Comparative Evaluation of Engineering Design Concepts Based on Non-Linear Substructuring Analysis

Abstract

The paper presents a novel approach to comparative evaluation of engineering design concepts that exhibit non-linear structural behaviour under load. The developed method has extended the substructures technique in order to apply the Finite Element Analysis (FEA) method to complex non-linear structural problems in the conceptual design phase. As conventional FE models based on substructures allow only linear analysis, it was necessary in this research to introduce a new algorithm capable of linearizing non-linear structural problems with sufficient accuracy in order to enable comparative evaluation of design concepts relative to each other under the given constraints and loading conditions. A comparative study with respect to model size, efficiency, accuracy and confidence was performed to validate the developed method. Obtained results indicate significant improvement over more traditional approaches to applying FEA in the conceptual design phase. The improvements achieved using the developed method compared to the traditional FE based approach are superior by a factor of 2.7 in efficiency and by a factor of 4.5 in confidence while not sacrificing the optimality of the solutions.

Effective practices for the concept design of electromechanical systems

“…This fourth stage is only effective if the third stage does not yield a clear winning concept, and supports the situation where definite performance measures are needed to aid final selection in some cases (Kajtaz et al, 2013). As Kihlander (2011) indicates, there has been a 'blurring' of the concept design phase with later activities in design making it longer, so the potential to accommodate more detailed analyses has increased, together with the design tools needed to facilitate these analyses being more rapidly executed.…”

“…Convergence on a final winning concept is achieved through a staged approach: first, a concept generation stage; second, decisions on the relevant criteria for selection; third, an evaluation of the concepts and a down-selection stage to minimise candidate solutions; and finally, if needed, a fourth and final quantitative assessment of remaining candidate solutions using key performance metrics through detailed analyses. This fourth stage is only effective if the third stage does not yield a clear winning concept, and supports the situation where definite performance measures are needed to aid final selection in some cases (Kajtaz et al, 2013). As Kihlander (2011) indicates, there has been a "blurring" of the concept design phase with later activities in design making it longer, so the potential to accommodate more detailed analyses has increased, together with the design tools needed to facilitate these analyses being more rapidly executed.…”

“…Verification of concept design methodologies is a very different undertaking from, say, that of DFX tools and techniques (Huang and Mak, 1997). This is mainly due to lack of tangible characteristics to measure and evaluate JEDT 14,3 (Kajtaz et al, 2013). The outcomes reflecting design decisions are, therefore, regarded as subjective when related to the potential performance of concept designs and satisfaction of the requirements.…”