Método paralelo de cálculo da trajetória de preenchimento em zigue-zague para manufatura aditiva

Abstract

Additive manufacturing (AM) is a manufacturing process by successively adding layers of material. Process planning is a pre-manufacturing computational step in which the three-dimensional CAD (Computer Aided Design) model is processed and converted into information that will be used by the machine. At this stage, the part is sliced into layers and, for each layer, the area to be filled with material is identified. The choice of the filling strategy and the definition of its calculation parameters is of fundamental importance for AM, as it confers different mechanical properties to the object, also impacting the time and cost of manufacturing. The development of AM machines, which allow the manufacture of larger parts with thinner layers, increases the computational demand of the process planning, which in some cases can take hours. Developing more efficient algorithms for process planning opens up a number of possibilities, such as parameter optimization via simulations. Currently, the parallelization of algorithms and the acceleration using graphics processors are techniques widely used in the literature. This work aims to develop a new parallel method for the calculation of one of the main filling strategies used in material extrusion technology (Fused Deposition Modeling – FDM), which is the zigzag (raster) strategy. Through computational testing and an application study, the method was evaluated to verify its ability to solve the zigzag calculation problem, to measure its speedup, and to evaluate the benefits of its use in optimizing the parameters of the trajectory. The results showed that the proposed parallel method is functionally correct and generates relevant computational gain, being up to 22 times faster than its serial counterpart. The application study showed that simulating path parameters results in up to 38% increase in average length of raster lines, reducing manufacturing time.