Page 242 - Fundamentals of Magnetic Thermonuclear Reactor Design
P. 242

First Wall Components  Chapter | 7    223


             ments for the cooling channels are strength and leak-tightness. We remember
             that in ITER, the coolant’s pressure limit is 4 MPa in normal conditions and
             6 MPa in test conditions. Channels with ∼2-mm walls seem to be suitable.
                For a flat FW with round-cross-section cooling channels, the maximum joint
             temperature is determined by the interchannel space and distance from a channel
             centre to the joint point equidistant from neighbouring channels. These distances
             must be as small as possible. To achieve this, it would be good to have cooling
             channels with a rectangular cross section. So, minimum joint temperature T , de-
                                                                       min
             pending on heat load, is in the 100°C–400°C window for different FW components.
                The material and thickness of the armour tiles are selected using the follow-
             ing criteria. Surface maximum temperature, Т max , is limited by the flux of parti-
             cles it emits, with evaporation not necessarily being a critical factor. For example,
             for graphite armour, which is prone to self-sputtering and chemical erosion, the
             maximum surface temperature must be within ∼1500°С (although it could well
             be increased to 2000°C if the mechanical strength considerations prevailed). If
             Be and W tiles are used, Т max  can be estimated based on requirements such as
             resistivity to thermal erosion and recrystallisation leading to loss of strength, as
             well as some technological factors. A series of thermal tests have proved that the
             acceptable temperature is 1000°С for Be tiles and 2500°С for W tiles.
                After selecting a T max  and ignoring the negligibly small temperature difference
             in the interfacial joint layer, we determine the greatest admissible tile thickness:
                                                  (
                                  h ~( T max  -T min ) ×  λ T )/  q                                                    h ∼(Tmax-Tmin)×λ(T)/q,
                                   1
                                                                                                                        1
             where λ(T) is the tile’s average thermal conductivity in the T max  – T min  range.
                Table 7.2 summarises the results of a thermal–physical modelling of the
             ITER components subject to the most severe heat fluxes, that is, the start-up
                              2
                                                                      2
             limiter (q = 8 MW/m ) and the divertor’s vertical target (q = 20 MW/m ).

               TABLE 7.2 Characteristics of Different Armour Materials
               Characteristic                    Material
                                                 Be        CFC       W
               Max allowable surface temperature (T max  (°С))  1000  1500  2500
               Average thermal conductivity in the T max –T min    110  180  110
               range (W/m·K)
               Min temperature of armour contacting with a heat sink panel (°C)
               q = 8 MW/m 2                      300       300       300
               q = 20 MW/m 2                     450       450       450
               Max allowable tile thickness (mm)
               q = 8 MW/m 2                      9.6       27        30
               q = 20 MW/m 2                     3         9.45      9.6
   237   238   239   240   241   242   243   244   245   246   247