Estimating the Seismic Behavior of Tunnel inside Sand Lens by the Extreme Machine Learning

Document Type : Research Article

Authors

1 Ph.D. Student, Department of Civil Engineering, Qazvin Branch, Islamic Azad University, Qazvin, Iran

2 Associate Professor, Department of Civil Engineering, Qazvin Branch, Islamic Azad University, Qazvin, Iran

3 . Assistant Professor, Department of Civil Engineering, Qazvin Branch, Islamic Azad University, Qazvin, Iran

Abstract

The presence of liquefaction in the soil has a great impact on the condition of the structures built in that area; therefore, studying and evaluating liquefaction is of great importance. Liquefaction of the sand lens can cause changes in the force and shape of the tunnel lining. These issues have been evaluated and investigated in this article. Asymmetric deformations and distortions in the tunnel lining lead to an increase in the ratio of dynamic torque to static torque and distance to diameter. The presence of this phenomenon causes changes in the state of effective stresses, pore water pressure, and settlement. In this thesis, the effect of the sand lens around the tunnel is investigated. Soil-structure interaction is also considered between the tunnel and the sand lens. To put it more simply, a sand lens is a piece of soil in the environment that has high liquefaction properties, so its investigation and evaluation is an important matter that has not been extensively studied so far. Estimation of peripheral and structural soil parameters in tunnel lining always requires software simulation and bulky and time-consuming studies. Providing a method to be able to present these parameters with appropriate accuracy and small computational effort in the fastest possible time has always been an engineering challenge. Therefore, the present study aims to present a machine learning-based method to predict some important properties such as liquefaction event, maximum bending stress of tunnel cover, settlement of subsurface tunnel, and pore water pressure under near- and far-field earthquakes. Hence, first, the three-dimensional finite-difference software with parameters such as soil-structure interaction between tunnel cover and sand lens has been used to simulate the tunnel cover model exposed to ground stimuli. Mohr-Coulomb and Finn models have also been used to consider clay sediment and sand lens liquefaction evaluation, respectively. Then, an extreme learning machine (ELM) is used to predict and estimate the quantities mentioned. The main purpose of this study is to introduce a new method using extreme learning machine to predict some important characteristics such as liquefaction event, maximum bending stress, settlement, and pore water pressure. The results of the studies indicate the proper performance and accuracy of the proposed method in estimating the mentioned parameters so that in the worst case, the estimation error was less than 6%. Also in this study, the effect of a liquefiable sand lens in a non-fluidic environment with different seismic waves and the results of the high influence of bending moment in the tunnel lining, effective stress, pore water pressure and the settlement along the axis of the tunnel in the presence of sand lens has been evaluated. The results demonstrate the great influence of the presence of the sand lens in bending moment parameters in the tunnel lining, effective stress, pore water pressure, and settlement along the tunnel axis. In the final part of the study, all the results obtained from the software are compared with the machine learning outcomes. Also, in the presence of a sand lens, the ratio of bending moment to the state without of sand lens in some cases is over 50%, which is a very significant value, and the maximum settlement occurred in places close to the tunnel axis.

Keywords

References

Andreotti, G., Calvi, GM., Soga, K., Gong, C., & Ding, W. (2020). Cyclic model with damage assessment of longitudinal joints in segmental tunnel linings. Tunnelling Underground Space Technology, 103, 103472.
Azadi, M., & Mir Mohammad Hosseini, S.M. (2010). The uplifting behavior of shallow tunnels within the liquefiable soils under cyclic loadings. Tunnelling Underground Space Technology, 25, 158-67.
Azadi, M., & Bryson, LS. (2018). Effect of width variation of liquefiable sand lens on surface settlement due to shallow tunneling. In: International Congress of Exhibition Sustainable Civil Infrastructure: Innovative Infrastructure Geotechnology, 155-163.
Beheshti, K. (1998). The Investigation of the Behavior of Saturated Sand Lenses within the Soil Deposits under Dynamic Loading. Dissertation, Faculty of Amirkabir University of Technology.
Cetin, K.O., Cakir, E., Ilgac, M., Can, G., Soylemez, B., Elsaid, A., & et al. (2021). Geotechnical aspects     of reconnaissance findings after 2020 January 24th, M6.8 Sivrice–Elazig–Turkey earthquake. Bulletin of Earthquake Engineering, 1-45.
Fattah, M.Y., Hamoo, M.J. & Dawood, S.H. (2015). Dynamic response of a lined tunnel with transmitting boundaries. Earthquakes and Structures, 8(1), 275-304.
Holchin, J., & Vallejo, L. (1995). The Liquefaction of Sand Lenses Due to Cyclic Loading.
Huang, G.B., Zhu, Q.Y., & Siew, C.K. (2004). Extreme learning machine: a new learning scheme of feedforward neural networks. In: IEEE International Joint Conference on Neural Networks (IEEE Cat. No. 04CH37541), 2, 985-990.
Jafarnia, M., & Varzaghani, M.I. (2016). Effect of near field earthquake on the monuments adjacent to underground tunnels using hybrid FEA-ANN technique. Earthquakes and Structures, 10(4), 757-768.
Kim, S., Tom, TH., Takeda, M., & Mase, H. (2021). A framework for transformation to nearshore wave from global wave data using machine-learning techniques: Validation at the Port of Hitachinaka, Japan. Ocean Engineering, 221, 108516.
Lin, SS., Shen, SL., Zhou, A., & Xu, YS. (2021). Risk assessment and management of excavation system based on fuzzy set theory and machine learning methods. Automation in Construction, 122, 103490.
Marcelino, P., de Lurdes Antunes, M., Fortunato, E., & Gomes, MC. (2021). Machine learning approach for pavement performance prediction. International Journal of Pavement Engineering, 22(3), 341-354.
Miranda, L., Caldeira, L., Serra, J., & Gomes, RC. (2020). Dynamic behaviour of Tagus River sand including liquefaction. Bulletin of Earthquake Engineering, 18(10), 4581-4604.
Mir Mohammad Hosseini, SM., & Azadi, M. (2012). Effect of the location of liquefiable sand lenses on shallow tunnels during earthquake loading. Arabian Journal of Science and Engineering, 37, 575-586.
Nguyen, H., Vu, T., Vo, TP., & Thai, HT. (2021). Efficient machine learning models for prediction          of concrete strengths. Construction and Building Materials, 266, 120950.
Pashangpishe, Y. (2004). Mechanism of Soil Deformation due to Double Lenses Liquefaction and Critical Depth Determination. Dissertation, Amirkabir University of Technology.
Seyrfar, A., Ataei, H., Movahedi, A., & Derrible, S. (2021). Data-driven approach for evaluating the energy efficiency in multifamily residential buildings. Practice Periodical on Structural Design and Construction, 26(2), 04020074.
Shokri, M. (1996). Evaluation of the Liquefaction Potential of Sand Lenses. Dissertation, Amirkabir University of Technology.
Taylor, E.J., & Madabhushi, S.P.G. (2020). Remediation of liquefaction-induced floatation of    non-circular tunnels. Tunnelling Underground Space Technology, 98,103301.
Tsinidis, G., de Silva, F., Anastasopoulos, I., Bilotta, E., Bobet, A., Hashash, Y.M.A., & et al. (2020). Seismic behaviour of tunnels: From experiments to analysis. Tunnelling Underground Space Technol, 99, 103334.
Vallejo, LE. (1998). Liquefaction of Sand Lenses During an Earthquake. Earthquake Engineering and Soil Dynamics II—Recent Advances in Ground-Motion Evaluation. Geotech. Spec. Publ., 493-507.
Zhao, K., Wang, Q., Wu, Q., Chen, S., Zhuang, H., & Chen, G. (2020). Stability of immersed tunnel in liquefiable seabed under wave loadings. Tunnelling Underground Space Technol, 102, 103449.

Files

History

  • Receive Date: 21 July 2022
  • Revise Date: 18 November 2022
  • Accept Date: 05 December 2022

Share

How to cite

Statistics