Recently, NATM (New Austrian Tunnelling Method) is becoming very familiar to Japanese civil engineers, because of its reasonable theory, in which ground is considered as a part of tunnel structure and strengthen actively using rock bolts and shotcrete. However, it is not easy to have a good tunnelling site condition in Japan and a lot of tunnelling with various difficulties has been tried in soft rocks and soils. Especially, soft rocks consisting of the Tertiary period mud stone make tunnelling by NATM difficult and the followings pose engineers many problems:
Difficulties on clarifying the mechanism of NATM.
Difficulties on evaluating the mechanical properties of soft rocks.
Under these circumstances, Finite Element Method attracts engineers because it is the one and only method which can analyze the mechanism of NATM and the complex mechanical properties of ground, appropreately. A lot of studies have been already done to establish the rational design method of NATM in Soft rocks. However, many problems still remain, such as;
Mechanical model of soft rocks
Evaluation of material constants
Estimation of -the 3-dimensional effect· of tunnel face progress
During tunnelling in soft rocks, field measurements shown in Fig. 2–1 were carried out. The tunnel was excavated by short-bench method. Examples of measured results are shown in Fig. 2–2 and 2–3. Fig. 2–3 shows the change of ground displacement measured by extensometer and Fig. 2- 2 shows the change of tunnel deformation measured by convergence meter. In these figures, horizontal axis indicates the distance between tunnel face and the measurement section, and it is normalized with the maximum tunnel width D.
From these results, it is found that the noticeable change of displacement occurs when the tunnel face approaches in the distance of -l.OD from the measurement section. When the tunnel face passes the measurement section and is 2.0D away from it, the noticeable change converges and a small but stable change remains. The tendency of these ground displacement may be caused by following factors; namely, the effect of tunnel face progress and the non-linear property of ground. In the subsequent sections, these factors are examined by comparing measurement results with the results of elastic finite element analysis.
The convergence of a circular tunnel, excavated in a 3-dimensional elastic continuum, is shown in Fig. 2–4 where the convergence is normalized with maximum convergence Umax which is not under the influence of tunnel face existence. As shown in Fig. 2–4, the influence of face is limitted in the section between -l.OD and 2.0D from tunnel face. In order to compare with Fig. 2–4, measurement results of convergence Cl and tunnel wall displacement by extensometer E2' E4 are normalized in the section between -l.OD and 2.0D. Result is shown in Fig. 2-S.
