Rapid Design Iteration Process for Spacecraft Kinematic Mounts Using Automatic Tet Meshing and Globa - PowerPoint PPT Presentation

1 / 39
About This Presentation
Title:

Rapid Design Iteration Process for Spacecraft Kinematic Mounts Using Automatic Tet Meshing and Globa

Description:

Automatic Tet Meshing and Global/Local Modeling Techniques. Hanson Chang ... Automatic tet meshing with efficient mesh control and convergence techniques ... – PowerPoint PPT presentation

Number of Views:173
Avg rating:3.0/5.0
Slides: 40
Provided by: long90
Category:

less

Transcript and Presenter's Notes

Title: Rapid Design Iteration Process for Spacecraft Kinematic Mounts Using Automatic Tet Meshing and Globa


1
Rapid Design Iteration Process for Spacecraft
Kinematic Mounts UsingAutomatic Tet Meshing and
Global/Local Modeling Techniques
  • Hanson Chang
  • MSC.Software Corporation

2
Acknowledgements
  • Co-author Chris Luanglat, TRW Stress analyst

3
Presentation Outline
  • Spacecraft program and kinematic mounts
  • Design challenges for kinematic mounts
  • Rapid design iteration process
  • Direct import of CAD solid geometry
  • Automatic tet meshing with efficient mesh control
    and convergence techniques
  • Global/local modeling techniques
  • Conclusions

4
Spacecraft Program Overview
  • EOS Spacecraft Aqua and Aura
  • Mission To study the Earth and its changing
    environment by observing the atmosphere, oceans,
    and land surface.
  • Launch dates
  • Aqua - 4/2002 Aura - 1/2004

5
Spacecraft Overview
  • Spacecraft Spec.
  • Dimensions 22 ft x 9 ft x 8 ft
  • Weight 6,500 lbs
  • All-composite spacecraft structures

6
Spacecraft FEM View 1
7
Spacecraft FEM View 2
8
FEM Exploded View
9
Load Sharing During Launch
10
Load Sharing On Orbit
11
Load Isolation Concepts
  • Statically determinant interface (6 DOF) isolates
    the instruments from the primary structure load
    path
  • This type of structural interface is called a
    Kinematic Interface
  • The attachment fittings used in this type of
    structural interface are called Kinematic Mounts

12
Releasing a Degree of Freedom
  • Sliding Design
  • Ball/socket, cup/cone, pin/slot, V block/groove,
    etc.
  • Relies on low and predictable friction
  • Flexure Design
  • Uses flexibility to isolate loads
  • Selected for TRW kinematic mount design

13
One-Axis Kinematic Mount (KM1)
14
Two-Axis Kinematic Mount (KM2)
15
Three-Axis Kinematic Mount (KM3)
16
Typical KM Arrangement
KM1
KM3
KM2
17
Typical Stiffness Matrix
Ideal KM2 Stiffness Matrix
18
Traditional Solution Space
19
Kinematic Mount Solution Space
20
Design Iteration Process
Strength SOL 101 Flexibility SOL 101
Stiffness SOL 103 Stability SOL 105
Fracture SOL101/FLAGRO
21
Rapid Design Iteration Process
  • Speeding up the iteration process
  • Direct import of CAD solid geometry
  • Automatic tet meshing with efficient mesh control
    and convergence techniques
  • Global/local modeling techniques

22
Geometry Import - Old Process
  • Clean up surfaces (slivers, tee, etc.)
  • Create B-rep solid from surfaces
  • Create solid geometry based on drawing

23
Geometry Import - New Process
CATIA Direct
CATIA Solid Geometry
MSC.Patran
  • Solid geometry directly imported into MSC.Patran
    as solid geometry with high success rate (95)
  • Sliver surfaces and short edges (dirty geometry)
    are best correct in the CAD package
  • Conferences held between designers and analysts
    to discuss how to identify and eliminate problem
    geometry

24
Meshing - Old Process
  • Hex element (8-node brick) is the preferred
    element
  • Created by manual meshing
  • Created by meshing 5 or 6-sided solids (simple
    solids) or sweeping 2D elements
  • Typical part must be broken into simple solids
    first

25
Meshing - Old Process (cont.)
Notched Regions
  • Hex meshing of above parts is labor intensive
  • Meshing time for typical KM is several days
  • Not acceptable the multiple design iteration
    environment

26
Meshing - New Process
  • Automatic tet meshing using TET10 elements
  • Can mesh arbitrarily-shaped solids
  • Meshing time for typical KM is 4 hours
  • Ideal for the multiple design iteration
    environment

27
Meshing - New Process (cont.)
  • Advantage of Tet Meshing
  • Fast
  • Quality of TET10 elements (linear strain) is
    compatible to HEX8 elements
  • Disadvantage of Tet Meshing
  • Larger model

28
Efficient Tet Meshing
  • Key to efficient tet meshing is mesh density
    control
  • Hitting the automatic tet mesh button without any
    mesh control typically results in excessively
    large modes
  • Correct density control puts a lot of elements in
    the area of interest and coarsens quickly away
    from this area

29
Efficient Tet Meshing (cont.)
  • Typical density control techniques
  • Surface mesh selected solid faces with TRIA6
    first to guide subsequent tet meshing
  • Curvature-based meshing
  • Break the part into multiple solids cookie
    cutter method

30
Cookie Cutter Method
  • Break the solid with planes or surfaces
  • Critical solid meshed first with a fine mesh
  • Sounding solids meshed with a coarse mesh

31
Cookie Cutter Method (cont.)
32
Cookie Cutter Method (cont.)
33
How to Achieve Convergence
  • 4 elements thru the thickness?
  • 8 elements thru the thickness?
  • Multi-pass convergence is time consuming
  • Single-pass convergence is fast but more
    subjective
  • Fringe plot with the difference option in
    MSC.Patran
  • Plots the stress jumps (discontinuities)

34
How to Achieve Convergence
  • Use a combination of both methods
  • For each type of notch geometry (circular,
    square, rectangular, etc.), a multi-pass
    convergence test is performed to establish the
    required number of elements thru the thickness
  • Each new part is then meshed using this rule of
    thumb and verified using the single-pass
    convergence test

35
Integrating the Models
  • Resulting model is unacceptably large

36
Global-Local Modeling
  • Use Static Reduction (Guyan Reduction) to reduce
    tet10 model to small stiffness matrix
  • Use ASET entry to specify boundary DOF
  • PARAM,EXTOUT,DMIGPCH to create DMIG entries
  • Use K2GG entry to assemble the KM matrices into
    Spacecraft model

37
Global-Local Modeling (cont.)
38
Conclusions
  • Rapid design iteration process
  • Direct import of CAD solid geometry
  • Automatic tet meshing with efficient mesh control
    and convergence techniques
  • Global/local modeling techniques
  • This process resulted in substantial cycle time
    reduction for the Aqua and Aura kinematic mounts

39
Conclusions (cont.)
  • The notched-column kinematic mount design
    configurations have been incorporated into the
    TRW Deployables Handbook

Merci beaucoup
Write a Comment
User Comments (0)
About PowerShow.com