Physically based simulation of fluid mixtures
Wu E.1
Source PublicationProceedings of 2007 10th IEEE International Conference on Computer Aided Design and Computer Graphics, CAD/Graphics 2007
AbstractIn our daily life, we may often see various natural phenomena of fluid mixtures, like pouring honey/ink into water, Coca Cola into strong wine etc. In fluid simulation, since fluids can not hold the shear stress, their shapes usually vary drastically in their dynamics. Interaction among multiple fluids in different features becomes even more difficult. Since the late 90s of the last century, graphics researchers have simulated lots of natural effects on fluid simulation for animating such as smoke, fire or water. However, little attention has been paid to demonstrating the effect of fluid mixtures. In the fluid mixture simulation, interactions among fluid mixtures in different kinds would generate comprehensive behavior. To investigate the interaction behavior in physically-based simulation of the fluids, it is of importance to build physically correct models to represent the varying interactions among the fluid mixtures. In this talk, stress will be put on the physically based simulation of fluid mixtures, in particular the mixture in two fluids, or binary mixtures. We will introduce the fundamental principle of the physics in the fluid mixture phenomena, and then the primary solutions to the simulation. The approaches we have proposed for the solution to the fluid mixture simulation will be introduced in detail, with various simulation results to be demonstrated. In some of the methods, advantage is taken with the Graphics Processing Unit (GPU) to achieve real-time computation for the simulation. Roughly in terms of the miscibility of binary mixtures, the fluid mixture can be categorized into immiscible mixtures such as bubbles in liquid or oil in water, and miscible mixtures such as honey dissolving in water. The visual appearance of the mixture physically roots in two opposite processes: the diffusing and the demixing. The diffusing process leads to the dissolving effect such as honey dissolving in water, while the demixing process leads to the separation (or decomposition) effect. As a matter of fact, these two processes are not absolutely isolated with each other, and it is their combination that determines what the mixture looks like. If the diffusing process is dominant, the mixture is miscible and the two fluid components dissolves into each other; otherwise, it is immiscible and the mixture is separated into two parts by a clear interface, but still with some fluid molecules diffusing into each other which can hardly be perceived by naked eyes. All these facts imply that there is a physical property which can be used to determine which process is dominant when two fluids are mixed together. In our solution, we choose the property, the miscibility of mixtures, to play such a role in the simulation. The miscibility can be represented with a parameter in an approach of two-fluid solution. With such a parameter, the miscible and immiscible mixtures can be simulated as a whole in a unified framework. As a relatively new and promising solution approach in Computational Fluid Dynamics (CFD) developed in recent years, it has the advantages on easy implementation, parallel processing, and easy handling of complex and moving boundaries. In our work, we mainly use a lattice Boltzmann method (LBM) to simulate the underlying dynamics of miscible mixtures in binary fluid simulation called TFLBM. However, it suffers from the limitation of only resolving mixture flows with low Reynolds number, and it would blow up when the Reynolds number gets higher. In order to resolve such mixture flows with higher Reynolds number, by further investigation, we proposed to use a subgrid method to stabilize the computation of two fluid mixtures. The idea of the subgrid method is to split the actual velocity field into large-scale (resolved) and small-scale (unresolved) components. The effect of the unresolved motion on the resolved one is included by introducing a so-called eddy viscosity, and the method is also referred to as large eddy simulation (LES). Though the method was only available on single fluid LBM, we extended the method to stabilize TFLBM by further extending the subgrid model to the binary fluid mixtures (STFLBM). Our experiments demonstrate the superior stability of STFLBM. With the method, we can simulate the free surface effect and add control forces to control the miscible mixture system, where the higher Reynolds number exists and TFLBM otherwise would become instable very quickly. By optimizing the memory structure and taking the advantage of those dual-core or multi-core systems, real time computation for a domain in 64 cells full of fluid mixtures can be achieved. © 2007 IEEE.
URLView the original
Fulltext Access
Citation statistics
Document TypeConference paper
CollectionUniversity of Macau
Affiliation1.Institute of Software Chinese Academy of Sciences
2.Universidade de Macau
Recommended Citation
GB/T 7714
Wu E.. Physically based simulation of fluid mixtures[C],2007:3-4.
Files in This Item:
There are no files associated with this item.
Related Services
Recommend this item
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[Wu E.]'s Articles
Baidu academic
Similar articles in Baidu academic
[Wu E.]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[Wu E.]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.