Calibration of the microparameters of the discrete element method using a relevance vector machine and its application to rockfill materials. (September 2020)

• Record Type: Journal Article
• Authors: Ma, Chunhui; Yang, Jie; Zenz, Gerald; Staudacher, Edwin Josef; Cheng, Lin
• Is Part Of: Advances in engineering software. Volume 147(2020)
• Publication Date: 2020-09
• Subjects: Rockfill slope -- Microparameter calibration -- Triaxial model test -- Constitutive model -- Contact bond model -- Relevance vector machine; Computer-aided engineering -- Periodicals; Engineering -- Computer programs -- Periodicals; Engineering -- Software -- Periodicals; Periodicals; 620.0028553
• DOI: 10.1016/j.advengsoft.2020.102833
• Languages: English
• ISSNs: 0965-9978

Abstract

Highlights: The relationship between the macro- and microparameters is established with a relevance vector machine and the memetic algorithm. A calibration example involving rockfill materials is studied with a refined 3-D numerical triaxial test. The law of the fabric parameters under different confining pressures and their micromechanism are explained. The failure process of the rockfill slope is analysed, and the exit of the flood discharge tunnel will not be blocked even under the most dangerous situation. Abstract: To accurately simulate the mechanical properties of rockfill materials under triaxial stress using the discrete element method (DEM), the microparameters of the rockfill material DEM model are calibrated based on the macroparameters of a constitutive model. For the calibration model, a refined 3-D numerical triaxial test is established to obtain the deformation curves of the rockfill under different confining pressures. Deformation curves with different microparameters in the contact model are then calculated by numerical triaxial tests to determine the macroparameters of the constitutive model. Then, a nonlinear relationship between the macroparameters of the E-B model and the microparameters of the contact bond model is built by a relevance vector machine. A memetic algorithm is employed to calibrate the microparameters of the rockfill, thereby establishing a calibration model based on a relevance vector machine and a memetic algorithm. Afterwards, a numerical triaxial test example of rockfill materials is used to verify the feasibility of this calibration model, and then the fabric microparameters under different confining pressures are analysed. In addition, to determine the influence region, a failure process of a rockfill slope is simulated with the calibrated microparameters of the DEM model. In summary, this calibration method substantially improves the numerical results, thereby enhancing the ability to determine the mechanical properties of construction materials and solve various problems in engineering.