Abstract: To solve the problems faced in porous structure modeling such as the design complexity of anisotropy and the low controllability of the mechanical properties,an enhanced orthotropy porous structure design method based on the improved k-nearest algorithm was proposed.The site density function was used to generate sites in the design domain by random sampling algorithm,and a desired global direction field was imported into the design domain.Then,by improving the k-nearest algorithm to connect k-nearest sites for each site,a 2.5D/3Denhanced orthotropy porous structure was built.Therefore,the orthotropy could be controlled by the stretching ratio of the Riemann ellipse or ellipsoid,the rotational directions of the orhtogonal axis were controlled by the direction field,and what's more,the gradient of the mechanics property of the porous structure could be controlled by the site density.The homogenization and compressive tests were respectively applied to a set of 3D printed orthotropy porous structures to analyze their mechanics properties,failure states,randomness and elasticity design space.The results showed that the orthotropy porous structures had excellent mechanics properties and large elasticity design space.

Key words: gradient porous structure, orthotropy, homogenization, mechanical property

摘要： 针对各向异性多孔结构设计复杂、力学性能可控性低等问题,提出一种基于k-nearest算法的增强型正交各向异性多孔结构设计方法。首先,通过站点密度场驱动随机采样算法,在设计域内生成站点。然后提出一种基于黎曼距离的改进k-nearest算法,叠加全域方向场,生成25D/3D增强型正交各向异性多孔结构。该方法不仅可以通过黎曼椭圆/球的拉伸比例控制局部正交各向异性程度,通过方向场控制各向异性的朝向,还可以通过站点密度场调控多孔结构力学性能的梯度分布。最后,分别通过数值均质化和压缩实验对3D打印的多孔结构的力学性能、失效状态、随机性和弹性设计空间进行分析。实验结果表明,采用所提方法设计的正交各向异性多孔结构具有更优良的弹性性能和更大的弹性设计空间

关键词: 梯度多孔结构, 正交各向异性, 数值均质化, 力学性能