National Aeronautics and Space Administration Wind-Induced Oscillations Parametric Wind Tunnel Test Sam Yunis, Donald Keller, Thomas Ivanco, Jennifer Pinkerton NASA Langley Presented at Spacecraft and Launch Vehicle Workshop June 20-22,
Issue NASA is attempting to address persistent questions on Wind Induced Oscillations (WIO) on Launch Vehicles What are realistic design conditions for WIO? Is lock-in a real-world event or a contrived wind tunnel artifact? Full-scale events seem to indicate that WIO does exist in the wild What wind tunnel conditions are necessary to match full scale observations? Different organizations are designing to different conditions NASA has been designing vehicles to lock-in based on clean uniform flow in wind tunnel tests (WTTs) Often a significant burden during design, especially rollout (Ares, SLS) Some others have been designing to turbulent flow and not to lock-in No complete/comprehensive data exists to quantify the best design condition Question dates back to 1960 s NASA is running a parametric WTT to address these questions Tunnel entry May 15-July 31, Time still exists to address questions 2
WIO 101 Flow across the launch vehicle creates lateral forces due to flow separation Driving forcing function Comes in many forms from laminar to turbulent to vortex shedding Flight vehicle is flexible which results in aeroelastic responses Aeroelastic coupling can be >10x factor over rigid magnitudes Second mode can be more critical than the first mode because it occurs at higher velocities/energies and because of the different load distribution Magnitude of aeroelastic response is determined by several factors: Type of unsteady flow forces (turbulent, vortices, laminar, etc) Proximity of the vortex frequencies to structural frequencies Damping 3
Definitions For purposes of this presentation: WIO Generic term for the response of a vehicle to unsteady flow Unsteady flow includes gust, turbulence, and vortex shedding Vortex shedding Quasi-periodic flow off the vehicle at certain velocities Lock-in An aeroelastic event where the oscillations of the vortex shedding are forced to coincide with the vehicle frequencies, resulting in aeroelastic amplification 4
Parameters of Investigation Reynolds number Flow turbulence Boundary layer effects, including pad height Surrounding structure Protuberances Damping 5
Full scale vehicle Re usually >10 6 Vortex shedding at Re = 10 4 looks similar to vortex shedding at 10 6 Concerns with Re mis-match Scaling for Re is very difficult Need to match correct magnitudes and velocities Protuberances impossible to scale because of boundary layer differences Tip shedding cannot be mimicked at the wrong Re WTT will address the importance of Re matching in WTTs Approach: Use air and heavy gas to span Re range Reynolds Number 6
Turbulence Real-world flow has turbulence Ground, buildings, butterflies Turbulence-induced vortex shedding and turbulent response is a design condition for fatigue and RMS response Turbulence may prevent large resonant WIO and lock-in Concerns with turbulence If lock-in can be achieved, turbulent response is not a peak-load design condition Resonant WIO/Lock-in remain the design condition(s) for peak loads until otherwise shown Example: Damping too high for aeroelastic response WTT will investigate impact of turbulence on lock-in Design: blocks on floor (based somewhat on method used at the Cermak, Peterka, Petersen (CPP) tunnel (spires, low wall, floor blocks) Goal is to generate typical turbulence levels ( 5% - 10%) 7
Earth Boundary Layer The boundary layer (wind profile) creates a nonuniform velocity profile along the length of the vehicle Thought 1: The boundary layer reduces the correlation length of lock-in across the vehicle, thereby reducing the aeroelastic responses Thought 2: The larger diameter of a fairing may match the boundary layer velocities, thereby actually increasing correlation length of lock-in Wind tunnel testing is not necessarily required to understand the impact of the boundary layer Improved/verified analytical methods may be sufficient to ID potential critical conditions (wind speed and direction) WTT will investigate the impact of the boundary layer by creating a wind profile Design: Blocking spires, multiple height pads 8
Surrounding Structure Surrounding structures, both upstream and downstream, can definitely influence the response of a vehicle The variety of these is too big for a complete study WTT will use a solid tower/building to demonstrate the effect in conjunction with the other parameters Design: Block building bigger than vehicle Location: Near the vehicle (2-5D away) 9
Protuberances Protuberances often instigate separation and vortex shedding, and have been shown to magnify results Protuberances often act within the flow boundary layer Protuberances are particularly susceptible to Re mis-match WTT will address this across Re numbers Design: Includes raceway/feedline and block protuberances 10
Damping WTT may attempt to vary damping Effect of damping on smooth-flow WIO is well established Effect of damping coupled with other parameters is not as well established WTT will measure structural damping for all configurations tested WTT will use a variable damper Design: In-work and unclear if this can be incorporated in time Trying to modify a previously-used design (Ares I-X) Targeting values C/Cc <1% to 2% 11
WTT Logistics Testing will take place at the Langley Transonic Dynamic Tunnel (TDT) Quick facts Mach: up to 1.2, Re: up to 10 x10 6 /ft, Medium: Air or R-134a Total pressure: near vacuum up to 1.0 atm. Test Section: 16 x 16 with cropped corners, Length: 12 30 GWL Testing: Large remotely controlled floor turntable, data system designed for dynamic testing Ares I-X Model in TDT - 2008 12
Dynamic models Single stick vehicles 1. Constant diameter Wind Tunnel Test: LV Design 2. Two diameters: Single diameter first/lower stage with large upper stage and fairing Modes: 1 st Bending, 2 nd Bending (4 modes in total) Ability to vary mass to achieve target frequencies May grit surface to validate simulation of higher Re on a 3D model 13
Wind Tunnel Test: Baselining Prior to parametric testing, the tunnel must show that it can predict full scale responses NASA has data from 2 full scale vehicles that experienced severe WIO, possibly lock-in: Wind speed and direction Launch vehicle response dynamic motions and loads Plan of attack is to figure out what parameters are needed to match the full scale events Baseline Uniform flow (normal TDT turbulence and B-L profile) Parameters will be varied to match the response Wind profile Turbulence Boundary Layer Damping 14
Wind Tunnel Test: Parametric Testing Parametric testing on a generic vehicle Lot of possible parameters and variations Limited by non-infinite time Parameters Vehicle Constant diameter stick and stick w/fairing at top Protuberances Damping may not vary, depends baseline testing Parameters - Flow a. Velocity: Target first and second modes b. Reynolds Number: 10 4 10 6 c. Turbulence: None to as large as we can get d. Boundary Layer (incl. Pad Height): None to theoretical curve e. Surrounding Structure: None to block tower Data Steady/unsteady pressures, acceleration response, base bending moment 15
Schedule August 2016: ATP November 2016: Briefing to general LV community for feedback May 15: Wind-on for turbulence and boundary layer checkout June 15: Baseline runs begin July 15: Parametric testing begins July 31: End? 16
Data Release The parametric data will be released to the community Assume it will be subject to ITAR rules Timeframe for release is unclear, but as soon as possible 17
Summary A parametric wind tunnel test for understanding WIO is underway Any input to the test matrix is welcome Very soon or it will be too late 18