Date of Award

Winter 1999

Document Type


Degree Name

Doctor of Philosophy (PhD)


Mechanical & Aerospace Engineering


Engineering Mechanics

Committee Director

Oktay Baysal

Committee Member

Osama Kandil

Committee Member

Ponnampalam Balakumar

Committee Member

Fang Hu


In the last several years, there has been a growing demand for mobility to cope with the increasing population. All kinds of transportation have responded to this demand by expanding their networks and introducing new ideas. Rail transportation introduced the idea of high-speed trains and air transportation introduced the idea of high-speed civil transport (HSCT). In this expanding world, the noise legislation is felt to inhibit these plans. Accurate computational methods for noise prediction are in great demand.

In the current research, two computational methods are developed to predict noise propagation in air. The first method is based on the finite differencing technique on generalized curvilinear coordinates and it is used to solve linear and nonlinear Euler equations. The dispersion-relation-preserving scheme is adopted for spatial discretization. For temporal integration, either the dispersion-relation-preserving scheme or the low-dispersion-and-dissipation Runge-Kutta scheme is used. Both characteristic and asymptotic nonreflective boundary conditions are studied. Ghost points are employed to satisfy the wall boundary condition. A number of benchmark problems are solved to validate different components of the present method. These include initial pulse in free space, initial pulse reflected from a flat or curved wall, time-periodic train of waves reflected from a flat wall, and oscillatory sink flow. The computed results are compared with the analytical solutions and good agreements are obtained. Using the method developed, the noise of Mach 2.1, perfectly expanded, two-dimensional supersonic jet is computed. The Reynolds-averaged Navier-Stokes equations are solved for the jet mean flow. The instability waves, which are used to excite the jet, are obtained from the solution of the compressible Rayleigh equation. Then, the linearized Euler equations are solved for jet noise. To improve computational efficiency, flow-adapted grid and a multi-block time integration technique are developed. The computations are compared with the experimental results for both the mean flow and the jet noise. Good agreement is obtained. The method proved to be fast and efficient.

The second computational method is based on the boundary element technique. The Helmholtz equation is solved for the sound field around a railway noise barrier. Linear elements are used to discretize the barrier surface. Frequency-dependent grids are employed for efficiency. The train noise is represented by a point source located above the nearest rail. The source parameters are estimated from a typical field measurement of train noise spectrum. Both elevated and ground-level train decks are considered. The performance of the noise barrier at low and high frequencies is investigated. Moreover, A-weighted sound pressure levels are calculated. The computed results are successfully compared with field measurements.


In Copyright. URI: This Item is protected by copyright and/or related rights. You are free to use this Item in any way that is permitted by the copyright and related rights legislation that applies to your use. For other uses you need to obtain permission from the rights-holder(s).