Assessing a soft soil tunnelling numerical model using field data

Abstract

The aim of this paper is to assess a numerical soft soil tunnelling model, using field data. The lining of the tunnel to be studied was a fairly thick, very rigid structure. Given the low level of applied stresses, lining deformation was slight, and the deformation around the tunnel essentially concerned the annular space created between the soil and the lining during tunnelling. This void was partially filled with injected grout. The primary purpose of this paper is to examine models that demonstrate soil–tunnelling interactions. An empirical ‘deconfinement factor λ’ is then used to represent stress reduction around tunnels during construction. In the first model it is simply assumed that soil stresses around the tunnel reduce linearly by a factor λ. The second analysis refines this notion and considers a temporary 'reconfinement' stage in order to represent both the grouting behind the lining once it is installed, and the hydrostatic pressure that acts on the impervious tunnel (pushing the lining up into the grout). Comparisons between these models and field data measurements obtained in three fully instrumented areas of the Lyons-Vaise metro site will be discussed. Cet article a pour but d'évaluer un modèle numérique pour la construction d'un tunnel dans un sol meuble, en utilisant des données de terrain. Le revêtement du tunnel à étudier était une structure assez épaisse et très rigide. Étant donné que les contraintes appliquées étaient faibles, la déformation du revêtement était 1égère et la déformation autour du tunnel était liée essentiellement a l;espace annulaire créé entre le sol et la paroi pendant la construction. Ce vide était partiellement rempli de coulis injecté. Cet article a pour objet, principalement, d'examiner les modèles qui montrent les interactions entre le sol et l;action de creusement du tunnel. Nous utilisons ensuite un ‘facteur de d6confinement λ, obtenu de mamère empirique, pour représenter la baisse de contrainte autour des tunnels pendant la construction. Dans le premier modèle, nous présumons simplement que les contraintes dans le sol autour du tunnel baissent de fa¸on linéaire d'un facteur λ. La seconde analyse raffine cette notion et considère une étape de ‘re-confinement’ temporaire afin de représenter à la fois l'injection de coulis derrière la paroi une fois que celle-ci est installée et la pression hydrostatique qui agit sur le tunnel étanche (déplacement vertical du revetement à l'intéreur du coulis). Nous ommentons la comparaison entre ces modèles et les mesures obtenues sur le terrain dans trois zones entièrement instrumentées sur le site du métro de Lyon-Vaise.

Tunnel modelling: Stress release and constitutive aspects

“…Bernat et al (1999) modelled the TBM excavations of the Lyons-Vaise metro by calibrating the partial stress release factors of an unlined tunnel with the measured tunnel crown displacements. The soil was represented by the CJS model that account for kinematic strain-hardening.…”

Contributions to Géotechnique 1948–2008: Tunnelling

“…Following from these observations, a new procedure for twodimensional FDM numerical analysis was proposed. Bernat et al (1999) presented and used data from the same monitoring sections to examine soil-tunnelling interactions using an empirical 'deconfinement factor', which represents stress reductions around the tunnel during construction, implemented in an FEM analysis. Good comparisons, for both surface and subsurface settlements, were obtained as long as accurate account was taken of the stratigraphy and deconfinement factor.…”

A 3D finite element simulation model for TBM tunnelling in soft ground

“…Respective monitorings for the tunnel crown, the springline and the invert are presented in Figure 18. For a qualitative comparison, Figure 19 shows pore pressure measurement data from two different tunnels [6,24]. When the TBM approaches the monitoring section, the cutting face support generates excess pore pressures in the soil, see Figures 18 and 19.…”

“…The computed stress paths and excess pore pressures in the soil around the tunnel, the predicted soil deformations and the lining behaviour are investigated in detail. The results are compared with in situ measurements taken from the literature [6][7][8][9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24] in order to investigate the capability of the model to reproduce the main characteristics of shield tunnelling.…”