Computer scientists at the University of California San Diego have developed a method for generating highly realistic computer-generated images of fluid dynamics in elements such as smoke.
This research, conducted by the UC San Diego Center for Visual Computing, was presented at the SIGGRAPH Asia 2024 conference, where it received a Best Paper Honorable Mention for its contributions to computer graphics and physics-based simulation. The paper is published in ACM Transactions on Graphics.
To demonstrate the power of their approach, the team compared an iconic photograph from the 1980 eruption of Mount Saint Helens volcano in Washington State to a computer-generated rendering of a volcanic smoke plume created using their new method. The resulting simulation captures the intricate, multi-scale billowing of the smoke plume, including its twisting, curling motion and delicate turbulence, which are hallmarks of realistic fluid behavior.
Such visually complex features are challenging to reproduce with traditional methods because they demand extremely high computational resolution to capture fine details accurately. Achieving this level of realism with conventional approaches would require impractically large amounts of computational power and time, making them unsuitable for many practical applications.
This work introduces a more efficient approach that opens the door to increasingly realistic simulations while significantly reducing computational costs. By preserving the physical properties of fluid motion, such as energy and circulation, the method allows for accurate representations of natural phenomena that can be used for scientific verification and analysis, such as understanding smoke dispersion or atmospheric dynamics.
At the same time, it provides a powerful tool for generating high-quality computer-generated imagery (CGI) for entertainment purposes, such as movies, video games, and virtual reality, where realism and efficiency are equally critical.
Realism through physics: Why it matters
This work is part of a larger effort in computer graphics to integrate the fundamental laws of physics into simulation algorithms. By respecting the physical principles that dictate fluid motion, the new method not only improves visual realism but also enhances the overall consistency and predictability of simulations. For phenomena like volcanic eruptions, where the shape and appearance of the smoke plume are determined by complex fluid dynamics, these physics-based techniques are essential for generating believable results.
Beyond visual effects, such realistic simulations have broader applications in scientific research, environmental modeling, and education. For example, accurate simulations of volcanic smoke plumes can provide insights into atmospheric dynamics and air quality predictions following eruptions.
Representations of natural phenomena can be used for scientific verification and analysis, such as understanding smoke dispersion or atmospheric dynamics. At the same time, they provide a powerful tool for generating high-quality computer-generated imagery (CGI) for entertainment purposes, such as movies, video games, and virtual reality, where realism and efficiency are equally critical.
Physics-preserving fluid simulation: CO-FLIP
The new technique, called Coadjoint Orbit FLIP (CO-FLIP), improves upon existing Fluid Implicit Particles (FLIP) methods by preserving two key physical properties of fluid motion: energy and circulation. These properties are critical for accurately capturing how fluids evolve over time, ensuring visually and physically consistent results.
One of the most remarkable features of the CO-FLIP method is its ability to produce high-quality results even at low resolutions. This efficiency is particularly important for applications in film production, virtual environments, and interactive simulations, where computational resources are limited, and real-time performance is required.
The team demonstrated CO-FLIP in both 2D and 3D simulations, showcasing its versatility across different dimensions and use cases. A video accompanying the paper provides striking visual examples, closely mirroring their real-world counterparts.
Discover the latest in science, tech, and space with over 100,000 subscribers who rely on Phys.org for daily insights.
Sign up for our free newsletter and get updates on breakthroughs,
innovations, and research that matter—daily or weekly.
Differential geometry
Differential geometry allowed this breakthrough. The method is not commonly used for computational fluid dynamics and fluid simulations in computer graphics, which also makes the research team’s approach unique.
Differential geometry is usually used to model physics on a curved space or curved spacetime. So how is it useful in fluid simulation in a flat space, such as simulating volcano eruptions? It turns out that there is an alternative, geometric approach to formulating fluid equations, which is lesser known to the mainstream computational fluid community. Instead of the usual approach where the equation is derived from Newton’s law of motion (f=ma), in the geometric approach the equation for fluids is modeled as the “shortest path on the space of all fluid deformations.”
The space of all fluid deformations forms a mathematical structure called a Lie group, and analyzing such objects requires differential geometry. There are many hidden mathematical structures in this Lie group that are revealed with this geometric approach, even after simplification for computation. These mathematical structures do have clear physical implications for the behavior and appearance of fluids, even in turbulent flows when everything seems so chaotic.
In this paper, the researchers focused on the concept of preserving “coadjoint orbit,” one of the more nuanced conservation laws in fluids for the new fluid simulator they created. As a result, they obtained much more preservation of the vortex and visual details that many previous methods struggled to maintain.
Beyond this project, the research team also plans to use geometry to tackle various computational physics and computer graphics problems. It is a mathematical technique that has repeatedly produced fundamental breakthroughs and new understandings.
More information:
Mohammad Sina Nabizadeh et al, Fluid Implicit Particles on Coadjoint Orbits, ACM Transactions on Graphics (2024). DOI: 10.1145/3687970
Provided by
University of California – San Diego
Citation:
Advanced method produces highly realistic simulations of fluid dynamics (2025, January 6)
