Hierarchical Parallel Processes of Genetic Algorithms for Design Optimization of Large-Scale Products

A large-scale machine system often has a general hierarchical structure. For hierarchical structures, optimization is difficult because many local optima almost always arise, however genetic algorithms that have a hierarchical genotype can be applied to treat such problems directly. Relations between the structural components are analyzed and this information is used to partition the hierarchical structure. Partitioning large-scale problems into sub-problems that can be solved using parallel processed GAs increases the efficiency of the optimization search. The optimization of large-scale systems then becomes possible due to information sharing of Pareto optimum solutions for the sub-problems.

www.wapip.com: A Special-Purpose Search Engine for Web Applications in Product Introduction Process

There have been widespread interests in the development and application of World Wide Web (WWW) to support decision-making activities in product design and manufacture. An increasing number of web applications are emerging and a large number of practitioners are keen on trying these web-based decision support systems. In the meanwhile, it becomes increasingly difficult to surf for appropriate web applications on the Internet with general-purpose search engines. This paper describes a web site, WAPIP, that has been developed specifically to support new product introduction activities. It provides databases for software vendors and researchers to register their web applications with the “wapip” search engine. It also provides facilities to support practitioners in product design and manufacture to search rapidly for the right web applications suitable for solving their problems. This paper discusses the various issues regarding the design, development and operation of this “wapip” search engine.

Post Optimal System Analysis Using Aggregated Design Impact Matrix

In this paper the concept of the Aggregated Design Impact Matrix, ADIM, is introduced. This is a tool to calculate and present the relative importance of whole components and subsystems (instead of individual parameters) on different system characteristics. Optimisation is a well-established procedure for system development. Here a non-gradient method is used. Although the sensitivities are not calculated explicitly they can be estimated from the sequence of parameter sets evaluated during the optimisation. The technique used here is recursive least square.

Design and Optimization of Stamping Process to Improve Manufacturing Feasibility

In this paper, we present a numerical procedure combining a finite element inverse approach (I.A.) [10, 14–18] for the simplified analysis of the stamping process with a mathematical programming technique (BFGS method) to optimize some process parameters. Our objective is to optimize the quality of the final workpiece, by minimizing the risk of rupture and wrinkles. The design variables of the present problem are the drawbead restraining forces in relation with the Forming Limit Diagram (FLD). The optimization procedure associated to the analytical sensitivities analysis technique based on the adjoint method is applied for the square cup of Numisheet’93 and the Twingo dashpot cup proposed by RENAULT [32]. The satisfactory results demonstrate the usefulness of this automatic optimization procedure in the preliminary design of deep-drawing process.

Automatic Determination of Tolerance Chains in Unidirectional Tolerancing

Herein is proposed an automatic method by which the all-tolerance chains of a mechanical assembly may be constructed. The method, implementable in a CAD system, is divided into two main steps. In the first step, we model the mechanical assembly with a graph, of which the vertices represent the parts and the edges, the contact between the parts. By « sliding » the parts, we may determine all the configurations of the mechanical assembly. The proposed model, which uses classical algorithms of the graph theory, allows control of the coherence of the mechanical assembly. In the second step, we introduce an extended syntax by which the functional constraints may be decoded without ambiguity. Then, using the syntax and the model, we show how to construct the tolerance chain for each functional constraint.

Determination of Optimum Sensor Positions for Defect Detection on Revolving Machines by a Finite Element Vibration Analysis

The vibration monitoring is largely used to detect the defects in the revolving machines. The determination of the best sensor positions is one of main research goals in the domain of the conditional maintenance. This paper proposes a numerical methodology based on a finite element model and a spectral analysis in order to find optimum sensor positions. The bearing is considered as a key component in vibration propagation from the moving parts to static ones. In this paper, we use an analytical bearing model and its numerical implementation in a FE code. The tangent stiffness matrix of the bearing element is calculated by the Newton-Raphson method and then introduced into the modal and spectral analysis. The technique of “Mode Shape Summation Plot” (MSSP) is adopted to find the most sensitive zones to usual defects. The proposed numerical approach gives good agreements with the experimental results. A real grinder modeling shows its interesting industrial applications.

Intelligent Optimal Design of a Beam During the Crash

We consider the optimal design of a beam. To improve the safety of a car during the crash, it is needed to dissipate the maximum of energy within a limited displacement but with a limited acceleration at the level of the driver/passengers. The beam may have different complex cells linked with continuous or spot solders. In a special office, they have to design a beam. In Germany, in United Kingdom, and also inside all the automotive industries, dedicated centers to test such beams were created. Other centers have developped numerical simulations. Usually, for each family of beams, a Design of Experiments is performed. It is necessary to find its optimal design. Until now, this was not possible, as the CRASH is a very complex problem and the tests or numerical simulations are too expensive to allow the succssive iterations.

“Intelligent” Optimal Design

The engineers have to face very important problems in the design, the test, the survey and the maintenance of their structures. These problems did not yet get full answer even from the best people in the world. Usually in these problems (such as no satisfactory constitutive modeling of materials, no real control of the accuracy of the numerical simulations, no real definition of the initial state and/or the effective loading of the structure), there is no solution and the experts do not understand the problem in its whole. Moreover, the available data may be not statistically representative (i.e. are in limited number), fuzzy, qualitative and missing in part. We propose a practical solution the «Intelligent Optimal Design of Materials and Structures» where the actual best knowledges of the researchers/experts are intelligently mixed to the results of experiments or real returns.

Implementation and Analysis of a Mechanics Simulation Module for Use in a Conceptual Design System

Previously, we have described the theory of a general mechanics model for non-rigid solids (Jansson, Vergeest, 2000). In this paper, we will describe and analyze the implementation, i.e. algorithms and analysis of their time complexity. We will reason that a good (better than O(n2), where n is the number of elements in the system) time complexity is mandatory for a scalable real time simulation system. We will show that, in simplified form, all our algorithms are O(n lg n). We have not been able to formally analyze the algorithms in non-simplified form, we will however informally discuss the expected performance. The entire system will be empirically shown to perform slightly worse than O(n lg n), for a specific range and typical input. We will also present a working prototype implementation and show it can be used for real time evaluation of reasonably complex systems. Finally we will reason about how such a system can be used in the conceptual design community as a simulation of traditional design tools.

Decision-Based Collaborative Optimization Under Uncertainty

In this research a Collaborative Optimization (CO) approach for multidisciplinary systems design is used to develop a decision based design framework for non-deterministic optimization. To date CO strategies have been developed for use in application to deterministic systems design problems. In this research the decision based design (DBD) framework proposed by Hazelrigg (1996a, 1998) is modified for use in a collaborative optimization framework. The Hazelrigg framework as originally proposed provides a single level optimization strategy that combines engineering decisions with business decisions in a single level optimization. By transforming the Hazelrigg framework for use in collaborative optimization one can decompose the business and engineering decision making processes. In the new multilevel framework of Decision Based Collaborative Optimization (DBCO) the business decisions are made at the system level. These business decisions result in a set of engineering performance targets that disciplinary engineering design teams seek to satisfy as part of subspace optimizations. The Decision Based Collaborative Optimization framework more accurately models the existing relationship between business and engineering in multidisciplinary systems design.