Hands-on turbulence modeling with SGS, DES/IDDES, and real-world CFD case studies

What you'll learn
- Derive the LES formulation from the Navier–Stokes equations using spatial filtering
- Understand subgrid-scale (SGS) stresses and the physical role of SGS models
- Implement and compare major SGS models: Smagorinsky, WALE, k-equation, dynamic, and hybrid RANS–LES (DES/IDDES)
- Set up and run practical LES simulations for turbulent flow past a square cylinder
- Analyze vortex shedding, wake dynamics, turbulence statistics, and mesh resolution requirements
- Compare LES results with k-ω SST RANS to evaluate accuracy and computational cost
- Apply guidelines for wall-resolved vs. wall-modeled LES in engineering problems

Requirements
- Basic understanding of fluid mechanics and turbulence fundamentals
- Familiarity with Reynolds-averaged Navier–Stokes (RANS) concepts
- Basic knowledge of Linux/Unix environment and command-line usage
- Prior exposure to OpenFOAM (case structure, running solvers, post-processing)
- Understanding of numerical methods (finite volume method preferred)
- Comfort with vector and tensor notation, calculus, and differential equations
- (Optional) Basic scripting skills for post-processing and automation

Description
Large Eddy Simulation (LES) is an advanced turbulence modeling approach that resolves large-scale turbulent structures while modeling the smaller scales. Compared to traditional RANS models, LES provides improved predictions for unsteady, separated, and wake-dominated flows, but requires careful modeling choices and mesh design. This course offers apractical and intuitive introduction to LES using OpenFOAM, focusing on physical understanding and correct application rather than detailed mathematical derivations.

The course provides aconceptual overview of how LES is derived from the Navier–Stokes equations, explaining spatial filtering, filter width, and the physical meaning of subgrid-scale (SGS) stresses. The emphasis is on understanding what is resolved, what is modeled, and why SGS models are needed, without going through step-by-step mathematical derivations.

You will learn how different SGS and hybrid RANS–LES models behave in practice, including:

Smagorinsky, WALE, and k-equation SGS models

Hybrid RANS–LES approaches such as DES and IDDES

A major part of the course is ahands-on LES workflowapplied to a real engineering benchmark: turbulent flow past a square cylinder. Using this case, you will set up and run LES simulations in OpenFOAM, compare different SGS models, and analyze vortex shedding, wake dynamics, and turbulence statistics. All LES results are compared with a baselinek–ω SST RANS simulationto highlight accuracy and computational cost trade-offs.

The course also providespractical modeling guidelines, including mesh resolution requirements, time-step selection, wall-resolved vs. wall-modeled LES concepts, and common pitfalls to avoid. Estimation of turbulence scales and interpretation of LES results are discussed from an engineering perspective.

To support hands-on learning, the course includescomplete OpenFOAM case filesfor all demonstrations, along withadditional PDF documentsthat summarize theoretical concepts, modeling guidelines, and best practices. By the end of the course, you will be confident in setting up, running, and evaluating LES and hybrid RANS–LES simulations for practical engineering applications.

Who this course is for:
- Graduate students studying fluid mechanics, turbulence, or CFD
- CFD engineers who want to move beyond RANS to LES and hybrid RANS–LES methods
- Researchers working on turbulent flows, wake dynamics, or bluff-body aerodynamics
- OpenFOAM users seeking hands-on experience with practical LES simulations
- Industry professionals who need accurate unsteady turbulence modeling for engineering applications
- Academics and PhD students looking to understand LES theory with real-case implementation
More Info