A unified derivation of Voronoi, power, and finite-element Lagrangian computational fluid dynamics
Daniel Duque

TL;DR
This paper unifies various Lagrangian fluid dynamics methods by deriving a framework that connects Voronoi, power diagrams, and finite-element approaches, revealing new force terms and improving simulation accuracy.
Contribution
It introduces a systematic derivation linking Voronoi, power diagrams, and finite-element methods within a unified framework, including new force terms enhancing existing models.
Findings
pFEM outperforms other methods in standard tests
Additional spring-like force improves simulation accuracy
Unified framework clarifies connections between different CFD approaches
Abstract
Most approaches in Lagrangian fluid dynamics simulations proceed from the definition of particle volumes, from which discrete versions of the spatial differential operators are derived. Recently, Gallou\"et and M\'erigot [1] simultaneously tackled physical dynamics and geometrical optimization, with the result that the pressure field is linked to a geometric feature: the weights of a power diagram. Their resulting dynamics, surprisingly, does not feature a pressure gradient, but spring-like forces between each particle and the centroid of its cell. Inspired by this work, both geometrical and mechanical optimization are here included within a framework due to Arroyo and Ortiz [2]. In a systematic way, we first find a connection with the smoothed particle hydrodynamics method. In what we will call the ``low-temperature limit'', we show that the requirement of zeroth order consistency…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsFluid Dynamics Simulations and Interactions · Fluid Dynamics and Vibration Analysis · Lattice Boltzmann Simulation Studies
