Low-Order Modeling and Machine Learning in Fluid Dynamics: Turbulence Modeling I
ORAL · X15 · ID: 2665375
Presentations
-
Epistemic Uncertainty Quantification of Deep Neural-Network Based Turbulence Closures
ORAL
–
Presenters
-
Cody Grogan
- Utah State University
Authors
-
Cody Grogan
- Utah State University
-
Som Dutta
- Utah State University
-
Mauricio Tano
- Idaho National Laboratory
-
Som Dhulipala
- Idaho National Laboratory
-
Izabela Gutowska
- Oregon State University
-
-
Improving Predicted Statistics of Velocity Gradient Closures using Parameterized Lagrangian Deformation Models
ORAL
–
Presenters
-
Criston M Hyett
- University of Arizona
Authors
-
Criston M Hyett
- University of Arizona
-
Michael Woodward
- Los Alamos National Laboratory
- LANL
-
Yifeng Tian
- Los Alamos National Laboratory (LANL)
-
Mikhail Stepanov
- The University of Arizona
-
Chris L Fryer
- Los Alamos National Laboratory (LANL)
-
Daniel Livescu
- Los Alamos National Laboratory (LANL)
-
Michael Chertkov
- University of Arizona
-
-
Neural Network-Based Closure Model of the Ensemble-Averaging Dynamics of Turbulent Puffs in Transitional Pipe Flow
ORAL
–
Presenters
-
Yu Shuai
- Princeton University
Authors
-
Yu Shuai
- Princeton University
-
Clarence W Rowley
- Princeton
-
-
RANN: A Neural RANS Closure Model for Physics-Informed Machine Learning on General Geometries
ORAL
–
Publication: FluidAI provides cloud-based real-time machine-learning enabled CFD and Fluid-Dynamics optimization systems for several industrial and commercial applications. The disruptive technology, based on physics-informed deep learning, has already impacted several high complexity simulation and design areas, from high energy plasma physics to complex non-rigid robotics actuation, and is now ready to be brought to bear in the CFD domain. FluidAI can achieve several orders of magnitude speed-up compared to standard CFD solutions, without compromising accuracy, as well as allow for interactive design of complex aerodynamic instruments. The cloud-based solution allows for scalability and ease of use, as well as drop in integration on several existing workflows.
Presenters
-
Matthew Uffenheimer
- FluidAI
Authors
-
Matthew Uffenheimer
- FluidAI
-
Luca Rigazio
- FluidAI
-
Eckart Heinz Meiburg
- University of California, Santa Barbara
-
-
Subgrid Stress Modeling with Data Driven Structured State Space Sequence Models
ORAL
–
Publication: Planned papers:
AIAA 2025 SciTech Conference Paper (Extended Abstract Submitted)
Journal paper in progressPresenters
-
Andy Wu
- Stanford University
Authors
-
Andy Wu
- Stanford University
-
Sanjiva K Lele
- Stanford University
-
-
The multiscale-based data-driven subgrid-scale model with physics constraints for enhanced prediction of unresolved scales in turbulent flow
ORAL
–
Presenters
-
Bahrul Jalaali
- Osaka University
Authors
-
Bahrul Jalaali
- Osaka University
-
Kie Okabayashi
- Osaka University
-
-
A dynamic recursive neural-network-based subgrid-scale model for large eddy simulation
ORAL
–
Presenters
-
Chonghyuk Cho
- Seoul Natl Univ
Authors
-
Chonghyuk Cho
- Seoul Natl Univ
-
Haecheon Choi
- Seoul National University
-
-
SGS backscatter effects in coarse-grid LES predicted by a machine-learning-based SGS model
ORAL
–
Presenters
-
Soju Maejima
- Tohoku University, Japan
Authors
-
Soju Maejima
- Tohoku University, Japan
-
Soshi Kawai
- Tohoku University, Japan
-
-
Toward machine-learning-based large eddy simulation of flow over a complex geometry
ORAL
–
Presenters
-
MYUNGHWA KIM
- Seoul Natl Univ
Authors
-
MYUNGHWA KIM
- Seoul Natl Univ
-
Haecheon Choi
- Seoul National University
-
-
Abstract Withdrawn
ORAL · Withdrawn
–