Experimental simulation investigation on rockburst induced by spalling failure in deep circular tunnels

https://doi.org/10.1016/j.tust.2018.07.035Get rights and content

Abstract

Spalling and rockburst are two common failure modes in deep hard-rock tunnels and they exhibit a strong correlation. In this paper, the process of rockburst induced by spalling is investigated. The uniaxial compressive strength and rockburst tendency of red sandstone are measured. Four different initial stress conditions are set up, and cubic red sandstone samples with a prefabricated hole are tested by a true-triaxial test system. During the experimental process, the failure of the hole sidewall is monitored and recorded in real time by a micro-camera. The process of rockburst induced by spalling damage in deep hard-rock tunnels is reproduced, and the mechanism by which spalling damage on induces rockburst is revealed. In addition, the evolution process and failure characteristics of rockburst induced by spalling damage are analysed. The experimental results indicated that the red sandstone has a moderate rockburst tendency, and its rockburst process can be divided into four periods: the calm period, the small grain ejection & spalling damage period, the slab buckling & fragment ejection period and the violent ejection period. The mechanism of spalling damage on inducing rockburst is mainly embodied in two aspects: promoting large buckling deformations (providing energy for rockburst) and to weakening the strength of the rock mass (creating conditions for the suddenly release of energy). The effect of lateral stress on spalling damage and rockburst is more obvious than that of axial stress, and the severity of rockburst can be significantly reduced by increasing the lateral stress. The diameter of the hole has a strength size effect on the sidewall damage, producing a certain inhibitory effect on the spalling damage and rockburst. Smaller lateral stress corresponds with greater depths of the V-shaped notch and much smaller width-to-length ratios in the rock fragments.

Introduction

With the development of deep underground civil and defence projects, rock engineering is gradually being applied in progressively deeper environments, such as mining, nuclear waste disposal, traffic tunnels, and underground plants and laboratories. After entering the deep underground environment, high-stress-induced failures appear in the hard and brittle rock mass under the impact of excavation disturbances. Such failures are very rare in shallow rock mass (Martin, 1997). Rock is destroyed when the excavation-induced stress reaches or exceeds its strength, and the high-stress-induced failures can occur in either stable or unstable form. There are two main modes of spalling damage and rockburst (Diederichs, 2007, Du et al., 2016), as shown in Fig. 1. Spalling is a progressive failure process that is prevalent in deep hard rock engineering and affects the long-term stability of an engineering structure. Rockburst refers to the sudden and severe damage that occurs frequently in high stress hard rock tunnels, and often causes a number of casualties, equipment damage, and construction delays (Tang et al., 2010, Zhou et al., 2015, Zhang et al., 2012a, Kaiser et al., 1996). Rockburst poses new challenges for the safe construction of deep rock engineering.

Spalling is closely related to the propagation and coalescence of internal tension cracks in the surrounding rock. Under the effect of tangential stress and the free surface, the surrounding rock in the tunnel is usually cut into thin rock slabs that are approximately paralleled to the excavation surface (Fairhurst and Cook, 1966, Germanovich et al., 1994, Cai, 2008). These thin slabs are buckled to form spalling, and ultimately form a V-shaped notch on the sidewall as the damage develops (Ortlepp and Stacey, 1994, Ortlepp, 1997). A case study of hard-rock mine pillar failure in Canada shows that the dominant failure mode is progressive spalling and flaking when the width-to-height ratio of the pillar is less than 2.5 (Martin and Maybee, 2000). Numerical simulations indicate that only the shear failure in the areas of high compression located on the periphery results in spalling either sides of the opening (Zhu et al., 2005). Laboratory experiments and theoretical analysis show that the intermediate principal stress has a significant influence on the crack propagation, and then affects the spalling of the rock around the periphery of the openings (Sahouryeh et al., 2002, Cai, 2008, Zhang et al., 2011). In addition, the fracture morphology and fracture mechanism of spalling are also affected by the radius of curvature of the tunnel section, which is manifested in two aspects: the scale effect and the structure effect (Zhou et al., 2016). The self-developed parallel RFPA3D code was used to study the failure mechanisms and fracture morphology of laboratory-scale rectangular prism rock samples under uniaxial compression (Li et al., 2015). This indicated that spalling failure was affected by the homogeneity of the rock samples and could easily occurred in the relatively homogeneous rock samples. Due to the influence of various factors, the stress threshold for spalling damage is much lower than the uniaxial compressive strength (UCS), and that of the granite is generally 0.35–0.45 times of the UCS (Diederichs et al., 2004).

However, a large number of studies have shown that there is a strong correlation between spalling and rockburst (Martin et al., 1999, Wu et al., 2010, Zhang et al., 2012a, Zhou et al., 2015, Du et al., 2016). Spalling is a precursor to rockburst and can occur in a stable form or as a violent rockburst (Martin et al., 1999). That is, spalling is closely related to the triggering of rockburst (Du et al., 2016), and the failure mode of the surrounding rock can be changed from superficial spalling to violent rockburst as the depth increases (Dowding and Andersson, 1986, Zhao et al., 2014, Mazaira and Konicek, 2015). When the rock slabs formed by spalling are broken and suddenly separated from the sidewall, rockburst occurs (Dyskin and Germanovich, 1993). Therefore, there is a significant correlation and essential connection between spalling and rockburst (Zhou et al., 2015). For example, spalling will occur when the ratio of the maximum tangential stress (σθmax) to the UCS (σc) reaches 0.35; weak to medium rockburst appears at σθmax/σc = 0.5, and a strong rockburst generally occurs when σθmax/σc > 1 (Dowding and Andersson, 1986).

Spalling can be violent or non-violent, and in some cases, can be a time-related slow failure process. It can occur before the rockburst, forming an unstable parallel thin rock slab that provides the conditions for the sudden release of energy characteristic of rockburst (Diederichs, 2007). Based on in situ investigations of the tunnels at Jinping II hydropower station, the mechanism of tensile spalling rockburst in deep-buried intact marble was analysed qualitatively (Hou et al., 2011). Combined with the typical failure phenomenon of spalling and the cases of rockburst in deep tunnels at Jinping II hydropower station, uniaxial compression tests were carried out using similar materials to analyse the mechanism of rockburst induced by spalling damage (Zhou et al., 2015). The spalling damage is thought to be caused by excavation unloading, and the action of the concentrated tangential stress will lead to a buckling deformation of slabs towards the excavation space as well as increasing the strain energy stored in the slabs. When the accumulated energy exceeds its energy storage limitation, or in the event of external disturbances, rockburst will occur with the characteristics of slabs being crushed and fragments ejected. Using the self-designed deep rockburst process test system, He et al., (2007) verified that the rockburst process could be divided into vertical spalling, vertical slab buckling deformation, and rockburst damage. The evolution and formation mechanism of buckling rockburst in deep tunnels was revealed by the spalling simulations of true-triaxial loading-unloading tests and analyses of rockburst cases (Qiu et al., 2014). Based on a coupled static-dynamic analysis, the mechanism of rockburst induced by dynamic disturbances around the deep underground opening was studied by RFPA dynamics (Zhu et al., 2010). It is deemed that dynamic disturbances are considered to have a very obvious triggering effect on rockburst, and the dynamic spalling of the surrounding rock could be one mechanism of rockburst.

The above results greatly enriched the knowledge of spalling damage and rockburst induced by high-stress environments, and this is of great significance in understanding spalling damage, the mechanism of rockburst, and the internal relationship between them. However, as the characteristics of rock and the stress environment vary according to external factors, the mechanism of spalling damage and rockburst is extremely complex, and there is still no clear understanding of their formation mechanisms and intrinsic relationship. As a precursory phenomenon of rockburst, spalling contains important information about rockburst (Zhou et al., 2015). Revealing the mechanism by which spallling damage induces rockburst is of great practical significance, as it may allow the accurate interpretation of precursory information contained in spalling damage and the accurate prediction of rockburst. In addition, the geostress conditions are a very important factor in controlling the stress distribution and destruction area around underground tunnels, and have an important influence on the spalling damage and rockburst of the surrounding rock (Zhu et al., 2010).

Therefore, in this work, the UCS and rockburst tendency of red sandstone are first examined, and then a true-triaxial test system is used to test red sandstone samples with a prefabricated circular hole under four different stress conditions. The damage to the hole is monitored and recorded in real time during the test process. The process of spalling damage and rockburst in deep-buried hard-rock tunnels is effectively reproduced by experimental simulations, and the mechanism whereby spalling damage induces rockburst is revealed. In addition, the process and failure characteristics of rockburst induced by spalling damage are analysed in depth and summarised. The research results provide a reference for the design of deep hard-rock tunnels and the prevention and control of spalling damage and rockburst.

Section snippets

Rock description and specimen preparation

In this paper, red sandstone was taken from Linyi city, Shandong province, China. A photomicrograph of a thin section of the rock, observed under plane polarized light and cross polarized light, is presented to examine the mineral composition and texture of the rock, as shown in Fig. 2. A quantitative mineralogical composition analysis showed that the rock is mainly comprised of approximately 42% quartz, 35% plagioclase, 9% calcite, 8% zeolite, 5% K-feldspar, and 1% opaque minerals. The red

Uniaxial compression test results

Fig. 8 shows the stress-strain curve of the red sandstone cylinder specimens under the uniaxial compression test. It can be seen that the stress-strain curves of the three specimens are generally similar. For example, the ultimate strength and ultimate strain of samples S1 and S3 are close, but there is a certain deviation between the two stress-strain curves. The ultimate strength and ultimate strain of samples S2 and S3 are different, but the relative deviation between the stress-strain

Failure stress threshold of sidewall

It is assumed that the in situ stresses of the rock mass before excavation are p, q, and w, respectively, where p is the vertical stress, and q, w are the horizontal stresses. Assuming that the tunnel axis is arranged along the w direction, the surrounding rock before failure can be regarded as a homogeneous, continuous, and isotropic elastic body that does not deform along the axial direction of the tunnel (i.e., regarded as a plane strain problem), as shown in Fig. 16.

For two-dimensional

Conclusions

Red sandstone has a moderate tendency for rockburst, and the evolution of rockburst induced by spalling damage was reproduced under the true-triaxial loading condition. The evolution process can be divided into four periods: the calm period, the small grain ejection & spalling damage period, the slab buckling & fragment ejection period, and the violent ejection period. This process includes micro-crack propagation, micro-crack coalescence (spalling damage), slab buckling deformation, rock

Acknowledgements

This work has been supported by the National Natural Science Foundation of China (Grant No. 41472269), the State Key Research Development Program of China (Grant No. 2016YFC0600706), and the Fundamental Research Funds for the Central Universities of Central South University (Grant No. 2017zzts167)

References (44)

  • T.R. Stacey

    A simple extension strain criterion for fracture of brittle rock

    Int. J. Rock Mech. Min. Sci.Geomech. Abstr.

    (1981)
  • C.A. Tang et al.

    Preliminary engineering application of microseismic monitoring technique to rockburst prediction in tunneling of Jinping-project

    J. Rock Mech. Geotech. Eng.

    (2010)
  • C.Q. Zhang et al.

    Estimation of in situ stress along deep tunnels buried in complex geological conditions

    Int. J. Rock Mech. Min. Sci.

    (2012)
  • W.C. Zhu et al.

    Simulation of progressive fracturing processes around underground excavations under biaxial compression

    Tunn. Undergr. Space Technol.

    (2005)
  • W.C. Zhu et al.

    Numerical simulation on rockburst of underground opening triggered by dynamic disturbance

    Tunn. Undergr. Space Technol.

    (2010)
  • B.J. Carter

    Size and stress gradients effects on fracture around cavities

    Rock Mech. Rock Eng.

    (1992)
  • M.S. Diederichs

    The 2003 canadian geotechnical colloquium: mechanistic interpretation and practical application of damage and spalling prediction criteria for deep tunnelling

    Can. Geotech. J.

    (2007)
  • K. Du et al.

    Experimental study of slabbing and rockburst induced by true-triaxial unloading and local dynamic disturbance

    Rock Mech. Rock Eng.

    (2016)
  • A.V. Dyskin et al.

    Model of Rockburst Caused by Cracks Growing Near Free Surface. Rockbursts and Seismicity in Mines

    (1993)
  • Fairhurst, C., Cook, N.G.W., 1966. The phenomenon of rock splitting parallel to the direction of maximum compression in...
  • L.N. Germanovich et al.

    Mechanisms of brittle fracture of rock with pre-existing cracks in compression

    Pure Appl. Geophys.

    (1994)
  • B.C. Haimson et al.

    Borehole breakouts and in situ stress

  • Cited by (236)

    • Mechanical behaviors of backfill-rock composites: Physical shear test and back-analysis

      2024, Journal of Rock Mechanics and Geotechnical Engineering
    View all citing articles on Scopus
    View full text