Ground response to twin-tunnellingS

Ground response to twin-tunnellingSubsurface settlement over twin tunnelsThe subsurface settlement is measured by the markers positioned in the clay at various elevations as illustrated in Fig. 5. Generally, the settlement over T1 (z1) are reasonably symmetrical and can be well fitted by Gaussian functions as listed in Table1. For instance, the maximum settlement at dt = 2.5D is about 0.02D. It is pronounced that the trough narrows and the maximum settlement increases with the closer distance to the tunnel, which is in line with the expected behaviour (Mair et al., 1993).Fig. 6 presents the subsurface settlement caused by T2 construction (z2) after T1construction completed. It is noticeable that the settlement profiles are asymmetrical to the centerline of T2 and move toward the side of T1 as a whole. Similarly, thetroughs can be expressed as Gaussian function with additional parameter xc as: where z = settlement at position x; zmax = maximum settlement of the trough; xc =horizontal eccentricity between the center of the trough and the tunnel axis; and i = trough width parameter. The parameters of the troughs are listed in Table 1. It is evident that the maximum settlement zmax caused by T2 is higher than T1, while the trough width parameter i does not change with an obvious tendency.By the fitted Gaussian curves listed in Table 1, as depicted in Fig. 7, the additional subsurface ground settlement za (za = z2 - z1) above T2 due to disturbance by T1 construction can be determined. It indicates that the positive additional settlement (i.e., settlement above T2 is increased by disturbance due to T1) occurs at the side close to T1, while negative at the other side. The maximumadditional settlements at all elevations occurs in the close proximity to the closer springline of T1. The occurrence of negative additional settlement may be unexpectedto the traditional knowledge of many researchers, which is measured in bothcentrifuge modelling (Divall and Goodey, 2015) and real projects (Chapman et al.,2004). The reason for this phenomenon is interpreted subsequently by the shearstiffness change.