Multi-criteria GA-based Pareto optimization of building direction for rapid prototyping

Abstract

Selection of a building direction is an important step for rapid prototyping regardless of the specific processes used to create the part. It involves the consideration of multi-factors that have influences on surface quality, build efficiency and support structure, etc. Contemporary approaches did not consider the global directional space to search for the optimal building direction. In this paper, we use a multi-sphere model for multi-criteria optimization of building direction in rapid prototyping. Each sphere represents the global directional space for one optimization criterion, and is obtained by uniformly discretizing the surface of a unit sphere. Optimization is then conducted over each discretized spherical surface for each criterion. Two objectives, theoretical volume deviation (TVD) and part height are simultaneously optimized using genetic algorithm. TVD is computed to evaluate the volumetric error along each building direction in a general way. The Pareto front is computed as well in order to study the competing effect from these two criteria. At the end of the paper, examples are presented to show the effectiveness of the method.

Personalized design of part orientation in additive manufacturing

“…(Ahn et al, 2007) developed an intelligent methodology to determine the optimal fabrication direction to minimize the required postmachining region. (Li and Zhang, 2013) proposed to use a multisphere model for multi-criteria optimization of building direction in RP. Two objectives, theoretical volume deviation and part height are simultaneously optimized using genetic algorithm.…”

Description Logic Ontology-Supported Part Orientation for Fused Deposition Modelling

“…A one-step approach generally adopts a search algorithm to directly search for an orientation that best meets certain requirements from an infinite solution space. The commonly used search algorithms include the exhaustive search algorithm [31][32][33][34][35][36] and heuristic search algorithms, such as the genetic algorithm [37][38][39][40][41][42], non-dominated sorting genetic algorithm II [43,44], particle swarm optimisation algorithm [45,46], electromagnetismlike mechanism algorithm [47], bacterial foraging optimisation algorithm [48], sequential quadratic programming algorithm [49], and S-metric selection evolutionary multi-objective algorithm [50]. A prominent feature of one-step approaches is that they are applicable to all forms of 3D models in theory.…”

A Multi-Criteria Decision-Making Approach to Optimize the Part Build Orientation in Additive Manufacturing

“…To solve the multi-objective optimisation issue using a genetic optimisation method, normalised weights are assigned to various objectives based on their relative relevance. Instead, several technologies were evaluated simultaneously by (Li & Zhang, 2013) to choose the optimal direction using a multi-objective approach. A multi-sphere system was formulated in the methodology to optimise the construction orientation of an AM component.…”