Comparison of Ground Movement near Buildings due to Underground Station Excavation with Analytical and Numerical Methods-Paper

Paper

Pdf Summary

The study compares analytical predictions, PLAXIS 2D simulations, and field measurements of ground and building movement during excavation of a 20 m-deep launching shaft for Bangalore Metro. The shaft was supported by secant pile walls and struts. Two nearby reinforced-concrete buildings, located 3 m and 13 m from the excavation, were monitored.<br /><br />Ground movements were estimated using methods by Bowles and Clough and O’Rourke. The numerical model used a Mohr–Coulomb soil model and represented the excavation sequence, retaining walls, and building foundations.<br /><br />PLAXIS 2D predicted a maximum wall deflection of 66 mm, compared with 44 mm measured on site—a 33% overprediction. The maximum settlement predicted analytically was approximately 1% greater than the numerical prediction, while the numerical settlement exceeded the analytical estimate beyond 6 m from the wall. At the buildings, however, measured settlements were substantially smaller than either prediction. Some markers indicated heave, but tilt meters recorded no building rotation.<br /><br />The authors attribute the differences partly to conservative design assumptions and the limitations of a two-dimensional model, which cannot fully capture support from the opposite walls and three-dimensional stress effects. A lower groundwater level during excavation may also have reduced wall movement and settlement. Overall, the monitored excavation proceeded in a controlled manner, and observed building movements remained limited.

Keywords

Bangalore Metro

excavation

launching shaft

PLAXIS 2D

secant pile walls

ground settlement

wall deflection

building movement

Mohr-Coulomb model

field measurements