We demonstrate several techniques to encourage practical uses of neural networks for fluid flow estimation. In the present paper, three perspectives which are remaining challenges for applications of machine learning to fluid dynamics are considered: 1. interpretability of machine-learned results, 2. bulking out of training data, and 3. generalizability of neural networks. For the interpretability, we first demonstrate two methods to observe the internal procedure of neural networks, i.e., visualization of hidden layers and application of gradient-weighted class activation mapping (Grad-CAM), applied to canonical fluid flow estimation problems -- $(1)$ drag coefficient estimation of a cylinder wake and $(2)$ velocity estimation from particle images. It is exemplified that both approaches can successfully tell us evidences of the great capability of machine learning-based estimations. We then utilize some techniques to bulk out training data for super-resolution analysis and temporal prediction for cylinder wake and NOAA sea surface temperature data to demonstrate that sufficient training of neural networks with limited amount of training data can be achieved for fluid flow problems. The generalizability of machine learning model is also discussed by accounting for the perspectives of inter/extrapolation of training data, considering super-resolution of wakes behind two parallel cylinders. We find that various flow patterns generated by complex interaction between two cylinders can be reconstructed well, even for the test configurations regarding the distance factor. The present paper can be a significant step toward practical uses of neural networks for both laminar and turbulent flow problems.
Curvature estimation from a volume of fluid indicator function for the simulation of surface tension and wetting with a free surface lattice Boltzmann method
New estimations of the added mass and damping of two cylinders vibrating in a viscous fluid, from theoretical and numerical approaches
本文研究了两圆柱在粘性滞止流体中的小振动问题。基于亥姆霍兹展开和双极坐标系,提出了一种新的理论方法来估计作用在两个物体上的流体力。通过粘性流体附加系数,我们证明了这些力是{ \textcolor{black}{cylinder accelerations} }和速度的线性组合。{ \textcolor{black}{To assess the validity of this theory, we consider the case of two equal size cylinders, one of them being stationary while the other one is forced sinusoidally} }。自加质量和阻尼系数都随着Stokes数和分离距离的增加而减小。交叉附加质量和阻尼系数随Stokes数和间距的增大而增大。与无粘性结果相比,粘性的影响是增加了一个校正项,其标度为$Sk^{-1/2}$。当分离距离足够大时,两个圆柱的行为就好像它们是独立的,并且孤立圆柱的Stokes预测被恢复。与以前的工作相比,本理论提供了一个简单而灵活的替代容易确定的流体力和相关的附加系数。据我们所知,这也是第一次提出了一种数值方法的基础上的惩罚方法的流体-结构相互作用的背景下,相对较小的斯托克斯数,并成功地与理论预测。
On the Hill's Spherical Vortex in Fluid and Plasma, its Generalization, and Stability
1894年,M.J.M.希尔发表了一篇文章,描述了一个球形漩涡移动通过一个固定的流体。使用圆柱坐标并假设方位角速度分量为零,希尔找到了描述这种流动的简单解。1987年,A. A. Bobnev和R. Kaiser和D.洛茨谁适用于设置模型的球照明。我们提出了一个更简单的推导希尔的球形涡使用布拉格-霍桑方程。特别地,通过使用移动参考系,欧拉方程简化为平衡流动,其等价于直到重新标记的静力平衡MHD方程。一个新的广义版本的希尔的球形涡与一个非零的方位角分量的推导。静力平衡MHD方程的物理解是通过在球坐标系中求解Grad-Shafranov方程来计算的。最后,希尔的球形涡的稳定性进行检查,通过执行轴对称扰动描述;它表明,希尔的球形涡是线性不稳定的,相对于某些种类的小扰动。
Well-Posed Two-Temperature Constitutive Equations for Stable Dense Fluid Shockwaves using Molecular Dynamics and Generalizations of Navier-Stokes-Fourier Continuum Mechanics
Machine-learned control-oriented flow estimation for multiactuator multi-sensor systems exemplified for the fluidic pinball
我们提出了第一个机器学习的面向控制的多输入多输出工厂的流量估计。起始点是具有开环致动命令的恒定致动,该开环致动命令导致具有同时记录的致动命令、传感器信号和流场的数据库。关键使能器是包括传感器信号和致动命令的估计器输入向量。从传感器信号和致动命令到流场的映射以分析简单的、以数据为中心的和一般的非线性方法实现。分析简单的估计推广线性随机估计(LSE)的驱动命令。以数据为中心的方法通过从数据库中插值来从估计器输入中产生流场-类似于Loiseau et al.(2018)的非强迫流。使用k个最近邻(kNN)执行插值。通过迭代训练方法从深度神经网络(DNN)获得从输入到流场的一般全局非线性映射。对流体弹球装置进行了估计器比较,这是一种多输入、多输出尾流控制基准(Deng等人,2020),在稳定控制下具有丰富的动态特性。我们得出结论,机器学习方法明显优于线性模型。对于周期动态,kNN和DNN估计器的性能相当。然而,当流是混沌的时候,DNN的表现总是更好。此外,一个彻底的比较关于复杂性,计算成本,和预测精度,以证明每个估计的相对优点。所提出的方法可以推广到闭环流量控制对象。
Computational hemodynamics in arteries with the one-dimensional augmented fluid-structure interaction system: viscoelastic parameters estimation and comparison with in-vivo data