CFD Analysis in Advance of the NASA Juncture Flow Experiment

Henry C. Lee, Thomas H. Pulliam, Dan Neuhart, Michael A. Kegerise
2017 47th AIAA Fluid Dynamics Conference   unpublished
even designed by committee members. These candidates were predominantly evaluated using Overflow. It became apparent that not all the initial design criteria could be met. Large side-of-body separation was heavily influenced by wing loading, and symmetric wings could not achieve that wing loading unless at very high angles of attack. Cambered twisted wings at even very negative angles of attack, still showed small side-of-body separation bubbles. Six wing configurations, which combined met all
more » ... he goals, were selected for further testing. Some of the wing candidates were designed with a leading-edge horn. Previous work by Barber 9 and Gand 6 showed that a leadingedge horn would change the strength of the horseshoe vortex and the resulting side-of-body separation. These 6 wing configurations will be discussed in section II. Up to this point CFD was the primary evaluator for the configurations, yet as described above, there are questions on how accurate CFD can be for juncture flows. The Juncture Flow Committee thus decided to commission a few smaller, low cost risk reduction experiments to help further evaluate the wing candidates for the final experiment. A series of three risk reduction tests were performed: 1. Fluid Mechanics Lab (FML) Test Cell 2 (TC2) at NASA Ames Research Center 3% semispan model, 2. Virginia Tech Stability Tunnel 2.5% fullspan model, 3. NASA Langley 14-by-22 Ft. Subsonic Wind Tunnel 6% fullspan model. CFD was run in conjunction with the tests to help better understand the flow, the wall effects, and possible differences in CFD and experimental methods. The TC2 risk reduction experiment was designed as a semi-span wall-mounted 3% model, targeting a lower Reynolds number of 0.62 million based on a Yehudi crank chord of 8.2 inches. 10 The correlation between the TC2's results and CFD were not as strong in both the side-of-body separation, and the wing-fuselage boundary layer surveys. The fuselage nose and wall junction produced a large horseshoe vortex that the CFD simulations could not fully capture. This test affirmed that the wall effects from a side-wall mounted model were nontrivial and could affect the side-of-body separation in a negative way. The Virginia Tech risk-reduction experiment was a sting-mounted fullspan 2.5% model test, at a low Reynolds number of 0.62 million based on a Yehudi crank chord of 6.9 inches. The CFD results did show stronger correlation with the Virginia Tech results, 11 which is probably due to the sting mounted model instead of a wall mounted model. The low Reynolds number did cause the wing tips to separate at the higher angles of attack, which both oil-flow and CFD results confirmed. This test confirmed a sting-mounted model was the ideal choice. The third risk reduction experiment was performed in NASA Langley's 14-by-22 Ft. Subsonic Wind Tunnel, with a 6% scale full-span model (versus the 8% final juncture flow experiment scale), at a higher Reynolds number of 2.4 million based on the Yehudi crank chord of 16.5 inches. At 6% scale, the wing span is roughly 98.3 inches, and the fuselage is approximately 8 feet long. The experiment will test the the 6 wings detailed in Section II. CFD simulations, using NASA's Overflow, 7 were performed as well in both free air and with wind tunnel walls. Some brief comparisons will be made between the CFD and the 6% risk reduction experimental results (for the rest of the paper, the 6% risk reduction experiment will be referred to as just the experiment). For this paper, results will be limited to predominantly CFD surface streamlines and experiment's oil flow results. Additional experimental results, including data from pressure taps and Kulite pressure transducers will not be shown, see Kegerise and Neuhart [12] for the full experimental results. This paper will first cover the selected wing configurations, with a brief description of each. The CFD setup and post-processing method will be presented next, followed by a brief description of the relevant 6% risk reduction experiment setup. The CFD and experimental results will be shown next, and a brief comparison of the results will be included.
doi:10.2514/6.2017-4127 fatcat:is2l2mbdsve25ihimzj5v6ibke