Setup Planning and Tolerance Analysis
6.4 Information Modelling
Information models are data structures that represent information content in part design and manufacturing. The main task of information modelling is to capture, describe, and maintain the information structure and information relationships in the CAMP. In this section, an object-oriented systems analysis (OSA) approach [6.23] is utilised to analyse and represent information models, and a systematic information modelling hierarchy is proposed to model both static and dynamic characteristics of information from a system perspective.
6.4.1.1 The Object-oriented Systems Analysis (OSA) Approach
Object-oriented (O-O) modelling is recognised as a powerful tool to model real- world systems. An object is an encapsulation of data and procedures (or methods) that operates on the data. A relationship establishes a logical connection among objects. An object could be an existing entity in the real world such as a part and a machine tool. The definition of objects implies the correlative relationships between the data and the procedures related to the data. The real world can be considered as a group of interacting objects. The interactions, including static relationships and dynamic relationships, are described according to the way that human beings think.
Therefore, the main task of O-O modelling for a system is to identify objects and analyse their interactions within the system.
In this section, an OSA approach is used to analyse the information in CAMP: an object-relationship model (ORM) is used to represent the static relationships between objects. An object-behaviour model (OBM) describes the behaviour of individual objects and how objects respond to dynamically occurring events and conditions. An object-interaction model (OIM) expresses the interactions between objects.
Object-relationship model (ORM). An ORM is created to represent the static relationships between objects, which is described by an ORM diagram. A rectangle represents an object, and the variables of the object are shown in the lower rectangle.
There are two basic relationships: Generalisation–specification relationship means the is-kind-of relationship, which is represented by a transparent triangle in an ORM diagram; and whole–part relationship indicates the is-part-of relationships, which is described by a solid-filled triangle. Users can define their own relationships with the specific relationship name attached to ORM diagrams.
Object-behaviour model (OBM). The objective of a behaviour model is to describe the way that each object in a system interacts, functions, responds, or performs. A behaviour model for an object is similar to a job description for an object. In this chapter, state nets are used to represent OBMs, which is composed of states, triggers and actions. A state, represented by a rounded rectangle, describes an object’s status, phase, situation, or activity. The events and conditions that activate state transitions are called triggers. The activity that an object performs is called an action. A rectangle is divided into two sections, the top section contains a trigger description and the bottom section contains the actions.
Object-interaction model (OIM). The ORMs describe the static relationships among objects. The OBMs describe the behaviour of an object, but in isolation from other objects. An OIM model is used to describe the interaction among objects.
One object interacts with another in many different ways. For example, an object
may send information to another, request information from another, alter another object, or cause another to do some actions. To understand object interaction, we must understand: 1) What objects are involved in the interaction; 2) How the objects act or react in the interaction; and 3) The nature of the interaction. Since objects are identified in ORMs, and the behaviour of each object is described in OIMs, a combination of ORMs and the state nets is used to create OIMs, in which a zigzag arrow is used to describe the interactions between objects. In general, the interactions indicated by zigzag arrows imply the user-defined correlative relationships associated to specific systems.
6.4.1.2 Systematic Information Modelling Hierarchy
When using the OSA approach to model a complex system, high-level abstraction of objects is applied to reduce complexity and make the information models easy to create, maintain, and display. A high-level object package groups relative objects and the relationships among the objects into a single object. The top-down approach is used to expand a high-level object into low-level objects and relationships. Figure 6.19 shows the hierarchical structure of system information models. The building of information models is split into three levels.
The definition of a system model contains domains that are subdivided into subsystems. The system model may be deduced from analysis of the system’s high- level object interaction models. The definition of an information model contains objects that are subdivided into states. The definition of the state model describes the behaviour of objects.
Domain Domain
Domain
Information model
Object Object
Object Object
State model System model
State State
State State
OIM
ORM
OBM Is generated from
Is generated from
Is generated from
Object Package
Object Package Object Package
Object Object
State State
Figure 6.19. Systematic information modelling hierarchy
6.4.2 Information Model of CAMP for Mass Customisation
The tasks of CAMP have been carried out by four functional modules, which are shown by the grey round rectangles in Figure 6.20 [6.17]. Part information is composed of features and the tolerance relationships between the features. Part information modelling extracts features from part CAD models, with feature manufacturing strategies associated with the features. Setup planning is carried out based on tolerance and manufacturing resource capability analyses. Multi-part
fixture design may be involved. Conceptual fixture design is used to determine part layout on the fixtures. Manufacturing plan generation is to determine the optimal process sequence and tool path. The information involved in CAMP is organised into three categories:
Manufacturing resource capability
Cutter Part Information Model
1. Manufacturing feature specification 2. FTG generation
3. Feature manufacturing methods Features
Feature manufacturing strategies
Setup plans
Manufacturing plans Object-oriented Information modeling
Setup Planning 1. Feature grouping 2. Setup generation
Conceptual Fixture Design 1. Fixture base determination’
2. Part layout on fixture base
Manufacturing Plan Generation 1. Machine tool selection 2. Process sequence generation 3.Cutting parameter & global tool path
Blackboards CAMP function modules Data & Knowledge base
CAMP Document Part CAD models
Combined features in part families
Combined features’
process models Feature
Level
Manufacturing knowledge
Fixture
Machine tool Part
level
System level Best of Practice
(BOP)
Figure 6.20. Tasks and information content in the CAMP
Manufacturing databases and knowledge bases. In CAMP, the information is considered and stored in the manufacturing data and knowledge bases. Combined features are defined based on particular part families. The parts in the same family may have the same type of combined features and feature relationships so that the part-family BOP can be used as the reference to generate new plans [6.17].
Combined features are associated with predefined manufacturing strategies, in which customised combination cutters, tool paths, and machine tool motion requirements are specified for particular part families. The designs of cutters and tool paths are based on prior experience and are stored in templates. Therefore, when the same combined feature is encountered, the existing experience can be reused.
Manufacturing resources include cutters, machine tools, and fixtures. Some of them are standard tools and can be brought from the market. The others are designed specifically for particular processes used in manufacturing plans. The capabilities of available manufacturing resources should be described and stored in a format that the CAMP can interpret and manipulate. Manufacturing rules and knowledge are extracted from BOP and applied in the automated reasoning mechanism such as automated determination of feature manufacturing strategy, setup planning, and manufacturing plan generation. Three levels of manufacturing knowledge are identified, general knowledge without regard to a particular shop, shop-level process details, and part-level information based on particular part-family production in a specific machine shop.
Best-of-practice (BOP). BOP for part families is the most important reference enabling engineers to design a new manufacturing plan. The specific decision- making strategies of part families are embedded in the BOP, which include strategies about how to deal with the correlative relationships between part design, part manufacturing, and the utilisation of manufacturing resource capabilities.
Therefore, the decision-making strategies in the BOP must be identified first, and then the BOP should be described in a format that is accurate, complete, and unambiguous, so that it can be used by the CAMP system. In this chapter, information in BOP is divided into three levels: feature level, part setup planning level, and machine level.
Blackboards.Blackboards are used to store the shared information generated by the modules of the CAMP. It is in the blackboards that computers deal with the manufacturing information that is represented by information models. There are four blackboards in CAMP, which store features, features’ manufacturing strategies, part setup planning and manufacturing plan information. The design of information models considers the following issues, information relationship, information integration, and information extendibility. The design of information models should pay attention to correlative relationships and try to avoid information redundancy in models. The design of information models should consider the overall information requirements of the CAMP system. Different functional modules may have different requirements for the same information model. With the consideration of the new demands in product development practice, the scope of the CAMP may change accordingly. Therefore, the information models should be extendable to accommodate more information content without damaging origin information content and information relationships.
P art o b je ct p ac k ag e s
M an u fa cturin g re so u rc e ca p a bility ob je c t
pa c ka g e s M a n u fa ctu rin g plan n in g
ob je c t p a ck a ge s
M a nu fa c tu rin g k n ow led g e o bjec t
p a c ka g es
Capability evaluation C on troltheplan generation
Capability inform
ation U pdatekno wledge
Part d esign inform
ation
U pd a te re s ou rc e s C h o ic e o f re s ou rce s M a n u fa ctu ra b ility e va lua tion
Figure 6.21. System models for the CAMP
The systematic information modelling methodology is used in this chapter to analyse and represent the information in the blackboards of the CAMP. As shown in Figure 6.21, four object packages are established to describe the primary information involved, as well as the interactions between these packages. The part information is the input, which is composed of features and the tolerance relationships between features, and the features’ manufacturing strategies are linked with features. The manufacturing planning package includes feature-level, part-level and machine-level decision-making strategies. The manufacturing knowledge package provides the knowledge constraint to control the manufacturing planning
behaviours. The manufacturing resource capability package provides the description of available manufacturing resources. Hence, breaking down these packages will result in a detailed study on the correlative relationships within the CAMP, which facilitate the use of the part-family BOP to help engineers rapidly design new manufacturing plans.
In part information representation, the procedure of establishing a process object for a feature is called feature-level decision making, whose OIM is shown in Figure 6.22. Similarly, the OIMs can be established for setup planning, manufacturing resource planning, and manufacturing planning generation.
Process undefined
Process underdefined Define cutter
Select a end mill from database
Define tool path Design a tool path according to
cutter and feature dimension
Process defined Process
Feature parameters
Feature parameters FTG
Feature ID
FTGEdge FeatureID1 FeatureID2 is_ordered
Tolerance ToleranceType Value
Dimension Value Maxvalue Minvalue Graph
Figure 6.22. OIM of part object and one of the process objects