University of Tasmania

File(s) not publicly available

FDEM simulation of tunnelling‚ÄövÑv´induced rock failure, fracture and collapse and their control in deep underground

posted on 2023-05-28, 00:11 authored by Han, H
In recent decades, more and more underground tunnels have been designed and constructed in worldwide to meet the multiple needs of modern society, such as for transportation, mining, and oil exploitation. For tunnelling in deep underground, the evaluation and prediction of the developments of rock failure, fracture and collapse around the tunnel are of significant importance in ensuring the quality of the tunnels constructed and in preventing potential accidents. However, in some complex practical cases the rock fracture and fragmentation mechanisms under the influences of multiple effective factors on the sites have not been well understood yet, which are worthwhile for further study. Recently, to numerically study the progressive rock fracture process from continuum to discontinuum behaviour, combined finite-discrete element method (FDEM) has been increasingly developed and applied in simulating and analyzing underground engineering problems. In this work, an in-house FDEM computer program parallelized using general-purpose graphic process unit (GPGPU) is further developed with some new features implemented in order to achieve realistic simulations of the whole rock failure, fracture and collapse progressive processes induced by tunnelling in deep underground and their control. Using the GPGPU-parallelized FDEM code, a series of studies based on practical tunnelling cases are conducted to reveal the underlying rock failure, fracture and collapse mechanisms in these cases. In each study, the rock fracture initiation and propagation, as well as the rock block and fragment expulsion, ejection and flyout that could be difficult to capture on the site or via other research methods are vividly reproduced in the FDEM numerical modellings, followed by the investigations of the influences of various factors on rock fracture patterns and damage evolutions. Through these studies, firstly the occurrences and developments of rock fractures and rockbursts around a deep tunnel excavated by tunnel boring machine (TBM) under high in-situ stresses are investigated. It is found that some crucial geological and geotechnical characteristics of the site have significant effects on the rockburst development. For the modelled tunnelling case in which the in-situ major principal stress is along the vertical direction, the pre-existing fault near the tunnel could aggravate the rockburst incident at the tunnel roof and the tunnel side close to the fault, while a lower in-situ lateral pressure coefficient could contribute to the alleviation of the rockburst development around the tunnel. Besides, the tunnel shape plays a significant role in determining the initiation locations and patterns of the stress concentration zones and accordingly the fracture initiation and propagation behaviours that result in rockbursts around the tunnel. The effectiveness and efficiency of the proposed method in the investigation of the rockburst development mechanisms in deep tunnelling are embodied through the study. Then, the rock fracture and fragmentation process and the development of resultant excavation damaged zone (EDZ) during tunnelling by drilling and blasting in deep underground are simulated. The combination of in-situ stress field, equation-of-state based blast loading, and fracturing in tension and shear with gas flow loading of fractures for modelling the complex dynamic interactions from multiple blast rounds is achieved successfully for the first time using FDEM. For the modelled tunnelling case involving multiple contour blast rounds, it is found that in-situ stresses suppress the propagation of long fractures and removing the in-situ stresses results in smoother sidewall fracturing with more damage in the crown and invert. Increasing rock heterogeneity, above a threshold, induces more fractures. Increasing the detonation timing between blast-holes induces more damage into the rock mass and fragmentation in the burden. Moreover, longer decay time ratios and higher decoupling ratios result in additional rock fracture and fragmentation around the tunnel. The proposed method is highly flexible in investigating the influences of various factors on the rock fracture pattern and EDZ development around the tunnel, and its capability to show the realistic rock fracture and fragmentation processes contributes to the better understanding of these influences. Finally, numerical simulations are conducted to investigate the rock dynamic fracture mechanism by destress blasting and its application in controlling the violent fracture of rock during tunnelling in deep underground. The effectiveness of the destress blasting application in preventing potential rockburst in deep tunnelling is further studied, and it is found that the success of the destress blasting application is highly dependent on its design itself such as the arrangement of the blast-holes and the explosive charges. The developed code is innovative and robust in finding the balance between multi-factors and reducing the tremendous efforts to successfully apply the destress blasting in controlling the tunnelling-induced rockbursts in deep underground. Outcomes of these studies show that the in-house GPGPU-parallelized FDEM code provides a powerful tool to realistically investigate complex rock failure, fracture and collapse processes and their control during tunnelling in deep underground, which overcomes the obstacles of the conventional numerical simulation approaches in modelling the progressive rock fracture processes from continuum to discontinuum behaviours in deep tunnelling and therefore fills the research gaps on this topic. The research outcomes in this thesis are also of guiding importance for certain tunnelling designs and constructions in practice, and the proposed method is expected to be used for the evaluation and prediction of the developments of rock failure, fracture and collapse in many other tunnelling scenarios in future studies.


Usage metrics

    Thesis collection


    No categories selected


    Ref. manager