基于格子Boltzmann方法表面形貌对微通道对流换热的影响

Effects of Surface Morphology on Convection Heat Transfer in Microchannels Based on Lattice Boltzmann Method

关键词:  格子Boltzmann方法  微通道  对流换热  表面形貌  分形理论

Key words:  Lattice Boltzmann method  Microchannel  Convective heat transfer  Surface topography  Fractal theory

摘要:

为推动微通道冷却技术在航空发动机中的应用，进一步提高航空发动机性能，通过扫描电子显微镜、粗糙度扫描仪等多方式扫描得到实际加工物体表面形貌，结合分形理论进行处理，得到真实、光滑、分形插值、康托尔集四类圆形截面管模型，管径100~400μm；采用格子Boltzmann方法进行数值模拟，首先通过与文献中实验结果的对比，验证了此方法的正确性，然后对空气在这四类圆形截面微通道中的对流换热特性进行了数值模拟，计算雷诺数Re=80~640。结果发现：表面形貌是微通道中对流换热不可忽略的因素，相对粗糙度越大越有利于换热，真实扫描管的换热性能要比相同条件下光滑管高2.22%；分形理论可用于微通道表面形貌的构建，在相同迭代次数下，分形插值模型与真实扫描形貌换热性能相差0.03%，而康托尔集模型与真实扫描形貌换热性能相差1.88%，分形插值模型较康托尔集模型更能真实地反映实际形貌；微通道中，随着雷诺数的增加Nu将不再是常数，而是有增加的趋势。

Abstract:

To promote the application of microchannel cooling technology in aero engine and further improve the performance of the aero engine， the surface morphology of the actual processed object was obtained by scanning electron microscopy， roughness scanner and other methods. Combined with the fractal theory， four types of circular cross-section tube models were obtained， such as real tube， smooth tube， fractal interpolation tube and cantor set tube. The diameter range of the tubes was 100~400μm. The Lattice Boltzmann method was used for numerical simulation. Firstly， the correctness of the method was verified by comparison with the experimental results in the literature. Then， the convective heat transfer characteristics of air in these four types of circular cross-section microchannels were numerically simulated， and the Re range was 80~640. The results show that surface morphology is a factor that cannot be neglected in convective heat transfer in microchannels， the larger the relative roughness， the better the heat transfer， and the heat transfer performance of the real scanning tube is 2.22% higher than that of the smooth tube under the same conditions. The fractal theory can be used to construct the surface morphology of microchannels. However， the fractal interpolation model can more accurately reflect the actual shape than the Cantor set model under the same number of iterations. The heat transfer performance of the fractal interpolation model differs from the real scan topography by 0.03%， while the heat transfer performance of the Cantor set model differs from the real scan topography by 1.88%. In the microchannel， as the Reynolds number increases， Nu will no longer be a constant， but has an increasing trend.