Cone penetration test (CPT) is a commonly used in-situ testing method used in geotechnical engineering for its practical efficiency. It is in particular useful for sandy soils when sampling is difficult. How to interpret CPT result to get reliable estimation of soil properties and design parameters is the key challenge in the development of CPT technique.  Results from CPT testing is affected by internal friction angle, density, compressibility, K₀, etc. Existing analysis methods are established based on the bearing capacity theory [1] or cavity expansion theory [2-4]. How to take account sand compressibility has not yet been solved. Numerical simulation have been widely used to study penetration process [5-6]. Most finite element simulation use the arbitrary Lagrangian-Eulerian (ALE) large deformation technique. Material point method [7] has also been used lately. And yet, the difficulty in sand constitutive modeling remains challenging. Discrete element, as an emerging simulation technique [8] has the advantage of modelling sand based particles and their interactions without use of soil constitutive models The difficulty however lies in the computation cost and efficiency. Recent years, multi-scale numerical approach has been developed rapidly [9], and  uses element responses from discrete element and use FEM framework to solve boundary value problems.
In this paper, a hierarchical multi-scale numerical simulation platform has been developed to achieve explicit FEM-DEM multi-scale simulation based on the general commercial software Abaqus and the open source software LIGGGHTS. The RVE results are used instead of the constitutive model at the Gaussian point, and the arbitrary Lagrangian-Eulerian (ALE) method is used to avoid the mesh distortion caused by large deformation, which effectively improves the computational efficiency. Based on this multi-scale platform, simulation of cone penetration calibration chamber test in sand is then carried out, the variations of cone tip resistance and lateral friction resistance with depth are obtained and compared with the calibration chamber test results, which verified the validity of the multi-scale model. The penetration simulations in sand under various cell pressure and void ratio are then carried out, the effects of factors such as soil compressibility and initial void ratio on the penetration process and results were studied, and the applicability of the existing analytical model was analysed and compared. Then, the results and micro-scale data provided by DEM simulations are fully utilized to analyze the stress and deformation, void ratio and anisotropy evolution of the soil around the cone in the cone penetration process, and to explain in depth the effect of cone penetration on the soil and its influence mechanism on the penetration resistance.
 Cone penetration test (CPT) is a commonly used in-situ testing method used in geotechnical engineering for its practical efficiency. It is in particular useful for sandy soils when sampling is difficult. How to interpret CPT result to get reliable estimation of soil properties and design parameters is the key challenge in the development of CPT technique.  Results from CPT testing is affected by internal friction angle, density, compressibility, K₀, etc. Existing analysis methods are established based on the bearing capacity theory [1] or cavity expansion theory [2-4]. How to take account sand compressibility has not yet been solved. Numerical simulation have been widely used to study penetration process [5-6]. Most finite element simulation use the arbitrary Lagrangian-Eulerian (ALE) large deformation technique. Material point method [7] has also been used lately. And yet, the difficulty in sand constitutive modeling remains challenging. Discrete element, as an emerging simulation technique [8] has the advantage of modelling sand based particles and their interactions without use of soil constitutive models The difficulty however lies in the computation cost and efficiency. Recent years, multi-scale numerical approach has been developed rapidly [9], and  uses element responses from discrete element and use FEM framework to solve boundary value problems.
In this paper, a hierarchical multi-scale numerical simulation platform has been developed to achieve explicit FEM-DEM multi-scale simulation based on the general commercial software Abaqus and the open source software LIGGGHTS. The RVE results are used instead of the constitutive model at the Gaussian point, and the arbitrary Lagrangian-Eulerian (ALE) method is used to avoid the mesh distortion caused by large deformation, which effectively improves the computational efficiency. Based on this multi-scale platform, simulation of cone penetration calibration chamber test in sand is then carried out, the variations of cone tip resistance and lateral friction resistance with depth are obtained and compared with the calibration chamber test results, which verified the validity of the multi-scale model. The penetration simulations in sand under various cell pressure and void ratio are then carried out, the effects of factors such as soil compressibility and initial void ratio on the penetration process and results were studied, and the applicability of the existing analytical model was analysed and compared. Then, the results and micro-scale data provided by DEM simulations are fully utilized to analyze the stress and deformation, void ratio and anisotropy evolution of the soil around the cone in the cone penetration process, and to explain in depth the effect of cone penetration on the soil and its influence mechanism on the penetration resistance.
 

cone penetration test, FEM-DEM, ALE, cavity expansion theory，macro and micro relations