Computing the Information Content of Decoupled Designs

The information content of uncoupled designs can be computed by summing the information content associated with each functional requirement. This paper proves that information cannot be summed for decoupled designs. To overcome this problem, this paper presents two algorithms for computing information content of decoupled designs. One algorithm is applicable to any joint probability density function for the design parameters; the second algorithm applies only to uniformly distributed design parameters. The algorithm for uniform distributions is based on a recursive procedure for computing the volume of a convex polytope in n -dimensional real space, where n is the number of design parameters. An engineering application of the algorithms is presented. The example demonstrates that summing information content can significantly over-estimate total information when compared to an algorithm that accounts for correlation. The example also demonstrates that decoupled designs can have lower information content than uncoupled systems with the same functional requirements and similar components.     

Structure and Matrix Models for Tolerance Analysis from Configuration to Detail Design

A substantial amount of all quality problems that arise during assembly can be referred back to the geometrical design, and especially the geometrical concept of the product, i.e. the way in which parts are designed and located with each other. Special emphasis should thus be put on geometry design, especially during the early design phases, to try to find robust concepts and avoid solutions that may cause down-stream production problems.  This paper presents a generic set of evaluation tools for robust geometry design encountering (i) potential tolerance chain detection in configuration design, (ii) assembly robustness evaluation in concept design, and (iii) tolerance sensitivity analysis in detail design. Special attention is given to the development of a new matrix-based evaluation tool for the configuration design part. The tool presented is based on a new way of representing geometry variation constraints in an enhanced function-means tree structure model. Different parts of the function-means tree that are of interest for analysis purposes are then extracted and converted to matrix representation. The reason for doing this is that the structure model is most suitable for modeling, but becomes unsuitable for analysis as the model complexity increases. For this latter purpose, the matrix representation is far better. The use of the different tools is demonstrated in the design of a new vehicle front system for which the geometry a priori is unknown.     

Tuning Parameter Tolerance Design: Foundations, Methods, and Measures

In this paper, a novel technique is presented to solve tolerance design problems. To achieve the desired performance tolerance, the technique uses a subtle, but significant, change in the design: the addition of a tuning parameter in place of an increase in component precision. Statistical models are used to develop a framework for the tuning parameter design method. Also developed is a new, dimensionless design metric which ranks candidate tuning parameters. A step-by-step method is developed for the application of tuning parameters using this metric. The step-by-step tuning parameter design method is applied to a heavy-duty manual stapler as a clarifying example.     

Numerical Modeling of a Post & Dome Snap-Fit Feature

Snap-fits are often designed using guides that rely on classic beam theory, with the basic assumption that the beam undergoes small rotations and displacements. This is a poor assumption, for they typically experience both large rotations and displacements due to loading offset from the neutral axis and axial loading. This paper investigates the performance of the post & dome feature, establishes its nomenclature, and derives the equations needed to intelligently design different variations of it. The post & dome feature was selected for analysis because it is a high performance snap-fit that is self-datuming and can withstand some shear loading in addition to retention. The design equations were generated in three steps. First, an experimental array was created using a design of experiments approach. Finite element methods and multiple regression techniques were used in lieu of beam equations models for each of the trials in the experimental array. Finally, response surface methods were used to develop response curves based on the performance data generated by the finite element models. Sensitivity data was plotted for both the main effects and selected variable interactions. The traditional benchmarks for defining high performance snap-fits are retention strength, insertion force, and insertion strain. This paper uses an expanded definition of these benchmarks that also includes locking ratio (the ratio of retention force to insertion force).     

Functional Interdependence and Product Similarity Based on Customer Needs

In this paper, related product functions are determined for a group of approximately 70 consumer products. Using customer need data, a new matrix approach is introduced to identify these relationships. Techniques are then created for determining product similarity. These techniques are clarified and validated through three case studies, including beverage brewers and material-removal products. The results of these case studies are argued to have significant impact on design-by-analogy procedures, benchmarking methods, mass customization strategies and modular design. The paper concludes with a discussion of applications and related procedures for product development.     

Arrow's Theorem and Engineering Design Decision Making

This article establishes that Arrow”s General Possibility Theorem has only indirect application to engineering design. Arrow”s Theorem states that there can be no consistent, equitable method for social choice. Many engineering design decisions are based on the aggregation of preferences. The foundation of many engineering decision methods is the explicit comparison of degrees of preference, a comparison that is not available in the social choice problem. This explicit comparison of preference levels is coupled with the choice of an aggregation method, and some forms of aggregation may be inadequate or inappropriate in engineering design.     

‘Signposting’, A Parameter-driven Task-based Model of the Design Process

This paper addresses the need for computer support in aerospace design. A review of current design methodologies and computer support tools is presented, and the need for further support is discussed, with particular reference to the early formative stages of the design process. A parameter-based model of design is proposed, founded on the assumption that a design process can be constructed from a predefined set of tasks. This is supported by knowledge of possible tasks in which the confidence in key design parameters is used as a basis for identifying, or signposting, the next task. A prototype implementation of the signposting model, for use in the design of helicopter rotor blades, is described and results from trials of the tool are presented. Further areas of research are discussed.     

A Dominance-based Design Metric in Multiattribute Robust Design

This paper outlines the development of an effective and consistent ‘designing-in-quality’ strategy that can be used to deal with concepts of uncertainty, quality and robustness in engineering design. Specifically, this paper presents a decision analysis-based robust design metric that seamlessly integrates objective evaluations on the goodness of a design alternative with the designer’s intent and preferences. This is achieved through the development of a set of performance-reflecting dominance indices for the attributes and their utilization in a preference-influenced multiattribute utility formulation. Such a knowledge feedback-based decision model development will be particularly useful when dealing with complex iteration-based engineering design process where little information on the expected outcomes may be known a priori, or where product performance is computationally expensive to evaluate. Application of this robust design metric in a multi-stage experimentation and modeling design process is presented. The characteristics of the proposed design metric and the effectiveness of the overall design procedure in dealing with constrained engineering design problems are examined with the aid of demonstrative case studies and the results are discussed.     

Framework for Modeling Dependencies in Collaborative Engineering Processes

This article describes an engineering process representation and modeling tool. The approach is especially suitable for describing large-scale, mature design processes involving numerous tasks, some of which may be performed by automated computer agents. The underlying representation is a graph of information-processing units with explicitly defined input and output feature elements. We show that this representation is more complete than those used in previous process modeling approaches and overcomes some of their limitations when dealing with design processes involving dependencies at multiple levels of detail. The representation is combined with rules for automatically operating upon the graph to preserve consistency when traversing to higher or lower levels of detail.     

Evaluation Metrics for the Rating and Optimization of Snap-fits

Current snap-fit design guides recommend sizing snap-fit features on the basis of insertion force and allowable strain during assembly. Retention force information in such guides is often inaccurate, although this is considered to be the primary attribute of the snap-fit after assembly. The authors contend that these (insertion force, allowable strain, retention force) are not the only critical performance criteria for snap-fit features. Designers have to contend with several other constraints and design requirements. Additional performance metrics for snap-fit features are proposed by drawing upon considerable experience with plastic part design issues. Locking ratio, dimensional and volumetric retention force, consideration of the characteristic dimension of the joint and snap-fit, feature stiffness, required over-insertion and consideration of snap-fit strength relative to part strength are proposed to supplement currently used metrics for evaluating and rating snap-fit designs. The applicability of these metrics is illustrated with real-life examples, and their merits and demerits discussed. A chart of achievable locking ratios for different snap-fit topologies is presented for use as a design tool for the initial selection of snap-fit topologies. Its use as a rational basis for selection and optimization of snap-fits is suggested. Adoption of proposed metrics will allow designers to better quantify, and thereby optimize the performance of, snap-fit features. These ideas will be built upon in the future, and used as a basis for a comprehensive snap-fit selection and detailed design tool.     

A Theory of Complexity, Periodicity and the Design Axioms

One of the topics that has received the attention of mathematicians, scientists and engineers is the notion of complexity. The subject is still being debated, as it lacks a common definition of complexity, concrete theories that can predict complex phenomena, and the mathematical tools that can deal with problems involving complexity. In axiomatic design, complexity is defined only when specific functional requirements or the exact nature of the query are defined. Complexity is defined as a measure of uncertainty in achieving a set of specific functions or functional requirements. Complexity is related to information, which is defined in terms of the probability of success of achieving the Functional Requirements (FRs). There are two classes of complexity: time-dependent complexity and time-independent complexity. There are two orthogonal components of time-independent complexity, i.e., real complexity and imaginary complexity. The vector sum is called absolute complexity. Real complexity of coupled design is larger than that of uncoupled or decoupled designs. Imaginary complexity can be reduced when the design matrix is known. As an example of time-independent imaginary complexity, the design of a printing machine based on xerography is discussed. There are two kinds of time-dependent real complexity: time-dependent combinatorial complexity and time-dependent periodic complexity. Using a robot-scheduling problem as an example, it is shown that a coupled design with a combinatorial complexity can be reduced to a decoupled design with periodic complexity. The introduction of periodicity simplifies the design by making it deterministic, which requires much less information. Whenever a combinatorial complexity is converted to a periodic complexity, complexity and uncertainty is reduced and design simplified.     

Improving Systems by Combining Axiomatic Design, Quality Control Tools and Designed Experiments

This paper presents an approach for solving design problems in existing designs. A design analysis with Axiomatic Design, called Design Object Analysis, describes a product or a system in terms of Customer Needs (CNs), Functional Requirements (FRs), Design Parameters (DPs) and Process Variables (PVs), as well as their associated Design Matrices (DMs). In this paper, the design analysis is combined with a thorough investigation of possible problems within the design, utilizing the seven quality tools, noise factor analysis, and designed experiments to form an approach for quality improvements and problem solving. The Design Object Analysis helps secure valid input-factors to the designed experiments, and the designed experiments correct or improve the assumptions made in the Design Object Analysis. Thus, a combination of product modeling by Axiomatic Design and designed experiments overcomes shortcomings of the two methods. The benefits of performing a Design Object Analysis, as compared to other methods, become clear when it comes to evaluating the results from the designed experiment, and preventing the problem. Once the critical parameters are confirmed, and the design matrices are updated, suggested design improvements can then be checked against the design matrices, and the system effect of a design-change-order can be estimated. The approach described in this paper was successfully applied and verified in a case study at a large automotive company.     

Building Agility for Developing Agile Design Information Systems

Agile manufacturing relies heavily on the quality of information that organizations have and on their ability to organize and reuse it. Constant inflow of information and knowledge is the fuel of agile manufacturing. In the process of forming virtual enterprises, these new organizations have to be equipped with information systems that integrate their present legacy technology and improve upon it. To support the quick formation of virtual organizations, one must have the ability to develop such systems quickly. Over the past few years we have evolved, through collaborative projects with industry, an approach composed of methods and an information infrastructure called n-dim that improves the ability of becoming agile manufacturers of information systems, by responding quickly to information needs of new and evolving organizations. Following an analysis of the requirements of information systems for agile design, we discuss this approach; describe some of the infrastructure features; and present several examples of simple applications that illustrate them. We summarize by discussing the advantages and limitations of our approach.     

On extreme values of orbit lengths in M/G/1 queues with constant retrial rate

In the design of waiting facilities for the units in a retrial queue, it is of interest to know probability distributions of extreme values of the orbit length. The purpose of this paper is to investigate the asymptotic behavior of the maximum orbit length in the queue with constant retrial rate, as the time interval increases. From the classical extreme value theory, we observe that, under standard linear normalizations, the maximum orbit length up to the nth time the positive recurrent queue becomes empty does not have a limit distribution. However, by allowing the parameters to vary with n, we prove the convergence of maximum orbit lengths to three possible limit distributions when the traffic intensity approaches 1 from below and n approaches infinity.

---

---

Received: October 7, 1999 / Accepted: November 21, 2000     

Analysis of a single stage production system with heterogeneous machines

The performance of a single stage production system with two heterogeneous machines and two classes of jobs is investigated. The machines have a common buffer with jobs of both classes waiting for service. The arrivals are assumed to follow a Poisson process and the service times to be distributed exponentially. The evaluated production system differs from a classical homogeneous multiple server queueing system with regard to inhomogeneities of the two machines. Time inhomogeneity – the service times of the two machines being unequal – and functional inhomogeneity – one of the machines can handle only one class (A) of jobs – are to be distinguished. In the case of time inhomogeneity the calculation of system performance parameters may be carried out using an explicit formula, whereas for the analysis of functional inhomogeneity a numerical solution has to be derived. The impact of time inhomogeneity is very small and decreases with the system workload. On the contrary, functional inhomogeneity leads to elevated cycle times of up to 40% depending on the degree of inhomogeneity (measured by the fraction of A jobs) and the workload. Therefore, in contrast to the time-inhomogeneous case, single stage production systems with functional inhomogeneity can only be approximated tolerably by a homogeneous multiple server queueing system if the fraction of one-machine jobs is less than 30%. The increased throughput times above 30% are supplied by the diagram developed from the numerical solutions. RID="*" ID="*" While this research was constructed the author was affiliated to Institute of Conveying Technology and Logistics. Correspondence to: C. R. Lippolt     

Evaluation of Product and Process Design Robustness

Critical design decisions are commonly made throughout the product development process assuming known material and process behavior. However, stochastic variation during manufacture can inadvertently result in inferior or unacceptable product performance and reduced production yields. Stochastic simulations have been developed to estimate the end-use performance distribution prior to the commitment of hard tooling. This article proposes a definition for integrated product and process robustness, and extends existing stochastic methods to model the important role of the manufacturing flexibility in elimination of defects and product optimization. The goal is to enable the designer to understand and account for not only the negative effects of manufacturing variation, but also the positive impact of manufacturing flexibility wherein instantaneous corrections in the manufacturing process can frequently improve the product quality and eliminate flaws in the product design. Then, a methodology is introduced and contrasted with conventional development methods in the evaluation of best practices for development of a molded plastic component.     

On the Influence of Functional Materials on Engineering Design

To obtain a better understanding of how to make use of functional materials in engineering design, the design-related behaviors of some of these materials have been analyzed and discussed with reference to a general design procedure model and the design tools currently available. The functional materials subjected to this examination are shape memory alloys, piezoelectric materials and magnetostrictive materials. The discussion has been carried out with a major focus on how benefits might be gained by selecting the functional material and utilizing complementary design tools at an early stage of the design process.     

Observations on Some German Contributions to Engineering Design In Memory of Professor Wolfgang Beitz

This paper, written in memory of Professor Wolfgang Beitz, discusses some of the influences of the work undertaken in Germany on systematic engineering design. It highlights differences between the language regions, and gives examples of design research and design education linked to Konstruktionslehre– the standard text on systematic engineering design for which Professor Beitz was most widely recognised outside Germany. The paper finishes with a plea for a greater exchange of ideas.     

Design Support Using Distributed Web-Based AI Tools

Currently, designers are faced with searching through a ‘sea’ of on-line knowledge to support their decision making activities. This paper describes WebCADET, which is a reimplementation of the stand-alone CADET – a Knowledge-Based System (KBS) for product design evaluation. WebCADET aims to provide effective and efficient support for designers during their searches for design knowledge. WebCADET uses the ‘AI as text’ approach, where KBSs can be seen as a medium to facilitate the communication of design knowledge between designers. The development of WebCADET to include practical support via World Wide Web-based functionality, which illustrates the potential of the ‘AI as text’ approach, is described in the paper.     

An application of mixed-integer linear programming models on the redesign of the supply network of Nutricia Dairy &; Drinks Group in Hungary

Between 1995 and 1998 Nutricia acquired a number of dairy companies in Hungary. Each of these companies produced a wide variety of products for its regional market. Although alterations had been made to the production system in the last few years, production and transportation costs were still substantial. This paper presents a research study with regard to the optimisation of the supply network of Nutricia Hungary using a mixed-integer linear programming model. Focussing on consolidation and product specialisation of plants the objective was to find the optimal number of plants, their locations and the allocation of the product portfolio to these plants, when minimizing the sum of production and transportation costs. The model is in line with traditional location/allocation models, with a modification concerning inter-transportation of semi-finished products between plants. The production costs used in this model are based on a Green field situation, taking into account new and more advanced technologies available today. The model is used by the Nutricia Dairy and Drinks Group as a decision supporting tool. Correspondence to: F. H. E. Wouda