ABSTRACT
Application of three different response surfaces and SORM method to the estimation of the reliability index associated with exceeding a certain allowable settlement level by shallow foundation is presented in this paper. Two random variables: the Young modulus and Poisson’s ratio, of lognormal and beta distributed respectively, in a single soil layer are taken into account. The linearly deformable model of soil is assumed which is most frequently used in engineering practice when the limit state of serviceability is considered. The main problem encountered using the parabolic response surface was existence of false design points that prevented coordinate calculation of the real ones. Two procedures were developed. The first one added a barrier to force the calculation algorithm to move out of the false design point region. The later consisted of widening of the area covered by the response surface (polynomial of the second degree) with an additional “oedometric” term. Inserting of the oedometric term improves quality of the fitting and enables extending the range of approximation. But it did not prevent the SORM algorithm to calculate the false design point in all cases analysed. The third response surface used is the one based on the neural network. Covering whole ranges of independent variables and utilising all data points, only the neural network response surfaced did not let a false design point to emerge.
