Digital holographic microscopy enables the recording of sample holograms which contain 3D volumetric information. However, additional optical elements, such as partially or fully coherent light source and a pinhole, are required to induce diffraction and interference. Here, we present a deep neural network based on generative adversarial network (GAN) to perform image transformation from a defocused bright-field (BF) image acquired from a general white light source to a holographic image. Training image pairs of 11,050 for image conversion were gathered by using a hybrid BF and hologram imaging technique. The performance of the trained network was evaluated by comparing generated and ground truth holograms of microspheres and erythrocytes distributed in 3D. Holograms generated from BF images through the trained GAN showed enhanced image contrast with 3–5 times increased signal-to-noise ratio compared to ground truth holograms and provided 3D positional information and light scattering patterns of the samples. The developed GAN-based method is a promising mean for dynamic analysis of microscale objects with providing detailed 3D positional information and monitoring biological samples precisely even though conventional BF microscopic setting is utilized.

Unsteady flow fields over a circular cylinder are used for training and then prediction using four different deep learning networks: generative adversarial networks with and without consideration of conservation laws; and convolutional neural networks with and without consideration of conservation laws. Flow fields at future occasions are predicted based on information on flow fields at previous occasions. Predictions of deep learning networks are made for flow fields at Reynolds numbers that were not used during training. Physical loss functions are proposed to explicitly provide information on conservation of mass and momentum to deep learning networks. An adversarial training is applied to extract features of flow dynamics in an unsupervised manner. Effects of the proposed physical loss functions and adversarial training on predicted results are analysed. Captured and missed flow physics from predictions are also analysed. Predicted flow fields using deep learning networks are in good agreement with flow fields computed by numerical simulations.

Convolutional neural networks (CNNs) have recently been applied to predict or model fluid dynamics. However, mechanisms of CNNs for learning fluid dynamics are still not well understood, while such understanding is highly necessary to optimize the network or to reduce trial-and-errors during the network optmization. In the present study, a CNN to predict future three-dimensional unsteady wake flow using flow fields in the past occasions is developed. Mechanisms of the developed CNN for prediction of wake flow behind a circular cylinder are investigated in two flow regimes: the three-dimensional wake transition regime and the shear-layer transition regime. Feature maps in the CNN are visualized to compare flow structures which are extracted by the CNN from flow at the two flow regimes. In both flow regimes, feature maps are found to extract similar sets of flow structures such as braid shear-layers and shedding vortices. A Fourier analysis is conducted to investigate mechanisms of the CNN for predicting wake flow in flow regimes with different wave number characteristics. It is found that a convolution layer in the CNN integrates and transports wave number information from flow to predict the dynamics. Characteristics of the CNN for transporting input information including time histories of flow variables is analyzed by assessing contributions of each flow variable and time history to feature maps in the CNN. Structural similarities between feature maps in the CNN are calculated to reveal the number of feature maps that contain similar flow structures. By reducing the number of feature maps that contain similar flow structures, it is also able to successfully reduce the number of parameters to learn in the CNN by 85\% without affecting prediction performances.

A computational methodology, which combines a computational fluid dynamics (CFD) technique and a computational structural dynamics (CSD) technique, is employed to design a deformable foil whose kinematics is inspired by the propulsive motion of the fin or the tail of a fish or a cetacean. The unsteady incompressible Navier–Stokes equations are solved using a second-order accurate finite difference method and an immersed-boundary method to effectively impose boundary conditions on complex moving boundaries. A finite element-based structural dynamics solver is employed to compute the deformation of the foil due to interaction with fluid. The integrated CFD–CSD simulation capability is coupled with a surrogate management framework (SMF) for nongradient-based multivariable optimization in order to optimize flapping kinematics and flexibility of the foil. The flapping kinematics is manipulated for a rigid nondeforming foil through the pitching amplitude and the phase angle between heaving and pitching motions. The flexibility is additionally controlled for a flexible deforming foil through the selection of material with a range of Young's modulus. A parametric analysis with respect to pitching amplitude, phase angle, and Young's modulus on propulsion efficiency is presented at Reynolds number of 1100 for the NACA 0012 airfoil.

A computational methodology, which combines a computational fluid dynamics (CFD) technique and a computational structural dynamics (CSD) technique, is employed to design a deformable foil whose kinematics is inspired by the propulsive motion of the fin or the tail of a fish or a cetacean. The unsteady incompressible Navier–Stokes equations are solved using a second-order accurate finite difference method and an immersed-boundary method to effectively impose boundary conditions on complex moving boundaries. A finite element-based structural dynamics solver is employed to compute the deformation of the foil due to interaction with fluid. The integrated CFD–CSD simulation capability is coupled with a surrogate management framework (SMF) for nongradient-based multivariable optimization in order to optimize flapping kinematics and flexibility of the foil. The flapping kinematics is manipulated for a rigid nondeforming foil through the pitching amplitude and the phase angle between heaving and pitching motions. The flexibility is additionally controlled for a flexible deforming foil through the selection of material with a range of Young's modulus. A parametric analysis with respect to pitching amplitude, phase angle, and Young's modulus on propulsion efficiency is presented at Reynolds number of 1100 for the NACA 0012 airfoil.

Tracks of typhoons are predicted using a generative adversarial network (GAN) with satellite images as inputs. Time series of satellite images of typhoons which occurred in the Korea Peninsula in the past are used to train the neural network. The trained GAN is employed to produce a 6-hour-advance track of a typhoon for which the GAN was not trained. The predicted track image of a typhoon favorably identifies the future location of the typhoon center as well as the deformed cloud structures. Errors between predicted and real typhoon centers are measured quantitatively in kilometers. An averaged error of 95.6 km is achieved for tested 10 typhoons. Predicting sudden changes of the track in westward or northward directions is identified as a challenging task, while the prediction is significantly improved, when velocity fields are employed along with satellite images.

Large-eddy simulations (LES) using explicit filtering are performed to obtain grid-independent solutions of turbulent wake flow behind a circular cylinder at ReD = 3900 on non-Cartesian type grids. A differential elliptic equation where the filter kernel is implicitly defined is discretized in an unstructured-grid solver to enable explicit filtering on non-Cartesian grids. The separation of filtering procedure from discretization is known to produce an LES solution of which error is mainly attributed to the capability of a sub-filter scale (SFS) model. Equipped with the differential filter and the Vreman SFS model, explicitly filtered LES on unstructured grids is shown to produce nearly grid-independent solutions for flow over a circular cylinder at a critical Reynolds number.

Turbulent flow over the DrivAer fastback vehicle model is investigated using large-eddy simulation with particular emphasis on flow separation, vortical structures and unsteady quantities. A systematic and detailed analysis of the flow field is made considering rotating wheels and moving ground floor. Overall features of vortical structures at the cowl top, behind side mirrors, near front and back wheels, at A-, B- and C-pillars and behind the rear end of the vehicle are revealed by investigating velocity and vorticity fields. The rear end is identified to be the main contributor to the pressure force acting on the vehicle, followed by back and front wheels and side mirrors. The resulting pressure force on the upper part of the vehicle, including A-, B-, and C-pillars but excluding the cowl top and side mirrors, is found to be only slightly higher than the contribution of the gap in the cowl top. Flow separation and resulting vortices do not only have an impact on automotive drag, it is also pointed out how unsteadiness in the flow field affects pressure fluctuations. High levels of surface pressure fluctuations are found near side mirrors and front wheels. Similarities in distributions of pressure fluctuations and turbulent kinetic energy are found.

Turbulent flow over the DrivAer fastback vehicle model is investigated using large-eddy simulation with particular emphasis on flow separation, vortical structures and unsteady quantities. A systematic and detailed analysis of the flow field is made considering rotating wheels and moving ground floor. Overall features of vortical structures at the cowl top, behind side mirrors, near front and back wheels, at A-, B- and C-pillars and behind the rear end of the vehicle are revealed by investigating velocity and vorticity fields. The rear end is identified to be the main contributor to the pressure force acting on the vehicle, followed by back and front wheels and side mirrors. The resulting pressure force on the upper part of the vehicle, including A-, B-, and C-pillars but excluding the cowl top and side mirrors, is found to be only slightly higher than the contribution of the gap in the cowl top. Flow separation and resulting vortices do not only have an impact on automotive drag, it is also pointed out how unsteadiness in the flow field affects pressure fluctuations. High levels of surface pressure fluctuations are found near side mirrors and front wheels. Similarities in distributions of pressure fluctuations and turbulent kinetic energy are found.

Microreactors are emerging as an efficient, sustainable synthetic tool compared to conventional batch reactors. Here, we present a new numbering-up metal microreactor by integrating a flow distributor and a copper catalytic module for high productivity of a commercial synthetic drug. A flow distributor and an embedded baffle disc were manufactured by CNC machining and 3D printing of stainless steel (S/S), respectively, whereas a catalytic reaction module was composed of 25 copper coiled capillaries configured in parallel. Eventually, the numbering-up microreactor system assembled with functional modules showed uniform flow distribution and high mixing efficiency regardless of clogging, and achieved high-throughput synthesis of the drug “rufinamide”, an anticonvulsant medicine, via a Cu(I)-catalyzed azide–alkyne cycloaddition reaction under optimized conditions.