Showing papers in "Manufacturing Research and Technology in 1994"
TL;DR: In this paper, a volume decomposition method called minimum convex decomposition by halfspace partitioning has been developed to recognize machining features, where the total volume to be removed by machining is obtained by subtracting the part from the stock.
Abstract: In a design by feature system, the database may contain features that are not known to a process planning system. Also, additional features arise from the shape and size of the workpiece from which the part is to be machined. Production rules are inadequate for dealing with these undocumented features. In order to get machining features, one must enhance and transform the design database. A volume decomposition method called minimum convex decomposition by halfspace partitioning has been developed to recognize machining features. First, the total volume to be removed by machining is obtained by subtracting the part from the stock. This volume is decomposed into minimum convex cells by halfspace partitioning at every concave edge. A method called maximum convex cell composition is developed to generate all alternative volume decompositions. The composing subvolumes are classified based on degree of freedom analysis. Manufacturing knowledge is characterized in fundamental terms, based on tool and workpiece motion (deep knowledge), by means of algebraic expressions. Inverse mapping of decomposed volumes into these expressions determines feasibility of various machining operations. The method deals effectively with feature interactions and results in alternative machining sequences.
75 citations
TL;DR: The spatial decomposition and composition method of feature recognition decomposes the space surrounding a solid model into minimal convex cells with the geometric surfaces of the solid model that has a pre-defined set of topological and geometric characteristics.
Abstract: A generic form feature recognition method called “spatial decomposition and composition” has been developed. The method can recognize all the features that are interpretable from the B-rep of a solid model regardless whether they intersect or not. Recognized features can be presented as both volume features and surface features. To specify and aid in understanding the type of volume features involved, this work first defines a concept of cavity volume to conform with our notion of cavity and then, a generic volume feature is defined as a cavity volume that has a pre-defined set of topological and geometric characteristics. A user can define a set of topological and geometric characteristics for each specific type of feature interactively with computer graphics. The spatial decomposition and composition method of feature recognition decomposes the space surrounding a solid model into minimal convex cells with the geometric surfaces of the solid model. Combinations of the cells are then composed into volumes and checked to determine if they are volume features. The exhaustive nature of the method enables the recognition of all the features, but also poses the problem of computational explosion. Some initial approaches to cope with computational explosion are introduced in this work. In addition, the on-going research on applying this feature recognition method to process planning is briefly discussed.
58 citations
TL;DR: In this article, the authors present a methodology for analyzing some of the manufacturability aspects of machined parts during the design stage of the product development cycle, so that problems related to machining can be recognized and corrected while the product is being designed.
Abstract: This chapter describes a methodology for analyzing some of the manufacturability aspects of machined parts during the design stage of the product development cycle, so that problems related to machining can be recognized and corrected while the product is being designed. Starting with the CAD design for a proposed part, our basic approach is to systematically generate alternative operation plans for machining the part, evaluating the capabilities of each operation plan to see which one best balances the need for efficient manufacturing against the need for a quality product. We anticipate that the information provided by this analysis will be useful both to provide information to the manufacturing engineer about alternative ways in which the part might be machined, and also to give feedback to the designer about problems that might arise with the machining.
51 citations
TL;DR: A novel method is described to recognize volumetric form features intrinsic to the shape of a given solid using ASVP decomposition, and how the form feature decomposition is converted into corresponding negative components to represent the removal volume for machining applications.
Abstract: A convex decomposition called Alternating Sum of Volumes with Partitioning (ASVP) is a hierarchical volumetric representation which is obtained from the boundary information of the given solid by exploiting convexity. We describe a novel method to recognize volumetric form features intrinsic to the shape of a given solid using ASVP decomposition. We also describe how the form feature decomposition is converted into corresponding negative components to represent the removal volume for machining applications.
45 citations
TL;DR: This chapter describes the automated recognition of form features from boundary representations of solid models using rule-based, graph-based and neural net based techniques.
Abstract: This chapter describes the automated recognition of form features from boundary representations of solid models. The methods covered include rule-based, graph-based and neural net based techniques. Each method is briefly described with specific instances of each referenced and pros and cons listed.
44 citations
TL;DR: The geometrical aspects of an automatic feature recognizer recently developed at the University of Southern California for the domain of parts that can be machined on 3-axis machining centers are discussed.
Abstract: Machining process planners, both humans and automata, reason in terms of features such as holes, slots and pockets. Some of today's CAD systems provide capabilities for defining objects through features. But these are essentially shape macros. They are not guaranteed to be machinable, and often do not correspond directly to machining features. To establish a bridge between CAD and CAM systems, machining features must be automatically recognized. Spatially interacting features are a major source of difficulties in feature recognition. Feature interferences alter the face and edge patterns associated with the features. This complicates both the identification of the features in a part, and the derivation of all the data about the features required for automated manufacturing planning. This paper discusses the geometrical aspects of an automatic feature recognizer recently developed at the University of Southern California for the domain of parts that can be machined on 3-axis machining centers. The recognizer is implemented in an environment consisting of the KnowledgeCraft TM AI shell coupled with either the Parasolid TM or the PADL-2 solid modelers, and running on Sun workstations. It automatically produces feature removal volumes and representations for feature interactions. The recognizer's ability to deal with interacting features is due mainly to an operation called feature completion, described in this paper. Completion is a surface-to-solid transformation. It produces the largest volumetric feature that is machinable and compatible with its traces present in the object's boundary.
42 citations
TL;DR: To support multi-view feature modeling, a combination of a design-by-features approach and feature conversion is proposed, where the system maintains the consistency between the different feature models.
Abstract: Different applications require the use of different feature sets, and hence multiple views and multiple feature models will have to be supported. One such application is assembly planning, for which a short description and two feature classes are given. Further an interactive design-by-feature approach is presented and illustrated with the GeoNode system. To support multi-view feature modeling, a combination of a design-by-features approach and feature conversion is proposed, where the system maintains the consistency between the different feature models.
26 citations
26 citations
TL;DR: In this paper, the authors present a review of computer-aided inspection planning (CAIP) within the context of evolving trends in manufacturing process planning, concurrent engineering and product design models.
Abstract: Planning inspection processes, of parts and products, is evolving as an important module of integrated manufacturing. Manufacturers are increasingly using in-process inspection to control production and achieve the desired quality rather than a means of acceptance or rejection at the end. This requires fast yet accurate inspection as well as effective integration with the product model and relevant data bases. The need for more automated inspection process planning and better decision support tools for the human planners increases as the complexity and variety of products increases and the product development cycle decreases. In this chapter Computer-Aided Inspection Planning (CAIP) is reviewed within the context of evolving trends in manufacturing process planning, concurrent engineering and product design models. A discussion, of the commonly used inspection technology and special inspection requirements, is used to bring to focus the unique aspects of inspection planning and the product model attributes needed to support it. These include tolerances and relationships between inspection features. A review of research efforts to date indicates that this domain is still in its infancy and points out the need for, and potential benefit from, more intelligent inspection planning.
21 citations
TL;DR: In this article, the authors describe techniques for automating the process planning of complex 2.5D components of the type often encountered in the aerospace and avionics industries, based on the ability to extract geometrical data and features from solid models of the components.
Abstract: Techniques aimed at automating the process planning of complex 2.5D components of the type often encountered in the aerospace and avionics industries are described. The work is based on the ability to extract geometrical data and features from solid models of the components. A characteristic of the simplified approach has been to use 2D data and Booleans where possible to interpret planning problems. The work describes algorithms which validate solid models as 2.5 single or double sided components and select suitable workpiece dimensions. Vise, machine clamping and frame bolting workpiece holding methods are examined. Algorithms capable of determining the best clamping approach are presented.
16 citations
TL;DR: XCUT can successfully plan for 80 percent of the features on moderately complex prismatic machined parts and reduce 2 – 4 hours of manual process effort to 15 – 30 minutes, and the automatic recognition and creation of rectangular depression features from the piece part solid model has been addressed.
Abstract: The expert system, XCUT, utilizes Artificial Intelligence (AI) technology for preparing production process plans, automatically selecting cutting tools, and automatically defining machinability parameters and NC tape requirements. XCUT includes the capability to analyze solid model product representations and recognize feature information required by subsequent automated manufacturing processes. A distributed architecture has been implemented where XCUT, the solid model, and the feature extraction interface all run as separate processes and communicate through inter-process mail boxes for integration. XCUT can successfully plan for 80 percent of the features on moderately complex prismatic machined parts and reduce 2 – 4 hours of manual process effort to 15 – 30 minutes. The automatic recognition and creation of rectangular depression features from the piece part solid model has been addressed. Also, XCUT can now automatically classify 80 percent of the feature solid model bodies as a particular instance type in the XCUT feature taxonomy and determines feature attributes (such as, bottom face, depth, width, length). Areas of future work have been identified that will enable the XCUT system to be used for production. These areas where further development is required are: geometric reasoning, representation of manufacturing features, dimensioning and tolerancing, feature extraction, distributed computing architecture, knowledge gathering, and user interfaces.
TL;DR: The interacting and interfeature relationships are shown to significantly facilitate the decision-making processes involved in automated GT coding.
Abstract: Automated Group Technology coding relies on a complete geometric description of the part, along with a description of the high level manufacturing features such as slots, holes and pockets, and the spatial relationships between these features. The research reported herein is focused on a method to perform automatic GT coding that utilizes the geometric model of the part, the feature model of the part, and a formal representation of the relationships among the features of the part. The representation includes a set of structural and geometric primitives, along with a methodology for modeling features in the context of interacting and interfeature relationships. This formalism is referred to as the intermediate geometry representation. The information represented includes interactions between features and geometric relationships among features. The interacting and interfeature relationships are shown to significantly facilitate the decision-making processes involved in automated GT coding.
TL;DR: A novel feature modeling system is described that introduces an incremental feature modeling approach by implementing a hybrid of feature-based design and feature recognition in a single framework that allows changes to a geometric model to be recognized as new or modified features while preserving previously recognized features that remain unchanged in the geometric model.
Abstract: A novel feature modeling system is described that introduces an incremental feature modeling approach by implementing a hybrid of feature-based design and feature recognition in a single framework. During the design process of a part, the user can modify interactively either the solid model or the feature model of the part while the system keeps the other model consistent with the changed one. This gives the user the freedom of choosing the most convenient means for expressing each needed operation. The system is based on an incremental feature recognizer, which allows changes to a geometric model to be recognized as new or modified features while preserving previously recognized features that remain unchanged in the geometric model. Each recognizable feature type is specified by means of feature definition language which facilitates the addition of new feature types into the system.
TL;DR: This chapter gives a brief overview of the Features Testbed, which has an open architecture in which users can define feature libraries for modeling their products and build knowledge bases for a variety of manufacturing applications, and discusses several applications that have been implemented.
Abstract: The A.S.U. Features Testbed provides an infrastructure for rapid prototyping of feature based applications. The system has an open architecture in which users can define feature libraries for modeling their products and build knowledge bases for a variety of manufacturing applications. The Testbed is organized into two shells: one for feature based product definition, the other for feature based product evaluation or manufacturing planning. The product modeler allows one to integrate features, dimensions, tolerances, assembly data, geometry, topology, and design rules into a unified product description. The applications shell facilitates the creation of knowledge bases and reasoning procedures desired for the applications. In this chapter, we give a brief overview of the system and then discuss several applications that have been implemented. These include GT coding for machining, machinability evaluation, and composite panel forming. The Testbed can be used for two different purposes: (1) Organizations that want to evaluate feature based technology in conjunction with knowledge based applications can quickly do so in this integrated system. (2) feature related techniques and algorithms can be evaluated by software developers. Customization of the Testbed does not require any programming or re-compiling because the system is data-driven.
TL;DR: In this paper, a feature relationship graph, derived from the product model, is presented to represent the geometry, dimensions, tolerances and manufacturing process of a component, and the graph supports the link between design and inspection.
Abstract: Computer aided inspection is one small but important component in the move toward a more automated design and manufacturing process. Previously reported research has concentrated on determining workpiece setup, probe selection and probe path planning for coordinate measuring machines. An area that has received little attention to date is the stage of inspection planning that determines which critical surfaces and tolerances of a mechanical component actually require inspection. The work reported in this chapter considers this problem. Evolving computer based systems developed to support engineering design and manufacture are adopting the concept of a Product Model as the central source for providing the data required by various applications in the product life-cycle. In a product model, the geometry, dimensions, tolerances and manufacturing process of a component can be described. The chapter describes a feature relationship graph, derived from the product model, that allows the geometry, dimensions and tolerances and planned processes to be represented consistently. The feature relationship graph supports the link between design and inspection.
TL;DR: Two aspects of parallel machines which impact on process planning are discussed, these are the part domain for parallel machines and the machine configuration.
Abstract: Parallel machines represent a new generation of machine tool. Through reducing the number of setups both the efficiency and the accuracy of the machining process is increased within the part domain. While Flexible Manufacturing Systems (FMSs) and Machining Cells (MCs) are said to be agile, the parallel machine is the first stand-alone machine which can claim to have this property. This makes them ideally suited for machining small batch sizes and for rapid prototyping. Unfortunately like FMSs and MCs these machines will be largely underutilized if agile data generation, processing and transfer mechanisms are not incorporated into CAD/CAM systems. One major hurdle to achieving this objective is the development of an automatic process planning system for parallel machines. This presents new challenges beyond those encountered in process planning for sequential machining. In this paper we discuss two aspects of parallel machines which impact on process planning. These are (1) the part domain for parallel machines and (2) the machine configuration.
TL;DR: A framework that allows a designer to create a design with preset geometric forms or features called C-loops and allows the designer to modify the shapes of these forms as needed is presented.
Abstract: The current article presents a framework that allows a designer to create a design with preset geometric forms or features called C-loops and allows the designer to modify the shapes of these forms as needed. An underlying grammar maps the geometry generated after every modified step to a higher-level set of features. Subsequent to the flexible design-by-features step, a feature recognition step identifies domain-specific features. These features may subsequently used to perform manufacturability analysis of a given design.
TL;DR: A design methodology and computational environment which is based on three sources of design knowledge: compiled heuristics or symbolic knowledge, functional and manufacturing process simulations and the cognitive capabilities of the designer through a user interface is presented.
Abstract: Many of today's “Design for X” methodologies rely heavily on compiled heuristics or symbolic knowledge to guide the iterative design process. It is argued that the solution of real world design problems requires design methodologies which draw on all possible sources of design knowledge, whether it be low-level numerical data obtained from a finite element simulation or high-level human-based cognition. We present in this paper a design methodology and computational environment which is based on three sources of design knowledge: 1) compiled heuristics or symbolic knowledge 2) functional and manufacturing process simulations and 3) the cognitive capabilities of the designer through a user interface. The methodology is tested in the domain of the preliminary design of injection-molded load-carrying three-dimensional components based on functional and manufacturing specifications. Several test case design problems are treated. Based on a hybrid evaluation technique for qualitative rating of preliminary designs, results indicate that the proposed methodology is a viable one and may offer an attractive alternative to other approaches based on more narrow sources of design knowledge.
TL;DR: A part program generator which is a part of a feature-based process planning system to generate automatically reliable part programs from sequenced process plans to provide a fast and relatively reliable test of the correctness of the generated code.
Abstract: This paper describes a part program generator which is a part of a feature-based process planning system The goal of the described system is to generate automatically reliable part programs from sequenced process plans Process plans are generated and sequenced by a feature-based process planning system The part program generator can be tailored by associating parametric part program definitions to primitive processes of the system, work-elements, which are related to classes of features These parametric part program definitions are filtered through a code generation algorithm, which produces segments of part program code with relevant coordinate and nominal values Generated code segments are glued together to form a complete part or pallet program for part family members The part program can be interpreted and the tool path visualized together with fixturing arrangement Visualization provides a fast and relatively reliable test of the correctness of the generated code Instead of correcting or adjusting the generated code the changes should be made into work-element knowledge base, ie, parametric part program definitions, when errors are found in the generated code