Performance Predictions of Hard Rock TBM in Subcritical Cutter Load Conditions.

In mechanized tunneling, the ability to accurately predict the performance of the tunnel boring machine (TBM) is critical, because of its impacts on the project schedule and cost. A substantial amount of research has been conducted on predicting the penetration rate of TBMs in rocks but sometimes the actual cutting process of the disc cutter has been ignored, as most of the available models have been developed based on the assumption that the chipping process is the dominant mode of rock breakage. This might be true in many cases, but there are also situations where tensile fractures do not extend long enough to form chips, instead the tracks get deeper, and fines and rock powder are generated instead of chipping. This situation happens when the cutter load is insufficient to form chips, or so called sub-critical loading occurs. Two of the recent tunneling projects where subcritical cutter load has been the dominant mode are the southern extension of Tehran subway Line 6 (SEL6), and the southern lot of Kerman water conveyance tunnel (KrWCT). The issue in SEL6 was insufficient thrust of the machine and in KrWCT the high strength of the rock. Evaluation of the FPIs in SEL6 shows that the FPIs were less than 100 (kN/cutter/mm/rev) and the prediction errors of the existing models are low. However, in KrWCT project in rocks with higher strength and higher FPIs, the prediction errors are considerable. This indicates that the common models cannot offer an accurate estimate of the performance of TBM in subcritical loading in rocks with high strength. This study attempts to use statistical and machine learning methods to develop a new relation for TBM performance prediction in such conditions based on actual machine operating data and rock engineering geological features. New equations are introduced for the estimation of FPI in subcritical loading conditions to achieve more reliable and accurate prediction of TBM penetration rate. Highlights: Issues of hard rock TBM performance predictions in subcritical cutter load conditions. Development of a new performance prediction model for subcritical loading conditions using regression analysis and AI algorithms. Consideration of geological and geotechnical parameters and machine specification in model development. [ABSTRACT FROM AUTHOR]