Controllable fractal modelling method based on soil statistical parameters

Internal structure of soil is disorder and with statistical self-similar fractal attribution. How to construct its micro-structure model is very important for the study of electrical properties, thermal properties and hydrodynamics. So, this paper presents a controllable modelling method for soil. This paper firstly introduces the development process from the classical soil fractal model (Sierpinski carpet model) to randomized Sierpinski carpet model, and analyses the difference between the latter and actual soil. A Lower-weighted Non-uniform Iterative Fractal (LNIF) method is proposed to reduce the influence weight of first fractal by introducing Initial Scale Reduction Factor(ISRF). X
²
test is used to verify that the porosity of soil model established by LNIF method conforms to normal distribution regularity when ISRF is large enough. Furthermore, this paper studies the control strategy of model parameters aimed to make LNIF model closer to measured parameters in actual soil. The numerical relationship between soil solid particle size, porosity and its deviation range and LNIF model input parameters is established. Then the paper proposes a fractal modelling strategy based on actual soil structural parameters, and presents a Controllable Fractal Modelling(CFM) method based on soil statistical parameters. This method can not only make soil model porosity conform to the statistical regularity of normal distribution, but also take the distribution of soil particle size, the average and error range of measured porosity as control variables to realize the control of solid particle size and porosity in the model, which leads to the established soil microstructure model have the same statistical regularity as the actual soil. Finally, this paper presents the specific operation process taking the Leached Cinnamon soil as an example.