Goals and Status of the NASA Juncture Flow Experiment

The NASA Juncture Flow experiment is a new effort whose focus is attaining validation data in the juncture region of a wing-body configuration. The experiment is designed specifically for the purpose of CFD validation. Current turbulence models routinely employed by Reynolds-averaged Navier-Stokes CFD are inconsistent in their prediction of corner flow separation in aircraft juncture regions, so experimental data in the near-wall region of such a configuration will be useful both for assessment as well as for turbulence model improvement. This paper summarizes the Juncture Flow effort to date, including preliminary risk-reduction experiments already conducted and planned future experiments. The requirements and challenges associated with conducting a quality validation test are discussed. 1.0 INTRODUCTION Most turbulence models in wide use in computational fluid dynamics (CFD) are incapable of accurately predicting the flow physics that occur in juncture flow regions (i.e., flow along the intersection of two walls). For example, CFD computations at past Drag Prediction Workshops (e.g., Vassberg et al. [1]) have produced very large variations in the predictions of separation, skin friction, and pressure near the side-of-body (SOB) wingfuselage juncture, close to the wing trailing edge. While some linear eddy-viscosity turbulence models have shown excessively large regions of side-of-body separation, others have shown small regions of separation that are closer to what is seen in wind tunnel experiments. However, grid and numerical dissipation characteristics are often confounded with model effects. For example, use of a thin-layer approximation for viscous terms tends to yield smaller corner separation than full viscous terms for Reynolds-averaged Navier-Stokes. The widelyused Spalart-Allmaras linear eddy-viscosity model [2] predicts large regions of separation; however, replacing the linear eddy-viscosity relationship with a quadratic constitutive relationship [3] produces small side-of-body separation. Unfortunately, very few experimental details are available for such flows to distinguish submodel effects, in part because of the difficulties inherent in measuring data in the flowfield very close to walls. Some recent efforts have been aimed at improving knowledge of these flow physics [4], but much more remains to be done. Because multiple juncture flows are present on practically all civilian and military air vehicles, there is strong motivation to improve CFD’s capabilities to predict them. Oberkampf and Smith [5] identified four general categories of experiments based on the goal of each experiment: (1) physical discovery experiments, conducted primarily to improve the fundamental understanding of a physical process; (2) model calibration experiments, conducted to improve or determine parameters in https://ntrs.nasa.gov/search.jsp?R=20160012016 2017-11-07T04:08:49+00:00Z