Numerical investigations on the deposition characteristics of lunar dust in the human bronchial airways

Abstract

Although it is widely acknowledged that lunar dust (LD) is toxic to the human health, its deposition characteristics in the bronchial airways remain unknown, which is significantly important to understand its toxicity. Therefore, this study employs computational fluid dynamics and machine learning algorithm methods to address this issue considering the difficulty of conducting the experiments of LD deposition. The major results are: (1) the deposition efficiencies (DE) of micrometer-sized LD in the terminal bronchioles vary significantly depending on the human body posture, with a notable difference of DE up to 29% between standing and lying flat postures; (2) LD deposition in various bronchial regions shows differences under activity intensities, with higher DE in segmental bronchi and terminal bronchioles under intense and lower intensive activities, respectively; (3) In predicting DE of LD, machine learning algorithms outperform fitting functions, achieving higher precision and smaller errors, reducing the root mean square error by approximately 60%–80%. These results indicate that LD deposition characteristics in the bronchial airways under lunar environment are also influenced by the combined factors of particle size, activity intensity, and body posture.