Reconstruction of “Crystal Caves” geometry from 3D scan data for engineering and geological applications - from point cloud to numerical simulation

Abstract

In geological engineering, 3D scanners are increasingly used to analyze complex structures. They are essential for developing detailed 3D models of irregular geological formations, as well as for creating terrain relief models, which are used in numerical simulations. Such analyses can be used to predict changes in caves and assess geological hazards such as deformation and stress in rocks. Common algorithms often struggle to capture complex structures, requiring manual adjustments that can reduce accuracy. The article focuses on reconstructing the geometry of the “Crystal Caves” in the Wieliczka Salt Mine for CFD/FEM analysis. The object is characterized by unusual complexity and irregularity of the surface, covered with transparent halite crystals. An original, author-developed procedure is presented, highlighting key aspects to consider during 3D scanning, point cloud development, and surface reconstruction. This will help to avoid costly and time-consuming errors in the reconstruction process. Crystal Caves geometry was reconstructed from a point cloud, confirming the effectiveness of the proposed procedure. Described approach ensures efficient meshing while preserving geometric features, making the model suitable for use in both CFD and FEM simulations. Despite the growing number of publications in 3D scanning and numerical simulation applications, there is a lack of studies presenting the complete procedure from a point cloud to a flow model. The novelty of this study lies in the development of a complete and reproducible workflow from raw point cloud data to geometry ready for CFD/FEM simulations, specifically tailored for environments with highly complex surface morphology.