With the improved moving least-squares (IMLS) approximation, an orthogonal function system with a weight function is used as the basis function. The combination of the element-free Galerkin (EFG) method and the IMLS a...With the improved moving least-squares (IMLS) approximation, an orthogonal function system with a weight function is used as the basis function. The combination of the element-free Galerkin (EFG) method and the IMLS approximation leads to the development of the improved element-free Galerkin (IEFG) method. In this paper, the IEFG method is applied to study the partial differential equations that control the heat flow in three-dimensional space. With the IEFG technique, the Galerkin weak form is employed to develop the discretized system equations, and the penalty method is applied to impose the essential boundary conditions. The traditional difference method for two-point boundary value problems is selected for the time discretization. As the transient heat conduction equations and the boundary and initial conditions are time dependent, the scaling parameter, number of nodes and time step length are considered in a convergence study.展开更多
文摘针对传统有限元法求解声学问题由于刚度矩阵过硬导致较大的色散误差,以及在较高波数和网格扭曲时计算精度过低甚至错误的问题,采用移动最小二乘权函数对传统有限元法的声压梯度进行加权重构,推导了梯度移动最小二乘加权(Gradient weighted by moving least-squares,GW-MLS)的二维声学计算公式。对声压梯度的加权重构使得GW-MLS模型的刚度相对于FEM模型得以软化,刚度更接近真实模型刚度。采用与有限元法相同的方式构造质量矩阵和边界积分矢量,保证质量矩阵和边界条件的正确施加和积分精度。通过二维管道声腔模型和二维车内声腔模型算例对所提出的算法进行验证,数值分析结果表明,GW-MLS有效地减少了色散误差的影响,提高了计算精度,尤其是对较高波数和网格扭曲时表现出良好的适应性。
基金the National Natural Science Foundation of China (Grant No. 11171208)Shanghai Leading Academic Discipline Project (Grant No. S30106)
文摘With the improved moving least-squares (IMLS) approximation, an orthogonal function system with a weight function is used as the basis function. The combination of the element-free Galerkin (EFG) method and the IMLS approximation leads to the development of the improved element-free Galerkin (IEFG) method. In this paper, the IEFG method is applied to study the partial differential equations that control the heat flow in three-dimensional space. With the IEFG technique, the Galerkin weak form is employed to develop the discretized system equations, and the penalty method is applied to impose the essential boundary conditions. The traditional difference method for two-point boundary value problems is selected for the time discretization. As the transient heat conduction equations and the boundary and initial conditions are time dependent, the scaling parameter, number of nodes and time step length are considered in a convergence study.