Limit Support Pressure and Critical Cover-to-Diameter Ratio of Shallow-Buried Tunnel Faces in Sand-over- Clay Composite Ground

Technical Summary: Limit Support Pressure and Critical Cover-to-Diameter Ratio of Shallow-Buried Tunnel Faces in Sand-over-Clay Composite Ground

Problem Statement
Shallow-buried tunnels in urban environments increasingly encounter complex composite ground conditions, specifically sand-over-clay stratigraphy. Unlike homogeneous ground, these layered systems exhibit contrasting strength and deformation characteristics that significantly influence tunnel face stability. While existing research has addressed tunnel face stability in homogeneous sandy or clayey soils, and some studies have examined layered ground, there remains a gap in understanding the specific failure mechanisms, earth-pressure release patterns, and the transition from global to local failure modes in sand-over-clay composite ground. Specifically, systematic comparisons of deformation differences between sand and clay layers, the progressive failure process relative to the soil-layer interface position (center vs. crown), and a quantitative criterion for the critical cover-to-diameter ratio ($C/D$) under varying shear strengths are lacking.

Methodology
This study employed a multi-faceted approach combining physical modeling and numerical simulation:

1. Centrifuge Model Tests: Conducted at 60g using a TLJ-3 geotechnical centrifuge. Four test cases were designed with a 6 m prototype diameter tunnel. The soil-layer interface was positioned either at the center of the tunnel face or at the tunnel crown, with cover-to-diameter ratios ($C/D$) of 0.8 and 1.2.
2. Monitoring Techniques: Particle Image Velocimetry (PIV) was used to analyze soil displacement fields and failure zone evolution. Laser displacement sensors measured face retraction, while earth pressure cells monitored stress release at the tunnel face and within the soil ahead of the face.
3. Three-Dimensional Lower-Bound Finite-Element Limit Analysis: A numerical model was developed to determine the minimum support pressure required for stability. This method utilized statically admissible stress fields satisfying equilibrium, boundary conditions, and the Mohr-Coulomb yield criterion. It was used to validate experimental results and perform parametric analysis on the internal friction angle ($\phi$) of the upper sand layer.

Key Contributions and Results

• Failure Mechanisms and Interface Position:

  • When the soil-layer interface intersected the center of the tunnel face, instability was governed primarily by the upper sand layer. The failure zone propagated upward and outward, eventually reaching the ground surface (global failure).
  • When the interface was located at the tunnel crown, the underlying clay layer effectively restricted the upward propagation of the failure zone. In these cases, the failure zone remained confined within approximately 1.5 tunnel diameters ahead of the face, and no continuous failure surface formed even at a face retraction of 30 mm.

• Progressive Failure Stages:
  The face retraction process was identified as occurring in three distinct stages:

  1. Local Deformation: Concentrated near the face.
  2. Progressive Failure: Upward and outward propagation of the displacement zone with continuous earth-pressure release.
  3. Residual Stability: Stabilization of support pressure and formation of a continuous displacement band.
     Cases D1 and D2 reached the limit state at a face retraction of approximately 10 mm.

• Earth-Pressure Release and Support Pressure:

  • Earth-pressure release in the sand layer was significantly greater than in the clay layer. For instance, in Case D1, the stress release ratio reached 66% in the sand layer versus 56% in the clay layer.
  • The limit support pressures for the failure cases were 12.7 kPa (D1, $C/D=0.8$) and 13.4 kPa (D2, $C/D=1.2$). Increasing $C/D$ from 0.8 to 1.2 resulted in only a 5.5% increase in limit support pressure, suggesting limited sensitivity to cover depth within this specific range for this interface configuration.
  • Cases D3 and D4 (interface at crown) did not reach a limit state within the test range, with support pressure ratios remaining above 0.85.

• Numerical Validation:
  The three-dimensional lower-bound limit analysis yielded limit support pressures with deviations of less than 10% compared to centrifuge test results (−9.4% for D1 and −4.5% for D2). The numerical model successfully reproduced the failure characteristics and confirmed that the clay layer provides a stabilizing effect when located below the tunnel face.

• Critical Cover-to-Diameter Ratio:
  Parametric analysis revealed a strong relationship between the internal friction angle ($\phi$) of the upper sand layer and the critical $C/D$ ratio required to transition from global to local failure.

  • As $\phi$ increased from 15° to 35°, the critical $C/D$ decreased from 2.20 to 1.10.
  • A linear regression established a relationship: $(C/D)_{crit} = 3.02 - 0.056\phi$, with a coefficient of determination $R^2 = 0.995$.
  • Higher shear strength in the sand layer enhances the soil-arching effect, allowing the failure zone to close ahead of the face at smaller cover depths.

Significance
The authors state that these findings provide a reference for determining tunnel-face support parameters and assessing the risk of global failure in shallow-buried tunnels constructed in sand-over-clay composite ground. The study highlights that the position of the soil-layer interface is a controlling factor for failure modes and that the transition from global to local failure can be quantitatively predicted based on the internal friction angle and cover depth. The validated lower-bound limit analysis method offers a reliable tool for identifying potential yielded zones and failure modes in similar complex ground conditions.