摘要
为了探究更高效的碳化硅(SiC)材料热导率的模拟方法,应用逆非平衡分子动力学(rNEMD)法及传统非平衡分子动力学(NEMD)法对β晶型SiC(β-SiC)材料的热导率进行模拟计算和对比; 2种方法的模拟过程均先建立横截面尺度小而轴向尺度大的棒状模型,采用周期性边界条件、应用修正嵌入原子法(MEAM)势函数,先后进行正则系综(NVT)的弛豫和微正则系综(NVE)内的动态沿轴向生成温度梯度的过程,分别利用傅里叶定律模拟计算得到SiC材料的热导率。结果表明:2种方法的计算结果均出现热导率随生成温度梯度的材料轴向尺度增加而增大的有限尺度效应,应用倒数拟合的外推法可以计算模拟体系沿轴向为无穷大时的宏观体相β-SiC材料的热导率; r NEMD法具有较高的计算效率,更适合热导率的模拟计算。
In order to find a more efficient method to simulate the thermal conductivity of SiC,reverse and conventional non-equilibrium molecular dynamics method were applied to investigate the thermal conductivity ofβ-SiC crystal respectively.In the simulation process of the two methods,simulation cells of cross sections with small area and axial lengths with large size were constructed.Periodic boundary conditions and modified embedded atom method(MEAM)potential were adopted and canonical(NVT)ensemble for equilibrating and microcanonical(NVE)ensemble for generating temperature gradient in axial direction were implemented,then the results of thermal conductivity ofβ-SiC crystalachieved from the two methods were deduced from Fourier's law.The results show that the finite-size effect of both methods is observed that the thermal conductivity increases with the increasing length in axial direction.An inverse fitting method is applied and bulk thermal conductivity ofβ-SiC crystal can be obtained by extrapolating the size in axial direction to an infinite system size.The reverse non-equilibrium method has a more effective computational process making it more applicable for the numerical simulation of thermal conductivity.
作者
毛亦尘
熊扬恒
岳亚楠
MAO Yiche;XIONG Yangheng;YUE Yanan(Hubei Provincial Key Laboratory of Accoutrement Technique in Fluid Machinery and Power Engineering,Wuhan 430072,China;School of Power and Mechanical Engineering,Wuhan University,Wuhan 430072,China)
出处
《济南大学学报(自然科学版)》
CAS
北大核心
2019年第1期15-21,共7页
Journal of University of Jinan(Science and Technology)
基金
国家自然科学基金项目(51576145)
关键词
非平衡分子动力学
碳化硅
热导率
有限尺度效应
外推法
non-equilibrium molecular dynamics
silicon carbide
thermal conductivity
finite-size effect
extrapolation