a demonstration of the sandia sierra mechanics simulation tools … › tfaws11 › proceedings ›...
TRANSCRIPT
National Aeronautics and Space Administration Jet Propulsion Laboratory California Institute of Technology
A Demonstration of the Sandia Sierra Mechanics Simulation Tools for Spacecraft Integrated Models Dr. Lee D. Peterson Principal Engineer Mechanical Systems Division (35)
Presented at the NASA Thermal and Fluids Analysis Workshop Hampton, VA
16 August 2011
© 2011 California Institute of Technology. Government sponsorship acknowledged.
2
National Aeronautics and Space Administration Jet Propulsion Laboratory California Institute of Technology
Acknowledgements
• JPL • Kevin Anderson
• Greg Agnes
• Scott Basinger
• Case Bradford
• Katherine Corrigan
• Claus Hoff
• Derek Lewis
• John Schiermeier
• Robby Stephenson
• Eric Sunada
• Sandia • [email protected]
• David Womble
• Ken Alvin
• Joe Jung
• Sam Subia
• The Sierra Team
• The DAKOTA Team
3
National Aeronautics and Space Administration Jet Propulsion Laboratory California Institute of Technology
Demonstration of the Sierra multiphysics tool suite on spacecraft integrated structural-thermal models • Motivation and Background
• What is Sierra
• What are its advantages?
• SWOT demonstration problem
• Telescope Benchmark problem
• Other applications
4
National Aeronautics and Space Administration Jet Propulsion Laboratory California Institute of Technology
New missions like Surface Water Ocean Topography challenge simulation technology
• Self-shadowing in LEO on a large, flexible structure
• Micron-scale dimensional error budget allocations
• Comparable magnitude for effects neglected by conventional tools
• Nonlinearity, thermal snap
• Ground test validation will need models to extrapolate to flight
Model and simulation errors may drive design when comparable to subsystem
error budgets.
5
National Aeronautics and Space Administration Jet Propulsion Laboratory California Institute of Technology
A spacecraft “Digital Twin” would predict flight performance under difficult-to-test conditions
Model V&V
Component Ground Tests
Subsystem Ground Tests
Flight
Digital Twin Flight System Ground Tests
Extrapolation
Demonstration
Model V&V
Model V&V
OSSE
Science Measurements
A Digital Twin would need to capture emergent, unexpected system
interaction not ordinarily simulated during design
6
National Aeronautics and Space Administration Jet Propulsion Laboratory California Institute of Technology
DOE labs faced similar challenges in their quest for model-based qualification without full system test • DOE advanced modeling,
simulation and model V&V technology over two decades
• Multi-$B investment in hardware, software and test methodologies
• Pervasive use of multiple physical domains, nonlinearity
Specialized to DOE weapon performance
• Rigorous model V&V practices
Complex Systems, Components, and Physics
Diverse Applications (ref: A. Ratzel, SNL, 5/19/09)
7
National Aeronautics and Space Administration Jet Propulsion Laboratory California Institute of Technology
The Sandia Sierra simulations have unprecedented resolution and fidelity
(ref: A. Ratzel, SNL, 5/19/09)
• Multiple codes can run as coupled, integrated simulations
• No “bucket-brigade” analyses
• Nonlinearity and generalized material properties
• Quantified uncertainty
• Mesh error quantification methods
• Efficient forms of non-Monte Carlo uncertainty quantification
• Architectural foundation is open-source
• Sierra Toolkit (STK)
• Trilinos
Animation
8
National Aeronautics and Space Administration Jet Propulsion Laboratory California Institute of Technology
Sandia’s Sierra creates simulations with multiple, interacting codes targeting separate physics • Aria: Fluid mechanics and
thermodynamics
• Adagio/Presto: Solid Mechanics
• Salinas: Linear statics, modal analysis, nonlinear transient vibration, structural-acoustic models
• Encore: Mesh transfer; error analysis; computed output interface
JPL has collaborated with Sandia to pilot the application of their tools
for space missions.
9
National Aeronautics and Space Administration Jet Propulsion Laboratory California Institute of Technology
Sierra architecture consists of multiple, interacting simulation “regions” of different physics
Sierra Multiphysics Model
Time Simulation Procedure
Mechanics Region A
Mechanics Region B
Transfers and Mappings
Analysis Region
Model Parameters
Values
System Response Quantities
(SRQ)
Mechanics Region C
10
National Aeronautics and Space Administration Jet Propulsion Laboratory California Institute of Technology
Meshes can be optimized for region-dependent physics and numerics
Sierra Multiphysics Model
Time Simulation Procedure
Mechanics Region A
Mechanics Region B
Transfers and Mappings
Analysis Region
Model Parameters
Values
Simulation Outputs (System
Responses)
Mechanics Region C
Transfers optimized for massively paralleled architectures
11
National Aeronautics and Space Administration Jet Propulsion Laboratory California Institute of Technology
Sierra plug-in capability enables user customized outputs
Sierra Multiphysics Model
Time Simulation Procedure
Mechanics Region A
Mechanics Region B
Transfers and Mappings
Science Region
Model Parameters
Values
Simulated Science
Measurement
Sensor Physics Region
We are adding output regions for optics, microwave sensors
12
National Aeronautics and Space Administration Jet Propulsion Laboratory California Institute of Technology
DAKOTA coupling enables uncertainty quantification of integrated simulation
Sierra Multiphysics Model
Time Simulation Procedure
Mechanics Region A
Mechanics Region B
Transfers and Mappings
Science Region
Model Parameters
Values
Science Measurement
Samples
DAKOTA Uncertainty Quantification
Software
Science Measurement Predictions with
Uncertainty Distribution
Sensor Physics Region
13
National Aeronautics and Space Administration Jet Propulsion Laboratory California Institute of Technology
Initial Piloted Application of Sierra to SWOT Concept
Integrated Sierra Structural-Thermal
Model
Thermal Structural
Model Development Team: K. Anderson, Thermal R. Stephenson, K. Corrigan, Structural
14
National Aeronautics and Space Administration Jet Propulsion Laboratory California Institute of Technology
Model Performance Highlights • Observed generally exceptionally
fast quasi-static solution speeds
• Sierra-Adagio solid mechanics solver (conjugate gradient with FETI preconditioner)
100,000 DOF on single processor in 18 seconds
1.2 MDOF on 8 processors in 120 seconds
• Sierra-Aria thermal solver (with contact)
40,000 DOF x 100 time steps on 8 processors in 37 seconds
Quasi-static 1-g sag solution
Quasi-static heat radiation and conduction
15
National Aeronautics and Space Administration Jet Propulsion Laboratory California Institute of Technology
Example IM Results: Transient Solar Heating of SWOT Reflectarray Panel
High density thermal and structural mesh
(~100K DOF)
Solar flux for 2<t<7 sec
Animation
16
National Aeronautics and Space Administration Jet Propulsion Laboratory California Institute of Technology
Integrated Thermal-Structural Model of SWOT KaRIn Array
Animation
17
National Aeronautics and Space Administration Jet Propulsion Laboratory California Institute of Technology
Telescope Benchmark Problem captures essential challenges in structural-thermal-optical modeling
Textbook Cassegrain telescope prescription
Thermal shroud and support structure not
shown
Low mass mirror with honeycomb core
Point force actuators for correcting figure
Thermal radiative heat transfer
Optical wavefront coupling to structural-
thermal simulation
0.5 meter diameter
18
National Aeronautics and Space Administration Jet Propulsion Laboratory California Institute of Technology
Primary mirror configuration is programmable to allow trade studies • Python script within the Sandia
Cubit preprocessor
• Allows DAKOTA to vary the CAD configuration
• Inputs STEP file from Code V optical prescription
• Variable mirror
• Thickness
• Honeycomb cell radius
• Modeled as surfaces to support shell elements
• Point force (“bed of nails”) actuators at each honeycomb corner
19
National Aeronautics and Space Administration Jet Propulsion Laboratory California Institute of Technology
Structural and thermal mesh densities also DAKOTA accessible parameters
64,609 shell elements 387,654 structural DOF
4,926 shell elements 4,264 thermal DOF 1,347,230 non-zero viewfactors
Small “workstation class” mesh
20
National Aeronautics and Space Administration Jet Propulsion Laboratory California Institute of Technology
Optical mesh is a regular 199x199 grid
Mesh is deformed in z direction (out of the page) to match the mirror surface
21
National Aeronautics and Space Administration Jet Propulsion Laboratory California Institute of Technology
Numerical Performance for Ambient Cooling (8 processor Mac) • Viewfactors
• 4.9K faces, 4.2K temperatures
• 1.3M VF
• Hemicube: 5 min 59 sec
• Row sum error ~ 1.e-6
• Thermal time step
• Radiosity solve: 0.24 sec
• Tolerance: 1.e-6
• GMRES sparse solver with Jacobi preconditioner
• Structural time step
• Static (nonlinear) solution: 35 seconds
• Conjugate gradient sparse solver with FETI preconditioner
Have also run problem with 250K faces, 124K
temperatures, >1B VF, 128 processors, 75 minutes
22
National Aeronautics and Space Administration Jet Propulsion Laboratory California Institute of Technology
Transient Simulation Animation
Front (mirror), e = 0.01 Core, e=0.2 Bottom, e=0.1
Partial enclosure 5K boundary
All front, core and back surfaces included in viewfactor calculation
Animation
23
National Aeronautics and Space Administration Jet Propulsion Laboratory California Institute of Technology
Transient Simulation Animation
Animation
24
National Aeronautics and Space Administration Jet Propulsion Laboratory California Institute of Technology
Transient Simulation Animation
Animation
25
National Aeronautics and Space Administration Jet Propulsion Laboratory California Institute of Technology
Additional Notes on Sierra-Aria Thermal Analysis Capabilities • Viewfactors
• Chaparral library
• Hemicube method
• Contour integration with pair-wise monte carlo method
• Contact and conductance
• Support incompatible meshes with conductances
• Wish list items
• Sun model (we are adding)
• Specularity
26
National Aeronautics and Space Administration Jet Propulsion Laboratory California Institute of Technology
Summary
• Sierra mechanics tools provide scalable, high performance multiphysics modeling capability
• Demonstration on structural-thermal-sensor simulations
Sierra Multiphysics Model
Time Simulation Procedure
Mechanics Region A
Mechanics Region B
Transfers and Mappings
Science Region Model
Parameters Values
Science Measurement
Samples
DAKOTA Uncertainty Quantification
Software
Science Measurement Predictions with
Uncertainty Distribution
Sensor Physics Region