Assessment of damage to an underground box tunnel by a surface explosion

https://doi.org/10.1016/j.tust.2017.04.001Get rights and content

Highlights

  • Numerical peak pressures of soil showed a good agreement with TM5-855-1 at a depth of 4 m with differences ranged between 9–14%, while pressure equations were derived for depths 6 and 8 m.

  • Tunnel lining thickness 250 mm appeared low damage levels at mentioned depths under surface explosion of 227 and 454 kg TNT charge weight.

  • Tunnel lining thickness 750 mm at depth 8 m was capable to resist 1,814 and 4,536 kg TNT charge weight.

Abstract

Recently, external terrorist activities have become one of the most influential events on tunnel structure safety because of the absence of proper mechanisms to detect these events in time to take preventive action. The present study used ANSYS/LS-DYNA software to investigate the damage behaviour of an underground box frame tunnel caused by a surface explosion of a sedan, van, small delivery truck (SDT), and container carrying 227, 454, 1814, and 4536 kg, respectively, of TNT charge weight. The Arbitrary Lagrangian Eulerian (ALE) technique was used to simulate and monitor the propagation of the blast pressure waves into the soil. The validation results indicated that the pressure waves propagated into the soil as hemispherical waves, and the peak pressure values closely matched the predicted values of the technical design manual TM5-855-1, except for large distances. Therefore, an equation was derived to calculate the values of the peak pressure at large distances for each explosion case. Intensive parametric studies were conducted to evaluate the interaction between the explosive charge weight, the tunnel lining thickness and the burial depth, which has a significant effect on tunnel safety. The assessment of the damage levels using the single degree of freedom (SDOF) approach proved that the tunnel experienced little damage when the explosive charge is a sedan or van with a lining thickness of 250, 500 or 750 mm at burial depths of 4, 6, or 8 m. However, tunnel collapse occurred when the lining thickness was 250 mm, and the tunnel was subjected to an explosion of an SDT or container at all investigated depths, as well as the case for a lining thickness of 500 mm at a depth of 4 m for the container explosion. The tunnel lining with a thickness of 750 mm appeared to be highly resistant to the explosion of an SDT or container for all the investigated depths, and the best resistance was achieved at a depth of 8 m, which should be considered by designers to ensure the safety of an underground box tunnel when subjected to an incredible surface explosion.

Introduction

Throughout the years, underground tunnels have offered a quick and cost effective alternative to address transportation requirements in many countries. Terrorist attacks, such as the bombing of the Moscow Metro in 2004, London Subway in 2005, and Belarus in 2011, highlight that these structures should be carefully designed to withstand such events. The main method terrorists used to implement these attacks is using a vehicle bomb because of its enormous charge power, high success rate and serious demolition (Kong et al., 2007).

Tunnel structures can be exposed to internal or external explosions. The internal explosions are less likely to occur because it is hard to get an explosive material inside a tunnel because of modern security and control systems inside subway tunnels that can easily detect such explosive devices. Nevertheless, scholars have studied the dynamic response of underground tunnels subjected to internal explosions (Feldgun et al., 2008, Feldgun et al., 2014, Jiang and Zhou, 2012, Papanikolaou and Kappos, 2014). On the other hand, external blasts are more likely to occur and cause more destruction because of the difficulties in monitoring and preventing such events before they occur. In this research field, there are no databases on full-scale field experiments to investigate the tunnel dynamic response for a surface blast because such an experiment is extremely risky, costly and unachievable in civilian research because of the requirement of a large quantity of explosives. However, some researchers (Kutter et al., 1988, Davies, 1991, Davies, 1994, Davies and Williams, 1992) used scaled-down centrifuge modelling techniques to determine the dynamic responses of tunnels. They observed that centrifuge tests are a valid method to model the effects of an explosion in soil. De (2012) and De et al. (2013) carried out several centrifuge tests to describe the influence of a surface detonation on a cylindrical structure buried in dry sand. These experiments provided quantitative data for the effect of surface explosions on buried tunnels.

Recently, numerical techniques have been used, and they represent a good alternative to provide valuable information in a timely and cost-effective manner. Wang and Lu, 2003, Wang et al., 2004 created a numerical three-phase soil model that is able to simulate an explosion wave propagation into soils. Lu et al. (2005) and Wang et al. (2005) utilized this model to study the response of buried structures subjected to underground explosions.

Luo et al. (2007) analysed the dynamic response of a circular underground tunnel to a surface explosions of 100 and 300 kg of TNT charges in sandy soils. The numerical results indicated that the top part and the centre at the bottom of the subway section are the most damaged zones. The tunnel was safe when 100 kg of TNT explodes at a height of 1.5 m. Yang et al. (2010) estimated the dynamic behaviour of a circular metro tunnel exposed to a surface detonation using the Arbitrary Lagrangian-Eulerian (ALE) method in which the Eulerian meshes were applied to the air, soil and TNT while the Lagrangian mesh was applied to the tunnel. The results showed that the upper part of the tunnel lining is more vulnerable compared to other parts, and the tunnel with a lining thickness of 350 mm is safe for depths greater than 7 m for surface explosions containing less than 500 kg of TNT.

Tiwari et al. (2014) performed a coupled Eulerian-Lagrangian (CEL) analysis using Abaqus software on circular underground tunnels subjected to an internal blast. The outcomes demonstrated that damage in a tunnel decreased as the thickness of the tunnel lining increased; the displacement of the 550 mm thick lining was 90% less than the displacement of the 350 mm lining. In addition, a positive relationship was observed between the charge weight and tunnel lining displacement, in which a 78% higher displacement is observed for a 50 kg explosive compared to a 25 kg explosive. Koneshwaran (2014) studied the effects of the tunnel depth, ground distance and various explosive weights of TNT on the dynamic response of a circular tunnel. The results showed that a tunnel is safe when placed at a depth of 6.35 m and exposed to a detonated charge weight less than 625 kg, whereas a tunnel at a depth of 9.52 m was safe from an explosion of 1125 kg.

Mobaraki and Vaghefi (2015) assessed the dynamic responses of several tunnel shapes buried at depths of 3.5, 7, 10.5 and 14 m subjected to a surface detonation of 1000 kg of TNT surrounded by a sandy soil. The results indicated that the Kobe box-shaped subway with a wall and roof thickness of 70 and 80 cm, respectively, was more resistant than circular and horseshoe shaped tunnels. Furthermore, the study revealed that the distances 0–2 m from the blast centre were more vulnerable compared to the distances 2–25 m from the blast centre.

In previous studies, there are few studies that focus on the dynamic behaviour of a box frame underground tunnel subjected to a surface explosion; therefore, the present work is performed with the finite element software ANSYS/LS-DYNA using the ALE technique to study the dynamic response of an underground box tunnel with a lining thickness of 250,500, and 750 mm at depths of 4, 6, and 8 m caused by a surface explosion of a sedan, van, small delivery truck (SDT), and container carrying 227, 454, 1814, and 4536 kg, respectively, of TNT. To ensure the accuracy of the analysis, the numerical results have been compared with the analytical formulas of the Technical Manual (TM5-855-1) (Station, 1986).

Section snippets

Geometry and meshing

The dynamic response of a box shaped underground railway tunnel constructed in Spain with the dimensions of 7.5 × 10 m and earth cover of 4 m has been evaluated (Perea et al., 2008). Because of the symmetry about the YZ and YX planes, a quarter of the field, which is 25 m × 25 m × 30 m is modelled based on the Alekseenko test (Henrych et al., 1979). However, the dimensions of 10 m × 10 m × 22.5 m showed an identical behaviour as the Alekseenko test field; therefore, these dimensions were utilized in the current

Validation of model results

Four verification models were created without a tunnel structure to monitor the propagation of shock waves inside a soil during a surface explosion of a sedan, van, SDT, and container at different times, as shown in Fig. 2a. The results indicated that the blast shock waves were transmitted into the soil as hemispherical waves and caused a crater in the ground that clearly matches the observed phenomenon during the explosion in the Ryongchon railway station in North Korea (2004), as shown in

Conclusions

The results of this study can be summarized as follows:

  • Validation of four numerical models without a tunnel structure indicated that the blast pressure waves induced by various vehicles explosion propagated into the soil as hemispherical waves, and the peak pressures in each explosion case showed good agreement with the analytical formula of the manual TM5-855-1 at a depth of 4 m with differences ranged between 9 and 14%, however, large differences were observed at depths 6 and 8 m, ranged

Acknowledgments

The authors would like to acknowledge the financial support from Dana Impak Perdana (grant number DIP-2014-019) and the Fundamental Research Grant Scheme (grant number FRGS/1/2015/TK01/UKM/02/4) to perform this research.

References (37)

  • V.K. Papanikolaou et al.

    Practical nonlinear analysis of unreinforced concrete tunnel linings

    Tunn. Undergr. Space Technol.

    (2014)
  • C. Perea

    Design of reinforced concrete bridge frames by heuristic optimization

    Adv. Eng. Softw.

    (2008)
  • Z. Wang et al.

    Numerical analysis on dynamic deformation mechanism of soils under blast loading

    Soil Dyn. Earthq. Eng.

    (2003)
  • Z. Wang

    A full coupled numerical analysis approach for buried structures subjected to subsurface blast

    Comput. Struct.

    (2005)
  • L. Xie

    Damage evolution mechanisms of rock in deep tunnels induced by cut blasting

    Tunn. Undergr. Space Technol.

    (2016)
  • Y. Yang et al.

    Numerical simulation of dynamic response of operating metro tunnel induced by ground explosion

    J. Rock Mech. Geotech. Eng.

    (2010)
  • A. AUTODYN

    Interactive Non-Linear Dynamic Analysis Software, Version 12, User’s Manual

    (2009)
  • A.C. Bailey et al.

    A soil compaction model for cylindrical stress states

    Trans. ASAE

    (1989)
  • Cited by (100)

    View all citing articles on Scopus
    View full text