Turbulence and turbulent heat or scalar transfer are essential processes in many fluid dynamics problems relevant to the design of sustainable buildings and cities. RANS turbulence models are commonly used in the engineering practice, but can introduce significant uncertainties in the result.
The objective of this project is to develop turbulence model-form uncertainty quantification methods for RANS simulations with heat transfer. The framework is based on both physical constraints and on mathematical properties of the governing equations, and is intended to be independent of any specific model formulation. Benchmark databases of turbulent statistics for typical engineering problems are provided by the results of wall-resolved large eddy simulations (LES), and are being used for development and validation of the method.