Analytical and numerical analysis of ground movement above a tunnel under seismic loading

Abstract

Numerous methods have beenused to construct a tunnel, Earth Pressure Balance Tunnel Boring Machine (EPB – TBM) method has been found to be more suitable than others. EPB – TBM has to balance earth and water pressure with the tunnel face pressure. This research deals with the movement of tunnel using EPB-TBM in homogenous sandy layer only. Ground conditions have beenmodelled in PLAXIS 3D software to estimate results for static (long term)and seismic loading. The site under consideration contains two layers of soil such as upper clay layer and lower sand layer respectively. Thickness of clay and sand layers are 3.5m and 24m respectively. Lateral and longitudinal dimensions of finite element models are 30m and 80m. Movement of tunnel depends on many parameters. Major parameters are relative depth (An/Dn) and tunnel length (yn). Movement of tunnel is affected by surface settlements. Strain induced volume lossis an important parameter for tunnel movement. In this research, five depths of tunnel crown (A1, A2, A3, A4 and A5), three diameters (D1, D2 and D3) and three lengths (y1, y2 and y3) variations have been considered to predict ground movement of tunnel under seismic and static loading. Many researches havealready been doneconsidering plain strain conditions. Previously, empirical and analytical formulas are developed based on plain strain (2D) assumption. But that method is not fully able to estimate the settlement correctly. The present researcherhas modifiedthese empirical and analytical formulas to estimate settlements. Empirical and analytical formulas results are compared with each other.Longitudinal, lateral and vertical surface settlements have varied with the variation of relative depths, diameters and phased construction modes of tunnel. Seismic loading has been appliedin the model in free field condition and duration of seismic shaking used is five seconds.

PLAXIS 3D software has been used for numerical modelling in the present research. Relevant model of existing literature has been validated using PLAXIS 3D. Maximum vertical surface settlement ofliterature data was 8.53mm. Present researcher reanalysedthe model of literature by PLAXIS 3D and obtain vertical surface settlement of8.54mm. Also, author modified empirical and analytical formulas have been validated by using previous researchers formulae based on case study related issue. Results of settlements obtained from present study and case study are close.Results present two different types of settlements such as surface and total settlement (bottom of tunnel). Settlements of three directions varied with the variations of relative depths of tunnel. Settlements values of static loading are higher than the seismic loading because of short duration ofseismic shaking. Minimum values of vertical, lateral and longitudinal surface settlements are 5mm, 2.5mm and 3mm respectively at A5/D1 location for static loading based on numerical analysis.Maximum value of major principle stress is 160 KN/m2 at A5/D1 location based on static loading. Minimum value of acceleration is 0.2 m/s2 at relative depth A5 and diameter D1 for dynamic time 0.48 seconds.Maximum vertical surface settlement is1.2mm at 0.6s atA5/D1 location during seismic loading. To obtain more accuracy of results, larger model sizes and more data are required. The finding from this research offer significant new information about settlements of segmental bored tunnel such as EPB – TBM and which as guide for future implementation in geotechnical applications.