Abstract

Direct SPT-based pile design methods are very popular these days despite the fact that many of such methods are based on small databases of pile load tests. Due to the dependence of soil behaviour on geological setting and site specific conditions, it is possible that some of the methods do not produce good prediction of pile capacity. Accordingly this paper presents the evaluation of two SPT-based pile design methods in residual soils against a pile load test database from the Southern African region. The methods include the (i) Franki-SA method reported in Byrne et al. (1995) and (ii) Decourt Method (1995). The pile load tests consist of 26 cases of bored piles in residual soil with each case accompanied by SPT measurements. The SPT measurements were used to calculate the predicted capacity in accordance with the procedure for each of the two methods while the pile load tests were used to determine the measured capacity. The findings of the evaluation indicate that the Decourt method is more reliable and accurate than the SA method. The poor performance of the SA methods suggests further studies to develop specific calculation factors for base and shaft capacities in residual soils.