Experimental research and statistical analysis on the dielectric properties of lunar soil simulators
To support the microwave brightness data retrieval of future China space-borne lunar exploration microwave radiometer, based on the collection of plentiful terrestrial basalts and anorthosites and their chemical compositions got by X-ray fluorescence, nine lunar soil simulators were prepared and made respectively into 0.8, 1.0, 1.2, 1.4 and 1.6 g/cm3 five densities each. We measured their relative dielectric constants over the range of 0.5-20 GHz with open-ended coaxial line model on the HP8722C Network Analyzer and then processed and analyzed the measurement data. This study shows that among the three parameters of density, frequency and composition, density has the strongest effect on the relative dielectric constants, frequency comes second, composition the least. The three parameters account for 45%, 33% and 22% respectively of the changed real part of a relative dielectric constant, and 55%, 27% and 19% respectively of the changed imaginary part. The real parts of the relative dielectric constants are linearly linearly correlated with density or frequency, and the imaginary parts have a linear relation with both approximately over the range of 0.5-10 GHz and tend to be poorly correlated with them in 10-20 GHz. The effect of composition on a relative dielectric constant seems very complicated, both probably do not follow a simple function relation, with the least correlation. Multiple regression analysis indicates that major element oxides SiO2, Al2O3, CaO, MgO, TiO2 and -Fe are correspondent to a one-order polynomial, and TiO2 or -Fe or TiO2+-Fe has not been proven to be the indicators in the contribution to the relative dielectric constants.
作 者： LI Dihui JIANG Jingshan WU Ji ZHANG Dehai ZHANG Xiaohui 作者单位： Center for Space Science and Applied Research, Chinese Academy of Sciences, Beijing 100080, China 刊 名： 科学通报（英文版） SCI 英文刊名： CHINESE SCIENCE BULLETIN 年，卷(期)： 2005 50(10) 分类号： P1 关键词： lunar soil simulator relative dielectric constant density microwave frequency composition regression analysis correlation coefficient