Elements towards Next-Generation Knowledge Representations and Product Modeling Techniques

1
Elements towards Next-Generation Knowledge
Representations and Product Modeling Techniques
Planning Meeting for Product, Lifecycle
Management, and Systems Engineering Models May
20, 2003 NIST Gaithersburg MD (via telecon)

• Russell.Peak_at_marc.gatech.eduhttp//itimes.marc.ga
  tech.edu/
• http//eislab.gatech.edu/projects/

2
Contents
Purpose Help identify comparison factorsand
encourage thinking about next-generation needs

• Multiple views in a knowledge representation
• Declarative thinking
• Object graph view of model interoperability
• Some factors for comparing knowledge
  representations
• Leveraging multiple standards
• Managing computing environments via systems
  engineering methods
• Elevated terminology thinking
• See backup slides for other examples references

3
Contents

• Multiple views in a knowledge representation
• Human-sensible computer-sensible
• Graphical, lexical, application-oriented
• Declarative thinking
• Multi-directional (non-causal)
• With derivable lower-level procedural approaches
• Object graph view of model interoperability
• Leveraging multiple standards
• Managing computing environments via systems
  engineering methods
• Elevated terminology thinking

Examples from Constrained Objects (COBs)
CAD-CAE Integration
4
COB Structure Graphical Forms Tutorial
Triangle Primitive
a. Shape Schematic-S
c. Constraint Schematic-S
b. Relations-S
Basic Constraint Schematic-S Notation
d. Subsystem-S (for reuse by other COBs)
Aside This is a usage view in AP210
terminology (vs. the above design views)
5
COB Structure (cont.) Lexical Form Tutorial
Triangle Primitive
e. Lexical COB Structure (COS)
COB triangle SUBTYPE_OF geometric_shape base,
b REAL height, h REAL diagonal,
d REAL area, A REAL RELATIONS r1
"ltareagt 0.5 ltbasegt ltheightgt" r2
"ltdiagonalgt2 ltbasegt2 ltheightgt2" END_CO
B
for reference c. Constraint Schematic-S
6
Example COB InstanceTutorial Triangle Primitive
Constraint Schematic-I
Lexical COB Instance (COI)
example 1, state 1.1
state 1.0 (unsolved) INSTANCE_OF triangle
base 2.0 height 3.0 area
? diagonal ? END_INSTANCE state 1.1
(solved) INSTANCE_OF triangle base
2.0 height 3.0 area 3.0
diagonal 3.60 END_INSTANCE
Basic Constraint Schematic-I Notation
7
Multi-Directional I/O (non-causal)Tutorial
Triangle Primitive
Constraint Schematic-I
Lexical COB Instance (COI)
state 2.1 (solved) INSTANCE_OF triangle base
2.0 height 9.0 area 9.0
diagonal 9.22 END_INSTANCE state 3.0
(unsolved) INSTANCE_OF triangle base
2.0 height ? area 6.0 diagonal
? END_INSTANCE state 3.1 (solved) INSTANCE_O
F triangle base 2.0 height 6.0
area 6.0 diagonal 6.32 END_INSTANCE
example 1, state 2.1
example 1, state 3.1
8
COBs as Building Blocks Tutorial Triangular
Prism COB Structure
a. Shape Schematic-S
c. Constraint Schematic-S
b. Relations-S
e. Lexical COB Structure (COS)
d. Subsystem-S (for reuse by other COBs)
COB triangular_prism SUBTYPE_OF geometric_shape
length, l REAL cross-section
triangle volume, V REAL RELATIONS
r1 "ltvolumegt ltcross-section.areagt
ltlengthgt" END_COB
9
Example COB InstanceTutorial Triangular Prism
Constraint Schematic-I
Lexical COB Instance (COI)
example 1, state 1.1
state 1.0 (unsolved) INSTANCE_OF
triangular_prism cross-section.base 2.0
cross-section.height 3.0 length 5.0
volume ? END_INSTANCE state 1.1
(solved) INSTANCE_OF triangular_prism
cross-section.base 2.0 cross-section.height
3.0 cross-section.area 3.0 length
5.0 volume 15.0 END_INSTANCE
Basic Constraint Schematic-I Notation
10
COB Modeling Languages Views
Structure Level (Template)
Instance Level
11
Contents

• Multiple views in a knowledge representation
• Declarative thinking
• Object graph view of model interoperability
• Include connections with lower-level models
  COTS tools
• Facilitate solution management reasoning
  control
• Some factors for comparing knowledge
  representations
• Leveraging multiple standards
• Managing computing environments via systems
  engineering methods
• Elevated terminology thinking

Examples from Constrained Objects (COBs)
CAD-CAE Integration
12
Constrained Object Panorama for Multi-Fidelity
CAD-CAE InteroperabilityFlap Link Benchmark
Example
13
Flexible High Diversity Design-Analysis
Integration Phases 1-3 Airframe ExamplesBike
Frame / Flap Support Inboard Beam
Analysis Modules (CBAMs) of Diverse
FeatureMode, Fidelity
Design Tools
Modular, Reusable Template Libraries
MCAD Tools CATIA v4, v5
XaiTools
Analysis Tools
1.5D
General Math Mathematica In-HouseCodes
LugAxial/Oblique Ultimate/Shear
Image API (CATGEO) VBScript
Analyzable Product Model
XaiTools
1.5D
FittingBending/Shear
Materials DB
FEA Elfini
MATDB-like
3D
Assembly Ultimate/FailSafe/Fatigue
Fasteners DB
FASTDB-like
Item not yet available in toolkit (all others
have working examples)
14
Usage of a COB-based Analysis TemplateCAD-CAE
Interoperability during Lug Strength Analysis
CAD-CAE Associativity (idealization usage)
Geometry
Material Models
Solution Tool Interaction
Boundary Condition Objects (links to other
analyses)
Model-based Documentation
Requirements
15
Convergence of Representations
Database Techniques (data structure, storage )
Software Development (algorithms )
Flow Charts
ER
OMT
EER
STEP Express
UML
Constrained Object - like Representations
COBs, OCL, ...
Constraint graphs
Objects
Rules
Artificial Intelligence Knowledge-Based
Techniques (structure combined with
algorithms/relations/behavior)
16
Contents

• Multiple views in a knowledge representation
• Declarative thinking
• Object graph view of model interoperability
• Include connections with lower-level models
  COTS tools
• Facilitate solution management reasoning
  control
• Some factors for comparing knowledge
  representations
• Leveraging multiple standards
• Managing computing environments via systems
  engineering methods
• Elevated terminology thinking

17
Dimensions of AssociativitySome Knowledge
Representation Comparison Factors
Associativity Relations among objects
r1
System Y
a
a
System X
r2
System Z
b

• Operand representation a, b
• Type numeric, logical, string, , general object
• Human-sensible vs. computer-sensible
• Computer-sensible Flattened vs.
  object/feature-oriented
• Other facets security, units, uncertainty,
  maturity, version history, (un)known/withheld,
• Relation representation r1, r2
• Relation type Math formula, geometric
  constraint, computable algorithm, computer system
  (e.g., FEA tool), higher order constraint,
  arbitrary human process, ...

electrical circuits analogy
18
Dimensions of Associativity (cont.)

• Relation representation (continued)
• Explicit vs. implicit vs. unrecognized vs.
  unknown
• Human-sensible vs. computer-sensible
• Computer-sensible Dumb string vs. smart string
  vs. object/feature-oriented relation
• Level instance, template (schema, structure),
  adaptable template
• Other facets priority, (in)active, plus similar
  facets as operands
• Relation directionality
• Uni-directional vs. multi-directional vs.
  iteratively multi-directional
• Relation duration
• Continuous (live) vs. event-controlled
• Relation granularity
• Coarse vs. fine (macro vs. micro)

• Associativity graph type
• Declarative vs. procedural
• Cyclic vs. acyclic
• Variable vs. fixed topology

19
Contents

• Multiple views in a knowledge representation
• Declarative thinking
• Object graph view of model interoperability
• Leveraging multiple standards
• Managing computing environments via systems
  engineering methods
• Including versioning configuration mgt. of
  meta-models, standards, and tools
• Elevated terminology thinking

20
Tool-Product Model Schema Relationships in
aStandards-Based Engineering Framework
Electrical CAD Tools
Systems Engineering Tools
Mechanical CAD Tools
Eagle
Doors
Pro/E
Traditional Tools
MentorGraphics
Slate
CATIA

AP210
AP203, AP214
AP233

• Smart Product Model
• Building Blocks
• Models meta-models
• International standards
• Industry specs
• Corporate standards
• Local customizations
• Modeling technologies
• Express, UML, XML, COBs,

AP210
AP2xx
XaiToolsPWA-B
LKSoft,
STEP-Book AP210, SDAI-Edit, STI AP210 Viewer, ...
XaiToolsPWA-B
EPM, LKSoft, STI,
pgef
Gap-Filling Tools
PWB Stackup Tool,
Instance Browser/Editor
Engineering Framework Tool
21
Primary Technologies for Schema-based
Engineering Frameworks
Based on Engineering Framework Interest Group
(EFWIG) emails from steve.waterbury_at_gsfc.nasa.gov
(dated July 13, 2002 wrt PGPDM directions) and
David Leal (dated November 26, 2002).
22
Contents

• Multiple views in a knowledge representation
• Declarative thinking
• Object graph view of model interoperability
• Leveraging multiple standards
• Managing computing environments via systems
  engineering methods
• Elevated terminology thinking

23
Needed Shifts in Engineering Thinking
Traditional Viewpoint

• Math-based models of physical behavior
• Learn mathematics as a modeling language

• Information models of physical objects
• Includes math-based models of physical behavior,
  but in their richer context
• Learn information representation as another type
  of modeling language

Note Information models have their roots in
modern mathematics (e.g. set theory).
24
Needed Shifts in Engineering Thinking (cont.)
Traditional Computing Viewpoint

• Tool usage
• Data / files
• Data exchange
• Translators
• Single tools
• Drawings documents
• Calculations

• Model creation interaction (using tools) -
  knowledge capture
• Information models knowledge representations
  (objects)
• Model connection, associativity, interoperability
  (often via equality relations)
• Interfaces
• Integrated submodels
• Views (submodels) connected to their richer
  models
• Usage of model operations

Objects (having structure and operations) that
are interrelated.
25
Summary
Purpose Help identify comparison factorsand
encourage thinking about next-generation needs

• Multiple views in a knowledge representation
• Declarative thinking
• Object graph view of model interoperability
• Some factors for comparing knowledge
  representations
• Leveraging multiple standards
• Managing computing environments via systems
  engineering methods
• Elevated terminology thinking
• See backup slides for other examples references

26
Other Slides for Reference
27
Procedural vs. Declarative Knowledge
Representations
h
b
A 1/2 bh
How does one compute h given A, b ?
28
Constrained Objects A Knowledge Representation
for Design, Analysis, and Systems Engineering
Interoperability

• Students Manas Bajaj, Injoong Kim, Greg Mocko
  Faculty Russell Peak

Objectives
Contributions
Chip Package Stress Analysis Template

• Develop better methods of capturing engineering
  knowledge that
• Are independent of vendor-specific CAD/CAE/SE
  tools
• Support both easy-to-use human-sensible views
  and robust computer-sensible formulations in
  a unified manner
• Handle a diversity of product domains,
  simulation disciplines, solution methods, and
  leverage disparate vendor tools
• Apply these capabilities in a variety of
  sponsor-relevant test scenarios
• Proposed candidates are templates and custom
  capabilities for design, analysis, and
  systems engineering

• To Scholarship
• Develop richer understanding of modeling
  (including idealizations and multiple levels of
  abstraction) and representation methods
• To Industry
• Better designs via increased analysis intensity
• Increased automation and model consistency
• Increased modularity and reusability
• Increased corporate memory via better knowledge
  capture

Constrained Object (COB) Formulations
Approach and Status
Resources Needed

• Approach
• Extend and apply the constrained object (COB)
  representation and related methodology based on
  positive results to date
• Expand within international efforts like the OMG
  UML for Systems Engineering work to broaden
  applicability and impact
• Status
• Current generation capabilities have been
  successfully demonstrated in diverse environments
  (circuit boards, electronic chip packages,
  airframes) with sponsors including NASA, Rockwell
  Collins, Shinko (a major supplier to Intel), and
  Boeing.
• Templates for chip package thermal analysis are
  in production usage at Shinko with over 75
  reduction in modeling effort (deformation/stress
  templates are soon to follow)

• Support for 1-3 students depending on project
  scope
• Sponsor involvement to provide domain knowledge
  and facilitate pilot usage

COB-based Airframe Analysis Template

• Additional Information
• 1. http//eislab.gatech.edu/projects/
• 2. Response to OMG UML for Systems Engineering
  RFIhttp//eislab.gatech.edu/tmp/omg-se-33e/
• 3. Characterizing Fine-Grained Associativity
  Gaps A Preliminary Study of CAD-E Model
  Interoperabilityhttp//eislab.gatech.edu/pubs/con
  ferences/2003-asme-detc-cie-peak/

Russell.Peak_at_marc.gatech.edu -- 2003-05-12
29
COB-based Libraries ofAnalysis Building Blocks
(ABBs)
Continuum ABBs
Extensional Rod
Material Model ABB
1D Linear Elastic Model
modular re-usage
Torsional Rod
30
Flap Link ExampleParametric Design Description
Extended Constraint Graph
COB Structure (COS)
31
Representing External Tools as COB
RelationsParametric FEA Model
FEA Tool
32
Constrained Object (COB) RepresentationCurrent
Technical Capabilities - Generation 2

• Capabilities features
• Various forms computable lexical forms,
  graphical forms, etc.
• Enables both computer automation and human
  comprehension
• Sub/supertypes, basic aggregates, multi-fidelity
  objects
• Multi-directionality (I/O changes)
• Reuses external programs as white box relations
• Advanced associativity added to COTS frameworks
  wrappers
• Analysis module/template applications (XAI/MRA)
• Analysis template languages
• Product model idealizations
• Explicit associativity relations with design
  models other analyses
• White box reuse of existing tools (e.g., FEA,
  in-house codes)
• Reusable, adaptable analysis building blocks
• Synthesis (sizing) and verification (analysis)

33
Constrained Objects (cont.) Representation
Characteristics Advantages - Gen. 2

• Overall characteristics
• Declarative knowledge representation (non-causal)
• Combining object constraint graph techniques
• COBs (STEP EXPRESS subset) (constraint
  graph concepts views)
• Advantages over traditional analysis
  representations
• Greater solution control
• Richer semantics (e.g., equations wrapped in
  engineering context)
• Unified views of diverse capabilities
  (tool-independent)
• Capture of reusable knowledge
• Enhanced development of complex analysis models
• Toolkit status (XaiTools v0.4)
• Basic framework, single user-oriented, file-based

34
An Introduction to X-Analysis Integration (XAI)
Short Course Outline

• Part 1 Constrained Objects (COBs) Primer
• Nomenclature
• Part 2 Multi-Representation Architecture (MRA)
  Primer
• Analysis Integration Challenges
• Overview of COB-based XAI
• Ubiquitization Methodology
• Part 3 Example Applications
• Airframe Structural Analysis (Boeing)
• Circuit Board Thermomechanical Analysis (DoD
  ProAM JPL/NASA)
• Chip Package Thermal Analysis (Shinko)
• Summary
• Part 4 Advanced Topics Current Research

35
Techniques for Complex System Representation
Model Interoperability (CAD-CAE)
http//eislab.gatech.edu/research/
a. Multi-Representation Architecture (MRA)
b. Explicit Design-Analysis Associativity
c. Analysis Module Creation Methodology
36
Circuit Board Design-Analysis IntegrationElectron
ic Packaging Examples PWA/B
Analysis Modules (CBAMs) of Diverse Mode
Fidelity
Design Tools
Modular, Reusable Template Libraries
ECAD Tools Mentor Graphics, Accel
Analysis Tools
XaiTools PWA-B
General Math Mathematica
1D, 2D, 3D
STEP AP210 GenCAM, PDIF
Solder Joint Deformation
FEA Ansys
PWB Stackup Tool XaiTools PWA-B
Analyzable Product Model
PWB Warpage
XaiToolsPWA-B
1D, 2D
Laminates DB
PTH Deformation Fatigue
Materials DB
1D, 2D
37
Iterative Design Analysis PWB Stackup Design
Warpage Analysis
PWB Stackup Design Tool
1D Thermal Bending Model
Quick Formula-based Check
Layup Re-design
PWB Warpage Modules
Analyzable Product Model
2D Plane Strain Model
Detailed FEA Check
38
PWB Warpage Modulesa.k.a. CBAMs COB-based
analysis templates
PWB Thermal Bending Model (1D formula-based CBAM)
Usage of Rich Product Models
APM
PWB Plane Strain Model (2D FEA-based CBAM)
39
Example Chip Package Products Source
www.shinko.co.jp
Quad Flat Packs (QFPs)
Plastic Ball Grid Array (PBGA) Packages
Wafer Level Package (WLP)
Glass-to-Metal Seals
System-in-Package (SIP)
40
Flexible High Diversity Design-Analysis
Integration Electronic Packaging Examples Chip
Packages/Mounting Shinko Electric Project
Phase 1 (production usage)
Analysis Modules (CBAMs) of Diverse Behavior
Fidelity
Modular, Reusable Template Libraries
Design Tools
Prelim/APM Design Tool
Analysis Tools
XaiTools ChipPackage
XaiTools ChipPackage
General Math Mathematica
FEAAnsys
Thermal Resistance
Analyzable Product Model
3D
XaiTools
PWB DB
Materials DB
ThermalStress
EBGA, PBGA, QFP
Basic 3D
Basic Documentation Automation
AuthoringMS Excel
Demonstration module
41
Typical Issues Knowledge Representation,Inter-Mo
del Associativity (Model Interoperability)
CAD Model bulkhead assembly attach point
CAE Model channel fitting analysis
material properties
detailed design geometry
idealized analysisgeometry
analysis results
42
Flexible High Diversity Design-Analysis
Integration Phases 1-3 Airframe ExamplesBike
Frame / Flap Support Inboard Beam
Analysis Modules (CBAMs) of Diverse
FeatureMode, Fidelity
Design Tools
Modular, Reusable Template Libraries
MCAD Tools CATIA v4, v5
XaiTools
Analysis Tools
1.5D
General Math Mathematica In-HouseCodes
LugAxial/Oblique Ultimate/Shear
Image API (CATGEO) VBScript
Analyzable Product Model
XaiTools
1.5D
FittingBending/Shear
Materials DB
FEA Elfini
MATDB-like
3D
Assembly Ultimate/FailSafe/Fatigue
Fasteners DB
FASTDB-like
Item not yet available in toolkit (all others
have working examples)
43
Explicit Capture of Idealizations (part-specific
template adaptation in bike frame case)
Idealized Features in CAE Model
G2
Detailed Features/Parameters Tagged in CAD Model
(CATIA)
zf
yf
te
yf
yf
xf
xf
zf
b
cavity3.base.minimum_thickness
xf
cavity3.width, w3
yf
zf
cavity 3
rib9
xf
G1
rib8
Tension Fitting Analysis
t8,t 9
rib8.thickness rib9.thickness
Often missing in todays process
Gi - Relations between idealized CAE parameters
and detailed CAD parameters G1 b
cavity3.inner_width rib8.thickness/2
rib9.thickness/2 G2 te cavity3.base.minimu
m_thickness
44
Todays Fitting Catalog Documentation from DM
6-81766 Design Manual
Categories of Idealized Fittings
Calculation Steps
Angle Fitting
Channel Fitting End Pad Bending Analysis
Channel Fitting
Bathtub Fitting
45
Modular Fitting TemplatesObject-Oriented
Hierarchy of Analysis Building Blocks (ABBs)
ABB - independent of specific products
- usable on many designs
ABB
Working Examples
Specialized Analysis Body
Specialized Analysis System
Fitting Casing Body
Fitting Bolt Body
Fitting Washer Body
bolt
Fitting System ABB
washer
casing
load
P
Fitting Wall ABB
Fitting End Pad ABB
Open Wall Fitting Casing Body
Channel Fitting Casing Body
Angle Fitting Casing Body
Fitting End Pad Bending ABB
Bathtub Fitting Casing Body
Fitting End Pad Shear ABB
Open Wall Fitting End Pad Bending ABB
Channel Fitting End Pad Bending ABB
-

)
2
(
t
e
K
C
3
b
1

K
K
C
2
1
1
46
Channel Fitting System ABBs
End Pad Bending Analysis
End Pad Shear Analysis
ABB analysis building block
47
Bike Frame Bulkhead Fitting Analysis
TemplateUsing Constrained Object (COB)
Knowledge/Info Representation
48
Bike Frame Bulkhead Fitting AnalysisCOB-based
Analysis Template - in XaiTools
Focus Point of CAD-CAE Integration
Detailed CAD data from CATIA
Library data for materials fasteners
Idealized analysis features in APM
Object-oriented spreadsheet
Modular generic analysis templates (ABBs)
Explicit multi-directional associativity between
detailed CAD data idealized analysis features
49
Cost of Associativity GapsReference
http//eislab.gatech.edu/pubs/reports/EL004/

• Categories of Gap Costs
• Associativity time labor
• - Manual maintenance
• - Little re-use
• - Lost knowledge
• Inconsistencies
• Limited analysis usage
• - Fewer parts analyzed
• - Fewer iterations per part
• Wrong values
• - Too conservative Extra part costs
  and performance inefficiencies
• - Too loose Re-work, failures, law
  suits

50
Information Capture GapsContent Coverage and
Semantics
Existing Tools
Tool A1
Tool An
...
Legend
Content Coverage Gaps
dumb information capture (only
human-sensible, I.e., not computer-sensible)

• Smart Product Model
• Building Blocks
• Models meta-models
• International standards
• Industry specs
• Corporate standards
• Local customizations
• Modeling technologies
• Express, UML, XML, COBs,

Content Semantic Gaps
Example dumb figures
51
Summary

• Tool independent model interoperability
• Application focus analysis template methodology
• Multi-representation architecture (MRA)
  constrained objects (COBs)
• Addresses fundamental gaps
• Idealizations CAD-CAE associativity
  multi-fidelity, multi-directional, fine-grained
• Based on information knowledge theory
• Structured, flexible, and extensible
• Improved quality, cost, time
• Capture engineering knowledge in a reusable form
• Reduce information inconsistencies
• Increase analysis intensity effectiveness
• Reducing modeling cycle time by 75 (production
  usage)

52
For Further Information ...

• Contact Russell.Peak_at_marc.gatech.edu
• Web site http//eislab.gatech.edu/
• Publications, project overviews, tools, etc.
• See X-Analysis Integration (XAI)
  Central http//eislab.gatech.edu/research/XAI_Cen
  tral.doc
• Engineering Framework Interest Group
  (EFWIG) http//eislab.gatech.edu/efwig/
• XaiTools home page http//eislab.gatech.edu/tool
  s/XaiTools/
• Pilot commercial ESB http//www.u-engineer.com/
• Internet-based self-serve analysis
• Analysis module catalog for electronic packaging
• Highly automated front-ends to general FEA math
  tools