| dc.description.abstract | Today, in the context of a globalized market, customers have high demands for products that are tailored to their individual needs and are offered at a price that is very close to that of mass-produced products. Engineering-To-Order (ETO) companies are forced to reduce costs and lead time to gain an advantage over the competition. These companies encounter two major issues that greatly affect the cost and the quality of their products. The first issue is the configuration complexity of ETO products. Many ETO companies are employing Knowledge Based Engineering (KBE) systems to manage the configuration complexity. These systems can be used to effectively capture knowledge by storing technical guidelines, "best practices", and even a company’s commercial and business rules. When it comes to complicated product configurations, the use of KBE tools is indeed an efficient solution automating configuration specification. However, ETO companies very often are confronted with "first-time" product configuration requirements. Since previous experience design rules are not adequate to cover the new configuration requirements, these companies are usually proceeding with experimental tests using full scale prototypes to check the structural integrity of the proposed design, spending time and raw materials. Also, since these tests must be performed, usually, in very tight lead times required by the customer, the design/engineering team has very limited time to achieve optimized material usage, reduced weight, etc. Thus, usually they end up with over-engineered solutions. ETO companies could gain significant benefits and achieve significant cost reduction if they could perform simulated experiments using Finite Elements Analysis models instead of using full scale prototypes. A major concern about using FE simulated experiments is the validation of the FE models in terms of their accuracy. FE model validation is even more complicated to be achieved for dynamic phenomena simulations.
The second issue is the time and the cost required for the product to be designed and engineered, and for manufacturing drawings to be published and launched to the shop floor. In most cases, these companies have a number of premade 3D models (and the corresponding manufacturing drawings) and modify them to adjust the dimensions, the function and/or the aesthetics of the product to the customer requirements. Unfortunately, this method is prone to human errors, and these errors may create extra remanufacturing costs. An ETO company would gain a significant advantage by using a tool that would create automatically 3D assembly models for its products.
In this thesis a framework that addresses both of these issues is presented. The present framework provides: a) usage of FE models and simulated dynamic experiments to deduce new design rules, instead of performing experiments with full scale prototypes, b) a methodology for
validating these finite element models for their accuracy in simulating dynamic experiments and c) a Mechanical-Design methodology based on the Automatic Assembly Synthesis Model (AASM), that links a KBE and a CAD system, and automatically generates and synthesizes the final 3D assembly model. | en_US |
