Framework para otimização dos movimentos de reposicionamento em manufatura aditiva por extrusão de material

• Ano de defesa: 2019
• Autor(a) principal: Weller, Tiago Rodrigues
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Tecnológica Federal do Paraná; Curitiba; Brasil; Programa de Pós-Graduação em Engenharia Mecânica e de Materiais; UTFPR
• Palavras-chave em Português: Impressão tridimensional; Ciclo de vida do produto; Produtos industrializados; Programação linear; Plásticos - Extrusão; Framework (Programa de computador); Heurística; Otimização combinatória; Three-dimensional printing; Product life cicle; Manufactures; Linear programming; Plastics - Extrusion; Framework (Computer program); Heuristic; Combinatorial optimization; CNPQ::ENGENHARIAS::ENGENHARIA MECANICA::PROCESSOS DE FABRICACAO; Engenharia Mecânica
• Link de acesso: http://repositorio.utfpr.edu.br/jspui/handle/1/4763

Building time is an important issue in material extrusion-based additive manufacturing due to the deposition principle and because head repositionings are always required between deposition segments (contours and rasters). The time waste in head repositioning, usually referred to as non-productive time can be minimized by applying optimization algorithms. In this work, the problem of repositioning movements optimization is detailed in order to increase its understanding, thus proposing ways to reduce its computational complexity. From the understanding of the problem a framework was proposed as a method to structure, decompose and simplify its solution. The framework has been divided into four (4) main steps and is designed to be used with different optimization methods. The first step was designed with some innovative methods to reduce the problem size. Mixed integer linear programming (MILP) models and tabu search metaheuristics were implemented to test the proposed framework. Three cases of different complexities were analyzed and compared with a classic greedy algorithm and the nearest neighbor insertion algorithm with 2-opt. The results show that the framework was effective to deal with this problem and, especially for the analyzed cases, it was possible to reduce the repositioning distance considerably. The simplification of the problem for more complex cases still needs to be analyzed.