www.nasa.gov
Underexpanded supersonic plume surface interactions: applications for spacecraft landings on planetary bodies
Manish Mehta, Ph.D.Aerosciences Branch, EV33
NASA Marshall Space Flight CenterAnita Sengupta, Ph.D.
NASA Jet Propulsion LaboratoryNilton O. Renno, Ph.D.
University of MichiganJohn W. Van Norman
Analytical Mechanical Associates, Inc.Peter G. Huseman and Douglas S. Gulick
Lockheed Martin Space Systems
April 18,2011“Approved for public release; distribution is unlimited.”
Outline
♦ Introduction• Motivation• Near-field and far-field plume flow fields• Scaling Theory
♦Phoenix Mars spacecraft• Experimental• Numerical• Data comparison/Flow Physics
♦Mars Science Lab (MSL) Descent Stage spacecraft • Experimental• Numerical • Data comparison/Flow Physics
♦Conclusion
2
Motivation
♦ Plume-surface interactions due to spacecraft landings• Spacecraft stability and survival
− Moments/Torques− Updraft plumes− Plume induced heating
• Cratering & dust lifting • Implications for manned and large payload landings to Mars,
moon, asteroids and other planetary bodies
♦ Theory, test data and numerical simulations were used to characterize the complex plume impingement physics and identify the environments observed due to spacecraft landings
3
1998 Mars Polar Lander Failure Report (Whetsel et al, 2000)
- MPL Project failed to conduct studies on plume-surface interactions - Recommended these investigations for future powered descent landing missions to Mars
These NASA spaceflight projects deemed it necessary to conduct these investigations thru University research partnership.
Phoenix: Pulsed-modulated descent system (Rocket Engine Module – REM)
Mars Science Lab: Sky-crane throttled landing system (Mars Landing Engine – MLE)
2007 2011
Last detailed study was completed in 1973 for Viking 1 and 2.
Motivation
4
Near-field flow –TEST DATA
Underexpanded Supersonic Jet
Overexpanded Supersonic Jet
Planar Laser Induced Fluorescence Imaging
Inmann et al., 2009
Important flow structures with implications to cratering, acoustics and spacecraft dynamics during descent
Far-field flow/Impingement zone – TEST DATA
Lamont and Hunt, 1976 5
( )
( ) 1MaMa
Re
PrRe1
][Re1
Fr1
22
22
22
2
−==
′+′′
′′+′∇′′⋅∇′=′′
′′
′⋅∇′+′∇′−′=′′
′
′⋅∇′′−=′′
γ
φβλρ
τρρ
ρρ
epe
e
epe
epeepep
Tcc
TcU
TcU
tDpDT
TcUT
tDTDc
pgtDvD
vtD
D
( ) ( ) ( ) Re
1Ma 1MaPrRe
1 22 φγβγλρ ′−
+′′
′′−+′∇′′⋅∇′=′′
′′tDpDTT
tDTDcp
Scaling theory for plume-surface interactions
6
( )
][
φβλρ
τρ
ρρ
++∇⋅∇=
⋅∇+∇−=
⋅∇−=
DtDpTT
DtDTc
pfDtUD
UDtD
p
( )
e22
e2
e
/ ,/ , ,/ ,/ ,/
, / ,/ , / , / ,/ ,/
µµµµφφβββλλλ
ρρρρ
=′=′∇=∇′=′=′=′=′=′−=′=′=′=′
eeeepepp
eeeee
UDDccc
TTTUpppUuvDtUtDxx
Schlichting and Gersten, 2001
Normalized parameters
Continuity
Conservationof Momentum
Conservationof Energy
Nondimensional Navier-Stokes Equations
Scaling theory for plume-surface interactions
7
( ) ratioexpansion jet 1 1
222 =−
=
−
=−
=′ ∞∞∞
∞ eU
ePU
PPP
UPPp
eeee
e
ee
efb ρρρ
eve
pe
e
pee
e
ee
e
ee
UfD
ccc
aU
gDUDU St , ,Pr ,Ma ,Fr ,Re ====== γ
λµ
µρ max
∞
−=P
PCα ∞
=PPe e 2)1( Mk −= γγ
Free surface boundary condition
Nondimensional numbers that satisfy dynamic similarity
Phoenix Mars Spacecraft
Courtesy of NASA/JPL/Lockheed Martin
8
Experimental and Numerical Methodology
9
-½ scale Phoenix nozzle (MR-107)-N2 test gas-10 Hz pulsing-Mars atmosphereUniversity of Michigan Thermal Vacuum Chamber
Numerical Methodology
-Two Navier-Stokes computational solvers were used for modeling and analyses
-ANSYS FLUENT-3-D & axisymmetric density based solver-Transient RANS-Time step – 1 us – explicit marching-Turbulent (RNG) model-Adaptive meshing for resolving shocks-Grid independence-2nd order upwind discretization scheme-2 million unstructured grid cells
-Aerosoft GASP-3-D density based solver-Transient & steady-state RANS eqns solved-Van Leer flux splitting-Laminar-Dual implicit time stepping-Single species – frozen flow-Grid independence-4 million unstructured grid cells
Experimental Methodology
Plemmons, Mehta et al, 2008
Ground pressure profiles – TEST DATA
Plemmons, Mehta et al, 2008
Overpressure
Temporal Profile Spatial Profile
Pc ~1200 kPa0 deg canth/de = 8.410 Hz pulsingN2 test gas
10
Ground pressureat centerline
Chamber stagnation pressure
Back pressure
Observed large transient overpressures during engine start-up and shut down
Observe a monotonic rise in ground pressurefollowed by a drop in centerline pressure
t = 0.112 sec
t = 0.136 sec
MARS ATMOSPHERE ~ 700 Pa
Plate shock dynamics - CFD
Mechanism deduced from experimental measurements, transient numerical simulations and theory.
Gulick et al., 2006
N2 test gasAxisymmetric transient simulation
Pc ~ 1200 kPa0 deg canth/de = 8.4
GASP
11
Plate (recovery) shock formation
MARS ATMOSPHERE ~ 700 Pa
Plate shock dynamics - CFD
Plate shock formation and collapse
Ground pressure spatial profiles
Plemmons, Mehta et al, 2008
Wall jet
Plate shockcollapse
- N2 test gas-Axisymmetric- Transient-10 Hz pulsing
Pc ~1200 kPa0 deg canth/de = 25
FLUENT12
MARS ATMOSPHERE ~ 700 Pa
Plate shock dynamics - CFD
Plemmons, Mehta et al, 2008
FLUENT
- N2 test gas-Axisymmetric- Transient
Pc ~1200 kPa0 deg canth/de = 25
13
Plate shock formation and collapse
3-D full-scale flow field and ground pressure profiles -CFD
14
3-D transient simulations of full-scale Phoenix plumes interacting at surface
Pc ~ 1200 kPa0 deg canth/de = 2510 Hz pulsingEquivalent plume gas
GASP
Gulick et al, 2006
Modeled as a 60 degree wedge to reduce computational resources
3-D numerical simulations show that ground pressure loads are asymmetric and develop overpressures during rapid engine startup and shutdown
1E4
1E3
1E2
(Pa)
MARS ATMOSPHERE ~ 700 Pa
Experimental and numerical data
Good agreement between experimental results and numerical simulations (CFD)
Gulick et al, 2006
GASP
FLUENTShadowgraphPc ~ 1200 kPa0 deg canth/de = 8.410 Hz pulsingN2 test gas 15
Dashed lines – surface pressureSolid lines – thruster inlet stagnation pressure
(chamber pressure)
Comparing spatial and temporal ground pressure profiles
Comparing plume shock structureTEMPORAL
SPATIAL
Red line – CFDDots – Test Data
MARS ATMOSPHERE ~ 700 Pa
MSL Descent Stage Spacecraft & Testbed
16
- ¼ scale MSL MLE nozzle-N2 test gas-steady operation-Mars atmosphere-NASA Ames Research Center Planetary Aeolian Laboratory
Courtesy of NASA/JPL/Caltech
Descent stage Testbed
Thruster
ImpingementPlate
Testbed
Numerical methodology
- NASA OVERFLOW 2.1- 3-D time-marching implicit code- structured overset grid- Navier-Stokes eqns solved over full domain and internal nozzle- SST turbulence model- compressibility correction- steady-state - frozen flow used for modeling rocket plume gases- 12 million cells
17
Temporal ground pressure profiles – TEST DATA
18
Earth AtmosphereSteady Supersonic Jet
Mars atmosphereSteady Supersonic Jet
Pc ~1700 kPa22.5 deg canth/de = 35
Repetitive overpressures not observed
N2 test gassteady operation
Spatial ground pressure profiles – TEST DATA
19
Pc ~1700 kPa22.5 deg canth/de = 35N2 test gassteady operation
Pg/Pc
Pg/Pc Ps = Pg
Shows that the plume is collimated and leads to large pressure gradients at Mars atmospheric pressure
Ground pressure and rise rate vs. jet expansion ratio –TEST DATA
20
−
−=
∆
∆=
minmax
minmax 11 rate riseggc
gg
c
g
ttPPP
tPP
ratioexpansion jet amb
e
eamb
ee
PP
APAPe ===
Overshoot = Max ground pressure – quasi-steady ground pressure
Plume flowfield and spatial pressure profiles - CFD
Pc ~ 1.7e3 kPa22.5 deg canth/de ~35
e =3.5
OVERFLOW
N2 test gassteady-state
Pc ~ 1.7e3 kPa22.5 deg canth/de ~35EquivalentPlume gassteady-state
21
Plume is collimated and does not dissipate at large axial distances – in agreement with test data
MARS ATMOSPHERE ~ 700 Pa
Experimental and numerical data
22
OVERFLOW
Pc ~ 1.7e3 kPa22.5 deg canth/de ~35N2 test gas
MARS ATMOSPHERE ~ 700 Pa
Jet expansion ratio
EARTH MARS MOON
Stitt, L.E. , 1963
23
TEST DATA
CFD – Mach Contours
e = 4.50 e = 0.02
All tests were done at steady engine operation
Max normalized ground pressure
Mars – moderately underexpandedplumes lead to max ground pressure loads due to collimated plume structure and development of a small areal plate shock
Earth – highly overexpanded plumes dissipate/no plate shock formation
Moon – highly underexpanded plumes leads to a large areal plate shock – decreases ground pressure
Other shock interaction effects during spacecraft landings
Altitude Effects Spatial Asymmetry
Pc ~ 1200 kPa0 deg canth/de = 25Steady-state
FLUENTGASP Gulick et al, 2006
N2 test gas
24
SURFACESHEAR STRESS SURFACE PRESSURE
MACH CONTOUR
Ground pressure vs normalized altitude
MARS ATMOSPHERE ~ 700 PaMARS ATMOSPHERE ~ 700 Pa
Conclusions♦ Moderately underexpanded jets demonstrate:
• collimated shock structures• large supersonic core lengths• plate shock dynamics • max pressure loads
♦ Plate shock dynamics leads to:• large pressure gradients• asymmetry • overpressure
♦ Ground pressure loads are highly sensitive to • jet expansion ratio • strouhal number• spacecraft altitude
♦ Scaling laws show that cold plume gases can simulate ground pressure loads and interaction physics due to rocket plumes provided dynamic similarity is satisfied
♦ How does this effect spacecraft landing? • Transient ground pressure loads translate to load perturbations at the spacecraft base which may lead to
destabilizing moments (observed to a minor degree on the Phoenix spacecraft, Gulick et al, 2006)• Pressure loads can lead to extensive cratering and dust lifting which can destabilize the spacecraft upon
touching down on the surface (observed at the Phoenix Landing Site, Mehta et al, 2011)• Dust lifting can erode important spacecraft sensors and science instrumentation (a concern for the MSL
mission, Mehta et al, 2011)• Propulsion systems of small scale landers show maximum ground pressure loads at Mars atmosphere at
relatively high altitudes (h/de ~35) (a concern for the MSL mission)
♦ Provided JPL with landing environments which they incorporated into their risk analysis models 25
MISSION SUGGESTION: Due to highly complex plume impingement physics, accurate landing environments are needed for future planetary manned and robotic spaceflight missions
Acknowledgement
♦ I would like to thank my supervisor Mark G. D’Agostino and my colleagues within the home organization – Aerosciences Branch (EV33) at NASA Marshall Space Flight Center for their support.
26
References
♦ Lamont, P. J., and B. L. Hunt (1980), The impingement of underexpanded axisymetrical jets on perpendicular and inclined flat plates, J. Fluid Mechanics, 100, 471-475.
♦ Carling, J. C., and B. L. Hunt (1974), The near wall jet of a normally impinging, uniform axisymmetric, supersonic jet, J. Fluid Mechanics, 66, 11-19.
♦ Clark, L. V. (1970), Effect of retrorocket cant angle on ground erosion – a scaled Viking study, NASA Technical Memorandum. NASA TM X-2075, NASA Langley Research Center.
♦ Inman, J.A., P.M. Danehy, R.J. Nowak and D.W. Alderfer (2008) Fluorescence imaging study of impinging underexpanded jets, 46th AIAA Aerospace Sciences Meet., AIAA 2008-619.
♦ Stitt, L.E. (1963), Interaction of highly underexpanded jets with simulated lunar surfaces, NASA-TN-D-1095, NASA Glenn Research Center, Cleveland, OH.
♦ Schichting, H. and K. Gersten (2001), Boundary Layer Theory, Springer: New York♦ Van Norman, J. W. and L. Novak (2009), CFD support of MSL MLE plume simulations. NASA Contractor Report, No. NNL04AA03Z.♦ Land, N., and H. Scholl (1966), Scaled LEM jet erosion tests, NASA Langley Working Papers, LWP–252, NASA Langley Research Center.♦ Clark, L., and W. Conner (1969a), Exploratory study of scaled experiments to investigate site alteration problems for Martian soft-lander
spacecraft, NASA LWP-765, NASA Langley Research Center, Hampton, VA.♦ Sengupta, A., M. Mehta et al. (2009), Mars Landing Engine Plume Impingement Environment of the Mars Science Laboratory, IEEE/AIAA
Aerospace Conf., 1349.♦ Huseman, P. G. and J. Bomba (2000), CFD Analysis of Terminal Descent Plume Impingement for Mars Landers, AIAA Thermophysics
Conf., 2000-2501.♦ Buning, P.G., W.M. Chan, K.Z. Renze, D.L. Sondak, I.-T. Chiu, J.P. Slotnick (1993) OVERFLOW User's Manual, Version 1.6ab, NASA
Ames Research Center, Moffett Field, CA, Jan.♦ Gulick, D.S. (2006), Phoenix Mars Lander Descent Thruster Plume/Ground Interaction Assessment, WS-06-002 Lockheed Martin
Interoffice Memo, Lockheed Martin Space Systems, Denver. ♦ Mehta, M., et al. (2011), Explosive erosion during the Phoenix landing exposes subsurface water on Mars, Icarus, 211, 1. ♦ Plemmons, D. H. , M. Mehta et al (2008), Effects of the Phoenix Lander descent thruster plume on the Martian surface, J. Geophys.
Res., 113, E00A11.♦ Whetsel, C., et al. (2000), Jet Propulsion Laboratory MPL Investigation Board, Report on the Loss of the Mars Polar Lander and Deep
Space 2 Missions, JPL D-18709.
27
Back-up Slide: Phoenix Entry, Descent and Landing Sequence
-200 Hz (Inertial Measureiment Unit) IMU data and 10 Hz Radar data
CurrentResearch Investigation
28
Back-up Slide: Terminal descent
Plumes interacted with the surface for less than 2 seconds (even less than predictedby landing simulations)
Lift loss occurred at ~4.5 m and ground effect started around ~3.5 m
Noticed a second bounce – could be the result of plume-surface interactions afterInitial contact.
Touchdown
29