Major civil engineering structures such as dams, bridges, tall buildings, towers, etc. are generally founded on a rock mass if intact or competent rook is located at a reasonable depth below the ground level. However, the ooncept of founding a structure on rock whenever the soil above the rook has inadequate engineering characteristics, end without regard for the nature, type end behaviour of rock, has become quite outdated (Kulhawy, 1980 and Goodman, 1980). This is because any rock mass generally contains discontinuities whose deformation behaviour under load is quite different from that of intact rock.
For important structures such as mentioned above, the settlement of foundations becomes a governing factor in design since excessive total and differential settlements endanger the safety or utility of the structure. Thus the ultimate bearing capacity at which the foundation rook will fail has less engineering significance (Farmer, 1978; Chappe11 end Maurice, 1980). The settlement comprises mainly of two parts. The first part is due to the deformability of intact rook material while the second part is due to the compressibility of discontinuities'. A major contribution to the settlement canes from the closure of discontinuities since discontinuities are usually more oompressib1e than intact rook. It i8 not on4r the presence of discontinuities that makes the prediction of foundation behaviour more difficult but also the random nature of their location, strength and deformation properties. The random variability of the deformation properties of intact rock is also important. If a design does not incorporate this complexity and variability it may prove either unsafe orover-conservative. In this context it is appropriate to quote Fairhurst (1976):"..rock does have characteristics of variability end, to a degree, unpredictability of variation, that we neglect at our peril".
Present work is concerned with incorporating the random variability in the aforesaid parameters to improve the predietion of settlement and Monte Carlo simulation method is adopted.
Several methods have been in use for the estimation of settlement of foundation on rook, a majority of which use the principles of elastioity either directly or indirectly (Timoshenko and Goodier, 1951; Hobbs, 1975; Chappell and Maurice, 1975; Farmer,1978; Hungr and Coates, 1978; etc.). Some methods estimate the foundation settlement by direct extrapolation from field load tests (Waldorf et.a1., 1963; Hobbs, 1975) or by empirical correlations between the modulus reduction factor and RQD(Peck et a1., 1974, Kulhawy 1978,etc.). However it is noticed that the current methods for the estimation of settlement do not incorporate the random variability in the geometrical and deformation properties of discontinuities and that in the deformation properties of intact rock.
Adopting second moment probability methods the authors in their earlier work (1980) have proposed two models to study the influence of random variability: in each parameter. In the first model the discontinuous rock mass is idealized as a system with equivalent continuum properties (Kulhawy,1978). The inherent random variability in the spac1rJgand the stiffness of horizontal discontinuities and that in the Young's modulus of rock material are considered.
