Issue 53

A. Namdar, Frattura ed Integrità Strutturale, 53 (2020) 285-294; DOI: 10.3221/IGF-ESIS.23.22 287 Step 4, statistical analysis: Each element of soil mechanical properties behaves differently in a group of selected data from those reported in the literature [1], and the interaction between these elements requires a suitable technique for selecting the range of safe bearing capacity in the soil mixture design. To establish a histogram for safe bearing capacity of soil, the class interval of 500 was selected and the frequency of safe bearing capacity computed. To find intensity of repeating safe bearing at each 500 interval, relative frequency is divided into classes. After drawing rectangle for each class in construct histogram, the probability of safe bearing capacity was draw for the prediction. The Central Limit Theorem was used for probability analysis. In application of the Central Limit Theorem, a large enough sample is needed. It was suggested sample size of 30 or more is sufficient for application of Central Limit Theorem [23]. In this study, 31 sample were selected from an experimental investigation, so the Central Limit Theorem is applicable for probability analysis and prediction of safe bearing capacity. For this work, the following equations are used: 1 n X X X n   = Sample mean (1) 1 n n S X X   = Sample observation (2) 2 ~ , X N n         approximately (3)   2 ~ , n S N n n   approximately (4) where X 1 , X 2 , ………., X n constitute the random sample from population, µ is the Mean and σ 2 is the Variance [23]. ANN is applied to predict soil foundation behavior considering the mechanical properties of the mixed soil. ANN contents three layers, which are the input, the hidden and the output layers. These three layers integrate assessment and prediction of safe bearing capacity mechanism of the soil foundation. In this study, one hidden layer has been used in the ANN analysis. The minimum hidden layer in the ANN is one and maximum hidden layers depend on the problem complexity. The safe bearing capacity prediction by the effect of optimum moisture content (%), density (kN/m 3 ), angle of friction (deg) and cohesive of soil (kN/m 2 ) was performed using ANN. For the cover effect of each element of the mechanical properties in the prediction of the safe bearing capacity, R 2 and RMSE of optimum moisture content (%), density (kN/m 3 ), angle of friction (deg) and cohesive of soil (kN/m 2 ) on the safe bearing capacity of the soil are separately analyzed in the first stage. In the second stage, R 2 and RMSE of the safe bearing capacity are analyzed considering all mechanical properties together. Figure 1: Flowchart for explaining stepwise in this investigation.