Realization of conceptual design for mechanical products
Advances in scientific technologies in the past few decades have brought forth evolutionary changes to conventional engineering design processes. One significant phenomenon resulting from this transformation is the drastic reduction in the time used for engineering design. Regarded as a major factor in incurring extra expenses for manufacturing, the lead time for designing a product has been cut dramatically since the introduction of computer-aided design tools. Despite being subject to such improving endeavor, design lead time still remains the biggest factor in escalating product costs.
Numerous projects were initiated to investigate and promote the design technology which would expedite the generation of an optimized product prototype. Yet with all the advancement of modern design sciences the technology by far still can not relieve designers from the constant involvement in the decision-making processes for a product's logistic resolutions. Therefore design lead time still constitutes a great percentage of the throughput time for contemporary manufacturing processes. In light of such a problem a number of commercial CAD systems provide mechanisms to alleviate designers from those repetitive and redundant decision-making processes. But oftentimes component selection which is based on previous design cases can not meet the specific nature of product functionality requirements. Therefore further human intervention is needed for the modification of design prototypes which were rendered under such circumstances. Furthermore because product geometry plays an important role in determining the marketability and the manufacturing costs of the product conceived, a substantial amount of resources and energies are always allocated to perfect a product's geometry. The ease to manipulate product geometry, therefore, becomes a crucial factor in the shortening of the design cycle in addition to the one for component selection.
A design approach which stems from the reasoning of a product's functional specification and then produces design prototypes through an optimizing vehicle is presented here together with an enhanced method for manipulating the geometry of three-dimensional CAD models. To speed up the natural selection process for various product components a parametric evaluation process and a 0-1 integer programming scheme are introduced in the component selection mechanism in this modified design approach. The association between related components is preserved by using the Bond graph method which depicts the physical characteristics of individual components and their interrelationships. To help the manipulation of three-dimensional product models the Newton-Raphson method is combined with the Doolittle's method (a matrix reduction method) to provide a mechanism in facilitating the solving for new product geometry.