Analysis of the behavior of the two-component grout around a tunnel segmental lining on the basis of experimental results and analytical approaches

Highlights

• Innovative laboratory tests were performed on a two-component grout.

• Stiffness and strength parameters were evaluated;

• An analysis of the filling material on the behavior of segmental lining was carried out;

• The analysis results were applied to a real case of a tunnel.

Abstract

The mechanized excavation of tunnels generally requires the injection of a grout material between the segmental lining and the ground. This material has an important role for the support system and governs the segmental lining loading, also because of the significant evolution of its mechanical characteristics over time. In this work, an accurate experimentation on the two-component material was conducted. Normally, hand mixing is performed in laboratory to test the mechanical performance of the two-component mix-design. In this paper an innovative instrumentation able to mix appropriately the components was used in order to obtain samples truly representative of the real site conditions. Laboratory compression tests performed on these samples allowed to accurately quantify the evolution of the strength and stiffness of the filling material during the setting time. In order to properly take into account the evolution over the time of the mechanical behaviour of the filling material and of the support system more generally, a new procedure was developed and illustrated in the work. The analysis procedure is based on the convergence-confinement method and is able to describe in some detail the reaction line of the support system during its loading. Due to the presence of the filling material, which improves its mechanical characteristics over time, the reaction line assumes a curved shape. The intersection with the tunnel convergence-confinement curve allows to determine the radial loads acting on the support system and to design the support system to ensure the stability of the tunnel. The innovative procedure was then applied to a real case of a tunnel excavated in Northern Italy, confirming the validity of the main project data adopted for segmental lining and filling material. A sensitivity analysis on the same case allowed to detect how some aspects of the support system loading have an influence, not only on the final radial load acting on the segmental lining, but also on the final stress state induced inside the filling material.

Introduction

In the mechanized excavation of tunnels with Tunnel Boring Machines (TBM), the filling of the annular gap, about 6 to 37 cm [41] but with typical values ranging between 13 and 20 cm [43], [2] is particularly important. The gap is created as the excavation progresses, between the TBM excavation profile and the extrados of the precast concrete segments (tunnel linings). The void is filled at the end of the TBM shield by injecting cement mixtures, which are sometimes referred to as annulus grout or backfill (Fig. 1).

The practical aim for the use of the annulus grout is to provide a uniform contact between the lining and the surrounding soil in order to avoid settlements at the ground surface, to ensure the homogeneous transmission of stress between the soil/rock mass and the lining, to avoid misalignments of the lining segments and to provide impermeabilization [7], [34]. The grout must be able to meet certain workability and mechanical parameters. For instance, the two-component grout should be water-tight, be pumpable, be workable, able to fill the void, not to shrink, to stiff quickly and to be wash-out resistant (e.g. [24]). Two-component grouts, consisting of a component A (cement, bentonite, water and a retarding agent) and a component B (accelerating additive), are increasing in popularity, because they combine an almost immediate development of the strength with the pumpability necessary for conveying the fluid from the mixing site (usually outside the tunnel) to the excavation face, where the injection takes place (e.g. [7], [24]). The typical mix-design in a m³ of two-component grout is very variable and depends strongly on the project specification but in general it consists of cement (280–450 kg), bentonite (30–60 kg), water (730–860 kg), retarder (3–5 kg) and accelerator, normally sodium silicate (60–80 kg). Further addition of other materials (e.g. filler) could be considered to achieve specific design requirements or to replace the bentonite with chemical viscosizer polymers. The accelerator (B component) is generally added just before the pumping phase of the mix of water, bentonite, retarder and cement (A component). In comparison with the mortar type grouts, the simultaneous backfilling with two-component grouts, keeps in general lower settlements during TBM excavation [16] and normally the lining pressures, few rings behind the TBM, do not change significantly in the long-term [14].

Several authors developed research activities on two-component grouts with the aim of defining standards that can allow the comparison of data from different laboratories and construction sites, [18], [44], [7]. Moreover, since it is well-known that the evolution over the time of the mechanical properties of the two-component grout changes based on the mix-design, relevant efforts the optimization of the mix design have been made in recent years [13], [45]. However, the literature on the behavior of the two-component grouts, both from experimental and numerical point of view, is scarce (e.g. [36], [40], [22], [45]). Besides, tail void grouting, as it cannot be directly observed and it is rarely possible to have samples and measurements coming directly from the tunnel construction and the real geometry, is often difficult to be simulated [3]. In literature, ground settlement in tunneling caused by ground loss (i.e. the difference between actual and theoretical excavation volume) considers also tail loss which occurs along the annular void between ground and concrete segmental lining (as a result of shrinkage or compression of backfill grout material).The gap model proposed by Lee et al. [19] is based on simple elastic equations for the squeezing of tunnel face and the contraction of the excavated cavity and it has some limitations and uncertainties for practical use [32]. The use of numerical modeling in tunneling, both with two-dimensional and three-dimensional methods [17] requires the definition of the characteristics of all the materials that are used as support structures or rock reinforcement interventions [35], [8], [9], [10], [11], [12]. Specifically, the tail void grouting is often simulated through a radial pressure distribution [6], [49] or through continuum elements [21], whose mechanical properties can be progressively modified according to the hardening of the grout [4], [5]. Shah et al. [40], Ochmański et al. [22] and more recently Ochmański et al. [23] performed a numerical analysis regarding the effects of the two-component grout on the tunnel settlement. Shah et al. [40] used a newly introduced hardening soil (HS) constitutive model for the Torino soil case and a linear elastic constitutive model for TBM lining segments. Ochmański et al. [22] and more recently Ochmański et al. (2020) modelled the mechanical response of the hardening grout using visco-plastic constitutive relations defined by Meschke et al. [20].

The role of the material considering the deformability and resistance values that characterize it during the loading phase of the segmental lining tunnel are not fully investigated. In the tunneling field, non-numerical solutions are already extensively used to successfully study phenomena of interaction between support or reinforcement systems and the ground (soil or rock mass) present around the tunnel [28]. In particular, the convergence-confinement method is used, considering both segmental lining and filling material in the reaction line of the support system.

In this work, laboratory tests able to accurately characterize the time evolution of the mechanical properties of the two-component material used as a filling material around the segmental linings are presented. A sophisticated equipment is used to effectively mix the grout components during the formation of the specimens. From the tests carried out, a curve able to describe the trend of the strength and stiffness of the material during setting – two key parameters for representing the behavior of the filling material – is obtained. Subsequently, a new analysis procedure based on the convergence-confinement method, which enables to evaluate the behavior of the filling material during the loading of the support system, is presented. This procedure allows to describe in some detail the interaction between the support system and the tunnel, taking into account the evolution of the grout mechanical properties over time, without resorting to more complex and sophisticated numerical methods. The procedure is also able to provide the final load acting on the support system and the stress state induced within the filling material, allowing careful design of all the components used in the construction of the support system. Finally, the application to a case study and a sensitivity analysis allowed to confirm the validity of the proposed procedure and to quantify the influence of some key parameters on the interaction process.