An immersed boundary method for efficient computational studies of nozzles designed to reduce jet noise

Nitin S Dhamankar, Purdue University


Turbulent jet mixing noise has been an active research area for the past 6 decades. Specially designed nozzles that can induce an earlier jet breakdown have shown potential to reduce the far-reaching low-frequency components of jet noise. In the last two decades, large eddy simulation (LES) and computational aeroacoustics (CAA) have become a significant complement to experimental research in quiet nozzle design. High-order finite-difference discretization-based solvers are better suited for CAA applications due to their excellent numerical dissipation and dispersion characteristics. However, geometrical complexities present in many nozzles pose a significant challenge to such solvers, which inherently require structured smooth grids. In the present work, an immersed boundary method (IBM) is implemented in an in-house, highly-scalable, finite-difference discretization-based LES code to tackle this challenge. In IBM, the surface geometry of a body under study is immersed in a non-body-conforming background grid and its effect on the flow is modeled indirectly. The impact of the implemented IBM on the order of accuracy of the base solver is quantified. The method is demonstrated to perform very well in predicting scattering of acoustic waves from a cylinder, and in producing realistic low-Reynolds number viscous flow over a sphere. To simulate experimental-scale high Reynolds number jets at an affordable computational cost, an equilibrium wall model is also implemented in the IBM. It is shown to produce a reasonably accurate outer region for a boundary layer, which is important for jet simulations. The resulting IBM is then applied to analyze the flow-field and noise signature of a round and a chevron nozzle at a subsonic operating condition. The nozzles are based on geometries previously investigated by experimentalists. The lack of experimental flow-field measurements inside the nozzles prevents an agreement of turbulence levels in the initial shear layer of the simulated jets with their experimental counterparts. This mismatch has resulted in systematic and justifiable differences between the experimental and simulated flow-fields and far-field noise. However, the simulations exhibit an excellent qualitative agreement with the experiments for the contrast between the flow-fields of the two nozzles. The quantitative difference in far-field noise levels of the two nozzles is captured very well by the simulations. The current IBM can therefore be used to study nozzle concepts that have not yet been tested experimentally. It can also serve as a versatile research tool for problems other than jet noise.




Lyrintzis, Purdue University.

Subject Area

Aerospace engineering

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server