Validation of any numerical solver requires validation with experiments. This is especially true of aerodynamics and fluid-flow solvers. The Research in Flight team has validated the capabilities of the FlightStream® solver for a variety of aircraft configurations. Some of these are provided here as a first look into what this solver can do.
The AIAA SciTech 2018 conference was held in January, 2018, in Kissimmee, Florida, USA. During this conference, Terrafugia and Research in Flight presented a validation study of the actuator models for steady-state propeller simulations in FlightStream®. This validation study was performed on the NASA XC-142 V/STOL geometry. The outer mold line for this aircraft is modeled in Vehicle Sketch Pad (OpenVSP) using a scripted parameter linking that allows tilt-wing modeling of the four engine nacelles and the main wing. Analysis results for the powered and unpowered test conditions were validated against wind-tunnel data for the XC-142 1/11-scale wind-tunnel powered model. And best of all, it generates these results (the entire AOA sweep, actually!) in under ten minutes on a standard workstation using nothing but a surface mesh and vorticity!
The AIAA High-Lift Prediction Workshop 2 (HLPW-2) was held on June 22-23, 2013, in San Diego, California, USA. During this workshop, the CFD community presented their results for the European High Lift Program (EUROLIFT) DLR F11 high-lift configuration (shown above). In 2015, when FlightStream® had matured as a flow solver, we at Research in Flight also began our own tests on the configuration to determine what was possible. What makes the DLR-F11 unique is that it is a genuine high-lift configuration, with deployed flaps and slats, as well as the appropriate wing configuration for such devices. Since FlightStream® works on surface vorticity, and more importantly, on shedding this vorticity to generate aerodynamic loads, the DLR-F11 represented a true test of our proprietary vorticity wake algorithms. We developed these algorithms to work for "difficult" flows such as those found on high-lift geometries, and work it did, as you can see in some of the same plots below. FlightStream® ran this configuration with highly resolved vorticity wakes, and so the run time was longer than what we typically like, but even so, it was under five minutes per angle-of-attack setting. The plots below were generated in FlightStream® in an hour on a standard workstation with quad-cores.
The Piper Pa-24 was tested at NASA Langley Research Center's Full-Scale Wind Tunnel, in 1969, and the results were presented in March 1970, in a technical report, titled NASA TN D-5700. These tests included both powered and un-powered flight. The Piper Pa-24 represents a standard validation model for us at Research in Flight, given the vast data that exists in the NASA Technical Report for this model in both powered and un-powered conditions. The CAD model for the Piper Pa-24 was created by DARcorporation engineers. This validation study was performed by DARcorporation engineers early in 2017 to test FlightStream®'s ability to use its propeller actuators on a steady-state aircraft and capture accurate aerodynamic effects, measured in loads and moments, as well as pressure distributions. The results of this validation study are to be presented by DARcorporation engineers at the AIAA SciTech Conference, to be held in January 2018. Three plots from this study are shown below for both powered and un-powered conditions. The Pa-24 takes less than a minute to run to convergence for each flow incidence angle on a standard laptop or workstation.
The NASA EET geometry serves as a practical geometry for validation of the FlightStream solver. The specific model of the EET, namely EET-AR-12, was used by NASA Langley Research Center to conduct an independent validation of FlightStream® under three different conditions: cruise (with flaps and slats deployed), take-off and landing (flaps and slats deployed). These three configurations are shown in the images on the left. The NASA study was presented at the 2015 NASA OpenVSP workshop in Langley, Virginia and presented at the AIAA SciTech conference in 2016 (AIAA-2016-0779). Research in Flight now uses the EET model as a standard validation model for enhancements in the vorticity solver. Two of the plots from the latest version of FlightStream® are shown below as examples of the type of validation work performed. A sample result is also shown for the upcoming vorticity-based flow-separation implementation.
The Cessna-210 model with the NASA NLF(1)-0414F airfoil wing is a standard validation case for FlightStream®. A conventional, contemporary civilian aircraft, the NLF(1)-0414F airfoil is nevertheless an advanced airfoil design that sheds light on the ability of the FlightStream® solver to model the flow over advanced airfoil shapes. Further, the effects of the single propeller on the nose of the aircraft is modeled in FlightStream® using the propeller actuator and provides the correct effect on the wing and horizontal stabilizer. Some of the validation plots are reproduced below. The solver run time for each angle-of-attack result is less than 3 minutes on a standard quadcore laptop.
The First AIAA Drag Prediction Workshop (DPW-1) was held on June 9-10, 2001, in Anaheim, California, USA. During this two-day workshop, the focus was on the prediction of aerodynamic forces, especially drag, on the DLR-F4 geometry. In 2012, we at Research in Flight applied FlightStream® to the test on the DLR-F4 as well. The DLR-F4 is a good model to use to simulate steady-state, high-compressible flows around an early-generation commercial airliner configuration. While a range of Mach numbers were tested, with good results, the image above and the plots below are shown for the Mach 0.6 condition. For a feel of what FlightStream® can do at higher Mach numbers, see the more-recent CRM model from the 4TH Workshop further up on this page. At Mach 0.6, most of the flow remains subsonic on the model, and this makes the DLR-F4 a standard test for us at FlightStream. As the plot below shows, FlightStream® predicts all aerodynamic forces in accordance with experimental data, until the wing stalls. And while stall represents the current limits on FlightStream®, rest assured we are already in work to solve it. In the meantime, FlightStream® generates these result sweeps in a few minutes!
The NASA Rotor Body Interaction (ROBIN) model is a series of NASA wind-tunnel tests on a generic helicopter fuselage with a fully articulated rotor, called the General Rotor Model System (GRMS). The tests were done in the Langley V/STOL tunnel in the late 1970s. The results are published in the openly sourced NASA Technical Memorandum 80051: "Fuselage surface pressure measurements of a helicopter wind-tunnel model with a 3.15 Meter diameter single rotor". The ROBIN model is used by Army and NASA rotorcraft groups for code and measurement validation.
The downwash induced by a rotor on a helicopter fuselage and its empennage generally produces an effect on the overall vehicle performance. These effects are generally most evident in hover and low-speed flight as fuselage downloads and yawing moments. The flow around the fuselage also has a secondary effect of inducing nonuniform downwash through the rotor disk. Consequently, it is also reproduced here, for validation of FlightStream®'s ability to model rotor downwash effects on a fuselage in a steady-flow environment, both in hover and in forward flight. As the plot below shows, FlightStream® predicts surface pressure with good accuracy in both the no-rotor and rotor conditions. The downwash effects are modeled for steady state using the potential theory actuator model in FlightStream®, which allows these solutions to be generated in less than a minute for each flight condition.
The NASA HL-20 model was generated by the Langley Research Center and is available to download from the VSP geometry hanger. The HL-20 is a winged re-entry vehicle designed to glide back to a landing upon return from orbit. This geometry is a challenging scenario for any numerical flow solver. There are large base regions with separated flow and thick wings with blunt trailing edges. FlightStream® models the flow without the base region on the rear of the vehicle and a semi-empirical correction for the flow in the base region corrects the aerodynamic loads as shown in the plot below. The solver run time is less than one minute for the results shown here. The experimental data is obtained from NASA TM-4117, which can be downloaded from the NASA Technical Server. The FlightStream® results for the HL-20 were presented at the NASA OpenVSP workshop in 2015.
The Generic Wing-Body model shown here is extracted from the NACA Technical Note No. 825, titled "Wind tunnel investigation of yaw on lateral stability characteristics. III - Symmetrically tapered wing at various positions on circular fuselage with and without a vertical tail". This document was authored by I. G. Recant and A. C. Wallace, from the Langley Memorial Aeronautical Laboratory, in September 1941. What makes this model unique in all the validation studies shown on this page, is that this is a relatively simple geometry focusing entirely on the lateral characteristics, allowing us, at Research in Flight, to validate FlightStream® for roll and yaw forces and moments. The flow for these experiments is broadly incompressible. As the plots below show, FlightStream® predicts all roll and yaw aerodynamic forces and moments against flow angles in accordance with experimental data, until the wing starts to stall. FlightStream® generates these sweeps in a few minutes for this model.
The AIAA High-Lift Prediction Workshop 1 was held in Chicago, June 26-27, 2010. This workshop featured the NASA Trap Wing configuration as a verification and validation model for the CFD community. The Trap Wing is a representation of the landing configuration on aircraft and features high-lift components, including a slat and deployed flap. The model had different configurations, which can be found on the workshop website. The model shown here in FlightStream® is Configuration-8. This geometry is relatively challenging for a numerical flow solver, considering that there are several regions of flow separation (behind the slats, under the wing TE, etc.). FlightStream® models the flow with ease, but cannot yet model flow separation at very high AOA, as shown in the plot below. The solver run time is less than 2 minutes for the results shown here. The experimental data is obtained from the workshop website.