We present a novel method for example-based simulation of fluid flow. We reconstruct fluid animation from physically based fluid simulation examples. Our framework shows how to decompose a given series of fluid motion example data into small units and then recompose them. We capture the properties of local fluid behavior by dicing the fluid motion example data into sequences of fragments, which have smaller volume and shorter length. We build a database out of these fragments, and propose a matching strategy to generate new fluid animation. To achieve highly efficient database query, we project our fragments onto lower dimensional subspace using Principal Component Analysis (PCA) approach, and construct our data structure as a kd-tree by treating each fragment as a point in this subspace. Our method has been implemented in synthesizing both two-dimensional (2D) and three-dimensional (3D) fluid’s velocity fields.

Fluid simulations for computer graphics applications have attracted the attention of many researchers and practitioners due to the enhanced realism that natural phenomena simulation adds to graphical applications. Vortex methods are receiving increasing attention from the computer graphics community for simple and direct modeling of complex flow phenomena such as turbulence. However, vortex methods have not been developed yet to the level of other techniques for fluid simulation in computer graphics. In this work we present a novel simulation framework to model inviscid flows using Lagrangian vortex particle methods. We introduce novel stable methods to solve the vorticity flow equations that produce highly detailed visual fluid simulations. We incorporate the full interplay of solids and fluids in our framework. The coupling between free-form solids, represented by arbitrary surface meshes and fluids simulated with vortex methods, leads to visually rich simulations. Previous vortex simulators only focus on modeling the solid as a boundary for the flow. We model solid boundaries using an extended potential flow at the solid surface coupled with a boundary layer simulation. This allows the accurate simulation of two processes of visual interest. The first is the introduction of surface vorticity in the main flow as turbulence (vortex shedding). The second is the motion of the solid induced by fluid forces, which is calculated from the dynamics of vorticity in the flow and the rate of vorticity creation at solid surfaces. We demonstrate high quality results of our methods simulating flows around solid objects and solid object propulsion due to flows. This work ameliorates one of the important omissions in the development of vortex methods for computer graphics, which is the simulation of two-way coupling of solids and fluids.

[pt] Neste trabalho foi feito um estudo investigativo sobre animação de fluidos utilizando sistemas de partículas. Baseado nas propostas apresentadas por Muller et al., esta dissertação objetiva investigar e compreender o uso do método Lagrangeano baseado em partículas, conhecido como Smoothed Particle Hydrodynamics (SPH), para simulação de fluidos. A validação do método foi feita através da implementação de uma biblioteca capaz de animar fluidos a taxas interativas. Para testar a eficácia e eficiência do método, a biblioteca desenvolvida permite a instanciação de diferentes configurações, incluindo o tratamento de colisões do fluido com obstáculos, o tratamento da interação entre dois fluidos distintos e o tratamento de forças externas exercidas pelo usuário via um mecanismo de interação. / [en] This work investigates the use of particle-based system for fluid animation. Based on proposals presented by Müller et al., the goal of this dissertation is to investigate and fully understand the use of a Lagrangian method known as Smoothed Particle Hydrodynamics (SPH) for fluid simulations. A library has been implemented in order to validate the method for fluid animation at interactive rate. To demonstrate the method effectiveness and efficiency, the resulting library allows the instantiation of different configurations, including the treatment of fluid-obstacle collisions, interaction between two distinct fluids, and fluid-user interaction.

This research develops an extendible framework for reproducing watercolor in a digital environment, with a focus on interactivity using the GPU. The framework uses the lattice Boltzmann method, a relatively new approach to fluid dynamics, and the Kubelka-Munk reflectance model to capture the optical properties of watercolor. The work is demonstrated through several paintings produced using the system.

Fluid simulations are computationally intensive and therefore time consuming and expensive. In the field of visual effects, it is imperative that artists be able to efficiently move through iterations of the simulation to quickly converge on the desired result. One common fluid simulation technique is the Smoothed Particle Hydrodynamics (SPH) method. This method is highly parellelizable. I have implemented a method to integrate a Graphics Processor Unit (GPU) accelerated SPH capability into the 3D software package Houdini. This helps increase the speed with which artists are able to move through these iterations. This approach is extendable to allow future accelerations of the algorithm with new SPH techniques. Emphasis is placed on the infrastructure design so it can also serve as a guideline for both GPU programming and integrating custom code with Houdini.

Art-directability is a crucial aspect of creating aesthetically pleasing visual effects that help tell stories. A particularly common method of art direction is the retiming of a simulation. Unfortunately, the means of retiming an existing simulation sequence which preserves the desired shapes is an ill-defined problem. Naively interpolating values between frames leads to visual artifacts such as choppy frames or jittering intensities. Due to the difficulty in formulating a proper interpolation method we elect to use a machine learning approach to approximate this function. Our model is based on the ODE-net structure and reproduces a set of desired time samples (in our case equivalent to time steps) that achieves the desired new sequence speed, based on training from frames in the original sequence. The flexibility of the updated sequences' duration provided by the time samples input makes this a visually effective and intuitively directable way to retime a simulation.

This thesis presents a method based on Smoothed Particle Hydrodynamics to simulate sparse particle systems with fluid like properties in real-time. The simulation supports interactions with terrain and objects and is scaled depending on activity of the fluid. We use a carpet method on the GPU to visualize the water surface with translucency, reflection, refraction and added topology. Splash effects and foam are imitated and added as a last step.

Two of the methods common in video game fluid simulation are SmoothedParticle Hydrodynamics (SPH), and Position Based Dynamics (PBD). They are both Lagrangian methods of fluid simulation. SPH has been used for many years in offline simulations and has truthful visuals, but is not as stable as the newer method PBD when using larger timesteps. SPH also tends to become unstable during compression. In this report both methods have been tested on different scenarios as the methods’ performance and visual depend on the scenario used. Additionally, the size of the particle radius was varied when comparing Compressible SPH (CSPH), Weak Compressible SPH (WCSPH), and PBD. From these tests, the conclusion could be drawn that CSPH performed slightly better than PBD regarding frames per second (FPS) in all cases except one. However, WCSPH and sometimes CSPH had stability issues. The stability of PBD and its possibility for larger timesteps with only minor FPS difference lead to the conclusion that PBD is overall the more suitable method for fluid simulation in video games. / Två av metoderna som är vanliga vid vätskesimulering i videospel är SmoothedParticle Hydrodynamics (SPH) och Position Based Dynamics (PBD). De är båda Lagrangiska metoder för vätskesimulering. SPH har använts i många år i offline-simuleringar och har realistiskt utseende, men är inte lika stabil som den nyare metoden PBD vid användning av större tidssteg. SPH tenderar också att bli instabil under kompression. Båda metoderna blev testade i olika scenarion eftersom deras prestanda och utseende beror på det använda scenariot. Storleken av partikelradien har också varierat när Compressible SPH (CSPH), Weak Compressible SPH (WCSPH) och PBD jämfördes. Från dessa tester kunde man se att CSPH presterade lite bättre än PBD gällande bilder per sekund (FPS) i alla fall utom ett. Däremot hade WCSPH och ibland CSPH stabilitetsproblem. Stabiliteten av PBD och dess möjlighet att ta större tidssteg med endast minimala FPS skillnader ledde till slutsatsen att PBD är överlag den mer lämpliga metoden för vätskesimulering i videospel.

An understanding of extreme waves is important in the design and analysis of offshore structures, such as oil and gas platforms. With the increase of interest in the shipping of LNG, the design of import and export terminals in coastal water of slowly varying intermediate depth requires accurate analysis of steep wave shoaling. In this thesis, data from laboratory experiments involving random wave simulations on very gentle slopes have been analysed in terms of a model of large wave events, and the results interpreted by observation of the shape and magnitude of the large wave events. The auto-correlation function of the free surface elevation time histories, called NewWave, has been calculated from the wave spectrum and shown to fit very well up to the point where waves start to break (when compared to the ‘linear’ surface elevation time history). It has been shown that NewWave is an appropriate model for the shape of the ‘linear’ part of large waves provided kd > 0.5. A Stokes-like expansion for NewWave analysis has been demonstrated to match the average shape of the largest waves, accounting for the dominant vertical asymmetry. Furthermore, an appropriate local wave period derived from NewWave has been inserted into a Miche-based limiting criterion, using the linear dispersion equation, to obtain estimates for the limiting wave height. Overall, the analysis confirms the Miche-type criterion applies to limiting wave height for waves passing over very mild bed slopes. A derivation of general Green-Naghdi (GN) theory, which incorporates non-linear terms in its formulation, is also presented. This approach satisfies the boundary conditions exactly and approximates the field equations. The derived 2-dimensional vertical GN Level 1 model, capable of simulating steep waves on varying water depth, is validated against solitary waves and their interactions, and solitary waves on varying water depth and gives good qualitative agreement against the KdV equation. The developed and validated numerical model is used to simulate focussed wave groups on both constant depth and gentle slope. In general, the behaviour of waves simulated by the numerical model is very similar to that observed in the experimental data. There is evidence of vertical asymmetry as the water depth is reduced, owing to the non-linearity. Although the main physics is still controlled by linear dispersion, the higher order harmonics become increasingly important for shoaling waves. The numerical results also show a slope-induced wave set-up that keeps on increasing in amplitude as the wave group travels on the gentle slope.

[pt] Neste trabalho, é apresentado uma abordagem para animação de águas rasas em aplicações interativas baseada em um modelo físico. Para a simulação, foi empregado o método Langrangreano conhecido como Smoothed Particle hydrodynamics (SPH). Com base no trabalho de Muller et al. (Muller et al., 2003), que utilizou SPH em computação Gráfica, e no trabalho de Rodriguez-Paz e Bonet (Rodriguez-Paz; Bonet, 2005) que propõe uma variação deste método para a simulação de águas rasas em aplicações de engenharia, propomos uma abordagem simples e eficiente para a simulação de águas rasas em jogos a influência de terrenos acidentados. / [en] In this work is present an approach to shallow-water animation on interactive applications based on a physic model. For the simulation, was employed a Lagrangian methos known as Smoothed Particle Hydrodynamics (SPH). Based on the work of muller et al. (Muller et al., 2003), which applied SPH in computer Graphics, and on the work of Rodriguez-Paz (Rodriguez-Paz; Bonet, 2005), wich proposes a variation of this method to shallow-water simulation on engineering applications, we have proposed a simple and efficient approach for shallow-water simulation on games under the influence of irregular terrains.