A structure and best practice for the management of composites data in engineering design and materials selection Dan Williams - Product Manager, Automotive Granta Design Ltd. 11 September 2012
A structure and best practice for the
management of composites data in
engineering design and materials selection
Dan Williams - Product Manager, Automotive
Granta Design Ltd.
11 September 2012
Contents
• Some history: the Materials Data Management
Consortium since 2002
• Challenges in managing composites data
• The MDMC schema for composites data
• Additional benefits to automotive industry of a
software-based approach:
• For screening and selection
• To support Computer Aided Engineering (CAE)
• Concluding thoughts
MDMC Consortium – since 2002
ASM International
AWE
Baker Hughes Boeing
Honeywell Aerospace
GE - Aviation
GE - Energy
Lockheed Martin
Los Alamos Nat Labs
NASA
Northrop Grumman
Oak Ridge Nat Labs
Raytheon
Rolls-Royce
Sandia Nat Labs
US Navy
US Army Research Labs
• Founded 2002 by ASM, NASA Glenn,
Granta Design
• Chair is Dr Steven Arnold (NASA)
• Now in third Phase
• Unique collaborative project focused
on developing and applying software to
manage mission-critical materials data
in the aerospace, defense, and energy
sectors.
• Resulting technology is being applied
in automotive, oil & gas, medical and
electronics industries
MDMC first phase
“I can say with complete confidence that the MDMC’s fatigue crack
growth module will result in at least an 80% reduction in the time it
takes me to do the analysis and roll up the data into a useful model”
Jeffrey Sickmeier, Materials and Processes Engineering, Rolls-Royce
Corporation.
Focus on lifecycle of materials data:
from test lab to design allowables
Materials lifecycle – as defined by MDMC
MAINTAIN
• Capture & maintain context
• Change management
• Security / access control…
ANALYZE
• Process test data
• Create approved models for FEA
• Forensic analysis
• Statistical process control
• …
CAPTURE
Underlying software system
Benefits of a software-based approach
Searchable, traceable, controlled, usable information
Composites data challenges
1. Complex hierarchy required to identify ‘pedigree’ of material
2. Additional data not found in ‘typical’ materials
• E.g. adhesive bond strengths, fiber tow count…
3. Anisotropy: more variables needed to qualify data
4. Process history & geometry dependent properties
How to describe a “material”? How to represent multi-dimensional data?
MDMC Composites Schema (2010)
Source: Warde S., Marsden W., ‘Data Management for Composite Materials’ (2010)
www.grantadesign.com/papers/
‘Pedigree’ data structure identifies all the
constituents and process history of the
laminate which is tested
‘Pedigree’ data structure identifies all the
constituents and process history of the
laminate which is tested
‘Design data’, including data to support CAE go here ‘Design data’, including data to support CAE go here
Some images from the database
Laminate datasheet
showing lay
sequence, cure cycle,
and references to
constituent materials
Laminate datasheet
showing lay-up
sequence, cure cycle,
and references to
constituent materials
Source: NCAMP Test Report Number: NCP-RP-2008-003 Rev B
History-
dependence:
capture of step-by-
step assembly
process
Some images from the database
Reinforcement
datasheet, showing
properties which are
unique to fibers
Reinforcement
datasheet, showing
properties which are
unique to fibers
Source: NCAMP Test Report Number: NCP-RP-2008-003 Rev B
Traceability: unique
designation codes to
support forensic analysis
Some images from the database
Part of a tension
test, showing test
conditions, test
Standards and
results
Part of a tension
test, showing test
conditions, test
Standards and
results
Source: NCAMP Test Report Number: NCP-RP-2008-003 Rev B
Anisotropy: properties as
function of direction
Conditioning: properties
as function of test
environment
Cost-Benefit Analysis in early R&D
• Automotive lightweighting is driving materials
development at an unprecedented pace
• Need a way to identify promising candidates quickly
Some tricky “material positioning” requirements:
• Fair, like-for-like comparison with metals /
traditional options
• Combination of engineering, cost and eco
objectives
• Holes in data
• How to handle hybrids? (sandwich panels,
honeycombs, etc.)
• Material choice is highly application-dependent
Some tricky “material positioning” requirements:
• Fair, like-for-like comparison with metals /
traditional options
• Combination of engineering, cost and eco
objectives
• Holes in data
• How to handle hybrids? (sandwich panels,
honeycombs, etc.)
• Material choice is highly application-dependent
Example analysis
Source: CES Selector 2012 Software
www.grantadesign.com
Supporting simulation (CAE) - challenges
RADIOSS
Abaqus
ANSYS
NASTRAN Manual, infrequent
publishing
CAE Community
RADIOSS
ANSYS
Static cards available on server
Materials Group
‘Material files’
• Islands of data
• Manual, infrequent updates
• Error-prone and slow
• Limited opportunities for material optimization
Most common challenges relating to materials data in CAE, as
validated by MDMC consortium:
Consortium validated approach
CAE Community Materials Group
Request a material and model
Get the latest data
Materials Database
One system for all materials data
Direct import or file export
CAE
• CAE models become part of overall composites data management schema
• “Discipline-agnostic” approach: CAE models can be at the composite system level or at the constituents level to support multi-scale modelling
• All data is traceable to original tests and pedigree
• Unique, persistent reference IDs assigned to CAD/CAE model
• Direct integrations with FEA pre/post processor
• “Open source” file export for FEA experts who need to tweak models
Example – direct Abaqus/CAE integration
• Composite materials pose new
challenges from a data perspective
• Complex, anisotropic properties; history
& component-specific
• Pedigree and traceability of design data
is vital
• A software-based approach to composite
data management exists and has been
validated by aerospace industry
• Any CAE integration must be discipline-
agnostic
• Benefits extend to design optimization
and early material selection
Concluding thoughts