Implicit Subgrid-Scale Modeling of a Mach 2.5 Spatially Developing Turbulent Boundary Layer.

Area of Science:

• Fluid Dynamics
• Computational Science

Background:

• Large Eddy Simulation (LES) is adapted for subgrid-scale (SGS) modeling using streamlined upwind/Petrov-Galerkin stabilization.
• This numerical method, originally for stability, now accounts for unresolved small-scale effects in turbulent flows.

Purpose of the Study:

• To assess the performance of implicit LES for adiabatic supersonic turbulent boundary layers at Mach 2.5.
• To compare numerical results with direct numerical simulation (DNS) data and experimental literature.
• To investigate compressibility effects and the validity of the Reynolds analogy.

Main Methods:

• Employed numerically implicit subgrid-scale modeling with streamlined upwind/Petrov-Galerkin stabilization.
• Utilized a dynamic rescaling-recycling approach for turbulent inflow conditions in high-speed flows.
• Performed direct comparison with DNS database and literature experiments for adiabatic supersonic boundary layers.

Main Results:

• Implicit LES showed good performance for mean flow characteristics (Cf, UVD+) on Mach 2.5 adiabatic flat plates.
• A longer power law behavior was observed for mean streamwise velocity in the outer region compared to the log law.
• While overpredicting peak stresses in the buffer layer, iLES achieved excellent agreement for turbulence intensities and Reynolds shear stresses in the outer region.

Conclusions:

• Implicit LES is a viable approach for simulating adiabatic supersonic turbulent boundary layers.
• The method demonstrates strong predictive capabilities, particularly for outer flow region turbulence statistics.
• Compressibility effects on thermal turbulent structures are weak, and the Reynolds analogy holds to a lesser extent than in incompressible flows.