Constructal T-shaped porous fins transfer better heat compared to the rectangular counterparts by improving the heat flow through the low resistive links. This type of fins can be used in aerospace engines which demand faster removal of heat without adding extra weight of the overall assembly. Here, in this study, three powerful nature-inspired metaheuristic algorithms such as particle swarm optimization, gravitational search algorithm, and Firefly algorithm have been used to optimize the dominant thermo physical as well as geometric parameters which are responsible for transferring heat at faster rates from the fin body satisfying a volume constraint. The temperature distribution along the stem and the flange has been plotted, and the effect of important parameters on the efficiency has been determined. Three different volumes are selected for the analysis, and the results have shown marked improvement in the optimized heat transfer rate. Particle swarm optimization has reported an increase of 0.81%, while Firefly algorithm reports 0.83% improvement as we increase the fin volume from 500 to 1000 and 0.19% (by PSO) and 0.4% (by FA) as the volume increases from 1000 to 1500. The paper also presents a scheme of reducing the computational effort required by the algorithms to converge around the optimum point. While a reduction of 14.36% computational effort has been achieved in particle swarm optimization’s convergence time, Firefly algorithm took 24.64% less time to converge at the near-optimum point. While particle swarm optimization has converged at better optimal points compared to Firefly algorithm and Gravitational search algorithm, Gravitational search algorithm has outperformed the two algorithms in terms of computational time. Gravitational search algorithm took 61.72 and 29.33% less time to converge as compared to particle swarm optimization and Firefly algorithm, respectively.

中文翻译：

---

对流环境下构造T形多孔翅片的群智能算法优化

与矩形对应物相比，结构性 T 形多孔翅片通过改善通过低电阻连接的热流来传递更好的热量。这种类型的翅片可用于航空发动机，这些发动机需要更快地散热，而不会增加整体组件的额外重量。在这里，在本研究中，三种强大的受自然启发的元启发式算法，如粒子群优化、引力搜索算法和萤火虫算法，已被用于优化主要的热物理和几何参数，这些参数负责以更快的速度传输热量。鳍体满足体积约束。沿着阀杆和法兰的温度分布已经绘制出来，并且重要参数对效率的影响已经确定。选择了三种不同的体积进行分析，结果显示优化的传热速率有显着改善。粒子群优化报告了 0.81% 的增加，而 Firefly 算法报告了 0.83% 的改进，因为我们将鳍体积从 500 增加到 1000 和 0.19%（通过 PSO）和 0.4%（通过 FA）随着体积从 1000 增加到 1500论文还提出了一种减少算法收敛于最佳点所需的计算工作量的方案。虽然粒子群优化的收敛时间减少了 14.36% 的计算量，但 Firefly 算法在接近最优点收敛的时间减少了 24.64%。虽然与萤火虫算法和引力搜索算法相比，粒子群优化已经收敛到更好的最佳点，引力搜索算法在计算时间上优于这两种算法。与粒子群优化和萤火虫算法相比，引力搜索算法的收敛时间分别减少了 61.72% 和 29.33%。