ROCK MASS-TUNNEL SUPPORT INTERACTION ANALYSIS

Abstract

The application of the rock mass-tunnel support interaction analysis in designing the tunnel support system is well known. An approach for quick and reliable determination of the ground reaction (response) and the support reaction curves, which are the two essential components of the rock mass-tunnel support interaction analysis, has been proposed. The proposed approach is based on the results of field instrumentation and other related field studies carried out in nine Indian tunnels. Description of the geology and its influence on the tunnelling conditions, as well as the details of the field studies, have been presented for these tunnelling projects located in the lower and middle Himalayas and in the peninsular region of India. As the first step towards obtaining the ground reaction curve, empirical correlations and a design chart have been proposed for the three types of tunnelling conditions, namely, self-supporting, non-squeezing, and squeezing, on the basis of the analysis of data obtained from the Indian tunnels and some of the case-histories reported by Barton et al. (1974). The correlations show that the ground condition depends on the rock mass quality (Q), height of overburden and the tunnel size. The correlations have important practical benefits, especially with regards to the possibility of achieving a favourable ground condition by changing the tunnel alignment to obtain a better rock mass quality, or a reduced overburden, or both. Alternatively, two or three smaller tunnels may be chosen instead of a larger tunnel to avoid squeezing ground condition, thereby IV reducing the support problems and the construction time. After predicting the expected ground condition, the next step is to determine the ground reaction curve for the predicted ground condition. Determination of the ground reaction curves for the self-supporting and the non-squeezing conditions depends on the modulus of deformation of the rock mass which is normally obtained from expensive and time-consuming uniaxial jacking tests, whose results often have a large scatter. An empirical correlation has, therefore, been proposed for prediction of the modulus of deformation of the rock mass. The correlation indicates that the modulus of deformation of the rock mass increases with RMR and the tunnel depth. This depth dependency of the modulus of deformation is likely to be more pronounced in weaker rock masses and almost absent in strong, brittle rock masses. For using the correlation for the modulus of deformation, the RMR value may either be obtained in the field or, if one prefers to use the Q-system, from a correlation proposed between RMR and Qm, where Qm is the modified Q (with SRF equal to 1). This modification has been carried out to overcome the uncertainities in determination of SRF. A semi-empirical correlation has been proposed for obtaining the value of cohesion of the rock mass. The correlation indicates the mobilisation of a much higher cohesion around the underground openings than the values suggested by Bieniawski (1979) on the basis of the field data of rock slopes. The observation of this 'apparant strength enhancement' on the basis of the field instrumentation data, is well supported by a number of laboratory tests conducted by several investigators on thick-walled hollow cylindrical samples. This apparant strength enhancement may be attributed to anisotropy in strength, statistical variation in strength and confining conditions around tunnels. As such, a strength enhancement factor of 4 to 6 is recommended which should be multiplied with the cohesion parameter from block shear tests on rock mass for obtaining the ground reaction curve in the squeezing ground condition. The behaviour of the steel rib-backfill support system has been studied at a number of tunnel sections, in order to propose an approach for determination of the support reaction curve. The study shows that the steel rib-backfill support system exhibits a non-linear behaviour under pressure, unlike the normally assumed linear elastic behaviour, due to the continuously changing backfill stiffness. The backfill, though not the main load carrying element, significantly influences the behaviour of the support system under pressure. The behaviour of three types of backfills, viz, concrete, gravel, and tunnel-muck, under pressure, has been studied. The conctete backfill provides a stiffer support than the other two types of backfills and is, therefore, preferable for the elastic ground condition. The tunnel-muck and the gravel backfills may be more suited to the moderately squeezing and the highly squeezing ground conditions respectively, as the latter is relatively more flexible. With the help of the proposed rock mass-tunnel support interaction analysis, the effect of charging of the water conductor system on the support pressure has been studied. The study has revealed that the additional support pressure on the final support due to the charging of the water conductor system, could be as high as 80 percent of the insitu stress in elastic ground condition. The proposed rock mass - tunnel support interaction analysis further shows that the short-term support pressure is practically independent of the tunnel size if the As/S and t^ values are increased in direct proportion to the tunnel size, where A£ is the cross-section area of the steel rib, S is the rib spacing and th is the backfill thickness. This explains the modern concept of support pressure (Barton et al., 1974 and Singh et al., 1992) based on extensive field observations. Knowledge of the stand-up time of an underground opening helps in determining the time by which the support installation may be delayed and it may, therefore, have a bearing on the selection of the support system. Empirical correlations have been proposed for determination of the stand-up time for underground openings with arch roof and flat roof. The correlations indicate that the stand-up time of an underground opening depends on its span, RMR, and height of overburden, with RMR having the most dominant influence. The influence of the opening size is more pronounced in openings located at deeper depths as compared to the shallow openings. The stand-up time is also influenced by the shape of the underground opening. An opening with an arch roof has a better stand-up time than that Vll with a flat roof for a given value of RMR. This difference, however, decreases with increase in RMR and ceases to exist for RMR > 65. Finally, it should be added that tunnelling is an art and adventure due to several uncertainities in exploration and behaviour of the rock masses, particularly in the Himalaya. The key to the management of the uncertainities lies in monitoring through instrumentation, contingency plans, and team spirit