Excavation Quality Evaluation of Highway Tunnels Using 3D Laser Scanning

Abstract

: Accurate evaluation of tunnel excavation quality is critical for subsequent support work, lining construction, and profile control. While 3D laser scanning captures dense and continuous point cloud data, extracting reliable geometric information from long, complex highway tunnels remains challenging due to local clutter and noise. To address this, a comprehensive evaluation workflow is proposed. Multi-station point clouds are first organized via spherical-target registration and preprocessed. Subsequently, a Polar coordinate Cloth Simulation Filtering (P-CSF) method is utilized for robust lining extraction, followed by cross-section extraction and continuous profile reconstruction. By comparing the reconstructed profiles with design contours, overbreak and underbreak are identified. The excavation quality is then comprehensively evaluated using average overbreak depth, profile roughness, longitudinal variation, and the Tunnel Contour Quality Index (TCI). Application to a real-world highway tunnel project with both standard and complex segments demonstrates that the proposed workflow effectively mitigates the influence of attached construction objects, yields highly stable reconstructed profiles, and provides a reliable geometric basis for segment-level excavation assessment in practical construction scenarios.