论文标题
用于设计拖放的多项式框架，以减少周期性的二维纹理表面
A Polynomial Framework for Design of Drag Reducing Periodic Two-dimensional Textured Surfaces
论文作者
论文摘要
储层计算是预测湍流的有力工具，其简单的架构具有处理大型系统的计算效率。然而，其实现通常需要完整的状态向量测量和系统非线性知识。我们使用非线性投影函数将系统测量扩展到高维空间，然后将其输入到储层中以获得预测。我们展示了这种储层计算网络在时空混沌系统上的应用，该系统模拟了湍流的若干特征。我们表明，使用径向基函数作为非线性投影器，即使只有部分观测并且不知道控制方程，也能稳健地捕捉复杂的系统非线性。最后，我们表明，当测量稀疏、不完整且带有噪声，甚至控制方程变得不准确时，我们的网络仍然可以产生相当准确的预测，从而为实际湍流系统的无模型预测铺平了道路。
Periodic and symmetric two-dimensional textures with various cross-sectional profiles have been employed to improve and optimize the physical response of the surfaces such as drag force, superhydrophobicity, and adhesion. While the effect of the height and spacing of the textures have been extensively studied, the effect of the shape of the textures has only been considered in qualitative manners. Here, a polynomial framework is proposed to mathematically define the cross-sectional profiles of the textures and offer a quantitative measure for comparing the physical response of the textured surfaces with various cross-sectional profiles. As a case study, textured surfaces designed with this framework are tested for their hydrodynamic frictional response in the cylindrical Couette flow regime in Taylor- Couette flows. With the reduction in torque as the objective, experimental and numerical results confirm that textures with height-to-half-spacing of lower and equal to unity with concave profiles offer a lower torque compared to both smooth surfaces and triangular textures. In addition, across multiple polynomial orders, textures defined by second order polynomials offer a wide range of responses, eliminating the need for considering polynomials of higher orders and complexity. While the case study here is focused on the laminar flow regime and the frictional torque, the same type of analysis can be applied to other surface properties and physical responses as well.
