Abstract:
A soil-soil engaging component interaction model with high accuracy can greatly contribute to the numerical simulation of soil engaging components. Traditional simulation methods cannot fully meet the harsh requirements of the soil- soil engaging components interaction under a complex field in recent years, including the Computational Fluid Dynamics (CFD), Finite Element Method (FEM), and Discrete Element Method (DEM). Thus, this study presents a new numerical model for the soil- soil engaging components interaction using Smoothed Particle Hydrodynamics (SPH). The cubic B-spline function was also chosen as the smooth kernel in the SPH framework. Specifically, the hypoplastic constitutive model was used to select the Navier-Stocks equations for the stress tensor of the soil flow particles, where the smooth length h = 1.2dini (dini was the initial distance between particles). The Monaghan-type artificial viscosity was applied to reduce the unphysical oscillations in the non-cohesive simulation, where the artificial terms α and β were 0.1, and 1.0, respectively. The boundary condition of the SPH model was the non-slip boundary with the three-layer virtual particles. A contact model was then established to calculate the contact force between the soil and blade particles. The optimal contact force was calculated when the distance between soil particle and soil engaging component particle dp was less than a threshold distance value d0. The contact force was then divided into the horizontal and vertical forces using the Fortran subprogram. Then, a sand collapse and a soil cutting test were used to validate the model. All the soil cutting experiments were performed in a soil box made of acrylic plates. Among them, the soil slipped naturally under the gravity in the sand collapse test. Once the soil was steady, there was a smooth free surface, accurately representing the large deformation of the soil. Some geometric parameters were also measured in the sand collapse test, such as the angle of the failure line (βa), the farthest distance of sand (df), and the height of sand in the final status (hf). The simulation βa, df and hf were 45°, 390 mm and 100 mm, respectively, whereas, the experimental βa, df and hf were 45°, 360 mm, and 100 mm, respectively, indicating the good consistence. The distribution of soil horizontal and vertical stress increased with the soil depth. There was a distinct shear rapture plane with the output of soil particles distance, while the soil shear stress was mainly concentrated in the soil where the lateral slip occurred. At the same time, the soil was vertically painted to the layers in the soil cutting test, where the cutting velocity was 20 mm/s. There was a fold-like displacement in the layers near the blade during the simulation. In addition, four cross-section shape (LC-parabolic, SC-parabolic, VER-vertical, and STR-straight) blades were simulated for the soil cut. The cutting simulation revealed that the force on the blade was similar when three types of sands had a similar density. The shear stress distribution was also simulated to clarify the mechanism of soil cutting for the four blades and the velocity of the blade LC. The cutting force variation during the simulation was better consistent with the experiment in the different shape blades, indicating the average relative value error of 9.885%. Consequently, the improved interaction model can be expected to accurately simulate the soil-tool interaction. The finding can also provide a new idea to establish the soil-soil engaging component interaction model for the optimal design of soil engaging components.