No abstract

No abstract

No abstract

No abstract

This paper presents recent achievements at Onera in a global program which objective is to improve the accuracy of numerical predictions of the aerodynamic noise generated by realistic landing gears. The program relies on the Gulfstream Partially Dressed -Closed Cavity Nose Landing Gear (PDCC-NLG), which has been tested at scale 1/3 in the BART aerodynamic windtunnel at NASA LaRC and in the UFAFF acoustic facility at the University of Florida, and is currently proposed in NASA's Benchmark for Airframe Noise Computations (BANC). Onera's numerical approach combines (i) an unsteady flow computation based on a "Delayed Detached Eddy Simulation" (DDES) method (unstructured solver CEDRE) and (ii) a farfield noise computation based on a Ffowcs Williams-Hawkings integral (FW-H) method (KIM solver). The research program covers the following aspects. Firstly, experimental installation effects are compared by using grids representative of the model in the BART (closed test section, case A) and in the UFAFF (open test section, case B). Secondly, two levels of CFD grid resolutions are compared, with grids of about 12M cells (resolution R1) or 70M cells (resolution R2). Thirdly, two different wall models are compared, namely a laminar model (labeled -l) and a turbulent model (labeled -t). In the first steps of the program, already presented in past papers, attention has been focused on, firstly, comparing installation effects with low resolution grids (computationsM1A-l vs. M1B-l) and, secondly, comparing resolutions R1 and R2 on the closed section configuration (computations M1A-l vs. M2A-l). All cited computations were restricted to the laminar wall model. Acoustic predictions have been achieved from data on the rigid surface and on several porous surfaces at various distances from the landing gear. In the present paper, two new DDES computations have been achieved with the highest resolution (R2), one for the "open jet" configuration (B) with the laminar wall model (computation M2B-l) and the other one for the "closed section" configuration (A) with the turbulent wall model (compuation M2A-t). Several parameters are studied, including (i) the influence of the grid resolution on the noise predictions in the open jet configuration, (ii) the sensitivity of the porous surface for FW-H integration to the far-field prediction and (iii) the sensitivity of the wall model to both aerodynamic and acoustic results.