当前位置: X-MOL 学术3D Print. Addit. Manuf. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Artificial Evolution and Design for Multi-Material Additive Manufacturing
3D Printing and Additive Manufacturing ( IF 3.1 ) Pub Date : 2020-12-16 , DOI: 10.1089/3dp.2020.0114
Wuxin Yang 1 , Emilio Calius 2 , Loulin Huang 1 , Sarat Singamneni 1
Affiliation  

Limitations of the traditional manufacturing methods often force engineered components to be made of single material systems. However, this is going through changes due to the advent of additive manufacturing (AM) methods, as the point-by-point consolidation allows for a possible change of the material constitution within a given part domain. This will give rise to a plethora of new material and property options for the designers, where just human perception may fail to realize the full benefits. Automated design tools integrating material choice, dispersion, analysis, and optimization algorithms need to be developed to assist in finding the optimal multi-material dispersion solutions achieving given performance criteria sets. Considering the fact that the multi-material manufacturing systems are only recently coming into use, design solutions targeting optimal placement of multiple materials are not common. This article addresses this gap, evaluating a numerical model integrated with different optimization schemes to find the optimal material solutions achieving certain preset performance criteria such as combinations of natural frequencies in different degrees of freedom. A case study of three different metaheuristic optimization schemes based on genetic algorithms indicates, first, that it is possible to create a beam with six uniformly spaced natural frequencies and to change these frequencies without modifying the structural geometry; and second that the basic genetic algorithm generally outperforms neural net-based alternatives for this problem. This tailoring of the structural resonance spectrum demonstrates that evolutionary computing combined with multi-material AM can be used to unlock previously unavailable structural functionality.

中文翻译:

多材料增材制造的人工进化与设计

传统制造方法的局限性通常迫使工程部件由单一材料系统制成。然而,由于增材制造 (AM) 方法的出现,这种情况正在发生变化,因为逐点整合允许在给定零件域内对材料构成进行可能的更改。这将为设计师带来过多的新材料和属性选择,而仅靠人类感知可能无法实现全部好处。需要开发集成材料选择、分散、分析和优化算法的自动化设计工具,以帮助找到实现给定性能标准集的最佳多材料分散解决方案。考虑到多材料制造系统最近才投入使用,以多种材料的最佳放置为目标的设计解决方案并不常见。本文解决了这一差距,评估了一个集成了不同优化方案的数值模型,以找到实现某些预设性能标准(例如不同自由度的自然频率组合)的最佳材料解决方案。基于遗传算法的三种不同元启发式优化方案的案例研究表明,首先,可以创建具有六个均匀间隔自然频率的梁,并在不修改结构几何形状的情况下改变这些频率;其次,对于这个问题,基本遗传算法通常优于基于神经网络的替代方案。
更新日期:2020-12-20
down
wechat
bug