Computation of turbulent wake flows in variable pressure gradient

Abstract

Transport aircraft performance is strongly in uenced by the e ectiveness of high-lift systems. Developing wakes generated by the airfoil elements are subjected to strong pressure gradients and can thicken very rapidly, limiting maximum lift. This paper focuses on the e ects of various pressure gradients on developing symmetric wakes and on the ability of a linear eddy viscosity model and a nonlinear explicit algebraic stress model to accurately predict their downstream evolution. In order to reduce the uncertainties arising from numerical issues when assessing the performance of turbulence models, three di erent n umerical codes with the same turbulence models are used. Results are compared to available experimental data to assess the accuracy of the computational results. I n troductionWithin the airframe industry, the optimal design of e cient high-lift devices for take-o and landing conditions is an important issue and one of the most challenging problems. For the current generation of commercial aircraft, such high-lift systems require a simplicity of design but a high level of eciency. A c hieving this goal demands improved Computational Fluid Dynamics CFD design tools that can realistically model the high-lift system ow eld for full scale free ight Reynolds numbers. Diculties are inherent to both the geometric complexity o f m ulti-element airfoils as well as limitations Research Associate, Royal Institute of Technology, KTH, Department of Aeronautics, SE-10044 Stockholm, Sweden, Member AIAA.y Senior Research Scientist, Subsonic Aerodynamics Branch, NASA LaRC, Hampton VA, Senior Member AIAA.z Senior Research Scientist, Aerodynamic and Acoustics Methods Branch, NASA LaRC, Hampton VA, Senior Member AIAA.x Senior Research Scientist, Aerodynamic and Acoustics Methods Branch, NASA LaRC, Hampton VA.Copyright c 1999 by the American Institute of Aeronautics and Astronautics, Inc. No copyright is asserted in the United States under Title 17, U.S. Code. The U.S. Government has a royalty-free license to exercise all rights under the copyright claimed herein for government purposes. All other rights are reserved by the copyright o wner.in ow p h ysics modeling. The ow eld surroundi n g a m ulti-element airfoil is dominated by v ery complex viscous ow phenomena such as boundary layer transition, laminar separation bubbles, con uent turbulent boundary layers, multiple viscous wake interactions, and possible shock boundary layer interactions. For this reason, the demands are high on the numerical methods used in this type of calculation, and several crucial aspects must be treated in an adequate way. One of these aspects is turbulence modeling.In recent studies, 1,2 analyses of high-lift multielement airfoil con gurations were performed to assess the predictive capability of di erent t ypes of turbulence models. The studies revealed three areas where turbulence model predictions were possibly de cient and which a ected the overall prediction of the ow eld. These areas were the inability to predict transition loca...

A framework to develop data-driven turbulence models for flows with organised unsteadiness

“…Numerous studies using this closure have reported differences in the quantities of interest such as the mean velocity and Reynolds stress profiles compared with experimental and HiFi data [22,23,24,25,26,27]. To better understand the reason for differences, the validity of the Boussinesq hypothesis was, for example, tested by Schmitt [28] through computing the alignment of the modelled anisotropy tensor with the reference one given by ρ RS = a re f i j a mod ji a re f mn a re f nm a mod pq a mod qp ,…”

Numerical study of turbulent trailing-edge flows with base cavity effects using URANS

“…To this end, numerical methods using RANS provide a time efficient and cost effective approach. Application of steady RANS with k-e turbulence model to the prediction of zero-pressuregradient wakes of flat plates and two-dimensional lifting bodies is widespread; see for example Patel and Scheuerer (1982), Patel and Chen (1987), Tummers et al (2007), Carlson et al (2001), Nguyen and Gorski (1991) and Mulvany et al (2004), all of which found this turbulence model predicts the near-wake features reasonably well compared to the experimental data. Moreover, a recent study by Iaccarino et al (2003) claims that unsteady RANS (URANS) is highly capable of predicting flows with gross unsteadiness, given that the unsteadiness is deterministic and that the frequency spectrum shows a spike at the vortex shedding frequency.…”

Consistency, Verification, and Validation of Turbulence Models for Reynolds-Averaged Navier—stokes Applications

“…One way is through the method of manufactured solutions (MMS), but only limited solutions are available for turbulence (see, e.g., Eca et al [3]). Demonstrating consistency using the same model in different codes can provide some level of assurance [4], but optimally the testing should be conducted by independent groups [5] to minimize the possibility of repeating the same mistakes.…”