The “communications blackout” phenomenon has bothered the aerospace industry for several decades. However, the unsatisfying numerical methods and the rapidly changing inflow conditions make the estimation of electrons’ density inaccurate. To lay the first step for model calibration and the anti-blackout design, the Artificial Neural Networks and the direct simulation Monte Carlo (DSMC) method with the quantum-kinetic (Q-K) model are brought together to perform the global sensitivity analysis and uncertainty quantification. The three-dimensional RAM-C II (the second flight of the Radio Attenuation Measurement experiments) head flows are simulated considering aleatory uncertainties (inflow uncertainties) and epistemic uncertainties (reaction parameters uncertainties). Under the inflow condition of an 81 km atmosphere and a velocity of 7.8 km/s, aleatory uncertainties are found to be the dominant type of uncertainty for the number density of electrons, especially the freestream velocity, which means the accurate measurement of the vehicle’s velocity is much more critical than the calibration of the model. The importance ranking is listed, and the “Three rules” for finding the essential reactions are proposed. The probability that the real value is smaller than the nominal value considering both uncertainties is 27%–68%.
Read full abstract