Ezgi S. Taskinoglu

Multi-Objective Shape Optimization Study for a Subsonic Submerged Inlet

The findings of a multi-objective shape optimization study conducted for a subsonic submerged air vehicle inlet are summarized. The objective functions of the optimization problem are distortion and swirl indices defined by the distribution of flow parameters over the exit cross section of the inlet. The geometry alteration is performed by placing a protrusion in the shape of a fin on the baseline inlet surface. Therefore, the design variables of the optimization problem are chosen to be the geometrical parameters defining the fin protrusion, namely, fin height, length, and incidence angle. The trade-off (also known as the e-constraint) method is employed for finding the Pareto optimal set formed by the nondominated solutions of the feasible design space. Because the flow-domain solution is required for every step along the line search, an automated optimization loop is constructed by integrating the optimizer with a surface modeler, a mesh generator, and a flow solver through which the flow parameters over the compressor face are computed. In addition, the trade study for fin protrusion, and the analyses and the comparison of the baseline and Pareto optimal solutions are presented, and observations concerning grid resolution and convergence behavior are discussed. Nomenclature DC(φ) = distortion index f (x) = objective function g(x) = constraint function hfin = height of the fin protrusion, m lfin = length of the fin protrusion, m

Computational fluid dynamics evaluation of bleed slot of purdue mach 6 quiet tunnel

Introduction T HE Purdue Mach 6 quiet tunnel is a Ludwieg tube that was designed to study boundary-layer transition at hypersonic speeds.1 For high-quality transition research, the noise level in the test section should be comparable to flight and an order of magnitude lower than in conventional wind tunnels.2 To achieve these low “quiet” noise levels, laminar boundary layers must be maintained on the nozzle walls. NASA Langley Research Center pioneered the development of such quiet tunnels, with features that include highly polished nozzles and a bleed slot to remove the contractionwall boundary layer upstream of the throat.3 However, the tunnel (Fig. 1a), which has been operational since 2001, has achieved quiet flow at high Reynolds numbers only very recently. The probable cause for the lack of quiet flow during 2001–2005 is thought to be fluctuations generated at the nozzle throat as a result of flawed bleed slot design (as in Benay et al.4). Klebanoff and Tidstrom5 showed experimentally that if the Reynolds number is sufficiently small, the presence of a separation bubble does not alter the stability characteristics of the boundary layer downstream of the bubble; however, for sufficiently high Reynolds number, the presence of a separation bubble destabilizes the boundary layer downstream of reattachment, leading to an earlier transition to turbulence. Thus, the original design of the bleed slot has been modified, and altogether eight different bleed slot designs have been built and tested using the highly polished electroformed throat.6 The objective of the study presented in this Note is to simulate the flow in the bleed slot of the Purdue Mach 6 quiet wind tunnel. In the course of this computational study, the bleed slot design

Data Driven Design Optimization Methodology Development and Application

The Data Driven Design Optimization Methodology (DDDOM) is a Dynamic Data Driven Application System (DDDAS) developed for engineering design optimization. The DDDOM synergizes experiment and simulation in a concurrent integrated software system to achieve better designs in a shorter time. The data obtained from experiment and simulation dynamically guide and redirect the design optimization process in real or near-real time, and therefore realize the full potential of the Dynamic Data Driven Applications Systems concept. This paper describes the DDDOM software system developed using the Perl and Perl/Tk programming languages. The paper also presents the first results of the application of the DDDOM to the multi-objective design optimization of a submerged subsonic inlet at zero angle of attack and sideslip.

Numerical Analysis of the Bleed Slot Design of the Purdue Mach 6 Wind Tunnel

One of the major challenges in hypersonic flow research is the accurate prediction of transition location. There is a wide range of scatter in transition data obtained from numerous experiments conducted since the 1960’s. It is believed that most of the experimental data obtained in conventional ground test facilities are contaminated by high levels of noise due to acoustic fluctuations from the turbulent boundary layers on the wind tunnel walls. One method to reduce noise is to delay boundary layer transition using, among other features, bleed slots before the nozzle throat. The Purdue Mach 6 quiet tunnel has been designed to reduce the noise level an order of magnitude below that in conventional facilities. However, in this tunnel quiet flow is achieved only for low Reynolds numbers. For high Reynolds numbers the level of noise measured in the test section is dominated by early transition of the boundary layer on the tunnel walls which now appears to be caused by laminar boundary layer separation at the bleed slot. This paper summarizes the numerical analysis performed for one of the bleed slot designs of the Purdue Mach 6 quiet tunnel. The objective of the numerical study is to understand the eect of bleed slot design on the wind tunnel performance where the slot is considered as an upstream perturbation source. The steady state numerical solutions for two dierent stagnation pressures are obtained and analyzed using a commercial flow solver GASPex, version 4.1.0+. To quantify the numerical error in the solutions, a grid sensitivity analysis is conducted. For comparison purposes the Langley Mach 6 quiet tunnel is also analyzed numerically. The presence of large separation on the bleed lip of the unsuccessful Purdue tunnel, and the absence of large separation on the successful Langley tunnel, suggests that the separations are the most likely cause of early transition in the Purdue facility.