Vasily Semiletov

On the effect of flap deflection on jet flow for a jet-pylon-wing configuration: near-field and acoustic modelling results

A computational investigation of jet-pylon-wing-flap interaction based on a model installed jet geometry for a co-axial jet at bypass ratio (BPR) 5 is presented. For numerical modelling of jet installation effects, Monotonically Integrated Large Eddy Simulations (MILES) are conducted with the CABARET method. For accurate sound predictions at downstream jet radiation angles, a penetrable Ffowcs Williams-Hawkings (FWH) technique with multiple closing disks is applied. The axial and vertical distributions of the mean axial velocity and turbulent kinetic energy obtained from the CABARET calculations are compared with the reference RANS/ILES solution of Lyubimov (2013) and with the experiment data for the axisymmetric jet case and for the nozzle-with-pylon case. The farfield acoustic spectra predictions are obtained for the jet flow with penetrable (for jetaiframe configuration) and impenetrable (for wing-flap alone) FWH formulations. The sound spectra computed is compared with the acoustic results for jet-pylon-wing-flap configuration of similar geometrical parameters at BPR 10 from the literature with the scaling based on the mixed out jet core parameters in accordance with Lush (1971) method.

CABARET GPU Solver for Fast-Turn-Around Flow and Noise Calculations

A GPU implementation of an aero-acoustic solver based on the low-dissipative, low-dispersive CABARET scheme is demonstrated. The implementation makes industrial relevant (10-50 million cells) LES studies of jet noise modeling possible in reasonable time (several days) using just a workstation computer, therefore avoiding the need for supercomputing facilities for these cases. Besides the solving of the turbulence, the post-processing relevant to the applications in mind can be done on the fly on the GPU as well.

Jet-wing interaction: computational modelling based on MILES CABARET and acoustic analogy

The computational model of a jet-wing configuration from the recent TsAGI experiment [1] that corresponds to a dual-stream co-axial subsonic jet under a swept lifting wing is considered. For the jet and wing calculations, Monotonically Integrated Large Eddy Simulations (MILES) are performed for which a modern parallel unstructured-grid computational code based on the high-resolution CABARET scheme [2],[3] is used. The computational domain includes the co-axial nozzle with a central body and a lifting wing section where round flowstoppers were installed on the wing tips to prevent downwash effects. For far-field noise predictions, the MILES CABARET solver is coupled with the Ffowcs Williams – Hawkings (FW-H) integral method. The FW-H solution corresponds to a large closed permeable control surface with multiple closing disks at the outlet side in accordance with the best practice. The results of the acoustic modelling are compared with the experiment. The differences, which can be associated with the unaccounted effects of sound refraction through a non-uniform free-stream flow are discussed.

Airfoil Flow and Noise Computation Using Monotonically Integrated LES and Acoustic Analogy

A new scalable CABARET MILES method has been applied for modelling of flow around NACA0012 airfoil. Results of computational modelling on several grids are compared with the experiment. The flow solver is coupled with the Ffowcs Williams – Hawking formulation for far field noise modelling. The CABARET-FWH results for a model problem of pulsating sphere are first demonstrated. For airfoil noise modelling, the computational results are compared with the experiment.

Jet and jet–wing noise modelling based on the CABARET MILES flow solver and the Ffowcs Williams–Hawkings method:

A co-axial subsonic unheated jet with and without a swept lifting wing at free-stream conditions from a recent jet–wing TsAGI experiment is considered. For computational modelling, Monotonically Integrated Large Eddy Simulations (MILES) are conducted based on the CABARET scheme which is implemented in a modern parallel unstructured-grid compressible Navier–Stokes computational code. The computational domain of the installed configuration includes a part of the nozzle and a wing section with round flow-fences present in the experiment to preclude the downwash effects. For the isolated jet, the same size of the computational domain is applied and two grid resolutions are considered to investigate the sensitivity of the current far-field noise predictions to the computational grid. The meanflow velocity profiles predicted downstream of the nozzle are compared with the flow data available. For far-field acoustic predictions, the Ffowcs Williams-Hawkings (FW-H) integral method is used. The Ffowcs Williams-Hawkings solutions correspond to a large closed permeable control surface with multiple closing disks at the outlet side set-up in accordance with the best practice to avoid pseudo sound in the acoustic modelling. The comparison of the acoustic predictions with the far-field spectra measured in the experiment for 30° and 90° observer angles to the jet flow are presented and discussed. It is shown that the current modelling robustly captures the same relative trends of the spectra behavior as observed in the experiment: while the presence of the wing does not lead to any significant change of sound spectrum in comparison with the isolated jet for 30°, there is a 8 - 10 dB sound amplification due to the jet–wing interaction at 90° angle to the jet.

Empiricism-free noise calculation from LES solution based on Goldstein generalized acoustic analogy: volume noise sources and meanflow effects

The hybrid CAA method with propagation equations corresponded to the Goldstein generalized acoustic analogy and noise sources extracted from time-resolved LES solution is considered. In the current implementation, which includes sound propagation meanflow effects, noise source information was extracted directly from LES without any assumptions of statistic properties such as fourth-order velocity correlations, which makes this tool applicable for investigation of jet mixing noise without using any intermediate assumptions. Far-field noise and noise source statistics have been calculated for two subsonic single-stream jets. One jet is isothermal and the other is heated corresponding to temperature ratio ( ); they have the same acoustic Mach number ( ) and correspond to jet conditions of model-scale tests recently conducted at the QinetiQ Noise Test Facility. For unsteady flow calculation, Monotonically Integrated Large Eddy Simulation (MILES) is used based on the high-resolution CABARET scheme. Flow field solutions and far-field noise predictions are compared with the experimental data available as well as with the results of the Ffowcs Williams Hawkings solution based on the same set of LES data.

A 3D frequency-domain linearised Euler solver based on the Goldstein acoustic analogy equations for the study of nonuniform meanflow propagation effects

For modelling sound propagation through non-uniform mean flows, a new scalable Adjoint Linearised Euler solver in the frequency domain is developed which extends the work of Karabasov and Hynes (2006) to fully three dimensional flows. The solver is based on second-order central finite differences and pseudo-time stepping based on iterative Adamstype. A special preconditioning technique based on the variable iteration time step is used to avoid numerical instability associated with shear-layer type flows. Numerical results are verified against the analytical solution corresponding to planar acoustic wave scattering from the parallel jet flow.

Similarity scaling of jet noise sources for low-order jet noise modelling based on the Goldstein generalised acoustic analogy

As a first step towards a robust low-order modelling framework that is free from either calibration parameters based on the far-field noise data or any assumptions about the noise source structure, a new low-order noise prediction scheme is implemented. The scheme is based on the Goldstein generalised acoustic analogy and uses the Large Eddy Simulation database of fluctuating Reynolds stress fields from the CABARET MILES solution of Semiletov et al. corresponding to a static isothermal jet from the SILOET experiment for reconstruction of effective noise sources. The sources are scaled in accordance with the physics-based arguments and the corresponding sound meanflow propagation problem is solved using a frequency domain Green’s function method for each jet case. Results of the far-field noise predictions of the new method are validated for the two NASA SHJAR jet cases, sp07 and sp03 from and compared with the reference predictions, which are obtained by applying the Lighthill acoustic analogy scaling for the SILOET far-field measurements and using an empirical jet-noise prediction code, sJet.