| Because of the complexity of the properties of rock materials and their environment,the study of the mechanical behavior of rock materials has been a very challenging field in the modern computational solid mechanics.The complex macroscopic mechanical properties of rock materials are caused by disordered structures and mesoscale defects.Thus,the bonded-particle model(BPM),which encompasses particle structure and a mesoscale constitutive model,provides a bridge linking the mesoscopic structure and macroscopic rock characteristics;this model has developed into a powerful tool for studying the physical and mechanical properties of rocks.When bonded particle model(BPM)is used to analyze,parameters calibration is the first problem that users need to face,and is also the key to obtain reasonable simulation results.However,because of the complexity of the particles system and the assumption of the ‘soft ball' of BPM,the relationship between the macroscopic and mesoscopic parameters of the model is difficult to derive from the theory,and the parameters calibration has been a tedious and difficult task.Thus,based on statistical analysis,experimental design and optimization model,two basic problems of BPM,such as the particle size effect and calibration method of mesoscopic parameters,are studied.The main research work and conclusions in this paper include the following aspects:(1)Based on the statistical analysis scheme of the single factor and multi level quantitative data,the particle size effect of the BPM is studied.The effect of particle size on the macroscopic mechanical parameters and failure modes of the BPM is obtained.Finally,combining the statistical results of mechanical parameters,the failure mode of the model and the computational efficiency,the L/R?200 should be more be suitable for the simulation of rock materials.(2)Based on orthogonal test design and central composite design,qualitative and quantitative relationship between macro and mesoscopic parameters of BPM on specific interval is studied and the influence of various mesoscopic parameters on the macro parameters is obtained.Finally,the quantitative relations between macro and mesoscopic parameters are fitted.It not only provides a basis and guidance for the calibration of meso parameters,but also provides a basis for establishing the optimization model of meso parameter calibration.(3)Based on the qualitative and quantitative relationships,the mesoscopic parameters calibration problem is transformed into an optimization problem.Then,a corresponding optimization model is constructed.A global search algorithm that does not depend on the initial value-genetic algorithm is used as optimization algorithm to solve the optimization model mentioned above.Add related constraints to control the final failure mode of the model.Finally,the calibration program of the mesoscopic parameters is written in the Matlab environment.The verification of the program shows that the demarcation program of the mesoscopic parameters has achieved good results.(4)The coupling calculation model is built on the basis of a diversion tunnel project,and the previous conclusions are verified.Based on the particle size effect of the rock model,the particle size of the cavern model is determined.Based on the above calibration program,a set of mesoscopic parameter selection schemes under different constraint conditions is given,and the effect on the deformation failure process and the final failure mode of the cavern model is studied.Finally,the influence of the mesoscopic parameters on the failure mechanism of the cavern model is obtained.In this paper,the parameter characteristics of the particle discrete element rock model are studied.The relation and law of the macro meso parameters of the rock model are given,and the calibration method of the mesoscopic parameters of the rock model is established.This can not only make the simulation and analysis of the rock material more reliable,but also greatly reduce the workload of the researchers in the early stage of modeling. |
PDF Full Text Request