On the development of a jet noise prediction methodology

Evangelos Konstantinos Koutsavdis, Purdue University

Abstract

Computational Aeroacoustics (CAA), has emerged as a very important discipline and attracted significant attention lately. Strict sound emission regulations for aircraft and helicopters dictate the need to accurately predict the generation of sound emissions and their propagation in the far field. New computational tools are required employing high-order scheme with low dispersion and dissipation errors. A study of a particular methodology, which relies on the use of highly accurate implicit compact finite difference schemes and implicit spatial filtering is presented in this work. A compact sixth-order method is employed. Spatial filtering is used as the means of rejecting unwanted numerical waves that emanate from the boundaries of the computational domain as a result of the discrete implementation of boundary conditions. State-of-the-art non-reflecting boundary conditions (i.e. Tam and Web's boundary conditions), that allow acoustic waves to freely exit the computational domain are used. The above numerical methodology is the core of a new parallel Direct Numerical Simulation (DNS) code for compressible jet to be used for aeroacoustic purposes. This study presents an extensive validation of the core numerical methodology using two and three-dimensional jets and comparisons with analytical results with excellent agreement.

Degree

Ph.D.

Advisors

Blaisdell, Purdue University.

Subject Area

Aerospace engineering

COinS