including surface roughness is 0.0145 dB/m. These calculations are based on a dc conductivity of mhos/m at 20°C temperature. The temperature variation of conductivity for aluminum is given by: If the broad wall under the thermal radiator (no insulation) is 220 C, the 7 conductivity of that wall decreases to 1. 93 x 10 mhos/m. The top wall attenuation of 0. 00095 dB/m increases to 0. 00127 dB/m and the total attenuation is 0. 0148 dB/m (0. 451 dB/100'). In the past, waveguide attenuation calculations have been based on de conductivity due to the lack of rf conductivity measurements at the frequencies of interest. The rf conductivity at 2.45 GHz may be slightly lower than the value used here. For example, rf conductivity of aluminum waveguide at 24 GHz is much lower, i. e. , mhos/m. A value of is probably typical for MPTS aluminum waveguide at 2.45 GHz. If the side walls are also at an elevated temperature, the attenuation would be greater. 6.7 MECHANICAL DESIGN AND ANALYSIS 6.7.1 THERMAL ANALYSIS AND CONFIGURATION Preliminary thermal analyses were conducted on the amplitron and klystron converter candidates mounted on the waveguide in the maximum packing configuration as appropriate for the central arrays of a high power system. These were done to establish radiator sizes for the converters and to determine thermal deflection characteristics and maximum temperatures of the waveguide. To obtain a reasonable temperature gradient along the waveguides, three different waveguide thicknesses were evaluated initially. These thicknesses were 0.01 inch (0.0254 cm), 0.02 inch (0. 0508 cm) and 0.03 inch (0. 0762 cm). Gradients and deflection decrease as wall thickness increases. Based upon considerations of temperature, weight and structural strength, the 0. 02 inch
RkJQdWJsaXNoZXIy MTU5NjU0Mg==