The Study And Application Of Soil-tool Interaction Numerical Model Based On SPH

China is a large agricultural country.Promoting the advancement of agricultural modernization is beneficial to the economic development of China.Agricultural mechanization can significantly improve agricultural productivity and efficiency.It is a necessary condition to accelerate the process of agricultural modernization.Agricultural machinery is the main tool and carrier of agricultural mechanization production.Soil-engaging components are an important part of agricultural machinery.The main energy consumption of agricultural machinery depends on the resistance force on the soilengaging components during the tillage process.Therefore,it is of great interest to study the resistance forces on soil-engaging components.The traditional design method of soil-engaging components mainly relies on empirical method or experimental method,which has low efficiency and long design cycle.With the advancement of agricultural modernization in China,the design of soil-engaging components of agricultural machinery should be with the help of numerical simulation methods which are more efficient and modern.In the numerical simulation of soi-tool interaction,finite element method(FEM)is widely used.But due to the limitation of FEM method,the large deformation of soil during the tillage process will lead to mesh distortion,which will influence the accuracy of simulation sharply and even terminate the simulation calculation.Another common method is discrete element method(DEM).However,there is a lack of unified parameter calibration standard in DEM simulation,and the model is not universal.Besides,DEM particles can’t represent the properties of the entire soil.It is significative to propose a high-precision numerical model for the interaction of soil and soilengaging components.The main work of this paper is to construct a soil-tool interaction model based on the smooth particle hydrodynamics(SPH)method with two soil constitutive models.It can accurately describe the variation of the cutting force of the soil-engaging components during tillage.And provide a new idea for numerical simulation method to the design of soil-engaging components.Firstly,this paper introduces the basic idea and concept of SPH method.Based on which the SPH form of Navier-Stocks equation controlling the motion of soil particles is derived.The common numerical methods in SPH simulation are analyzed and improved for the working scenario of soil-tool interactions.Monaghan-type artificial viscosity is applied to deal with numerical dissipation and suppressing unphysical oscillations in the simulation.Artificial stress is applied to solve the phenomenon of the tensile instability.Courant-Friedrichs-Levy condition is used to calculate the time step,which is used to calculate the variable integration using the leap frog method.The no-slip boundary is applied in the simulation.Finally,two cases of soil stress anomalies due to numerical calculation errors are discussed and solutions are proposed.The soil-tool interaction model based on SPH proposed in this paper is constructed in 2 main ways: on the one hand using 2 soil constitutive models to describe non-cohesive and cohesive soils.The soil elastoplastic constitutive model is based on the Drucker-Prager damage criterion.It can accurately simulate both non-cohesive soil and cohesive soil.In addition,we analyze the relationship between cohesion and accumulative plastic strain on the basis of the traditional soil elastoplastic constitutive model and constructs a shear damage model for cohesive soil.The revised model is more objective and realistic for the simulation of cohesive soil.Another soil constitutive model is the soil hypoplastic constitutive model.It has significant advantages for the simulation of anisotropy in noncohesive soil.Besides,soil hypoplastic constitutive model has the advantage of a clear structure of the equations and is easy to code.In this paper,the soil constitutive model in the form of SPH is derived to calculate the soil stresses through time integration and then solve the Navier-Stocks equation to achieve an accurate description of the soil particle movement.On the other hand,the contact force between the soil and the soil-engaging components is studied.The soil-engaging components is considered as a rigid body composed of a series of tool particles.The contact force is calculated according to the distance between the soil particles and the soil-engaging components particles.The calculation method of surface vectors for soil-engaging components is discussed.The direction problem of surface vectors is analyzed and a solution to confirm the direction is proposed.In this paper,the proposed soil-tool interaction models are validated using physical tests.Firstly,a non-cohesive soil collapse test was conducted to verify the accuracy of these 2 soil constitutive models in the SPH framework by comparing the free surface morphology and stress-strain distribution of the non-cohesive soil during the collapse process.Some quantitative indicators is developed to compare with physical tests.Then a soil cutting test set was designed for non-cohesive soil and cohesive soil cutting.The accuracy of the soil-tool interaction model based on SPH and its applicability to the simulation of different shapes of soil-engaging components were verified by comparing the internal changes of the soil during cutting,the form of damage and the pattern of cutting forces.Based on verifying the accuracy of the model,numerical simulations are carried out for different cutting angles,cutting depths and soil plane.The influence of each working condition on the cutting force of the soil-engaging component is analyzed.The numerical simulation results are of some guidance for the design and production of the soil-engaging components.Finally,to improve the efficiency of the model,the soil-tool interaction model is integrated into the Dual SPHysics.It is a SPH open-source framework,which can call on the graphic process unit(GPU)to accelerate the computation.In terms of the number of particles computed per second,the GPU acceleration makes the SPH program about 250 times more efficient than a pure CPU program.GPU acceleration can make the SPH programs more efficient and useful.