A single PINN approximates velocity, pressure, and stress fields for rarefied gas flow through a non-convex cylinder array across a range of Knudsen numbers, using DSMC data and physics-based loss terms.
Realistic air filter media performance simulation. Part II: Beyond finite -volume computational fluid dynamics procedures,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
physics.comp-ph 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
Physics-Informed Neural Networks for microflows: Rarefied Gas Dynamics in Cylinder Arrays
A single PINN approximates velocity, pressure, and stress fields for rarefied gas flow through a non-convex cylinder array across a range of Knudsen numbers, using DSMC data and physics-based loss terms.