ENGINEERING AND CFD

Motivation:

The numerical simulation of flow around aircraft at the boundaries of the flight envelope is important. The availability of trustworthy numerical solutions would allow a better understanding of flow processes taking part during critical flight phases, leading to the optimization of design processes while enhancing flight safety and increasing operational ranges.

One of the major limiting factors of the flight operational range of transport aircraft is the inlet separation of engine. Inlet separation may affect considerably the engine's operation of the subsequent stages and hence the overall flight operation safety. Therefore, the overall aim of this project is to provide an efficient numerical method able to compute the large range of spectral scales present in flows during the separation process.

Methodology:

To accomplish this aim, the research group FOR 1066 funded by DFG (“Deutsche Forschungsge-meinschaft”) has been extensively working in the direction of DES approaches (Detached Eddy Simulations), aiming to efficiently solve the attached boundary layer with advanced RANS models (Reynolds Average Navier-Stokes) and the detached flow regions with LES (Large Eddy Simulations).

The advanced JHh-RSM ADDES (Algebraic Delayed Detached Eddy Simulation) [1] has shown excellent results in the computation of the development of the attached boundary layer up to the separation onset, providing a precise switch from RANS to LES at the separation onset location. However, it fails after separation due to the long transition area required in the LES domain to restore the flow turbulence content missed across the RANS/LES interface.

The present project aims to mitigate this extended grey area by implementing a physically based synthetic turbulence generator. The generated velocity field of fluctuations is added to the resolved flow field as a source term of the governing equations that is activated in a volumetric domain downstream of the RANS/LES interface. In addition, the research partners from DLR Göttingen have developed a new low dissipative configuration for the spatial discretization that has also been tested for the JHh-RSM ADDES with synthetic turbulence along the current project. All the computations were carried out with the DLR-TAU code.

Main results:

The implementation of the synthetic turbulence forcing was tested in a ZPG flat plate and over the HGR-01 airfoil. Computations of these test cases performed with the JHh-RSM ADDES without applying synthetic turbulence have shown a very long adaptation distance where no resolved turbulence is developed downstream of the RANS/LES interface. The flat plate was used mainly for calibration purposes while the HGR-01 was applied for validation of the approach against experimental data.

The obtained results were successful, providing an important amount of resolved turbulence immediately after the forced domain (Figure 1 and Figure 3). In addition, the application of the new low dissipative setting made able to considerably improve the estimation of the mean skin friction without needing to refine the grid spacing and hence saving considerably computational cost (Figure 2).

Results obtained for the HGR-01 airfoil show that not only considerably amount of turbulence is developed immediately downstream of the forced domain but also the computed resolved turbulence agrees nicely with the experimental data (Figure 4).