Identification and Analysis of Fracture Zones in Tunnels Based on GPR Wave Characteristics

With the growing complexity of mountain tunnel construction and the increasing engineering demand for rapid detection, ground-penetrating radar (GPR) has become a core technique for engineering-scale fracture detection. Fracture zones are not only a major obstacle to tunneling through complex geological sections but also a key trigger of tunnel hazards, posing a serious threat to construction safety. Consequently, the effective identification of fracture zones and the investigation of their development characteristics remain central challenges in advance geological prediction for mountain tunnels. Owing to the complex morphology of fracture zones, existing approaches—including simple model simulation, single-parameter identification, and integrated prediction methods—cannot adequately characterize core attributes such as connectivity and development degree. To address this challenge, this study constructed a fracture attribute model and established a multidimensional collaborative identification system covering fracture scale, connectivity, and density, integrating time-domain wave-frequency morphology with the two instantaneous attributes in the spatial domain, namely instantaneous frequency and instantaneous amplitude. A collaborative analysis scheme based on multiple GPR statistical attributes was adopted to perform a qualitative comparison of fracture development characteristics by tracking multidimensional parameter trends. The results reveal that fracture zones in different development states exhibit certain correlation trends between their internal structural features (e.g., connectivity and compactness) and the frequency-related physical attributes of electromagnetic waves (e.g., wave-frequency morphology and instantaneous frequency). Based on the simulation results, the analysis of waveform characteristics, spatial variations of the two instantaneous attributes, and multi-attribute evolution trends clarified the correlation trends between the wave-frequency response patterns and the fracture development degree. These findings provide scientific and theoretical guidance for geological prediction in tunnel engineering in Hunan, China, and lay a foundation for the GPR-based identification of fracture zones in tunnels.