View Video Presentation: https://doi.org/10.2514/6.2022-2307.vid In this study, the influence of the turbulence model, grid resolution and flow solvers on the results of RANS simulations is investigated. The work is part of the NATO AVT Task Group AVT-316 (Vortex Interaction Effects Relevant to Military Air Vehicle Performance). Simulations have been carried out for a generic transonic missile configuration by various organizations and compared with data from wind tunnel experiments for a Mach number of M = 0.85, a roll angle of $\boldsymbol{45^\circ}$ and total incidences in the range of $\boldsymbol{15^\circ\leq\sigma\leq17.5^\circ}$. In the first part of the study each organization used their own computational mesh and best practice for the grid generation. In the second part a common mesh family (three grids) were used for the simulations. It was shown that even for the common mesh family, the results of the different flow solvers in some cases showed large variations. This applies in particular to the rolling moment. Here, the accurate flow solver-dependent prediction of the complicated flow topology on the large wings and the accurate prediction of the location of the leeward vortices have a particularly strong effect. Only one flow solver predicted this flow in a form that gave good agreement with the wind tunnel results over the entire angular range ($\boldsymbol{15^\circ\leq\sigma\leq17.5^\circ}$).

Mesh adaptation methods for the Reynolds-averaged Navier--Stokes (RANS) equations have matured for production aerodynamic analysis with Computational Fluid Dynamics (CFD). However, the pace of improvements that address known weaknesses in RANS turbulence modeling have slowed and may stagnate, which may limit the application of RANS prediction tools. Wall-modeled large eddy simulation (WMLES) and hybrid RANS/LES (HRLES) offer the potential of improved modeling capability, but require specialized expertise to construct appropriate meshes and are considered too expensive for routine practical use. The realization of the CFD Vision 2030 Study includes geometry linkage, mesh generation/adaptation, turbulence modeling/resolving methods, and management of uncertainties to enable certification or qualification by analysis. The blending of RANS and LES in HRLES is used to control interpolation error anisotropically for modeled turbulence and isotropically for resolved turbulence. A floor on spacing that is appropriate for the wall model is imposed for WMLES. Qualitative results are presented for viscous-dominated flat plate flows and quantitative results are presented for pressure-dominated high-lift flows. Research and development opportunities are identified to advocate for continuing investments that may allow HRLES and WMLES to enter routine practical use as a tool for aerospace vehicle analysis and design.

Limiters commonly used in the simulations of flows with discontinuities are compared with the new limiter function proposed by Nishikawa using idealized test cases in two dimensions as well as complex three-dimensional problems. The Nishikawa limiter is observed to be consistently the least dissipative in idealized test cases as well as complex practical problems, for both steady and time-dependent simulations. In the case of steady simulations, its iterative convergence characteristics are either similar or better than other limiter functions. The nearfield sonic boom signature of a low-boom demonstrator is computed to demonstrate the utility of the Nishikawa limiter function for realistic supersonic aircraft configurations.

This paper reports an angle-of-attack (alpha) sweep study with mesh adaptation for two high-lift configurations: the JAXA Standard Model (JSM) and the High-Lift Common Research Model (CRM-HL) from the 3rd and the 4th AIAA High-Lift Prediction Workshops, respectively. Flow solutions are obtained by solving Reynolds-averaged Navier-Stokes equations with negative Spalart-Allmaras turbulence model (SA-neg), using a finite-element solver, FUN3D-SFE. The method of alpha continuation, where the flowfield at a certain angle-of-attack is initialized from the solution obtained at the previous angle-of-attack, is applied to adapted unstructured meshes. Since the meshes are continuously adapted, flowfield initialization on a particular mesh for a certain angle-of-attack�is done through linear interpolation of the solution from the previous mesh. Interpolated initial conditions from a lower angle-of-attack solution usually prevent massively separated solutions that may otherwise occur with impulsive initial conditions. For both JSM and CRM-HL models, forward alpha continuations (increasing alpha) are performed. The resulting solutions and the meshes, at selected �, are further adapted to larger meshes to get mesh converged force values. Starting from a post-stall solution, a reverse alpha�continuation (decreasing alpha) is also done for the JSM model, resulting in the formation of a hysteresis loop. For both models, the skin frictions and streamlines are compared with the wind-tunnel oil flow images, and the lift and pressure coefficients are compared with the wind-tunnel values.

View Video Presentation: https://doi.org/10.2514/6.2022-1176.vid A research program has been underway for four years to study vortex interaction aerodynamics that are relevant to military air vehicle performance. The program has been conducted under the auspices of the NATO Science and Technology Organization (STO), Applied Vehicle Technology (AVT) panel by a Task Group with the identification of AVT-316. The Missile Facet of this group has concentrated their work on the vortical flow field around a generic missile airframe and its prediction via computational methods. This paper focuses on mesh-related effects and RANS simulations. Simulated vortex characteristics were found to depend strongly on the properties of the employed mesh, in terms of both resolution and topology. Predictions of missile aerodynamic coefficients show a great dependence on mesh properties as they are sensitive to computed vortex dynamics. Key suggestions about the desired mesh characteristics have been made. Based on these, a shared mesh was constructed to perform common analyses between the AVT-316 Missile Facet members. Mesh based uncertainties of the aerodynamic coefficient predictions were estimated via Richardson Extrapolation method.

Within the framework of the NATO Science and Technology Organization Applied Vehicle Technology Task Group AVT316 calculations have been made of the supersonic flow around a slender body with wings and fins. In this paper a synthesis of the results obtained using the Reynolds Averaged Navier-Stokes equations are presented. The results show significant sensitivity to the choice of turbulence model. Whilst the gross features of the flow are similar, details of the development of the leeward wake are different. Simple linear eddy viscosity models predict vortices that rapidly decay, resulting in weak interactions with the downstream fins and relatively small rolling moments. This is attributed to an over production in turbulence quantities that results in excessive effective turbulent viscosity. Interventions that limit the production of turbulence, for example the SST limiter or curvature corrections, results in vortices that grow more slowly, changing the nature of the downstream interactions resulting in increased rolling moment. The use of more complex formulations, such as Reynolds stress models, that are inherently more capable for highly strained flows, further limits the rate of growth of the vortex cores leading to rolling moment predictions that are 2-3 times greater than those obtained with the simplest models.

This paper describes in detail the FUN3D Mesh adaptation simulation results for Case 1a of the 4th AIAA High-Lift Prediction Workshop (HLPW-4). The Stabilized Finite Element library with FUN3D is used along with the Negative Spalart-Allmaras One-Equation Model to simulate fully turbulent free-air flow around the High Lift version of the Common Research Model at three flap settings. Meshes are adapted using multi-scale and goal-oriented metrics. The results are presented and the discussion focuses on the predicted differences between the three flap settings. Coefficients of lift, drag, and pitching moment are predicted within 7% of values measured in the QinetiQ 5m Wind Tunnel at each flap setting. Predicted surface pressures and skin friction distributions show inconsistent discrepancies with measurements across the flap settings. Consequently, flap increments for these figures of merit are not well predicted. These predictions are consistent with submissions to the HLPW-4 from other RANS-SA methods. Visualizations of Liutex vortex cores colored by angular velocity are presented as possible new quantitative and qualitative perspective for solution evaluation.

Wall-modeled large-eddy simulation (WMLES) capability has recently been implemented into FUN3D, an unstructured, node-centered, finite-volume solver developed at the NASA Langley Research Center. In this paper, WMLES is assessed for two configurations that are representative of high-lift applications. The first configuration is a nominal two-dimensional multielement airfoil that has been extensively studied in the literature. WMLES solutions are computed for four angles of attack and compared with previously reported solutions. Good agreement of integrated forces, surface pressures, and boundary-layer velocity profiles is shown with available experimental data especially at lower angles of attack. WMLES solutions are also computed for the NASA High-Lift Common Research Model over a large range of angles of attack. Forces, pitching moments, and pressure distributions are favorably compared with the experimental data up to the maximum lift, including the angle of attack where the maximum lift is obtained experimentally. Eddy visualization techniques of q-criterion and density-gradient magnitude illustrate the resolved content.

View Video Presentation: https://doi.org/10.2514/6.2022-1178.vid Vortex dominated flows provide a challenge for modern computational fluid dynamics (CFD) solvers to accurately predict and capture the true physics of the flow. Modeling the formation and propagation of vortices downstream, as well as interactions with other vortices and shocks, are affected by a number of decisions including but not limited to the mesh generated, numerical scheme employed, and modelling assumptions. A research program has been underway for four years to study vortex interaction aerodynamics that are relevant to military air vehicle performance. The program has been conducted under the auspices of the NATO Science and Technology Organization (STO), Applied Vehicle Technology (AVT) panel by a Task Group with the identification of AVT-316. The paper at hand will look at the OTC1 test case [1], which is comprised of a generic missile configuration in supersonic flow, and the influence of the numerical scheme on the predicted results. The decision of the spatial discretization scheme, order of the turbulent flux, and choice of limiter were all shown to have strong influence on the predicted rolling moment of the OTC1 test case.

Discretization error is a common source of uncertainty in computational fluid dynamics (CFD) analyses. Traditional means of controlling discretization error through fixed-mesh refinement studies have proved to be difficult particularly when modeling complex geometries and flowfields. One reason for this is that mesh generation in today’s production CFD workflow is often a labor-intensive process that is heavily dependent on user judgment. Unstructured mesh adaptation is known to be an efficient way to control discretization errors in CFD. Adaptive methods replace user-based decision-making with automated processes that optimize a mesh to reduce discretization error. This paper compares the application of multiple solution adaptive techniques in combination with multiple flow solvers to solve for the flowfield about a 2-D airfoil section of the NASA High-Lift Common Research Model. By driving the adaptive mesh processes to a similar level of mesh convergence, the ability to achieve consistent results between multiple adaptive techniques and flow solvers is demonstrated. Mesh convergence for the various adaptive mesh approaches is compared, identifying potential areas for improvement and providing mesh generation guidance for future workshops.

Unstructured anisotropic mesh adaptation is known to be an efficient way to control discretization errors in computational fluid dynamics simulations. Method verification is required to provide the confidence for routine use in production analysis. The current work aims at verification of anisotropic mesh adaptation for Reynolds-averaged Navier-Stokes simulations over the ONERA M6 wing. The present verification study is performed using four different flow solvers, three different implementations of the metric field, and three mesh mechanics packages. Two of the flow solvers use stabilized finite element discretizations (FUN3D-SFE and general geometry Navier-Stokes), one uses finite volume discretization (FUN3D-FV), and the last one uses mixed finite volume and finite element discretizations (Wolf). The mesh adaptation is based on an error estimator that aims to control the quadratic error term in the linear interpolation of the Mach number. Two sets of adaptations were performed; the first one controls the interpolation error in the L2 norm, and the second one controls the interpolation error in the L4 norm. Convergence studies were performed on the forces and the pitching moment using all four solvers, and the results are compared with previously verified convergence studies on fixed (nonadapted) meshes. Both the forces and pitching moment on adapted meshes are found to be converging to the fine-mesh values faster than those on fixed meshes. In addition to forces and moments, the convergence of surface pressure and skin-friction coefficients at various measurement locations on the wing are also presented. Adapted-mesh surface pressure distributions agree with the fine fixed mesh pressure distributions. Adapted-mesh skin-friction distributions contain high-frequency noise with mean values approaching the fixed mesh pressure skin-friction distributions.

The stabilized finite element solver, FUN3D-SFE, along with the grid mechanics package \refine are verified for aerospace applications of laminar and turbulent flow simulations. The current verification exercise represents an extension of previous research using FUN3D-SFE with adjoint-based mesh adaptation to generate highly anisotropic adapted meshes for inviscid problems. Adaptations are performed using a solution-based approach that controls the $L^p$ norm of Mach number interpolation error and an adjoint-based approach that controls the error in some output functional. Adaptive results are shown for laminar subsonic flow over a delta wing, laminar subsonic flow over an ONERA M6 wing, inviscid supersonic flow over a sonic boom test case, and a turbulent flow over a high lift configuration (JAXA Standard Model). Mesh convergence results are also compared with results from FUN3D-FV (Finite Volume) whenever available. The presented comparisons between FUN3D-FV and FUN3D-SFE concern only the accuracy in terms of the number of degrees of freedom, and not on the computational cost. It is observed that for the same number of degrees of freedom, FUN3D-SFE gives more accurate results than FUN3D-FV for all the cases where both FUN3D-SFE and FUN3D-FV results are available.

Thanks to recent advances in the fields of anisotropic grid adaptation, error estimation, and geometry modeling, a sketch-to-solution work flow is now possible for viscous computational fluid dynamic (CFD) simulations. With this work flow, a CFD application engineer provides geometry, boundary conditions, and flow parameters; and the sketch-to-solution process yields a CFD simulation through automatic, error-based, grid adaptation. To explore the benefits of this nascent capability, a conventional manual grid generation work flow is compared to this new automatic grid generation work flow for a given engineering question: What are the aerodynamic interactions caused by the reaction control system (RCS) on an entry vehicle? This case study indicates that while the automatic grid generation sketch-to-solution process is not yet mature, it is preferred over a manual grid generation work flow because it greatly reduces manual labor, eliminates many opportunities for human error, and provides grid sensitivity information.

Mesh adaptation methods for the Reynolds-averaged Navier–Stokes (RANS) equations are rapidly maturing and beginning to impact the design of aerospace vehicles. RANS turbulence modeling improvements have slowed and may stagnate. Wall-modeled large eddy simulation (LES) and hybrid RANS/LES (HRLES) may provide an improved modeling capability, but require specialized expertise to construct appropriate meshes and are considered be too expensive for routine practical use. The realization of the CFD Vision 2030 Study includes improving geometry linkage, mesh generation/adaptation, and turbulence modeling/resolving methods for automated management of errors and uncertainties of physics-based, predictive modeling that can set the stage for ensuring a vehicle is in compliance with a regulation or specification (i.e., certification or qualification by analysis). An exploration of mesh adaptation for HRLES is performed to document synergies and challenges between mesh adaptation and HRLES. Vortex breakdown over a delta wing is examined to show the improvement of HRLES over RANS turbulence modeling approaches. A high lift configuration is shown to demonstrate complex geometry capability. Research and development opportunities are identified to advocate for continuing investments that may allow HRLES to enter routine practical use as a tool for aerospace vehicle analysis and design.

Unstructured grid techniques have the potential of minimizing discretization errors for production analysis workflows where the control of errors is critical to obtaining reliable simulation results. Recent progress has matured a number of independent implementations of flow solvers, error estimation methods, and anisotropic grid adaptation mechanics. Here, the interoperability of several separately developed unstructured grid adaptation tools is verified using analytic functions and the Code Comparison Principle. Three analytic functions with different smoothness properties are adapted to show the impact of smoothness on implementation differences. A scalar advection-diffusion problem with an analytic solution that models a boundary layer is adapted to test individual grid adaptation components. While optimal asymptotic error convergence rates are achieved with many grid adaptation tool combinations for the scalar problems, the scalar problems also illustrate known differences in grid adaptation component implementations and a previously unknown interaction between components. Laminar flow over a delta wing is verified with multiple, independent grid adaptation procedures to show consistent convergence to fine-grid forces and pitching moment. These verification efforts form the nucleus of a benchmark to verify the integration of unstructured grad adaptation components and support production analysis workflows.

The computational fluid dynamics (CFD) prediction workshops sponsored by the AIAA have created invaluable opportunities to discuss the predictive capabilities of CFD in areas in which it has struggled, e.g., sonic boom prediction. While there are many factors that contribute to disagreement between simulated and experimental results, such as modeling or discretization errors, quantifying the errors contained in a simulation is important for those who make decisions based on the computational results. The linearized error transport equations (ETE) combined with a truncation error estimate is a method to quantify one source of error. The ETE are implemented with a complex-step method to provide an exact linearization with minimal source code modifications to CFD and multidisciplinary analysis methods including atmospheric propagation of sonic boom signatures. Uniformly refined grids from the 2nd AIAA Sonic Boom Prediction Workshop demonstrate the utility of ETE for multidisciplinary analysis with a connection between estimated discretization error and (resolved or under-resolved) flow features.

A summary is provided for the Second AIAA Sonic Boom Workshop held on January 8-9, 2017, in conjunction with AIAA SciTech 2017. The workshop used four models of increasing complexity: an axisymmetric body, a wing-body, and a wing-body with tail configuration with either flow-through nacelles or propulsion boundary conditions. The last configuration was optional. The models were designed with similar nearfield signatures to isolate geometry and shock/expansion interaction effects. These cases are more relevant to vehicles with lower ground loudness and have the potential for lower annoyance than those in previous workshops. Eleven international participant groups submitted nearfield signatures. In addition, noise levels were computed by propagating the nearfield signatures to the ground. This allowed the evaluation of grid convergence and statistical distribution with respect to the noise metric. Prediction dramatically improved for the wing-body with tail and flow-through nacelle case between the first and second workshops. The models for the second workshop with quieter ground noise levels than those in the first workshop exposed weaknesses in analysis, particularly in convective discretization. While progress is documented since the first workshop, improvement to the analysis methods for a possible subsequent workshop is also provided.

Free AccessSecond Sonic Boom Prediction WorkshopIntroduction to the Special Section on Second AIAA Sonic Boom Prediction WorkshopMichael A. ParkMichael A. ParkNASA Langley Research Center, Hampton, Virginia 23681*Research Scientist, Computational AeroSciences Branch, 15 Langley Blvd. MS 128. Associate Fellow AIAA.Search for more papers by this authorPublished Online:31 May 2019https://doi.org/10.2514/1.C035323SectionsRead Now ToolsAdd to favoritesDownload citationTrack citations ShareShare onFacebookTwitterLinked InRedditEmail AboutDecades of sustained research in areas that support supersonic commercial transport technology have resulted in an increased interest in concept development and flight testing. The Second AIAA Sonic Boom Prediction Workshop (SBPW2) was held 8–9 January 2017 in conjunction with AIAA SciTech 2017 to examine the sonic boom simulation capabilities of the international community. The data and analysis methods of the workshop are intended to contribute toward the discussion of replacing the prohibition of overland supersonic flight with a certification standard. This special section provides summaries and statistical analysis of the submissions to the nearfield Computational Fluid Dynamics (CFD) and atmospheric propagation workshop test cases. Detailed method and submission descriptions from a number of the participants are also included in this special section.While the first Sonic Boom Workshop (SBPW1) [1] focused on nearfield sonic boom prediction, the SBPW2 included both nearfield prediction and atmospheric propagation test cases. The nearfield cases were based on three geometric models of increasing complexity: an axisymmetric body, a wing body, and a wing body with tail and flow-through nacelle. An optional wing body with tail and propulsion boundary conditions was also provided. These models were designed with similar nearfield signatures to isolate geometry and shock/expansion interaction effects. Eleven international participant groups submitted nearfield CFD signatures. Statistics and grid convergence of these nearfield signatures were compiled. These submissions were propagated to the ground, and noise levels were computed. Analysis of the noise levels allowed grid convergence and statistical distribution to be assessed.Significant progress was documented between SBPW2 and SBPW1, where the wing body with tail and flow-through nacelle case was quieter with similar or lower levels of variation. While significant progress was documented since the first workshop, the optional case with propulsion boundary conditions motivate the continuing need for development of methods that increase the reliability of sonic boom simulation.SBPW2 introduced atmospheric propagation test cases to compliment the evaluation of nearfield CFD methods. Nearfield pressure waveforms for two cases were supplied by the workshop committee and ground signatures at multiple azimuthal angles and their corresponding loudness metrics were supplied by ten participants from three countries. The propagation cases used temperature and pressure profiles from the US Standard atmosphere, coupled with two different humidity profiles. The propagation cases also included atmospheric profiles with measured data including winds. These profiles were selected using radiosonde balloon data at multiple geographically spread locations.The atmospheric propagation submissions confirm that the ground sonic boom waveforms, locations, and loudness values depend greatly on the atmospheric profile used in the prediction. Results with measured profiles differ significantly from those with a standard atmosphere profile, particularly at off-track locations. The measured atmospheric profiles produced larger variation in computed noise measures than the US Standard atmosphere profile. The impact of varying humidity profiles was small compared to varying wind or temperature profiles.In this special section, seven articles related to SBPW2 were published in Volume 56, Issue 3 of the Journal of Aircraft. Summaries and statistical analysis of the nearfield CFD [2] and atmospheric propagation [3] submission are provided. Detailed methods and submission descriptions are included for participants that used Cart3D [4], DLR-TAU [5], USM3D [6], Wolf [7], and Linearized Error Transport Equations [8].The SBPW2 participants and organizers would like to acknowledge the support of the AIAA staff and Applied Aerodynamics Technical Committee in organizing both Boom Prediction Workshops and Conference Special Sessions. The other AIAA workshops provided an invaluable source of support and guidance in organizing and documenting SBPW2. We hope that these workshops and this special section continue to foster continued discussion and research interest in eliminating technical barriers to practical commercial supersonic flight. References [1] Park M. A. and Morgenstern J. M., “Summary and Statistical Analysis of the First AIAA Sonic Boom Prediction Workshop,” Journal of Aircraft, Vol. 53, No. 2, 2016, pp. 578–598. doi:https://doi.org/10.2514/1.C033449 LinkGoogle Scholar[2] Park M. A. and Nemec M., “Near Field Summary and Statistical Analysis of the Second AIAA Sonic Boom Prediction Workshop,” Journal of Aircraft, Vol. 56, No. 3, 2019. doi:https://doi.org/10.2514/1.C034866 Google Scholar[3] Rallabhandi S. K. and Loubeau A., “Propagation Summary and Statistical Analysis of the Second AIAA Sonic Boom Prediction Workshop,” Journal of Aircraft, Vol. 56, No. 3, 2019. doi:https://doi.org/10.2514/1.C034805 LinkGoogle Scholar[4] Anderson G. R., Aftosmis M. J. and Nemec M., “Cart3D Simulations for the Second AIAA Sonic Boom Prediction Workshop,” Journal of Aircraft, Vol. 56, No. 3, 2019. doi:https://doi.org/10.2514/1.C034842 LinkGoogle Scholar[5] Kirz J. and Rudnik R., “DLR TAU Simulations for the Second AIAA Sonic Boom Prediction Workshop,” Journal of Aircraft, Vol. 56, No. 3, 2019. doi:https://doi.org/10.2514/1.C034819 LinkGoogle Scholar[6] Elmiligui A., Carter M. B., Nayani S., Cliff S. and Pearl J., “USM3D Simulations for 2nd Sonic Boom Workshop,” Journal of Aircraft, Vol. 56, No. 3, 2019. doi:https://doi.org/10.2514/1.C034831 LinkGoogle Scholar[7] Loseille A., Frazza L. and Alauzet F., “Comparing Anisotropic Adaptive Strategies on the 2nd AIAA Sonic Boom Workshop Geometry,” Journal of Aircraft, Vol. 56, No. 3, 2019. doi:https://doi.org/10.2514/1.C034840 LinkGoogle Scholar[8] Derlaga J. M., Park M. A. and Rallabhandi S. K., “Application of Exactly Linearlized Error Transport Equations to Sonic Boom Prediction Workshop,” Journal of Aircraft, Vol. 56, No. 3, 2019. doi:https://doi.org/10.2514/1.C034841 LinkGoogle Scholar Next article FiguresReferencesRelatedDetailsCited byIntroduction to the Special Section on the Third AIAA Sonic Boom Prediction WorkshopMichael A. Park and Elizabeth M. Lee-Rausch20 May 2022 | Journal of Aircraft, Vol. 59, No. 3 What's Popular Volume 56, Number 3May 2019Special Section on Second Sonic Boom Prediction Workshop CrossmarkInformationCopyright © 2018 by the American Institute of Aeronautics and Astronautics, Inc. The U.S. Government has a royalty-free license to exercise all rights under the copyright claimed herein for Governmental purposes. All other rights are reserved by the copyright owner. All requests for copying and permission to reprint should be submitted to CCC at www.copyright.com; employ the eISSN 1533-3868 to initiate your request. See also AIAA Rights and Permissions www.aiaa.org/randp. TopicsAerodynamicsAeronautical EngineeringAeronauticsAircraft Components and StructureAircraft DesignAircraft Design SoftwareAircraft Operations and TechnologyAircraftsCFD CodesComputational Fluid DynamicsFluid DynamicsShock Waves KeywordsSonic BoomsCFDStatistical DistributionsSupersonic FlightFlight TestingNacellesSupersonic Commercial TransportApplied AerodynamicsData AnalysisRadiosondesPDF Received29 October 2018Accepted29 October 2018Published online31 May 2019

The Common Research Model wing/body configuration is investigated with the k-kL-MEAH2015 turbulence model implemented in FUN3D. This includes results presented at the Sixth Drag Prediction Workshop and additional results generated after the workshop with a nonlinear quadratic constitutive relation variant of the same turbulence model. The workshop-provided grids are used, and a uniform grid refinement study is performed at the design condition. A large variation between results with and without a reconstruction limiter is exhibited on "medium" grid sizes, indicating that the medium grid size is too coarse for drawing conclusions in comparison with experiment. This variation is reduced with grid refinement. At a fixed angle of attack near design conditions, the quadratic constitutive relation variant yielded decreased lift and drag compared with the linear eddy-viscosity model by an amount that was approximately constant with grid refinement. The k-kL-MEAH2015 turbulence model produced wing-root junction flow behavior consistent with wind-tunnel observations.