Statistical Modeling of Mechanical Properties of Blast-Compacted Soils

Abstract

Mathematical modeling of compaction of subsidence soils by explosion leads to the need to assess the mechanical properties of soils depending on their physical indicators. The angle of internal friction and specific adhesion of soil from moisture content are estimated. Correct use of the apparatus of mathematical statistics, namely the tools of correlation and regression analysis, made it possible to establish the dependencies of mechanical and physical properties of soil. It is shown that the studied soil properties are mutually correlated. A negative correlation is established between the angle of internal friction and specific adhesion from soil moisture content. Regression modeling made it possible to construct the dependencies of specific adhesion and the angle of internal friction from the moisture content of compacted soil. An increase in the values of strength properties with a decrease in the values of moisture content is recorded. The reliability of the approximation results is determined by the statistical coefficient of determination equal to 0.83–0.84. The modulus of soil deformation depends on the values of the porosity coefficient and the diameter of the camouflet. The obtained dependencies of mechanical on physical properties make it possible to predict the behavior of geological systems as a result of compaction by explosion. The practical implementation of the obtained results is applicable at the stage of engineering and geological design of construction projects erected on subsidence soils.