Investigation of vibrations created during TBM excavation and rock cutting

Abstract

Tunnel Boring Machines (TBMs) are widely employed in infrastructure projects due to their high advance rates and providing a safe working environment. The TBMs should be operated in accordance with the geological conditions to achieve optimal excavation rates. The majority of contractors, especially the ones working in remote locations, predominantly depend on the geological information provided during the tender phase or undertake a limited number of supplementary geotechnical investigations prior to starting projects. Especially in remote locations that characterized by a limited geological data, the expertise of the crew becomes highly significant. Although the most effective way of assessing geological conditions during tunnel excavation involves examining the excavation face, contractors often hesitate to halt excavation for face inspections unless deemed necessary or forced by contractual obligations. Furthermore, the application of closed-face soft ground TBMs often renders face inspections impractical. Although alternative techniques such as probe drilling or geophysical methods could be applied in the TBMs, these approaches require considerable amount of time, with some geophysical methods still undergoing development. Given the reluctance of contractors to allocate resources towards new geological studies and the necessity for continuous geological information it becomes evident that a system capable of providing real-time information to the TBM crew regarding the encountered geological conditions is required. Since the TBMs vibrate during excavation, utilizing vibrations to understand the geological conditions ahead seems practical. In this context, field measurements of the vibrations generated during excavation by TBMs have been investigated for ground identification in various studies. However, TBMs include too many influential factors and noise sources that impact the vibrations (e.g., multiple cutters on the cutterhead, motors and other parts in shield and backup area, etc.), requiring a deeper understanding of vibrations specifically arising from the rock cutting process within a more controlled environment. Even some authors focused laboratory analyses and investigated the vibrations generated during rock cutting tests, these experiments remain fairly limited. The principal objective of this study is to understand the vibrations generated during the TBM excavation and provide a system that could give information to the site crew regarding the geological conditions on the excavation face. To reach this goal, a special vibration data recording system has been developed and installed onto four Earth Pressure Balance Tunnel Boring Machines (EPB TBMs), excavating in three different project sites; that are Mecidiyeköy-Mahmutbey Metro Line, Umraniye-Atasehir-Goztepe Metro Line and European Region Potable Water Transmission Tunnel. The employed vibration recording system comprises accelerometers, analog-digital converters, and a dedicated computer running custom-designed data recording software tailored for the purpose of this project. Initial deployment and testing of the system took place at the Mecidiyekoy-Mahmutbey Metro Line project site. Following this initial implementation, the system underwent further refinement, with the addition of new features and hardware enhancements. A total of 6 km tunnel excavation was monitored, and vibration data continuously recorded in three orthogonal directions (x, y, and z). The influence of geological conditions including soils and hard rocks, EPB TBM operating parameters (thrust, torque, face pressure, cutterhead rotational speed, penetration rate), cutter damages/breakages, positions and types of accelerometers, and TBM diameters on vibration were investigated. During vibration analysis, primarily Root Mean Square (RMS) and interval RMS vibrations, together with different vibration statistics were used. A comprehensive analysis was conducted on the collected TBM vibration data, incorporating geological information, site shift reports, and disc changing logs. This analysis revealed a significant influence of geological and lithological variations on the vibration patterns generated during excavation. In particular, moderately strong to strong correlations were observed between interval RMS acceleration values and rock mass parameters such as Rock Quality Designation (RQD) and Geological Strength Index (GSI). The analysis yielded a clear trend, an increase in rock strength corresponds to rise in TBM vibrations, while vibrations greatly reduced in the case of weathered rocks and soils. The presence of short weathered zones was distinguishable in the vibration data. This characteristic suggests that the measurement has the potential to serve as an early warning system for detecting geological transitions during TBM excavation. Furthermore, an acceleration threshold value of 1g, representing the percentage of data points surpassing this limit, is established as an effective marker for distinguishing between different lithological units. It is observed that the vibrations recorded in TBMs could exceed 10g. The analysis also revealed an influence by the operational parameters of TBMs and conditions of the cutters on the vibrations. Notably, it was observed that broken disc cutters produce distinct high-amplitude vibrations. Considering the noise sources in TBMs, it is acknowledged that there is a need for a better understanding of vibrations created during the rock cutting process under more controlled environment. With this objective, laboratory rock cutting tests were conducted using a Full-Scale Linear Rock Cutting Machine (FLCM) equipped a 17-inch constant cross section (CCS) disc cutter. To record vibrations, FLCM was equipped with five accelerometers. Three different rock samples were used (siltstone, mudstone and beige marble) having uniaxial compressive strength values ranging from 21 to 82 MPa. During the experiments, it was discovered that the beige marble sample exhibited two distinct mineralogical characteristics on its left and right sides.