Automation and optimisation of Family Mould Cavity and Runner Layout Design (FMCRLD) using genetic algorithms and mould layout design grammars

Family Mould Cavity Runner Layout Design (FMCRLD) is the most demanding and critical task in the early Conceptual Mould Layout Design (CMLD) phase. Traditional experience-dependent manual FCMRLD workflow causes long design lead time, non-optimum designs and human errors. However, no previous research can support FMCRLD automation and optimisation. The nature of FMCRLD is non-repetitive and generative. The complexity of FMCRLD optimisation involves solving a complex two-level combinatorial layout design optimisation problem. Inspired by the theory of evolutionary design in nature "Survival of the Fittest" and the biological genotype-phenotype mapping process of the generation of form in living systems, this research first proposes an innovative evolutionary FMCRLD approach using Genetic Algorithms (GA) and Mould Layout Design Grammars (MLDG) that can automate and optimise such generative and complex FMCRLD with its explorative and generative design process embodied in a stochastic evolutionary search. Based on this approach, an Intelligent Conceptual Mould layout Design System (ICMLDS) prototype has been developed. The ICMLDS is a powerful intelligent design system as well as an interactive design-training system that can encourage and accelerate mould designers' design alternative exploration, exploitation and optimisation for better design in less time. This research innovates the traditional manual FMCRLD workflow to eliminate costly human errors and boost the less-experienced mould designer's ability and productivity in performing FCMRLD during the CMLD phase. (C) 2013 Elsevier Ltd. All rights reserved.