Date of Award

Spring 2014

Document Type


Degree Name

Doctor of Philosophy (PhD)


Aerospace Engineering

Committee Director

Oktay Baysal

Committee Member

Duc Nguyen

Committee Member

Onur Bilgen


When turbulent flow generates unsteady differential pressure over an aircraft's structure, this may generate buffeting, a random oscillation of the structure. The buffet phenomenon is observed on a wide range of fighter aircraft, especially fighters with twin-tail. More research is needed to better understand the physics behind the vortical flow over a delta wing and the subsequent tail buffet.

This dissertation reports the modeling and simulation of a steady-state one-way fluid-structure interaction for the tail buffet problem observed on a F/A-18 fighter. The time-averaged computational results are compared to available experimental data. Next, computations are extended to simulate an unsteady two-way fluid-structure interaction problem of the tail buffet of a F/A-18 fighter.

For the modeling herein, a commercial software ANSYS version 14.0, is employed. For the fluid domain, the unsteady Reynolds-averaged Navier Stokes (URANS) equations with different turbulent models are utilized. The first turbulence model selected is the modified Spalart-Allmaras model (SARRC) with a strain-vorticity based production and curvature treatment. The second turbulence model selected is the Non-linear Eddy Viscosity Model (NLEVM) based on the Wilcox k–ω model. This model uses the formulation of an explicit algebraic Reynolds stress model. The structural simulation is conducted by a finite element analysis model with shell elements. Both SARRC and NLEVM turbulence models are in ANSYS software.

The experimental data used for validation were conducted on a simplified geometry: a 0.3 Mach number flow past a 76-deg delta wing pitched to 30-deg. Two vertical tails were placed downstream of the delta wing.

The present work is the first ever study of the tail buffet problem of the F/A-18 fighter with two-way fluid-structure interaction using the two advanced turbulence models. The steady-state, time-averaged, one-way fluid-structure interaction case of the present investigation indicates that simulations employing the NLEVM and SARRC turbulence models do not match the experimental data. These results are somewhat expected for the steady-state, one-way simulation, because it involves no force and displacement transfer between the fluid and structural domains.

For the unsteady two-way fluid-structure interaction case, both models result in more favorable agreement with the experimental data by optimizing the available computational resources particularly when compared to prior simulations by other researchers. Results from the NLEVM model produce improved pressure predictions on the tail as compared to the results from the SARRC model.

Based on the simulation results, it is concluded that the buffet problem should be simulated as a two-way fluid-structure interaction. The NLEVM turbulence model is recommended in predicting vortical flow characteristics over a delta wing. The NLEVM turbulence model is necessary to predict the pressure distribution not only over the aircraft surface but also the tails since they experience the wake of vortices.