| To explore the Engineering and Dynamic characteristics of Sulphate Saline soil,three types of tests are performed.According to the early theoretical data collected,relevant procedures of shear test,compression test,and dynamic test parameters are known.The shear and compression properties of saline soil are two important indexes to describe the engineering properties of saline soil.Cohesion can well reflect the shear strength index of soil,and compression modulus can well reflect the compressibility index of soil.Through the triaxial shear test and compression test,the relationship between cohesion and compression modulus of soil and salt content and salt type of saline soil is studied.By controlling different moisture content and salt content conditions,indoor dynamic triaxial tests are carried out.The test results show that the dynamic stress and dynamic strain curves are nonlinear and have certain hysteresis and deformation accumulating.Dynamic elastic modulus of Sulphate saline soil decreases with the increase of moisture content.The salt content condition for sulfate saline soil is initially decreased and then increases with the increase of salt concentration.Maximum dynamic elastic modulus enhances with the increase of moisture content and salt content.The relationship between the damping ratio and the strain is in the scatter distribution.The size of the overall changes between 0-0.3,the influence of moisture content and salt content are not evident.The results of the dynamic strength test show that the dynamic strength of sulfate saline soil under these conditions emerges a decreasing trend with the increasing destruction of vibration time,a nearly linear relationship.In another aspect,the dynamic strength decreases with the increasing moisture content and salt content.The more concentration of salt content and moisture content,the easier it is destroyed under the same damage conditions of vibration times,and the lower the dynamic strength is.Using artificial intelligence technique,MATLAB tool ANN is used to predict dynamic properties of soil.A feedback propagation artificial neural network model(FBPANN)is constructed.The comparison between measured and predicted values by Root mean Square error shows the eligibility of the model. |