Fundamentals of a Methodology for Predictive Design Analysis

Abstract: The rapid development of computer-based design analysis tools and methods such as the finite element method (FEM) within computational structural mechanics (CSM), computational fluid dynamics (CFD), and multi-body systems (MBS) during recent decades has fundamentally changed the way in which products are designed and developed. Among the many advantages observed in industrial practice, one can mention improved understanding of product properties and behavior, the possibility to optimize critical design parameters and to optimize parts as well as the entire product at a system level, the possibility to explore a design space during the synthesis activity, and a decrease of the need for physical testing and thus of the number of physical prototypes needed. However, all of these opportunities to improve the performance and durability of the product-to-be come with challenges. In order to utilize these tools as efficiently and effectively as possible it is necessary to facilitate their integration into the engineering design process, and in product development and product innovation as a whole. Where computer-based design analysis replaces more traditional evaluation, validation and prediction methods, there is also a demand for greater confidence in the design analysis process and in the results obtained. The objective set out for the research project presented in this thesis is to outline the fundamentals of a methodology for predictive design analysis (PDA), a computer based design analysis methodology allowing for increased confidence in the predictions resulting from the design analysis activities regarding the critical design parameters and their influence on the behavior of the product-to-be (artifact) throughout the entire design and development of the artifact. The methodology is articulated around the generic design analysis (GDA) process model and how the activities within the phases of this process are influenced by factors emanating from the environment in which the design analysis task is originated and executed. Furthermore, a number of confidence appraisal activities (CAAs) are established to ascertain the confidence in the predictions made for the design analysis task during the analysis process. These activities provide for an increase in confidence in the design analysis process as well as in the results obtained as input to the subsequent engineering activities within the development project. The fundamentals in terms of constituent parts of a methodology for PDA, introduced in this thesis, are developed at a level of concretization that makes them directly applicable in an industrial setting.