Investigation of the factors affecting the design of piled raft foundations using heuristic methods

In this thesis, first of all, a literature review on piled raft foundations was made. Then, by adopting the Randolph mathematical model of piled raft foundations, the maximum settlement values and the size effect were investigated over the basic system properties changed with the written calculation code. In the analysis, the data of a project whose real-size construction was completed were used. In the analyses, the spacing between the piles, the number and placement of the piles and the uniform load on the raft were changed and the settlement amounts were examined. The maximum settlement values obtained as a result of the analysis are defined graphically. In the second part of the thesis, optimization codes were written for two different algorithms in line with the determined objective function and limiting functions, and optimum dimensions were determined in the determined design space. Soil and building parameters of 3 different real-size projects, whose construction was completed, were used in the analyses. The algorithms in which the objective function and limiting functions are coded are "Artificial Bee Colony Algorithm" and "Harmony Search Algorithm". In the objective function, minimizing the concrete volume is adopted. Since the optimization of the dimensions was investigated, optimization analyzes were carried out with 8 different limiting functions. The minimum foundamental concrete volume was calculated according to the optimum dimensions and the analysis results were presented graphically. In addition, the accuracy analyzes of the Randolph calculation code and the algorithm-based optimization codes written were made. Accuracy analyzes were compared with studies in the literature. According to the analysis results; Increasing the distance between the piles and the pile diameter resulted in a decrease in the amount of settlement, and an increase in the amount of settlement occurred when the stress on the raft was increased. On the other hand, in the volume of concrete used in the optimization analysis, an economy of 27% for the full-size first project with YAKA analyzes, 31% for the HAA analyzes, 39.5% for the second project with YAKA analyzes, 42% for the HAA analyzes and for the third project an economy of 24% was obtained with the YAKA analysis and 27% with the HAA analysis. Comparing the results of the algorithms, consistent results were observed between YAKA and HAA analyzes and more optimum results were obtained in terms of material efficiency with HAA analyzes.