1. Cooling of a hot elastic plate by using hybrid channel-jet impingement system with ternary nanofluid and efficient computations by using ANN assisted CFD.
- Author
-
Selimefendigil, Fatih and Oztop, Hakan F.
- Subjects
- *
ARTIFICIAL neural networks , *ELASTIC plates & shells , *COMPUTATIONAL fluid dynamics , *JETS (Fluid dynamics) , *NUSSELT number , *JET impingement - Abstract
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 1 0 5 and 1 0 8 ), 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. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF