Autonomic closure for turbulence simulations

Area of Science:

• Computational Fluid Dynamics
• Turbulence Modeling
• Adaptive Systems

Background:

• Traditional turbulence models require predefined structures, limiting adaptability.
• Existing dynamic models struggle with complex, nonequilibrium flow characteristics.

Purpose of the Study:

• To present a novel, fully adaptive, and self-optimizing turbulence closure method.
• To eliminate the need for predefined turbulence models in simulations.

Main Methods:

• Developed an autonomic closure solving a nonlinear, nonparametric system identification problem.
• The simulation determines optimal local, instantaneous relations for unclosed terms.
• Implemented a simple version for large eddy simulations (LES).

Main Results:

• The autonomic closure adapts freely to nonlinear, nonlocal, and nonequilibrium turbulence.
• Demonstrated remarkably more accurate results in a priori tests compared to dynamic traditional closures.
• Showcased the effectiveness of a simple implementation for LES.

Conclusions:

• The proposed autonomic closure offers a more flexible and accurate approach to turbulence modeling.
• This method overcomes limitations of traditional and dynamic turbulence closures.
• It paves the way for more robust and adaptive CFD simulations.