6585845000

211

South Africa

Fig. 8. Evaluation of in Dutch static cone penetration test.

(10)

ult

Flats areas, however, indications are that the angle of shearing resistance of the fine to medium sands determined in this way is sometimes rather less than would be expected. Because of the difficulties of taking undisturbed samples of fine, virtually cohesionless, sands from below ground-water level, laboratory shear strength tests have been carried out on representative recompacted materials and resulting values of 4>, over a rangę of densities, tend to confirm this. The values of <f>' determined from the resistance to penetration of the cone are, ne-vertheless, often used to assess the al-lowable bearing capacity of shallow spread foundations. For this purpose it is com-mon practice to employ the Terzaghi bearing capacity factors which, in cohesionless soils, are functions of <J>*. For a sguare footing, B x B, for example, the ultimate bearing capacity is given by the semi-empirical expression

^qult ⁼ Y(DfN_q + 0,4BN_y)    (8)

where N_q and Ny are dimensionless bearing capacity factors, y is the bulk density of the soil and Df is the depth of founding. Numerical values of the bearing capacity factors are given by Terzaghi (1943).

A factor of 3 is usually employed in arri-ving at the allowaole soil bearing pres-sure, q_a. When there is a variation in strength of the soil beneath a foundation it is necessary to base the value of q_a on the weakest soil. Altematively, if a weak soil occurs as a buried stratum its bearing capacity is assessed by cal-culating the pressure induced on its sur-face using the Boston Codę method or the Kógler method described by Taylor (1948).

Although the theory advanced by De Beer enables values of c* and to be determined, Sanglerat (1972) points out that the results are not exact because of the necessity of assuming a value of <J>. Also, in the stratified estuarine sediments va-riations in the clay often preclude pene-trometer testing at two significantly different levels as is necessary for the determination of c'. For these reasons the shear strength parameters of clays and clayey sands are usually obtained from the results of laboratory triaxial compression tests, or box shear tests, on tubę samples of materiał recovered from boreholes.

A preliminary estimate of the undrained shear strength c_u can be derived from available correlations with q_c. In both the Durban and Richards Bay areas it is found that the ratio q_c/c_u increases with increasing value of q_c, average values being

q_c/c_u = 12    (9A)

when q = 1500 kN/m² and

C

q_c/c_u = 37    (9B)

when q_c = 5000 kN/m²

The ratio appears to increase roughly linearly with increase in q_c. There is a scatter of about 30 percent in the ra-tios of q_c/c_ur which were obtained for clays varying from normally Consolidated to lightly overconsolidated in the depth rangę 5 to 18 m.

The ultimate bearing capacity of shallow spread footings in the fine to medium grained sands of the Durban and Cape Flats areas is sometimes examined using the formula, developed by De Beer,

Po^Vo ⁺ V^B in which p_Q is the effective overburden pressure at the same level as the bottom of the footing, y is the bulk density, B is the least width of the footing, and V_Q and V_g are dimensionless bearing capacity factors which depend on <$>' as shown in Table 3, and which are analogous to the Terzaghi bearing capacity factors N_q and

Ny.

Allowable bearing capacity is obtained by applying a factor ranging from 3,0 for strip footings to 2,5 for sguare footings. Equation (10) applies to soils which, in terms of the basie Prandtl equations from which they are derived, undergo negli-gible volume change during loading. A thorough knowledge of the soils to be stressed by the foundation is therefore required before these equations can be employed with confidence. The rela-tionship between cone resistance and applies only in cohesionless soils below ground-water level.