共 50 条
Hierarchical sheet triply periodic minimal surface lattices: Design, performance and optimization
被引:1
|作者:
Xu, Hong
[1
,2
]
Zhang, Yu
[1
,2
]
Mei, Yuheng
[1
,2
]
Wu, Zhiyuan
[1
,2
]
Zhang, Yuan
[1
,2
]
Ma, Mengxin
[1
,2
,3
]
Liu, Xiaohu
[1
,2
]
机构:
[1] Huazhong Univ Sci & Technol, Dept Mech, Wuhan 430074, Hubei, Peoples R China
[2] Hubei Key Lab Engn Struct Anal & Safety Assessment, Luoyu Rd 1037, Wuhan 430074, Peoples R China
[3] Xian Inst Electromecflan Informat Technol, Xian 710000, Shanxi, Peoples R China
关键词:
Triply periodic minimal surface;
Hierarchical lattices;
Thermal-hydraulic performance;
Deep reinforcement learning;
HEAT-TRANSFER ENHANCEMENT;
SINK;
D O I:
10.1016/j.applthermaleng.2024.125187
中图分类号:
O414.1 [热力学];
学科分类号:
摘要:
Increasing power consumption of critical components requires the development of more compact and efficient heat sinks. Triply periodic minimal surfaces have emerged as promising solutions for advanced thermal management. However, traditional approaches to enhancing the thermal-hydraulic performance of these heat sinks often relay on reducing cell size and increasing porosity. While effective, these approaches compromise cell wall thickness, leading to diminished mechanical properties and challenges in additive manufacturing. Inspired by natural materials, this study introduces a bionic hierarchical design for triply periodic minimal surface structures to address these limitations. The thermo-hydraulic characteristics of a novel class of heat sinks based on hierarchical structures were analyzed and compared with those of single-scale lattices. Hierarchical structures demonstrated more complex fluid flow patterns, including a greater number of high-velocity vortices, which promote enhanced mixing and heat transfer. Comparative analysis revealed that, while single-scale structures with smaller cell sizes achieved superior heat transfer performance at equivalent porosity, hierarchical structures offered significant advantages by reducing friction factors. Moreover, hierarchical structures achieved overall thermal-hydraulic performance comparable to that of single-scale structures while enabling thicker cell walls, which improve mechanical strength and reduce manufacturing precision demands. Hierarchical designs with larger overall porosities and cell size ratios exhibited particularly superior overall thermal-hydraulic performance. Additionally, deep reinforcement learning was employed to optimise the the hierarchical lattice design. The friction factor and j-factor were used to evaluate hydrodynamic and heat performances, respectively. Optimisation results indicated that the hierarchical Primitive structure reduced the friction factor by 14.7%, improved the j-factor by 46.3%, and increased the wall thickness by 165.4%. Similarly, the hierarchical I-WP structure achieved a 40.3% reduction in friction factor, a 17.0% improvement in j-factor, and a 119.3% increase in wall thickness. These findings highlight the potential of hierarchical TPMS designs to optimise thermal and mechanical performance while enhancing manufacturability.
引用
收藏
页数:22
相关论文