A Scalable OpenLB LAMMPS Framework for Fully Resolved Simulations of Hindered Settling of Arbitrary Non-Spherical Particles
This paper presents a scalable OpenLB-LAMMPS framework coupling Lattice Boltzmann and Discrete Element methods to simulate fully resolved hindered settling of up to 100,000 arbitrary non-spherical particles, revealing that cubic particles exhibit distinct coordination shell dynamics and system-size-dependent velocity correlations compared to spheres.
Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
Imagine a bucket of sand pouring into water. The grains do not fall at the same speed they would if they were alone; they slow each other down, colliding and pushing against the fluid in a complex, crowded dance. This phenomenon, known as hindered settling, is a fundamental process in nature and industry, governing everything from how sediment builds river deltas to how pharmaceuticals are mixed in a factory. For decades, scientists have understood this behavior well when the particles are perfect spheres, like tiny marbles. However, the real world is rarely made of perfect spheres. Most particles are irregular, jagged, or blocky, and their strange shapes create unique ways of touching and sliding past one another. Predicting how these odd shapes settle has been a stubborn computational challenge because the mathematics of their collisions and the way they drag fluid around them are incredibly difficult to solve simultaneously.
To tackle this, a team of researchers has built a new digital framework that couples two powerful open-source software tools to simulate these complex interactions. They combined a fluid solver called OpenLB, which models how liquids move, with a particle simulator called LAMMPS, which tracks how solid objects bump and grind against each other. The key innovation lies in how they represent the solid objects. Instead of trying to model a complex shape like a cube with a single, rigid mathematical definition, they break the object down into a cluster of overlapping spheres, much like a snowman made of three balls. This "clump" allows the computer to easily calculate how the object touches its neighbors. At the same time, the fluid solver sees the object not as a collection of balls, but as a solid block with a specific shape, using a technique called voxelization to map the object's surface onto the fluid grid. This separation allows the team to simulate the contact mechanics of the particles and the flow of the fluid with high precision, even when dealing with massive numbers of objects.
The researchers tested their system by simulating the settling of cubes in a fluid, comparing the results against known behaviors of spheres. They ran simulations with up to 103,823 cubes, a scale that pushes the limits of current computing power. Their findings revealed a distinct difference between how cubes and spheres behave as they settle. While spheres tend to cluster together through contact-dominated clustering, the cubes formed pronounced coordination shells without face-parallel contact. This suggests that the sharp edges and corners of the cubes create a specific type of spacing that prevents them from packing as tightly as spheres do. Furthermore, the team discovered that the speed fluctuations of the particles were not random or local; instead, the movement of one particle influenced the movement of others across the entire system. Even in the largest simulations, the correlation between particle velocities stretched across the whole container, indicating that the collective motion of the crowd is a unified phenomenon rather than a series of isolated events.
By validating their model against experiments with spheres and then applying it to cubes, the authors demonstrated that particle shape plays a critical role in suspension dynamics. The simulations showed that the settling speed of cubes differs from that of spheres, and the internal structure of the settling cloud is fundamentally different. The work confirms that ignoring the shape of particles can lead to inaccurate predictions in fields ranging from geotechnical engineering to environmental science. The team's framework provides a robust tool for studying these large-scale flows, offering a way to see the hidden order within the chaos of settling particles. While the current simulations did not include the thin film of fluid that lubricates the gap between particles when they are extremely close, the method successfully captured the broader dynamics of the system. The results suggest that for non-spherical particles, the way they settle is governed by a complex interplay of shape, contact, and long-range fluid interactions that simple models of spheres cannot capture.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.