Comparison of Non-Nominal Geometry Models Represented in Physical Versus Virtual Environments

In the automotive industry today, virtual geometry verification activities are conducted with nominal models in the early design phases. Later in the design process when the first physical test series are made, are concepts verified in a non-nominal manner. Errors detected at this stage can result in expensive post-conceptual changes. By combining Computer Aided Tolerance (CAT) simulation tools with Virtual Reality (VR) tools, virtual environments for non-nominal geometry verification can be utilized. This paper presents the results from a study, conducted at Volvo Cars, that investigates the perceptional aspects that are related to verification of quality appearance, using non-nominal virtual models. Although a realistic non-nominal model is created, the interpretation, i.e. how the model is perceived, must be clarified. This would represent a validation of the model from a perceptional point of view. Since the effect of geometric variation is a specific application, with high demands on realistic and detailed representation, perceptional studies are needed to ensure that VR and other virtual representations can be used for this kind of application. The question is whether it is possible to evaluate aspects like flush, gap and see-through in virtual environments. In this paper, two environments are compared, one physical and one corresponding virtual environment. Three adjusted physical vehicles are mapped to the virtual environment and compared using non-immersive desktop VR in a visualization clinic with test subjects from the automotive industry. The study indicates that virtual objects are judged as less good looking compared with physical objects. There is also a higher degree of uncertainness when judging virtual objects.

Application of Finite Volume Method for Solid Mechanics

Polymers constitute a large class of nearly incompressible solid materials (i.e., Poisson’s Ratio near 0.5). These materials are often used as passive vibration isolators. Accurately modeling vibration isolators made of nearly incompressible materials has been extremely difficult with standard finite element analysis. This paper provides an alternative to the specialized finite element formulations currently used to model incompressible materials. The finite volume methodology of computational fluid dynamics is employed in this paper to solve the Hooke’s Law equations in solid mechanics. Test cases have been performed to evaluate the performance of finite volume method applied to solid mechanics problems. The formulation has been coded in Matlab for practical use. Based on the preliminary test case results, the finite volume formulation compares favorably to finite element method.

CORBA-Based Task Requirement Driven Design System

Today’s design environment has become more distributed and professional. Efficient design management can greatly improve the ability of a company’s competition. To increase efficiency of a design process coordination of Computer-Aided Design and Analysis tools are very important, especially for large complicated systems. In this paper, we introduce the concept of a requirement driven system. Design process usually involves fulfillment of requirements from top-end customer. Adopting requirement driven mechanisms will provide more convenience for design coordination automation and help us find the most resource saving solutions for specific product design. A CORBA framework is discussed to facilitate the implementation of methodologies for requirement driven design coordination. System architecture and modules for the framework are introduced to support a requirement publishing and responding service. Distribution of the tasks is determined by “stigmergy” algorithm, which makes the decision using the performance history of each team and designers. An example of a coffeemaker product design based on the framework, is presented to demonstrate the application of new design system.

Three-Dimensional Finite Element Analysis of Melting in an Alloy Irradiated With a High Energy Laser Beam

Lasers are being widely used in the material processing industry lately. In this work, study is performed on the resluting temperature and stress distribution, the width and depth of the melt pool in the heat-affected zone during material processing with a high energy Gussian laser beam. A three-dimensional enthalphy-based mathematical model is developed to study the effect of high energy laser beam on the formation of molten pool. The mathematical model is based on a three-dimensional transient heat equation taking into consideration the power intensity of the Gussian laser beam and phase diagram of the material. A computational algorithm is implemented to evaluate the temperature distributions as well as shape and size of molten pool. The numerical solution with the applied heat flux and convective boundary conditions is obtained usign a 3-D finite element code.

Modeling Time-Dependent Systems Using Multi-Stage Bayesian Surrogates

Multi-Stage Bayesian Surrogate Models (MBSM) are meta-models, constructed using data obtained from different sources, which have the ability to integrate information and responses with different levels of accuracy. In applications of surrogate models for time-dependent systems, the data obtained from physical or computational experiments is usually a sequence of response values over time, measured for different combinations of design parameters. For such applications, the traditional MBSM approach is impractical to incorporate all the observed data in a single model of the system, mainly due to the prohibitive computational effort involved. In this paper, we propose a framework for building surrogate models for time-dependent systems, based on the cokriging technique. The proposed framework regards the observations as a set of time-correlated spatial processes, with a stationary, separable cross-covariance structure of known functional form. Results show that for time-dependent systems, the proposed methodology outperforms joint space-time models built with the traditional MBSM approach both in terms of accuracy and computational effort.

Inverse Solution of a Heat Conduction Problem Using Evolutionary Data Segregation Techniques

Engineering problems are typically solved by direct solution. For the direct solution of engineering problems the boundary conditions and physical properties of the domain are given, and the dependent variable is calculated throughout the domain. In contrast to this, for inverse engineering problems the dependent variable is known at select locations in the domain, and the material properties and/or the boundary conditions need to be determined. This paper will present a novel technique for the inverse solution of a heat transfer engineering design problem in which the temperature profile and materials are known, but the placement of these materials and the heat flux on the boundaries are unknown. This technique uses evolutionary optimization in the form of the Adaptive Modeling by Evolving Blocks Algorithm (AMoEBA) to determine the material configurations. The material configurations, geometry, and properties are defined by evolving binary trees. The evolved domains are solved directly and then compared with the known temperature profile. Fitness of the new designs is determined by the least squared error between the proposed and the known profile. When this fitness reaches a defined level, the material placement scheme of the real system is found, and boundary conditions matching the problem definition are identified.

GL-Link: A Novel Telerobotics-Based Platform Supporting Distributed Mechatronic Research Via the Internet

Mechatronics and robotics research efforts of large complexity are increasingly interdisciplinary involving collaboration between software, hardware, controls, and scientific teams. Traditionally, the level of integration has either required repeated site-visits or location of the teams at a common site. As the teams become increasingly diverse and disperse, there is a need for distributed operations platform that not only facilitates smooth communications, but also allows for remote experimentation and control of a common robot or device. By separating the principal design functions, a modular communications platform was developed to support the distance learning and experimental requirements of ambitious mechatronic development projects. This separation results in a modular system that is scalable and customizable to the particular conditions governing an experiment. The platform leverages off-the-shelf hardware and software and the presence of Internet connectivity. Where possible, open-source options were used to make the platform extensible to a variety of platforms and applications. The system is modular and consists of: a video observation/conferencing module, a file-transfer module, and a robot teleoperation module. This allowed multiple teams to test the operation of a robot independently and asynchronously without corrupting the work being conducted by another team member. It also allowed for new forms of interaction and reduced the need for travel between the multiple geographically-distributed research teams. Novel features of this work include a modular multiplatform architecture and an integration of basic telerobotics principles to extend PC-based collaboration/conferencing technologies from a basic communications platform to a means for supporting multi-site (robotics) research experiments. This paper describes the design considerations and evaluations associated with the development of the Great Little Inter link (GL-Link) architecture. This platform was motivated by robotics research ongoing between Stanford and Ohio State Universities. The platform was tested over several months as part of the design of a high-speed quadruped robot. Results from this trial highlight the impact of highly sensitive audio and video inputs and show the need for robustness to bandwidth fluctuations.

Generation of Rough Design Plans for Mechanical Products Using a Module Library

In this paper, a preliminary design support system for mechanical products is proposed. The preliminary design stage, in which rough design plans are generated, can be supported by a mechanism for automatic translation from functional requirements to modules and by optimization of arrangements in a module combination set. A module library is proposed as a mechanism for automatic translation. In this library, relationships among functional requirements, primitive functions and modules are represented in a hierarchical class description. The module combination sets that satisfy the functional requirements are generated from selected modules by referencing the module library. A genetic algorithm is applied to optimization of module arrangements in each combination set. Several module arrangements which satisfy functional requirements are obtained. These arrangements offer a rough design plan for the product. A trial implementation of the above procedure is applied to “Lego Mindstorms” a building block kit for toy robots.

Virtual DesignWorks: Designing 3D CAD Models Via Touch Interaction

Virtual reality (VR) devices, such as haptic (force feedback) devices, provide users with virtual environments where they can interact with digital models in 3D. Haptic devices show great promise for use in design. However, current haptic systems are used primarily to verify rather than to interact with CAD systems to modify a design. In order to use haptic devices in the design of CAD models, we use component technology (COM+) to develop a novel haptic model system—which we have termed super-object modeling. The novel haptic model solves several crucial problems for seamlessly integrating CAD models with haptic models, such as the efficient exchange of complex models in real-time, real-time model updating, and the issue related to CPU power etc. Based on this super-object modeling, we have developed a virtual reality system, called Virtual DesignWorks. The system provides a new approach to 3D design—Designing CAD models via touch interaction. Engineers or designers can, in virtual space, directly touch native B-rep CAD models, feel and deform the CAD models rather than just verify a design. Super-object modeling represents the first development of haptic geometric models based on component technology. It demonstrates significant advantages compared to the traditional haptic models. With this technology, touch interaction has the potential to become a critical interface of design, and force feedback gives designers the greatest flexibility for 3D design.

Evolutionary Optimization Using Graph Based Evolutionary Algorithms

Graph based evolutionary algorithms (GBEAs) are a novel evolutionary optimization technique that utilize population graphing to impose a topology or geography on the evolving solution set. In many cases in nature, the ability of a particular member of a population to mate and reproduce is limited. The factors creating these limits vary widely and include geographical distance, mating rituals, and others. The effect of these factors is to limit the mating pool, reducing the rate of spread of genetic characteristics, and increased diversity within the population. GBEAs mimic these factors resulting in increased diversity within the solution population. When properly tuned to the problem and the size of the population set, GBEAs can result in improved convergence times and a more diverse number of viable solutions. This paper examines the impact of the fitness landscape, population size, and choice of graph on the evolutionary process. In general, it was found that there was an optimal population size and graph combination for each problem. This optimal graph/population was problem dependent.