components increases, so does their weight. In turn, this leads to higher manufacturing and operating costs for fusion devices. At the same time, significant oversizing of components also increases the complexity of support systems, requiring the development of sophisticated actuators, timeconsuming adjustment of the same, significant maintenance costs and extensive training for technicians. These challenges become increasingly significant for large-scale fusion devices of industrial interest, as the electromagnetic loads grow with the size of the components.

[0012] Summary

[0013] It is an object of the present invention to mitigate, or even eliminate, the drawbacks of the prior art. In particular, it is an object of the present invention to provide a device for reflecting microwave radiation during electron-cyclotron resonance heating which allows for a reduction of electromagnetic, and thus mechanical loads, during both normal operating conditions and unforeseen disruptive events, yet is effectively prevented from significant material damage and overheating.

[0014] A first aspect of the present invention provides a device for reflecting microwave radiation during electron-cyclotron resonance heating (ECRH). The device comprises a bulk portion and at least one reflective layer. The bulk portion includes dielectric material and the at least one reflective layer is made of an electrically conductive material which has an electrical conductivity in a range of 0.1 - 70 MS / m. The thermal expansion coefficient of the bulk portion differs from the thermal expansion coefficient of the at least one reflective layer by an amount in a range of 0 - 20x10-6 / K. Furthermore, a cooling means is provided at or at least partially within the bulk portion

[0015] Preferably, the cooling means is an elongate cooling channel having an inlet and an outlet.

[0016] In general, the device according to the invention is suitable for use in microwave injection systems for fusion plasmas where high-power microwave beams need to be reflected and subsequently injected into the plasma. The device according to the invention is particularly useful as a launching mirror in Electron Cyclotron Resonance Heating (ECRH) systems, which are installed, for example, in tokamak, stellarator or other machines whose operation is based on the principle of magnetic confinement.

[0017] When the device according to the invention is used in a typical ECRH system, for example as a launching mirror, it is immersed in an external magnetic field. To decrease surface aberrations and misalignments of the device, a bulk material with a low thermal expansion coefficient is selected. The thermal expansion coefficient should preferably be in a range of -10x10-6 / K to +10x10-6 / K. Comparing to conventional devices made of metals, it is advantageous to use a dielectric material, for example a ceramic, such as silicon carbide (SiC) or aluminum nitride (AIN), with high thermal conductivity for the bulk portion of the device. In metals, a reduction of electrical conductivity, desirable to reduce electromagnetic effects, is accompanied by the less desirable decrease in thermal conductivity, which leads towards cumulation of heat in the components and therefore increase of the operating temperature of the mirrors. On the other hand, using a dielectric instead of a metal for the bulk portion of the device allows for a decoupling of the thermal conductivity from the electrical conductivity and has the potential of providing adequate cooling performance while not contributing to magnetic torques. It is worthwhile to note that in dielectrics, thermal conductivity can reach high values (up to 200 Wnr1K'1for SiC and AIN), even in the total absence of electrical conductivity.

[0018] If the bulk portion of the device comprises dielectric material, it is transparent to microwaves and needs to be coated with a micrometric layer of conductive material to restore reflection properties of the device. To minimize induced stresses at the interface between the two materials during thermal cycling, it is necessary to choose a coating material with a thermal expansion coefficient as close as possible to that of the dielectric bulk. For this reason, as well as for its excellent thermal stability, tungsten is a suitable candidate for the coating layer. For example, a dielectric material, such as SiC and AIN, has a thermal expansion coefficient between 4 to 5.5x10-6 / K , while that of tungsten is about 4.3x10-6 / K. Other conductors are also suitable as a coating material, providing that the thermal expansion coefficient of the reflective layer differs from the thermal expansion coefficient of the bulk portion by an amount in a range of 0 to 20x10-6 / K.

[0019] The invention is of great benefit because the combination of properties of dielectric and conductive materials makes it possible to eliminate the electromagnetic loads induced on the device without affecting the efficiency of heat exchange and cooling capacity.

[0020] In addition, the dielectric bulk ensures structural integrity of the devise and superior heat exchange with the aid of the cooling means.

[0021] For example, the dielectric material used for the bulk portion may be selected from the group consisting Germanium (Ge), Boron nitride (BN), Magnesia (Mg), Silicon, Silicon carbide (SiC), Aluminum nitride (AIN), Aluminum oxide (AI2O3), Beryllium oxide (BeO) and diamond (C)., whereas the material used for the reflective layer may be selected from the group consisting of Tungsten, Osmium, Platinum, Ruthenium, Tantalum, Iridium, Rhodium, Chromium and Molybdenum..

[0022] If the bulk portion comprises a ceramic as a dielectric material, it may become damaged due to its brittleness when exposed to very high temperatures in larger fusion machines. Failure of ceramics is difficult to predict even for known operating conditions and known material properties, as it occurs rather randomly due to the presence of small defects in the ceramic. This poses a significant risk as failure would result in malfunctioning of the device with associated costs, or higher maintenance costs due to relatively frequent substitutions (a brittle mirror should be subjected to less thermal cycles than a ductile one).

[0023] Thermal deformations are the main source of internal stress in a mirror tailored to handle high- power beams. To reduce such deformations, a dielectric bulk material with low thermal expansion coefficient, preferably in a range of -10x10-6 / K to +10x10-6 / K, is used. Then, working temperatures should be minimized, e.g. 400 °C using a Tungsten coating. Lower working temperatures can be achieved by reducing the power of the incident beam(s), by increasing the size of the component, by selecting a dielectric bulk material with high thermal conductivity, for example 25 W / mK, or by providing active cooling. Active cooling can be achieved in many ways. For example, cooling means in the form of a liquid layer with an inlet and an outlet may be provided between two dielectric layers. The liquid flow would only be contained by aspects of the device, for example a lateral solid perimeter of the device, and it would find its way from the inlet to the outlet unguided. However, the most effective way of cooling the device is by means of cooling channels, since this allows for forced convection to be exploited more efficiently and at locations very close to the reflective surface of the device.

[0024] The heat load absorbed by the device can easily reach a value of 15 kW, and the absorbed heat flux can amount to values in the order of MW4n2 in the center of its Gaussian distribution . Therefore, an effective heat conduction within the device and efficient heat exchange between the bulk portion and the cooling means, preferably the liquid within a cooling channel is one of the key conditions for device functionality.

[0025] Providing active cooling, preferably by means of a cooling channel having an inlet and an outlet at or at least partially within the dielectric bulk portion, helps mitigating or even eliminating the brittleness issue of the dielectric bulk portion. Introducing cooling channels at or at least partially within the bulk portion decreases costs of the device because ceramics may be used that are less expensive, like Alumina or Magnesia. Moreover, replacement of the device due to a damaged bulk portion needs to be done less frequently.

[0026] The preferred cooling means is a cooling channel which may have different geometries, such as linear-form, circular-form, spiral-form, zig-zag, meander-form etc. A preferred cross-sectional shape of the cooling channel is cylindrical. However, other cross-sectional shapes are conceivable. The cooling channel may have one or more bifurcations and may have one or more inlets and one or more outlets. The cooling channel may provide a continuous cavity for the cooling fluid to flow in a guided manner from the inlet to the outlet. The cooling channel may share one or more surfaces with one or more dielectric bulk portions. The depth and / or the width of the cooling channel may vary along the length of the cooling channel, thereby following a certain (depth or width) function, such as one or more Gaussians, sinusoid, or a combination thereof. The interior of the cooling channel may exhibit one or more protrusions which protrude inwardly into the interior or cavity of the cooling channel to deflect the fluid flow through the cooling channel and / or to generate turbulences to increase heat exchange. Moreover, the internal wall or walls of the cooling channel may be polished or be rough. There may also be a plurality of cooling channels, each of which may be configured as described above.

[0027] Another important advantage of having cooling channels at or at least partially within the bulk portion is that steady-state operation of the fusion machine can be achieved more easily, and that operating temperatures are lower, since thermal equilibrium of the device is obtained at lower temperatures, for example below 300 C°. This not only achieves steady-state operation when using dielectric bulk material but also yields an economical advantage, since steady-state operation of the device at lower temperatures requires less frequent maintenance operations and substitutions, as the brittle failure risk is reduced.

[0028] Cooling liquids can be chosen such that they are compatible, for example in terms of corrosion, with the dielectric bulk. SiC is particularly corrosion-resistant, so it is compatible with almost any cooling fluid. Non-limiting examples of cooling fluids are water, glycol water or gaseous helium.

[0029] In an embodiment, the cooling channel is provided entirely within the bulk portion. If the cooling channel is entirely within the bulk portion, the cooling channel is embedded within the bulk portion with no portion of the cooling channel being exposed to the vacuum chamber, thus improving the cooling efficiency.

[0030] In an embodiment, the cooling channel spirals radially outwardly. Cooling fluid may thus be introduced into the cooling channel at the center of the bulk portion where temperatures are highest. As the cooling fluid propagates through the cooling channel, it warms up and reaches the periphery of the bulk portion, where temperatures are lowest. Hence, cooling efficiency is further optimized.

[0031] In an embodiment, the cooling channel has a depth and / or width which varies along the length of the cooling channel, which allows to selectively increase the velocity of the cooling fluid through the cooling channel, thus allowing for selective and varying heat transfer from the bulk portion to the cooling fluid at selected areas of the bulk portion. The depth, the width and the length of the cooling channel are understood to extend along the axes of a Cartesian coordinate system.

[0032] In an embodiment, the varying depth and / or width of the cooling channel is defined by a depth / width function. The depth / width function comprises at least one Gaussian, a sinusoid or a combination thereof. The Gaussian function optimizes heat exchange when incident beam(s) have a Gaussian power density distribution, while the sinusoidal function optimizes heat exchange when the angle of incidence is different from 0°. In both cases, heat exchange enhancement is due to selective increase of fluid velocity in the areas that are subject to the highest heat load.

[0033] In an embodiment, the cooling channel is made in one piece with the bulk portion. If the cooling channel is made in one piece with the bulk portion, heat transfer from the bulk portion to the cooling fluid within the cooling channel is improved as well as manufacturing costs are reduced, as no prefabricated cooling channel, prefabricated for example from another material different to that of the bulk portion, needs to be implemented within the bulk portion.

[0034] In an embodiment, the bulk portion and the cooling channel are made by molding or by additive manufacturing. Production of the bulk portion may be done by means of additive manufacturing techniques, such as 3D printing. In particular, commercial solutions are available for 3D printing of Si-SiC compounds, in which the presence of moderate percentages of silicon (15% - 30%) prevents the formation of porosity. Additive manufacturing advantageously delivers a monolithic product with varying shapes and facilitates the manufacture of cooling channels in the bulk portion comprising dielectric materials like SiC, Si-SiC, AIN, AI2O3.

[0035] Production of the bulk portion by means of molding techniques allows the use of a wide variety of dielectric materials and yields high-density bulk material.

[0036] In an embodiment, the bulk portion is made entirely of dielectric material and has an electrical conductivity in a range of 0 - 0.1 MS / m. Such dielectric materials are advantageous in terms of very low sensitivity of the material to magnetic fields, with consequent strong reduction of mechanical loads resulting from magnetic induction, reduction in mechanical stresses and reduction of mirror deformation or misalignments.

[0037] In an embodiment, the thickness of the bulk portion is in a range of 1 - 30 mm, thus providing sufficient rigidity to the entire device. Moreover, bulk portions with a larger thickness allow for the accommodation of a larger cooling channel with a higher cooling capacity as well as provide the device with increased structural stiffness.

[0038] In an embodiment, the at least one reflective layer includes a plurality of reflective layers. The sum of the product of the thickness times the electrical conductivity of each reflective layer over the plurality of reflective layers divided by the sum of the total thickness of the plurality of reflective layers is in a range of 0.1 - 70 MS / m. Including a plurality of reflective layers can improve adherence between the at least one reflective layer and the bulk portion, for example by providing a gradual change in thermal expansion coefficient along the thickness of the at least one reflective layer.

[0039] In an embodiment, the at least one reflective layer is arranged to one side of the bulk portion, which greatly facilitates manufacture and reduces costs of the device.

[0040] In an embodiment, the thickness of the at least one reflective layer and / or the thickness of the bulk portion is uniform across the device. Again, manufacture of the device is facilitated, and manufacturing costs are reduced.

[0041] In an embodiment, the thickness of the at least one reflective layer is in a range of 0.1 - 100 pm, preferably in a range of 1 .7 - 7.5 pm. The minimum value is governed by the fact that transparency to the incident beam is to be avoided, and the maximum value assures that electromagnetic forces and torques are reduced to acceptable values.

[0042] In an embodiment, the at least one reflective layer is deposited via magnetron sputtering on the bulk portion. Deposition of the reflecting layer can be realized through a technique called magnetron sputtering, which is suitable for relatively large surfaces, like those of ECRH mirrors, with thicknesses of a few micrometers. In an embodiment, the material used for the bulk portion is selected from the group consisting of Germanium (Ge), Boron nitride (BN), Magnesia (Mg), Silicon, Silicon carbide (SiC), Aluminum nitride (AIN), Aluminum oxide (AI2O3), Beryllium oxide (BeO) and diamond (C).

[0043] In an embodiment, the material used for the at least one reflective layer is selected from the group consisting of Tungsten, Osmium, Platinum, Ruthenium, Tantalum, Iridium, Rhodium, Chromium and Molybdenum. For example, the function of a tungsten coating is to ensure microwave reflection. The choice of tungsten is aimed at minimizing thermal stresses at the interface with the bulk portion, by virtue of the low thermal expansion coefficient. The minimum coating thickness depends on the microwave frequency used and is in the order of a few micrometers. Its maximum thickness depends on the acceptable mechanical load during the disruption phase. Indeed, since the tungsten coating is conductive, it is affected by slight electromagnetic forces during the disruptive events of fusion machines, which are in any case 3 or 4 orders of magnitude lower than those that would be discharged on a metallic mirror. On the other hand, electromagnetic forces resulting from normal machine operation are negligible.

[0044] In an embodiment, the device is a launching mirror used in an Electron Cyclotron Resonance Heating system.

[0045] A second aspect of the present invention relates to the use of the device in an Electron Cyclotron Resonance Heating system for reflecting microwave radiation during electroncyclotron resonance heating of fusion plasma.

[0046] The invention is not strictly limited to what has been written above but may include combinations of the above features not specifically disclosed in that combination above.

[0047] These and other aspects and advantages of the invention will become apparent from the detailed description of preferred embodiments of the invention and will be elucidated with reference to the drawings.

[0048] Brief description of the drawings

[0049] Figs. 1 a - 1f show a device for reflecting microwave radiation, which may be employed as a mirror for an ECRH system, according to preferred embodiments of the invention;

[0050] Fig. 2 shows an example of a dielectric bulk 200 with a multi-layer tungsten coating 300b-1 , 300b-2, according to a preferred embodiment of the invention;

[0051] Figs. 3a, 3b show further alternative of multi-layer reflecting conductive coatings 300c-1 , ..., 300c-N according to further preferred embodiments of the invention; Fig. 4 shows a table with thermodynamical data and mechanical stress data;

[0052] Fig. 5a, 5b, 5c shows a device for reflecting microwave radiation, which may be employed as a mirror for an ECRH system, according to another preferred embodiment of the invention; and

[0053] Fig. 6 shows a cooling channel having a zig-zag form.

[0054] Detailed description of preferred embodiments

[0055] With reference to Fig. 1 a-1 f, two devices 100 according to preferred embodiments of the invention are shown.

[0056] The devices 100 shown in Fig. 1 a-1 f includes a dielectric bulk 200 and at least one reflecting layer 300a. The expression “bulk” is understood to mean that the dielectric bulk 200 makes up for the majority of the mass and / or volume of the device 100. The dielectric bulk 200 is not simply a thin layer or coating, but instead gives the device 100 sufficient structural rigidity and stability and serves as a substrate for the at least one reflecting layer 300a.

[0057] The bulk portion 200 includes dielectric material and the at least one reflective layer 300a is made of an electrically conductive material which has an electrical conductivity in a range of 0.1 MS / m - 70 MS / m. The thermal expansion coefficient of the bulk portion 200 differs from the thermal expansion coefficient of the at least one reflective layer 300a by an amount which is in a range of 0 - 20x10-6 / K.

[0058] In a preferred embodiment of the invention, the bulk portion 200 is made entirely of dielectric material and has an electrical conductivity in a range of 0 - 0.1 MS / m. The thickness of the bulk portion 200 is preferably in a range of 1 - 30 mm.

[0059] The thickness of the at least one reflective layer 300a and / or the thickness of the bulk portion 200 is preferably uniform across the device 100, and the thickness of the at least one reflective layer 300a is preferably in a range of 0.1 - 100 pm, and more preferably in a range of 1 .7 - 7.5 pm.

[0060] Fig. 1 b is a front view of device 100. It shows the shape of the device 100 which in this case is elliptical. The invention, however, is not limited to elliptically shaped devices 100. The device 100 may also be circular, rectangular, polygonal or may have any other shape that makes the device 100 suitable for reflecting microwave radiation during electron-cyclotron resonance heating (ECRH). Fig. 1c is a cross-sectional view of the device 100 shown in Fig. 1 a-1 b along line A-A. Fig. 1c shows a cooling channel 400a, 400b having an inlet 402 and an outlet 401 . The inlet 402 may also be used as an outlet 401 inasmuch as the outlet 401 may be used as an inlet 402. The cooling channel 400a, 400b is embedded within the dielectric bulk 200 and spirals radially outwardly. The cooling channel 400a, 400b may be made in one piece with the dielectric bulk 200 but may also be a separate element from the dielectric bulk 200.

[0061] Fig. 1d and Fig. 1f show different configurations of the cooling channel 400a, 400b.

[0062] According to Fig. 1 d, the cooling channel 400a has a variable depth along the length of the cooling channel 400a. The depth extends in a direction of the thickness of the device 100. In addition, the width (perpendicular to the depth) of the cooling channel 400a may also vary. The depth and / or the width of the cooling channel 400a may follow a Gaussian curve as it spirals radially outwardly, i.e. , the depth and / or width is smallest in the center of the device 100, and greatest in proximity of the outer peripheral edge of the device 100. The Gaussian function, and hence the varying depth and / or width of the cooling channel 400a images or is at least similar to the microwave power density that is reflected by the device 100 in operation. The varying width helps maximizing the exchange area and the varying depth helps to increase the local efficiency of heat exchange.

[0063] Other functions from which the varying depth and / or width of the cooling channel 400a can be determined are conceivable, such as for example a sinusoidal function or a combination of a Gaussian and a sinusoidal function. The depth may thereby follow a different function than the width. In case of a sinusoidal depth / width function, the depth and / or width varies sinusoidally in a circumferential direction, i.e., along the length of the cooling channel 400a. The sinusoidal function is useful in that it optimizes the heat exchange when the angle of the incident beam is different from 0°.

[0064] Fig. 1 e illustrates the depth and / or width variation of the cooling channel. The inlet 402 and the outlet 401 of the cooling channel are shown in Fig. 1 e as well. The depth and / or the width can be seen to extend between two planes 403a, 403b. The distance of the planes 403a, 403b indicates the variation of the depth and / or the width. The median between the planes 403a, 403b may be considered the central axis of the cooling channel. Hence, the cooling channel may be seen to be more or less planar and to extend, with a maximum variation which is equal to the distance of the planes 403a, 403b, in a plane within the dielectric bulk portion 200.

[0065] According to Fig. 1 f, the depth and / or width of the cooling channel 400b may also remain constant.

[0066] There are two advantages of using a cooling channel 400a with a variable depth within the dielectric bulk 200. First, use of a cooling channel 400a with variable depth allows to accelerate the cooling fluid along the length of the cooling channel 400a. Therefore, it avoids hotspots along that direction. Second, it helps to decelerate the cooling fluid near the peripheral edge of the dielectric bulk where the incident power is low. Moreover, pressure drops and mechanical overloads of the bulk material can be avoided, while also avoiding oversizing of other aspects of the cooling system, such as the pumps, tubes and so on. Both conditions contribute to increasing cooling efficiency of the device 100.

[0067] In the case of a cooling channel 400b with constant depth, as shown in Fig. 1 f, it was found that the maximum temperature at the reflecting surface 300a is 138.0 °C. The maximum temperature of water as the cooling fluid, found at the points of contact with the solid bulk part 200, is 116 °C. The pressure drop between inlet 402 and outlet 401 is 1250 kPa.

[0068] In the case of a cooling channel 400a with variable depth, such as shown in Fig. 1 d, the maximum temperature at the reflecting surface 300a is 135.4 °C. The maximum temperature of water, found at the points of contact with the solid bulk part, is 98.5 °C. The pressure drop between inlet 402 and outlet 401 is 470 kPa.

[0069] Therefore, using a cooling channel 400a with variable depth contributes to deceasing the temperature at the reflecting surface 300a from 138.0 °C to 135.4 °C. At the same time, it contributes to significantly reducing the pressure drop between the inlet 402 of the cooling channel 400a in the center of the device 100 and the outlet 401 near the peripheral edge of the device 100 from 1250 kPa to 470 kPa.

[0070] In case of a cooling channel 400a with variable depth, the maximum local tension in the ceramic material is 45 MPa, reached during thermal equilibrium phase, while maximum local compression of -66 MPa is reached during the heating phase. Both values are compatible with material specifications, which in the case of SiC and AIN is around 200 MPa for the stress limit.

[0071] In a concept study, a device 100 was studied as shown in Figs. 1 a - 1 e, the cooling channel 400a was made in one piece with the dielectric bulk 200, i.e. , the dielectric bulk 200 was carved out to form the cooling channel 400a. The spiral cooling channel 400a was delimited by a 2 mm thick solid wall, and the path of the 2 mm thick solid wall may be described by the following curve: ach= 68.5 mm r = 1.913 y = 1.16 T] = 1.1 8X= 0.0285 8y= —0.05 n = 9.5 i90= n

[0072] The varying depth of the cooling channel 400a is indicated by h and is given by the following function: he= 6 mm / i0= 2.5 mm A / i0= 2 mm ns= 6 KS= 20 a = 69 mm b = 132 mm

[0073] The simulated device 100 had a thickness of 10 mm (Fig. 1 a), and an elliptical peripheral edge (surface or shape) with diameters of 138 mm and 264 mm (Fig. 1 b). The device 100 had internal cooling channels 400a which are embedded in the dielectric bulk 200 (Fig. 1c).

[0074] The device was exposed to the following operational conditions in order to test different dielectric bulk materials 200, reflecting coating layers 300a and the efficiency of different cooling channel 400a configurations.

[0075] In numerical simulations, which last for a time period of 100 s, the device 100 of the present invention was used as a launching mirror which was subjected to thermal loads of three different types: irradiation with 1 MW 170 GHz microwave beams, for an absorbed power of about 9.5 kW with Gaussian surface intensity profile; infrared radiation from the plasma for an absorbed power of about 5.5 kWwith uniform surface intensity profile; neutron irradiation of negligible intensity.

[0076] The formula below quantifies the power density absorbed by the launching mirror:

[0077] Amw= 1%

[0078] Qmw= 1 MW w = 43 mm umwJOApi = 50% qpl= 0.5 MW / m2 35°

[0079] The numerical simulation setup comprised steady-state condition of the operation, i.e. thermodynamic equilibrium. The materials selected for the ceramic substrate 200 are silicon carbide (SiC) and aluminum nitride (AIN). Furthermore, suitable commercial Si-SiC materials for additive manufacturing with thermal conductivity greater than 150 Wnr'K'1and flexural mechanical strength greater than 200 MPa were used in the simulations. Use of those materials provides magnetic field neutrality and therefore avoids the induction of eddy currents and magnetic torques during the operation in a strong magnetic field, which can reach values in the range of 0.1 T to 4 T.

[0080] Finite element thermo-fluid-structural analyses were performed to quantify the distribution of temperature, fluid velocity, pressure and stress resulting from the application of thermal and hydraulic load. A thermal conductivity of 100 Wm1K'1of the mirror was considered for these analyses.

[0081] The cooling fluid provided in the cooling channels 400a, 400b is demineralized water, with a flow rate of 10 l / min, an inlet temperature of 50 °C and an outlet pressure of 400 kPa.

[0082] In one experiment, a specimen was used which included a bulk portion 200 that was made of Si- SiC, which had following material properties: Young’s modulus is 340 GPa, Poisson ratio is 0.18; thermal expansion coefficient is 4x10x-6 / K; mechanical resistance is 220 MPa; fracture toughness is 2 MPa^ / m. In this experiment, the reflective coating layer 300a was made from tungsten, which was deposed on the bulk 200 via a magneton sputtering deposition technique. The thickness of the tungsten coating 300a was 2 pm and is a value sufficient to ensure microwave reflection (with transparency losses of less than 100 W for 1 MW incident power) and at the same time makes the effect of electromagnetic forces negligible.

[0083] As shown in Fig. 2, another device 100 according to another preferred embodiment of the invention includes two reflecting layers 300b-1 and 300b-2. They may be deposited by magnetron sputtering as follows: a first interface layer 300b-2 with a thickness of 0.5 pm and with reduced relative density (70%), i.e. porous Tungsten to improve adhesion of the first interface layer 300b-2 to the substrate 200, and a second surface layer 300b-1 made from dense Tungsten with a thickness of 1 .5 pm at maximum relative density (100%) to ensure reflectivity. Such a coating with two layers 300b-1 , 300b-2 provides for effective microwave reflection while increasing adhesion of the coating to the dielectric substrate.

[0084] Tungsten has the following material properties: Young’s modulus is 400 GPa; Poisson ratio is 0.28; thermal extension coefficient is 4.4x10-6 / K; tensile strength is 750 MPa. In this experiment, the thermal expansion of bulk (4x10-6 / K) was similar to that of the coating (4.4x10-6 / K), which guarantees structural integrity and mechanical stability of the mirror.

[0085] Further multi-layer reflective coatings are shown in Fig. 3a and Fig. 3b.

[0086] Fig. 3a shows a reflective layer which includes a multiplicity of reflective layers 300c-1 , 300c-2, ..., 300c-N. The reflective layers 300c-1 , 300c-2, 300c-N may have the same thickness or may have a variable thickness. A multi-layer reflective coating offers the advantage that a large spectrum of differing reflective properties using different metals may be obtained. Moreover, it has the advantage that adhesion of the coating to the substrate could be improved.

[0087] Fig. 3b shows an embodiment, in which the reflective layers 300c-1 , 300c-2, ..., 300c-N are or alternatively only one reflective layer 300a is embedded or buried inside the dielectric bulk 200a, 200b, i.e. , two dielectric bulk portions 200a, 200b. Such a configuration offers the advantage that the risk of detachment of one or more reflective layers 300c-1 , 300c-2, ..., 300c-N (alternatively 300a) from the dielectric bulk 200 due to thermo-mechanical loads is reduced. Furthermore, it allows the use of a liquid electrically conductive material (e.g. a liquid metal) as a reflective layer.

[0088] Using a multi-layer reflective coating, such as shown in Fig. 3a and Fig. 3b, the sum of the product of the thickness times the electrical conductivity of each reflective layer 300c-1 , 300c-2, ..., 300c- N over the plurality of reflective layers 300c-1 , ..., 300c-N divided by the sum of the total thickness of the plurality of reflective layers 300c-1 , ..., 300c-N is preferably in a range of 0.1 - 70 MS / m.

[0089] Fig. 4 shows a table with thermodynamical data and mechanical stress data in case the dielectric bulk 200 is made of SiC and comprises spiral cooling channels 400a with variable depth, as shown in Fig. 1d. The thermodynamical data include temperature distribution T [°C], absorbed power density P [MW / m2], and fluid velocity distribution V [m / s] as a function of position X[mm], Y[mm] on the front surface 300a of the device 100, which is the most critical region. The table reports maximum and minimum principal stresses Stens [MPa] and Scomp [MPa] at each position. Principal stresses are the eigenvalues of the stress tensor. That is the reason why more than one stress value coexists at the same position (the middle principal stress is not reported, as it is not critical). The values of stresses Scomp [MPa] (compression values) and Stens [MPa] (tension values) are also taken at the front surface 300a.

[0090] The compression values Scomp [MPa] are the most meaningful numbers, since they reflect the compression of the device (mirror) 100 due to its thermal expansion. Tension values Stens [MPa] are the result of stress concentration at sharp edges and are strongly dependent on the details of the design of the device 100. The tension values represent mirror design evaluation criteria: in a good design, Stens [MPa] should be significantly lower than the compression value Scomp [MPa].

[0091] Figs. 5a - 5c show an alternative configuration of a cooling channel 400c. Here, the cooling channel 400c is located at the bulk portion 200, i.e., external of the bulk portion 200. In other words, the external cooling channel 400c is attached to a surface of the bulk portion 200 which is opposite to the surface of the bulk portion 200 on which the at least one reflecting layer 300a is deposited.

[0092] Fig. 5a shows a front view of the device 100 with the reflecting layer 300a. Fig. 5b shows a side view of the device 100 with the external cooling channels 400c on one side of the dielectric bulk 200 and the reflecting layer 300a on an opposite side. Fig. 5c shows a back view of the device 100 with the external cooling channel 400c having an inlet 402 and an outlet 401 .

[0093] The external cooling channel 400c may be configured in the same way as the internal cooling channels 400a, 400b.

[0094] Having a cooling channel 400c outside of the dielectric bulk 200 facilitates the manufacturing process of the device 100 and allows the use of larger range of materials for the dielectric bulk portion 200.

[0095] Fig. 6 shows another embodiment of the cooling channel 400a, 400b, 400c. Here, the cooling channel has a meander or zig-zag shape. The meander or zig-zag shape of the cooling channel may be used for both the internal cooling channel 400a, 400b and the external cooling channel 400c, and may be configured in the same way as the internal cooling channels 400a, 400b.

[0096] The dielectric bulk portion 200 and the cooling channel 400a, 400b, 400c in all embodiments of the invention are made by molding or by additive manufacturing.

[0097] In all embodiments of the invention, the material used for the bulk portion 200 is selected from the group consisting of Germanium (Ge), Boron nitride (BN), Magnesia (Mg), Silicon, Silicon carbide (SiC), Aluminum nitride (AIN), Aluminum oxide (AI2O3), Beryllium oxide (BeO) and diamond (C), whereas the material used for the at least one reflective layer 300a is selected from the group consisting of Tungsten, Osmium, Platinum, Ruthenium, Tantalum, Iridium, Rhodium, Chromium and Molybdenum.

[0098] The device according to the invention may most advantageously be used as a launching mirror used in an Electron Cyclotron Resonance Heating system.

Claims

Claims1 . Device (100) for reflecting microwave radiation during electron-cyclotron resonance heating (ECRH), the device (100) comprising a bulk portion (200) and at least one reflective layer (300a, 300b, 300c), wherein the bulk portion (200) includes dielectric material and the at least one reflective layer (300a, 300b, 300c) is made of an electrically conductive material which has an electrical conductivity in a range of 0.1 MS / m - 70 MS / m, wherein the thermal expansion coefficient of the bulk portion (200) differs from the thermal expansion coefficient of the at least one reflective layer (300a, 300b, 300c) by an amount in a range of 0 - 20x10-6 / K, and wherein a cooling means (400a, 400b, 400c) is provided at or at least partially within the bulk portion (200).

2. Device according to claim 1 , where the cooling means is an elongate cooling channel (400a, 400b, 400c) having an inlet (402) and an outlet (401 ).

3. Device (100) according to claim 2, wherein the cooling channel (400a, 400b) is provided entirely within the bulk portion (200).

4. Device (100) according to claim 2 or 3, wherein the cooling channel (400a, 400b) spirals radially outwardly.

5. Device (100) according to any one claims 2 to 4, wherein the cooling channel (400a) has a depth and / or width which varies along the length of the cooling channel (400a).

6. Device (100) according to claim 5, wherein the varying depth and / or the varying width of the cooling channel (400a) is defined by a depth / width function, and wherein the depth / width function comprises at least one Gaussian, a sinusoid or a combination thereof.

7. Device (100) according to any one claims 2 to 6, wherein the cooling channel (400a, 400b) is made in one piece with the bulk portion (200).

8. Device (100) according to any one of claims 2 to 7, wherein the bulk portion (200) and the cooling channel (400a, 400b, 400c) are made by molding or by additive manufacturing.

9. Device (100) according to any one of the preceding claims, wherein the bulk portion (200) is made entirely of dielectric material and has an electrical conductivity in a range of 0 - 0.1 MS / m.

10. Device (100) according to any one of the preceding claims, wherein the thickness of the bulk portion (200) is in a range of 1 - 30 mm.11 . Device (100) according to any one of the preceding claims, wherein the at least one reflective layer (300a) includes a plurality of reflective layers (300b-1 , 300b-2, 300c-1 , 300c-2, ..., 300c-N), and wherein the sum of the product of the thickness times the electrical conductivity of each reflective layer over the plurality of reflective layers (300b-1 , 300b-2, 300c-1 , 300c-2, ..., 300c-N) divided by the sum of the total thickness of the plurality of reflective layers (300b-1 , 300b-2, 300c-1 , 300c-2, ..., 300c-N) is in a range of 0.1 - 70 MS / m.

12. Device (100) according to any one of the preceding claims, wherein the at least one reflective layer (300a, 300b, 300c) is arranged to one side of the bulk portion (200).

13. Device (100) according to any one of the preceding claims, wherein the thickness of the at least one reflective layer (300a, 300b, 300c) and / or the thickness of the bulk portion (200) is uniform across the device (100).

14. Device (100) according to any one of the preceding claims, wherein the thickness of the at least one reflective layer (300a, 300b, 300c) is in a range of 0.1 - 100 pm, preferably in a range of 1 .7 - 7.5 pm.

15. Device (100) according to any one of the preceding claims, wherein the at least one reflective layer (300a, 300b, 300c) is deposited via magnetron sputtering on the bulk portion (200).

16. Device (100) according any one of the preceding claims, wherein the material used for the bulk portion (200) is selected from the group consisting of Germanium (Ge), Boron nitride (BN), Magnesia (Mg), Silicon, Silicon carbide (SiC), Aluminum nitride (AIN), Aluminum oxide (AI2O3), Beryllium oxide (BeO) and diamond (C).

17. Device (100) according to any one of the preceding claims, wherein the material used for the at least one reflective layer (300a, 300b, 300c) is selected from the group consisting of Tungsten, Osmium, Platinum, Ruthenium, Tantalum, Iridium, Rhodium, Chromium and Molybdenum.

18. Device (100) according to any one of the preceding claims, wherein the device (100) is a launching mirror used in an Electron Cyclotron Resonance Heating system.

9. Use of the device (100) according to any one of the preceding claims in an Electron Cyclotron Resonance Heating system for reflecting microwave radiation during electroncyclotron resonance heating of fusion plasma.

Images