In the present study, cooling of a hot elastic plate by using combined channel and jet impingement cooling is explored numerically by using finite element method. The operational parameters of the channel cooling system, such as flow rate, have an impact on the lower portion cooling because the distance between the elastic surface and the nozzle varies as a result of the object’s deformation. The coupled interaction of different cooling systems have impacts on the thermal performance of the elastic plate. Channel flow Reynolds number (Rec between 300 and 1500), jet flow Reynolds number (Rej between 100 and 400), elasticity of the hot object (E between and ), vertical distance from the jet to the elastic plate (Hj between 5wj and 10wj; wj being the slot width), and the distance between slots in jet impingement cooling (sj between 3wj and 12wj) are all taken into consideration when conducting the numerical analysis. The upper and lower sections of the elastic plate’s cooling performance are affected by the elasticity of the plate caused by the coupled system’s fluid structure interaction in the jet flow domain. When Rec is increased to its maximum value, the average Nusselt number (Nu) increases by 38.5% and 42%, respectively, for elastic and stiff plates. When the maximum Rec is reached, the lower surface Nu increases by approximately 13.5%. At Rec 1500, where the elasticity is at its maximum, the average Nu increment (decrement) for the upper (lower) section of the plate reaches 19%. When jet impingement cooling parameters are considered, the average Nu increases by roughly 35% and 33% for elastic plate and rigid plate arrangements while cooling performance of the upper plate portion increases by around 4.5% in the elastic case. As the vertical distance from the slot to the plate surface grows, the average Nu for the bottom part of the plate decreases, but the effect of the slot-slot distance on thermal performance is less sensitive. When the distance is raised to H=10wj, the average Nu for the elastic case reduces by roughly 15%, but for the rigid case, it decreases by 20% for the bottom part. With artificial neural network assisted computational fluid dynamics, an effective computational approach is proposed without using fluid–structure interaction. Accurate and fast results are obtained as compared to high fidelity 3D fully coupled computations.

中文翻译：

---

使用三元纳米流体混合通道射流冲击系统冷却热弹性板，并使用 ANN 辅助 CFD 进行高效计算


在本研究中，利用有限元方法对采用组合通道和射流冲击冷却的热弹性板进行冷却进行了数值探索。通道冷却系统的操作参数（例如流量）对下部冷却有影响，因为弹性表面和喷嘴之间的距离会因物体的变形而变化。不同冷却系统的耦合相互作用对弹性板的热性能产生影响。通道流雷诺数（Rec在300到1500之间），射流雷诺数（Rej在100到400之间），热物体的弹性（E在和之间），射流到弹性板的垂直距离（Hj在5wj和10wj之间进行数值分析时，均考虑了射流冲击冷却中的槽缝间距（sj 为 3wj 至 12wj）。弹性板上下段的冷却性能受到耦合系统在射流域中的流固相互作用引起的板弹性的影响。当 Rec 增加到最大值时，弹性板和刚性板的平均努塞尔数 (Nu) 分别增加 38.5% 和 42%。当达到最大Rec时，下表面Nu增加约13.5%。在 Rec 1500 处，弹性达到最大，板上（下）部分的平均 Nu 增量（减少）达到 19%。当考虑射流冲击冷却参数时，弹性板和刚性板布置的平均 Nu 增加大约 35% 和 33%，而弹性情况下上板部分的冷却性能增加约 4.5%。 随着槽到板表面的垂直距离增大，板底部的平均 Nu 减小，但槽与槽距离对热性能的影响不太敏感。当距离增加到 H=10wj 时，弹性情况下的平均 Nu 大约减少了 15%，但对于刚性情况，底部的平均 Nu 减少了 20%。通过人工神经网络辅助计算流体动力学，提出了一种不使用流固耦合的有效计算方法。与高保真 3D 全耦合计算相比，可以获得准确、快速的结果。