This book describes and explains the fundamental physical processes involved in bubble dynamics and the phenomenon of cavitation. It is intended as a combination of a reference book for those scientists and engineers who work with cavitation or bubble dynamics and as a monograph for advanced students interested in some of the basic problems associated with this category of multiphase flows. A basic knowledge of fluid flow and heat transfer is assumed but otherwise the analytical methods presented are developed from basic principles.

The book begins with a chapter on nucleation and describes both the theory and observations of nucleation in flowing and non-flowing systems. The following three chapters provide a systematic treatment of the dynamics of the growth, collapse or oscillation of individual bubbles in otherwise quiescent liquids. Chapter 4 summarizes the state of knowledge of the motion of bubbles in liquids. Chapter 5 describes some of the phenomena which occur in homogeneous bubbly flows with particular emphasis on cloud cavitation and this is followed by a chapter summarizing some of the experiemntal observations of cavitating flows. The last chapter provides a review of the free streamline methods used to treat separated cavity flows with large attached cavities.

/pdf/cavitation-and-bubble-dynamics-1vfop84jck.pdf

Cavitation and Bubble Dynamics

In recent years, many accurate decision support systems have been constructed as black boxes, that is as systems that hide their internal logic to the user. This lack of explanation constitutes both a practical and an ethical issue. The literature reports many approaches aimed at overcoming this crucial weakness, sometimes at the cost of sacrificing accuracy for interpretability. The applications in which black box decision systems can be used are various, and each approach is typically developed to provide a solution for a specific problem and, as a consequence, it explicitly or implicitly delineates its own definition of interpretability and explanation. The aim of this article is to provide a classification of the main problems addressed in the literature with respect to the notion of explanation and the type of black box system. Given a problem definition, a black box type, and a desired explanation, this survey should help the researcher to find the proposals more useful for his own work. The proposed classification of approaches to open black box models should also be useful for putting the many research open questions in perspective.

https://dl.acm.org/doi/pdf/10.1145/3236009

A Survey of Methods for Explaining Black Box Models

Image representations, from SIFT and Bag of Visual Words to Convolutional Neural Networks (CNNs), are a crucial component of almost any image understanding system. Nevertheless, our understanding of them remains limited. In this paper we conduct a direct analysis of the visual information contained in representations by asking the following question: given an encoding of an image, to which extent is it possible to reconstruct the image itself? To answer this question we contribute a general framework to invert representations. We show that this method can invert representations such as HOG more accurately than recent alternatives while being applicable to CNNs too. We then use this technique to study the inverse of recent state-of-the-art CNN image representations for the first time. Among our findings, we show that several layers in CNNs retain photographically accurate information about the image, with different degrees of geometric and photometric invariance.

/pdf/understanding-deep-image-representations-by-inverting-them-1od2jxrw8o.pdf

Understanding deep image representations by inverting them

Cavitating flows entail phase change and hence very large and steep density variations in the low pressure regions. These are also very sensitive to: (a) the formation and transport of vapor bubbles, (b) the turbulent fluctuations of pressure and velocity, and (c) the magnitude of noncondensible gases, which are dissolved or ingested in the operating liquid. The presented cavitation model accounts for all these first-order effects, and thus is named as the full cavitation model. The phase-change rate expressions are derived from a reduced form of Rayleigh-Plesset equation for bubble dynamics. These rates depend upon local flow conditions (pressure, velocities, turbulence) as well as fluid properties (saturation pressure, densities, and surface tension). The rate expressions employ two empirical constants, which have been calibrated with experimental data covering a very wide range of flow conditions, and do not require adjustments for different problems. The model has been implemented in an advanced, commercial, general-purpose CFD code, CFD-ACE+

Mathematical Basis and Validation of the Full Cavitation Model

Image representations, from SIFT and Bag of Visual Words to Convolutional Neural Networks (CNNs), are a crucial component of almost any image understanding system. Nevertheless, our understanding of them remains limited. In this paper we conduct a direct analysis of the visual information contained in representations by asking the following question: given an encoding of an image, to which extent is it possible to reconstruct the image itself? To answer this question we contribute a general framework to invert representations. We show that this method can invert representations such as HOG and SIFT more accurately than recent alternatives while being applicable to CNNs too. We then use this technique to study the inverse of recent state-of-the-art CNN image representations for the first time. Among our findings, we show that several layers in CNNs retain photographically accurate information about the image, with different degrees of geometric and photometric invariance.

Understanding Deep Image Representations by Inverting Them

For modeling the 3D world behind 2D images, which 3D representation is most appropriate? A polygon mesh is a promising candidate for its compactness and geometric properties. However, it is not straightforward to model a polygon mesh from 2D images using neural networks because the conversion from a mesh to an image, or rendering, involves a discrete operation called rasterization, which prevents back-propagation. Therefore, in this work, we propose an approximate gradient for rasterization that enables the integration of rendering into neural networks. Using this renderer, we perform single-image 3D mesh reconstruction with silhouette image supervision and our system outperforms the existing voxel-based approach. Additionally, we perform gradient-based 3D mesh editing operations, such as 2D-to-3D style transfer and 3D DeepDream, with 2D supervision for the first time. These applications demonstrate the potential of the integration of a mesh renderer into neural networks and the effectiveness of our proposed renderer.

/pdf/neural-3d-mesh-renderer-25u8r7x1rr.pdf

Neural 3D Mesh Renderer

A new cavity model that can explain the interaction between viscous effects including vortices and cavitation bubbles is presented in this study. This model, which is named a bubble two-phase flow (BTF) model, treats the inside and outside of a cavity as one continuum by regarding the cavity as a compressible viscous fluid whose density changes greatly. Navier–Stokes equations including cavitation bubble clusters are solved in finite-difference form by a time-marching scheme, where the growth and collapse of a bubble cluster is given by a modified Rayleigh's equation. Computation was made on a two-dimensional flow field around a hydrofoil NACA0015 at angles of attack of 8° and 20°. The Reynolds number was 3 × 105. The experiments were also performed at the same Reynolds number for comparison. The computed results by the BTF cavity model can express the feature of cloud-type cavitation shed from the trailing edge of the attached cavities when the angle of attack is 8°. It shows the mechanism of cavitation cloud generation and large-scale vortices. The boundary layer separates at the cavity leading edge. Then it rolls up and produces the cavitation cloud. In other words, the instability of the shear layer may produce the cavitation cloud. When the angle of attack is 20°, the flow was fully separated from the leading edge of the hydrofoil and vortex cavitation occurs in the separated region. The BTF cavity model can also express the generation of such vortex cavitation and the effect of cavitation nuclei in the uniform flow.

https://www.cambridge.org/core/services/aop-cambridge-core/content/view/S002211209200003X

A new modelling of cavitating flows : a numerical study of unsteady cavitation on a hydrofoil section

Generation mechanism of cloud cavitation on a hydrofoil section was investigated in a sequence of experiments through observation of cloud cavitation by high-speed video and high-speed photo as well as pressure measurements by pressure pick-ups and a hydrophone. The mechanism was also investigated by controlling cloud cavitation with an obstacle fitted on the foil surface. From the results of these experiments, it was found that the collapse of a sheet cavity is triggered by a re-entrant jet rushing from the trailing edge to the leading edge of the sheet cavity, and consequently, the sheet cavity is shed in the vicinity of its leading edge and thrown downstream as a cluster of bubbles called cloud cavity. In other words, the re-entrant jet gives rise to cloud cavitation. Moreover, cloud cavitation could be controlled effectively by a small obstacle placed on the foil. It resulted in reduction of foil drag and cavitation noise

Hiroharu Kato

Papers

Neural 3D Mesh Renderer

A new modelling of cavitating flows : a numerical study of unsteady cavitation on a hydrofoil section

Neural 3D Mesh Renderer

Mechanism and Control of Cloud Cavitation

Unsteady Structure Measurement of Cloud Cavitation on a Foil Section Using Conditional Sampling Technique