The Geotechnical Design of the Townline Road–Rail Tunnel: Discussion

Abstract

The geotechnical design of the project described by Conlon et al. (1971) has several interesting features. I t embodies a very comprehensive investigation of the shear behavior of the stratified deposits from which block samples were recovered. Case records of slides along the existing canal are available for analysis so that confidence may be gained on the design assumptions. For the k s t time in a major project, the concept of strength anisotropy was used for design, following earlier work (Lo and Milligan 1967). The writer had the opportunity of being involved in the slope stability problems of the project, and would propose to discuss some aspects of the mechanics of stability analysis in stratified clays and progressive failure in the slope design of this interesting project. For both the interpretation of laboratory tests for undrained shear strength and effective stress parameters, c' and +', and for insertion of strength values in computer analysis, it is necessary to relate the strength measured to the angle of inclination, i, of the bedding to the base of the sample. The relationship has been discussed by Lo and Milligan (1967) and Ladanyi (1967), and a more complete solution was obtained by Lo and Vallte (1970) for a sample (or soil stratum) having parallel planes of weakness. The results of the solution relevant to the stratified clay in this project are shown in Fig. 1. In Fig. 1, the strength ratio, defined as the strength measured at any value of i to the minimum strength measured at the critical value of i, is plotted against the angle 2i +',v, where +' , denotes the effective angle of internal friction of the plane of weakness. The curves were computed for the appropriate strength parameters of the stratified clay for different values of the pore pressure coefficient, Af. The c' and +' values are the values corresponding to shearing across the layers and c', and +', corresponding to shearing along the weak layers. The short horizontal lines denote constant strength cutoff corresponding to different mean values of effective stress 0'0 (= 4 -I0'3) at failure. While several conclusions may be drawn from Fig. 1, the pertinent one is that there is little increase in the strength ratio, R, for a variation of i, say 5. It follows, therefore, that: (a ) for laboratory tests it is not essential to know the critical value of i or to orientate the layering in the most critical orientation accurately, since the error involved in the strength measured will be small, and (b) a range of angle from the horizontal may be ascribed to the failure plane in the stratified clay for computer analysis. In fact, a 5 variation from the horizontal was used in the computer program to search for the failure surface. It turns out, however, for the geometry of the stratigraphy and slope, the horizontal surface along the lower boundary of the stratified layer is critical. Right from the early stage of design, the possibility of progressive failure has been of concern. The analyses of the slides along the existing canal, particularly of the successive slides between stations 1098-1 108 have shed considerable light on this problem. It is important to note that about half of the extent of the July 1939 slide, and the whole extent of the October 1939 slide, was embodied in the 1932 slide which occurred at end of construction. Two possibilities therefore arise: (a) The subsequent slides have their sliding surface wholly or partly coincident with that of the first time slide. Any decrease of strength with time may then be attributed to decrease in strength with large strains, or the residual strength is approached. (b) New slide surfaces were formed in each slide. Any decrease in strength may then be attributed to some form of progressive failure. The possibility (a ) may be designed for by having an adequate factor of safety to prevent a first time slide. For possibility (b), the rate of operative strength decrease with time must