CAJ | 학술논문

利用离散元方法对颗粒材料的细观力学特性研究，目前确定数值计算模型的细观力学参数大多数通过反复调试获取，效率低、可重复性差。本文采用开源的颗粒离散元程序 LMGC开展了土体双轴压缩数值试验，通过25组土体细观力学参数计算得到相应的宏观力学参数，建立了 BP人工神经网络反演系统。利用土体物理试验得到的土体宏观力学参数，输入 BP神经网络，反演得到土体的细观力学参数。将所得细观力学特性参数输入所建立的土体数值计算模型，得到土体破坏过程中的应力-应变关系曲线，以及土体颗粒的力链图和旋转变形云图。所建立的土体数值试验模型能够较好地模拟土体变形破坏过程，利用 BP神经网络反演细观力学参数以及数值模型计算得到的土体宏观力学参数与物理试验吻合较好，误差在10％左右，土颗粒间力链云图以及旋转变形云图较好地揭示了土体变形破坏的机理。
이용리산원방법대과립재료적세관역학특성연구，목전학정수치계산모형적세관역학삼수대다수통과반복조시획취，효솔저、가중복성차。본문채용개원적과립리산원정서 LMGC개전료토체쌍축압축수치시험，통과25조토체세관역학삼수계산득도상응적굉관역학삼수，건립료 BP인공신경망락반연계통。이용토체물리시험득도적토체굉관역학삼수，수입 BP신경망락，반연득도토체적세관역학삼수。장소득세관역학특성삼수수입소건립적토체수치계산모형，득도토체파배과정중적응력-응변관계곡선，이급토체과립적력련도화선전변형운도。소건립적토체수치시험모형능구교호지모의토체변형파배과정，이용 BP신경망락반연세관역학삼수이급수치모형계산득도적토체굉관역학삼수여물리시험문합교호，오차재10％좌우，토과립간력련운도이급선전변형운도교호지게시료토체변형파배적궤리。
The micromechanical parameters of numerical model are obtained by repeated tests.Researches onparticle material based on DEMare done.This article builds a numerical model of soil using LMGC,an open sourcesoftware.And the 25groups of macromechanical and micromechanical parameters calculated by the numerical modelbuild a BP neural network.The BP neural network can backcalculates the micromechanical parameters if themacromechanical parameters of real physical tests are inputted into the neutral network.The numerical model canget the stressstrain line,stress train and particle rotation graph if the micromechanical parameters are set.Theresults show that the numerical model can well simulate the soil destruction process.The stressstrain line ofnumerical model matches well with the result of physical test.And error of macromechanical parameters calculatedby BP neural network is about 1 0%.Besides,the stress train and particle rotation graph reveal the mechanism ofsoil destruction process.