3D Scanning and Modelling of Regolith Simulant to Characterise Surface Structures and Micro-Shadowing

3D SCANNING AND MODELLING OF REGOLITH SIMULANT TO CHARACTERISE SURFACE STRUCTURES AND MICRO-SHADOWING. M. Amorós Trepat ([email protected])1, J. N. Brecher1, A. Peschel1 and P. Reiss1. 1Technical University of Munich, Ottobrunn, Germany Introduction: The study of volatile elements in the solar system provides valuable insight into its formation and evolution. Airless solar system bodies present an ideal environment to investigate volatiles, since they are directly influenced by the planetary surface without any hydrodynamic processes to disturb them [1]. The Moon is one of the primary bodies for such studies [2], and ever since the discovery of hydrogen in lunar pyroclastic glasses [3], the investigation of lunar volatile water has reached unprecedented significance. Remote sensing results of surficial water near the poles [4] heightened the focus; however, many open questions remain on the origin of the water and other volatiles observed and their concentration on different positions on the Moon’s surface [5]. Previous research on volatile water on the Moon revealed that the local concentration strongly correlates with the surface temperature [6]. The surface structure greatly influences the temperature distribution, both on scales of topography, as seen in the Diviner Lunar Radiometer Experiment [7], as well as on smaller roughness scales, where the complex geometry leads to an irregular surface illumination. This changes non-thermal effects like photonstimulated desorption by introducing a sunward directionality [8] and thermal effects like sorption kinetics by casting micro-shadows and creating additional colder surface areas [6]. The project VOLARIS aims to improve the knowledge of the lunar water cycle by combining laboratory experiments with theoretical simulations on local to global scales. Therefore, we investigate the surface roughness of lunar regolith simulants on a thermally relevant spatial scale, using 3D surface scans to obtain information on the distribution of micro-shadows. The derived information can then be utilized to investigate the effect of surface roughness on the transport and accumulation of volatiles to take the first step towards a better (probabilistic) model of lunar surface temperatures in multiscale investigations. Methodology: To study the formation of microshadows, we investigated the surfaces of samples with 10 cm diameter of the lunar mare regolith simulant JSC-1A. We sieved the simulant for two size fractions, 250–500 µm and 500–1000 µm, to investigate the influence of particle size on the occurrence of micro-shadows. We also took measurements of particles <250 µm but encountered experimental limitations in the resolution with the available equipment. The samples were scanned using a Keyence VL700 optical 3D coordinate measuring system with a precision of ±10 µm and the possibility to increase the resolution by merging scans from multiple viewing angles. Here, twelve different positions were utilised, resulting in an average distance of ~55 µm between the recorded measurement points. The measurements were exported as a triangulation, to obtain an approximation of the surface. Given this triangulation, we computed the shadowed regions created by the roughness of the material. We casted rays simulating the solar insolation to each vertex of the triangles to assess if there was another structure obstructing its sunlight. Due to the small scales studied, the same angle of incidence was assumed for all vertices of the surface. Many raytriangle intersections were needed to compute the casted shadows, as our triangulations contain ~5 million triangles. To make the operation computationally feasible, we sorted the triangles depending on the ray direction to minimize the intersection operations needed. Moreover, parallelization was utilized to reduce the execution times. The result of this process can be seen in Figure 1, where the triangulation is rendered with shadows depicted in black. Figure 1: Rendering of casted shadows over a JSC1A sample with particle sizes of 500–1000 µm, with a solar incidence angle of 60º. Each vertex was then casted to the mean plane, and we performed a Delaunay triangulation of the casted points. Each shadowed vertex contributes to the total shadowed surface by one-third of the area of each triangle that it is part of. Thus, the percentage of the shadowed surface can be approximated by adding up each shadowed vertex’s contribution. Results: The percentage of shadowed surface over different incidence angles for both investigated size distributions are shown in Figure 2. The results for negative and positive angles mirror each other, indicating a homogeneous and representative sample. The larger particle size distribution results in a higher percentage of shadowed area at all incidence angles compared to the smaller one. A non- proportional relation between both curves is observable. For one, this could be due to the two size distributions not spanning over the same total size range, meaning that the difference in the shadowed area might not only be due to the particle sizes. Additionally, both size distributions in the respective fractions of JSC-1A are not perfectly uniform [9]. Considering this, the two fractions might not be directly comparable. Further, smaller particles might tend to agglomerate, creating agglomerates that lead to an increase in surface roughness and, thus, a higher percentage of shadowed area. Future Work: To increase our dataset and overcome experimental limitations, we will perform particle packing simulations in the next step to obtain the surface roughness profile of different size distributions. From this, we will gain further insight into the dependency of the shadow casting on parameters such as the range of the size distribution and the particle size. We then plan to derive a more generalized model for micro-shadows, which will be verified with 3D-scanned measurements. Furthermore, we will compute the mean and RMS slopes of the measured and simulated surfaces, which are key parameters in roughness modelling [6]. The obtained values will be compared to the ones found in the literature, previously obtained from the analysis of remote sensing data (e.g. [10]). References: [1] Hodges, R. et al. (1973) LPSC Proceedings 4, 2855. [2] Lucey, P. G. et al. (2022) Geochemistry 82, 3. [3] Saal, A. E. et al. (2008) Nature 454, 7201. [4] Reach, W. T. et al. (2023) PSJ 4, 3. [5] Reiss, P. (2024) PNAS. [6] Prem, P. et al. (2018) Icarus 299, 31-45. [7] Williams, J.-P. et al. (2017) Icarus 283, 300–325. [8] Sarantos, M. and Tsavachidis, S. (2023) Zenodo Dataset, 10.5281/ZENODO.7761817. [9] Sakatani, N. et al. (2018) Icarus 309, 13-24. [10] Bandfield, J.L. et al. (2015) Icarus 248, 357–372. Figure 2: Percentage of shadowed surface of two sieved JSC-1A size fractions, with different incidence angles. For each incidence angle, the shadowed area cast under eight different azimuth angles was calculated and averaged. The azimuth angles were distributed evenly over 360°. The lines and their bands represent the mean and standard deviation, respectively.