Aiming at the problems of a long design period and imperfect surrogate modeling in the field of airfoil design optimization, a convolutional neural network framework for airfoil design and performance prediction (DPCNN) is established based on the deep learning method. The airfoil profile parameterization, physical field prediction, and performance prediction are achieved. The results show that the DPCNN framework can generate substantial perfect airfoil profiles with only three geometric parameters. It has significant advantages such as good robustness, great convergence, fast computation speed, and high prediction accuracy compared with the conventional machine learning method. When the train size is 0.1, the predicted results can be obtained within 5 ms. The prediction absolute errors of physical field of most sample points are lower than 0.002, and the relative errors of aerodynamic performance parameters are lower than 2.5%. Finally, the optimization attempt of operating parameters is completed by gradient descent method, which shows good stability and convergence. Overall, the DPCNN framework in this paper has outstanding advantages in time cost and prediction accuracy.

With the rapid development of microelectromechanical systems (MEMS) in IT industry, the heat flux in microchannel has reached a high level which demands preferable cooling technology. Water cooling has become a favor cooling approach in electronic microdevices due to better thermal performance than air cooling method. In the present paper, thermal performance in microchannels with grooves and obstacles are investigated numerically. The height and width of the rectangular microchannel are 200 and 50 μm, respectively. As a simple modification of dimple/protrusion, the groove/obstacle diameter is 100 μm and the depth is 20 μm. Different arrangements of grooves and obstacles are considered on Reynolds range of 100–900. The numerical results show that groove/obstacle structure is effective for cooling enhancement in microchannel. Among the cases in this research, the normalized Nusselt number Nu/Nu₀ is within the range of 1.446–26.19, while the pressure penalty f/f₀ has a much larger range from 0.86 to 110.18 depending on specific orientation. Field synergy analysis and performance evaluation plot are adopted to discuss the mechanism of heat transfer enhancement and energy saving performance integrating the pumping performance. From the viewpoint of energy saving, groove on single surface (case 1) has the best performance. Furthermore, performances of grooved microchannels are compared with that of dimpled microchannels which were discussed in the author’s previous research. The results indicate grooved microchannels have larger range of both Nu/Nu₀ and f/f₀ and some grooved cases possess high TP than dimpled microchannels.

The heat transfer enhancement and entropy generation of Al₂O₃-water nanofluids laminar convective flow in the microchannels with flow control devices (cylinder, rectangle, protrusion, and v-groove) were investigated in this research. The effects of the geometrical structure of the microchannel, nanofluids concentration φ(0%–3%), and Reynolds number Re (50–300) were comparatively studied by means of performance parameters, as well as the limiting streamlines and temperature contours on the modified heated surfaces. The results reveal that the relative Fanning frictional factor f/f₀ of the microchannel with rectangle and protrusion devices are much larger and smaller than others, respectively. As the nanofluids concentration increases, f/f₀ increases accordingly. For the microchannel with rectangle ribs, there is a transition Re for obtaining the largest heat transfer. The relative Nusselt number Nu/Nu₀ of the cases with larger nanofluids concentration are greater. The microchannels with cylinder and v-groove profiles have better heat transfer performance, especially at larger Re cases, while, the microchannel with the protrusion devices is better from an entropy generation minimization perspective. Furthermore, the variation of the relative entropy generation S′/S′₀ are influenced by not only the change of Nu/Nu₀ and f/f₀, but also the physical parameters of working substances.