基于多尺度再生核质点法的h型自适应分析

基于多尺度再生核质点法(RKPM)的多分辨分析特性，给出了适于求解二维问题的h型自适应分析方法，讨论了不同高梯度指标对检测高梯度区的影响，同时还提出了基于邻节点搜索(SNN)及局部Delaunay三角分解(LDT)技术的适于任意节点分布的无网格自适应分析节点加密技术．通过对二维线弹性平面应力及平板弯曲问题的h型自适应分析，说明了不合适的高梯度指标将对自适应分析的收敛造成不良影响，且LDT节点加密技术较SNN技术的效率更高．数值算例的结果说明自适应分析的有效性、稳定性及良好的收敛特性．

h-Adaptivity Analysis Based on Multiple Scale Reproducing Kernel Particle Method

An h_adaptivity analysis scheme based on multiple scale reproducing kernel particle method was proposed, and two nodes refinement strategies were constructed using searching_neighbor_nodes (SNN) and local_Delaunay_triangulation (LDT) techniques, which were suitable and effective for h_adaptivity analysis on 2_D problems with the regular or irregular distribution of the nodes. The results of multiresolution and h_adaptivity analyses on 2_D linear elastostatics and bending plate problems demonstrate that the improper high_gradient indicator will reduce the convergence property of the h_adaptivity analysis, and that the efficiency of the LDT node refinement strategy is better than SNN, and that the presented h_adaptivity analysis scheme is provided with the validity, stability and good convergence property.