3D reconstruction-based numerical modeling of irregular‑shaped geo‑objects using digital images: a novel approach

摘要

Naturally occurring rock mass have rough surface with irregular geometry and highly unpredictable material properties, making them almost impossible to mathematically represent in computer-aided design (CAD)-based modeling. Simulating the responses of irregular geo-objects under naturally occurring loading conditions is a matter of extensive research. In this context, the contribution of this research is twofold: (i) to propose a 3D reconstruction-based numerical modeling technique that can be used for accurate simulation of irregular geo-objects and (ii) to demonstrate the effectiveness of the proposed modeling approach by testing its accuracy with respect to lab level experiments in simulating strength of intact rocks. Here, image sequences of irregular rocks are used to build 3D reconstructed models that incorporate the objects' exact geometry. These models are then simulated in numerical modeling software to predict the mechanical behavior of those objects. Validation of the proposed approach was done by comparing the laboratory obtained breaking strengths of the rock samples against their numerical modeling received strengths. It has been shown that the proposed methodology can predict the uniaxial and tensile strength of standard rock samples with 95-99% accuracy. The proposed method is more robust, simplistic, and as-accurate as the existing CAD-based numerical modeling approach, making them readily applicable in modeling the mechanical behavior of geo-objects having irregular surface geometry.