12' Future CADCAM Technologies 12'1 Objectives

1
12. Future CAD/CAM Technologies12.1 Objectives

• Soonhung Han
• http//me.kaist.ac.kr/newcourse/index.php?codeMAE
  548

2
12. Future CAD/CAM Technologies 12.1
Objectives 12.1.1 Design Process 12.2 Tools for
Preliminary Design 12.2.1 Expert Systems
12.2.2 Top-Down Design 12.2.3 Function-Based
Design 12.2.4 Bond Graphs 12.3 Collaborative
Design Tools 12.3.1 Systems Based on Design
History 12.3.2 Configuration Design 12.4
Qualitative Reasoning Methods 12.5 Epilogue
3
Ch.12 Future CAD/CAM Technologies

• Feature-based methods need to be extended for the
  early stages of design, where the designer is
  more concerned with the function and overall
  structure than its detailed geometry

4
12.1.1 Design Process

• During the early phases of design, desired
  characteristics and overall function of the
  product are made.
• In the later phases efficient manufacturing
• Design information are gradually elaborated from
  vague and fuzzy sketches to detailed descriptions

5
Importance of Early Design
Success or failure is determined in the early
phase of the project
RISK
RISK
Committed Cost
Impact of change on Cost
Accumulated Cost
Most of the submarine cost will be committed in
the first 18 months
Early Design
Support
Detailed Design.
Build
Time
6
?? ?? ??? ????? ?? ? ??? ?
Committed Cost vs. Lifecycle
Source DARPA RaDEO Project

• ?? ?? ??? 70? ?? ???? ??
• ????? ??? ???? ?? ??? ?? ?? ??? ??

7
5 stages of design process

• Functional design
• Conceptual design
• Embodiment design
• Detail design
• Engineering analysis (?? ??? ??)

8
12.2 Tools for Preliminary Design

• Soonhung Han
• http//icad.kaist.ac.kr/feature

9
12.2 Tools for Preliminary Design

• Early stage of design (functional, conceptual,
  embodiment) are not well supported by the present
  product modeling methods
• Function, behavior, structure (FBS)

10
12.2.1 Expert system

• Separate domain knowledge from the inference
  mechanism
• ICAD knowledge-based sys geometric modeling
• Design dynamic link between a knowledge-based
  system shell and AutoCAD

11
12.2.2 Top-Down Design

• High level abstraction stressing the function of
  the assembly.
• Detailed models of the individual components
• Transition between abstraction levels

12
(No Transcript)
13
Top-Down Design . . .

• Several levels of abstraction
• Represent design intent
• Support abstract geometry intermediate design
• Assembly design
• Fig.12.2 components and connections joined at
  common terminals
• ATMS (assumption-based truth maintenance system)
  allow accommodating different parallel designs

14
Common terminals denote shared variables
15
Top-Down Design . . .

• Part graphs (P-graphs) for assembly modeling
• Nodes represent objects or subassemblies, and
  edges represent constraints and relationships
• Multiple ways of representing the same assembly
  depending on the type of viewpoint

16
Part graph
17
Top-Down Design . . .

• Edinburgh Design System (EDS) 2 types of
  modules concrete modules and interface modules
• Concrete modules contain parameters and
  variables which are explicitly specified
  Features
• Interface modules Constraints between parameters
  and variables Spatial relationship of features

18
(No Transcript)
19
Top-Down Design . . .

• Logic-based mechanism Generate geometry's on the
  basis of incomplete constraint-based descriptions
• Tight constraints must strictly be satisfied
• Loose constraints specifies properties that are
  attractive for manufacturability and product
  quality Optimization

20
Strength of constraints tight, loose
21
Top-Down Design . . .

• Contour generation problem in 2 stages
• 1) Initial contour from tight constraints
• 2) Loose constraints, as far as possible
• Envelope method Contains all the holes, tight
  constraints, required segments
• Configuration method Rough shape of skeleton

22
Contour generation in 2 stages 1) initial
contour tight constraints envelope method or
configuration method 2) contour refinement
23
(No Transcript)
24
12.2.3 Function Based Design

• Modeling of function and behavior of mechanical
  devices
• Objective of function-based design is to support
  the early stage of design.
• Decompose functions hierarchically into
  sub-functions
• In the early creative design stages, a designer
  knows the required functionality of the device
  but has no idea about its implementation and
  appearance.

25
Function Based Design . . .

• 1. The designer determines the function.
• 2. The designer determines how the functional
  hierarchy can be realized by physical entities.
• 3. The designer should be able to evaluate the
  created model to check that all expected
  functions or phenomena occur and to observe
  possible unexpected side effects (behavior).
• 4. Functional design produces a functional
  structure.

26
(No Transcript)
27

• Fig 12.9

28
SYSFUND

• Framework for representing function and behavior
• Has knowledge base dealing with functions,
  behaviors, structures (FBS)
• Function prototypes are mapped into physical
  feature.
• Physical phenomena, entities, relations

29
functional decomposition function prototype
entities phenomena relations
30
12.2.4 Bond Graphs

• Bond graphs is to use the concept of energy.
• Bond graphs treat any kind of energy flow in
  terms of a force-like quantity and a flow-like
  quantity, referred to as effort(e) and flow(f).
• (e) volt (f) current (e) force (f) speed
• Bond graph has mainly been used for computer
  simulation of dynamic systems.
• The behaviors of various types of systems can be
  uniformly modeled.

31
Ports in Bond Graph

• Storage process
• spring, flywheel, moving mass, represented by
  (C), releasing (I)
• Dissipative process (R)
• Source
• Conversion process
• TF ( Between Effort energy and Flow energy)
• GY (Energy conversion between two deferent
  domain)
• Distribution process Kirchhoffs law

32
(No Transcript)
33
Fig 12.12
??? ??? ???? ???, ???? ??
34
12.3 Collaborative Design Tools

• Soonhung Han

35
12. Future CAD/CAM Technologies 12.1
Objectives 12.1.1 Design Process 12.2 Tools for
Preliminary Design 12.2.1 Expert Systems
12.2.2 Top-Down Design 12.2.3 Function-Based
Design 12.2.4 Bond Graphs 12.3 Collaborative
Design Tools 12.3.1 Systems Based on Design
History 12.3.2 Configuration Design 12.4
Qualitative Reasoning Methods 12.5 Epilogue
36
12.3 Collaborative Design Tools

• Key issues for product modeling in collaborative
  design
• Standardization, Modularity, Quality, Reuse,
  Compatibility, Constraints, Change management,
  and Change propagation

37
12.3.1 Systems Based on Design History

• Eng. Design History Step-by-step account of the
  events and states through which a design artifact
  evolved
• Design reuse Design history system (DHS)
• Browsing Replay of design, Reuse of a portion of
  the design
• Structural queries
• Advanced queries Explain a design process,
  Analyze a design history

38

• Analysis how, why
• Design learning CBR
• Version and configuration management
• Design maintenance Backtrack and parameter
  change
• Fig. 12.13 classification
• Process, model, versioning, knowledge, OODB
• Problem Burden to designer, otherwise
  unstructured design results
• Only few DHS exist today Querying and
  interpreting design histories

39
(No Transcript)
40
12.3.2 Configuration Design

• Trend since 1970s in AI, link with CAD systems.
• Basic Concepts
• Complex product, thousands of parts and
  subassemblies BOM, SCM
• Product design rules, product configuration rules
• Product configurator
• -gt Tool for automating the product configuration
  process.
• Expressed in product configuration model
• Fig. 12. 14

41
(No Transcript)
42

• Rules Permissible combination. Check validity of
  a configuration.
• Proper treatment of changes during and after the
  configuration process Specification, components
• -gt Major challenge to configuration modelers
• Long term maintenance of configuration
  information components, parameters, rules

43

• State of the Art and Current Work
• XCON Difficult to Use
• Need of general models and methods for
  configuration modeling
• AI research has the central role in
  configuration.
• - Nexpert Object of Neuron Data Inc.
• Intelligent CAD System
• - ICAD
• - Design
• - SalesBUILDER

44
12.4 Qualitative Reasoning Methods

• QR methods came from the desire to support
  reasoning needed at the conceptual design level.
• Involves a common sense (??) understanding of the
  behavior of devices at an abstract level.
• Qualitative models can be used when quantitative
  models are too complex.
• Bobrow, Forbus, DeKleer propose QR approach
• QR models a system in terms of some key
  parameters which can assume only qualitative
  values.
• Real Number -, 0, -gt inc, 0, dec
• QR methods may be incorporated in conceptual
  design system in the future

45
12.5 Epilogue

• Advanced product modeling
• Feature technology beyond solid modeling

46
cadcam ??? ??

• ??? e-Engineering, Collaborative Commerce (CPC,
  PLM)
• VR lt 3D lt 2D
• ???? KBE (knowledge-based engineering),
  Intelligent CAD

47
??? ??
48
???? ??
?? ??? ??? ????? ???
??? ???? ??? ?????? ?? ??