Document
Static and Dynamic Testing of Blunt Bodies in a Subsonic Magnetic Suspension Wind Tunnel 1 2 3 1 1 Mark Schoenenberger , Colton Finke , Colin Britcher , David Cox and Timothy Schott NASA Langley Research Center NASA Dr ive , Hampton, VA, 23666, United States Texas A & M University College Station, Texas , 77843 , United States Old Dominion University Hampton Blvd, Norfolk , VA , 23529 , United States ABSTRACT The MIT 6 - inch magnetic suspension wind tunnel is used in two configurations to measure lift forces of two blunt bodies and produce free - to - pitch oscillations driven by capsule static stability and dynamic instabilities. Lift tests show that static aerodynamic data can be measured on a magnetically levitated model without moment control. Free - to - oscillate results show that magnetic suspension balance system ( MSBS ) can produce capsule dynamics suitable for extracting static and dynamic stability data.
1. Introduction free to rotate about the N - S axis. This feature was The MIT 6 - inch magne tic suspension wind tunnel [1] has employed to create a set of models held at fixed angles been refurbished through a partnership between NASA of attack to measure lift. The MSBS was then rotated 90 and Old Dominion University . The tunnel is being used degrees to demonstrate that models can pitch freely to develop test methods to measure pitch damping about the magnetizing axis. Figure 1 shows the two test characteristics of blunt entry vehicles. Wind - on levitation configurations. A new smaller test section (2.3 7 5 - in. H x was achieved in the fall of 2017 . This was followed by 2.664 - in . W ) was fabricated that could pass through the calibration of the MSBS and characterization of different side viewing ports of the MSBS . The rotation meant that core materials as well as the first s tatic drag the duties of the side and drag force coils are swapped.
measurements [2 ] . Test procedures are being developed to measure lift forces with models suspended at angles of attack and dynamic testing where the model is free to pitch, while constrained in the other degrees of freedom.
Work has focused on d emonstrating the f easibility of static and dynamic testing without moment control of the test articles.
2. MSBS Operation Levitated models are controlled using a state estimator and Linear Quadratic Regulator (LQR) feedback design, with integral feedbac k to reject steady disturbances. Rotational states of the model ar e not sensed or commanded, and so were not part of the Figure 1 MSBS orientations for force and oscillation testing regulator design. The control system was implemented in Simulink, and autocoded to create software 4 . Wind Tunnel Models for Lift and Pitching compatible with a real - time computer operating at 1000 Models of the Stardust capsule [3 ] and an approximation Hz . P osition feedback comes from an Electromagnet ic of the Orion entry vehicle were 3D printed with PLA Position Sensor (EPS) system. The intended use for this plastic . 0.75 - inch diameter x 0.75 - i nch length cylindrical tunnel is to measure the dynamic stability of blunt NdFeBo magnets were located inside the models, canted bodies by measurin g free oscillation histories .
at 8 and 16 - degree angles relative to the model axes of Separating dynamic aerodynamic moments from static symmetry . When levitated in the baseline MSBS aerodynamic moments and magnetic moments with a configu ration, the m agnets would align with the full 6 - DoF controller was deemed impractical .
magnetizing field, parallel to the tunnel freestream , 3. Model Test Configurations holding the models at 8 and 16 degrees angle of attack.
To achieve free oscillation a number of model materials Dynamic pressure sweeps were run to measure the lift (permanent magnets and iron alloys) were assessed to coefficient at these angles.
see if a spherical core would rotate freely when Another Stardust model was 3D printed to levitated . The coercive forces of a ll materials tested accommodate a 0.75 - inch diameter NdFeBo sphere, were too high to achieve free oscillations. However, which allowed for a smaller model to fit around the e ach levitated core would freely rotate about the magnetic core so as to reduce blockage in the smaller magnetizing field. In normal operation this results in free - to - pitch test section. The poles of the magnet were models rolling freely in the flow . N eodymium - iron - oriented to be orthogonal to the spin axis of the model, boron (NdFeBo) permanent magnet cores were shown to though a small sideslip was observed when levitated. I n stiffly align wit h the magnetizing field, while remaining Corresponding author: Mark Schoenenberger E - mail address : mark.schoenenberger@nasa.gov the transverse tunnel, the model was free to pitch and the magnet centroid) . Future work will use this model to oscillated in tunnel flow due to its static pitch stability. extract static and dynamic moment coefficients.
Figure 4 Free - to - pitch Stardust capsule in transverse tunnel Figure 2 Lift and Free - to - Oscillate Wind Tunnel Models (dimensions in inches) 5 . Lift Tests Lift coefficient data from the Stardust and Orion canted - core models are plotted against reference wind tunnel data in Fig 3 . Agreement is good for this prelimin ary assessment. The wind - off levitation current was Figure 5 Planar oscillation model fit to Stardust video data subtracted from the lift current history and root - sum - squared with the side force current to determine the total This free - to - oscillate configuration was intended as a lift force. The variation due to dynamic pressure needs proof of concept test to demonstra te that models would to be investigated further. Signal - to - noise improved as oscillate freely and capsule oscillations would grow or dynamic pressure was increase d. During testing the decay due to dynamic damping properties of the models roll ed about the tunnel centerline. The 16 - degree capsule. This demonstration was successful, al though models showed more roll and lateral transl ation due to several sources of error must be addressed before the the ir increased lift, limiting the max imum dynamic accurac y of aerodynam ic coefficients can be pressure .
determined . Model blockage affected dynamic pressure measurements during this test . The active control of the MSBS acting at a (small) distance from the model cg can introduce non - aerodynamic oscillation growth or decay. The MSBS control inputs can produce plunging motions that affect the angle o f attack history, compl icating pitch damping measurements as well.
Conclusion Static lift forces were measured and free pitch oscillations recorded to extract damping information of blunt bodies. Future work will assess error sources and measure capsule pitch damping with uncer tainties.
Lessons learned from the transverse configuration can be applied to future MSBS designs for dynamic testing.
References Figure 3 MSBS lift compared to historical wind tunnel data [1] Stephens T., “Design, construction, and evaluation of a magnetic suspension and balance system for wind 5 . Free to Oscillate Tests tunnels,” NASA CR - 66903; 1969.
The small Stardust free - to - pitch model is shown [2 ] Schoenenberger et al, “ Preliminary Aerodynamic levitating in t he transverse tunnel in Fig. 4 . Video data Measurements from a Magnetic Suspension and Balance at a small dynamic pressure ( estimated to be System in a Low - Speed Wind Tunnel ,” AIAA 2018 - 3323, a pproximately 50 Pa) was recorded and digitized to 2018 .
determine the attitude history. The planar attitude [3] Mitcheltree et al, “Aerodynamics of Stardust Sample history can be approximated by the equation shown in Return Capsule,” AIAA - 97 - 2304 , 1997 .
[4 ] Schoenenberger and Queen, “Limit Cycle Analysis Fig. 5 [4 ] . This model was fit to the measured attitude Applied to the Oscillations of Decelerating Blunt - Body history to identify the oscillation frequency and nose - Entry Vehicles,” NATO RTO - MP - AVT - 152 - 006, 2008.
down trim angle (the model cg was slightly forward of