The visualization techniques for the aeroacoustics field of a supersonic jet flow are introduced and its capability is discussed by comparing with the characteristics of aeroacoustics fields acquired by computations. Aeroacoustics fields of supersonic jets (Re=10⁵, 10⁶) were visualized by means of the particle image velocimetry (PIV) and the schlieren. The single-pixel ensemble correlation was applied for PIV images to obtain the time-averaged velocity distribution with high spatial resolution. The flow fluctuation which correlates with the Mach wave emission can be observed using a proper orthogonal decomposition (POD) of schlieren images because the POD analysis can eliminate the noise modes of schlieren images. In addition, frequency-domain POD analysis of unsteady schlieren images can visualize the propagation pattern, sound pressure level, and propagation direction of acoustic waves whose peak frequency is clearly observed.

A number of demonstrative results in use of ultrasound for multiphase flows are shown. Feature of the technique is of capturing fluid interfaces in flowing state with time-resolved signal processing. For gas-liquid two-phase flows, methods of measurement for void fraction, interfaces, velocity of each phase, and volume flow rate are explained. As these extension, three-phase flow and turbidity current are also measured.

Drag reduction mechanism in turbulent flow field with polymer additives is investigated by simultaneous particle image velocimetry (PIV) / laser-induced fluorescence (LIF) measurement. Fluorescently labeled polymer is synthesized by chemically bonding fluorescent material to a polymer to realize precise visualization of the polymer and used in the measurement. Polymer area is clearly observed by the LIF measurement. By combining the result of LIF measurement with the flow field calculated from the PIV measurement, the influence of the polymers on the flow field is investigated. Turbulent energy is relatively high in the polymer areas and decreases with the distance from the polymer areas. Thus, the interaction between the polymer and flow field is experimentally revealed.

We have developed a multiscale multiphase flow simulation model and a multiscale weather simulation model. Those models have been applied to high-resolution simulations of atmosphere and cloud turbulence. The obtained large-size dataset has been visualized particularly in human’s eye-view. Such visualization images match with the snapshot camera images, which have been analyzed by means of the deep neural network technology, i.e., deep-learning technology. We show that those visualizations can bridge the simulation and data sciences.

Visualization by using reflective flakes enables us to see the qualitative feature of flow. It is, however, difficult to obtain quantitative information based on the visualized pattern. In this paper, we summarize the fundamental knowledge of flow visualizations by using reflective flakes. Main conclusions are followings. (1) The visualized pattern is due to the inhomogeneity of the orientation of flakes. (2) The temporal evolution of the orientation of a flake is the same as the one of a material surface element at the same location. (3) The orientation of a flake depends, in general, on its initial orientation. In turbulence, however, it is independent of the initial condition and it is a function of the position and time. More precisely, even if initial orientations are randomized at a position, they tend to a single orientation after the duration of about 25 Kolmogorov times. (4) Since the temporal evolution of flake orientation is determined by the velocity gradients, it is strongly affected by Kolmogorov-scale coherent vortices in turbulence. Therefore, we can subtract the quantitative information of these vortices from visualized patterns.