Development of Additive Strategy Generator for Metal Additive Manufacturing Build Prediction Using Laser Path Generation Algorithm

The Laser Powder Bed Fusion method is widely used in industrial additive manufacturing, which uses a laser as energy source to melt and build up powder materials. Additive process parameters such as the laser path and power, speed, layer thickness, and rotational additive strategy influence the quality of final products. In this study, in order to evaluate the build performance according to these additive process parameters during additive manufacturing, an algorithm that automatically creates a laser path based on the geometric information of the STL (stereolithography) model was developed. In addition, an algorithm that extracts section and layer information from the STL file of the sample model and automatically creates a laser beam path according to the desired additive process parameters was developed. Based on this, a program that can predict additive build performance by using commercial software Simulia Abaqus as a solver had been developed. Using the proposed algorithm and additive build prediction technology, build performance can be evaluated at the part level, and process optimization can determine additive manufacturing conditions. Numerical simulations have shown that the developed program can predict various build performances, using the cantilever model, which is widely used for extracting eigen strains for predicting part-scale build performance, as an example. The proposed program was developed to enable eigen strain prediction without the need for additional postprocessing because the variables and environment used in the actual PBF equipment can be used for the thermal-structure coupling analysis of Simulia Abaqus as it is. Therefore, by using the algorithm and the part-level build performance prediction technology that developed in this study, it will be possible to save a lot of time and money spent on experiments to determine the best additive process conditions, and it is predicted that it is possible to expand to the simulation-based prediction technology, such as digital twin, in the future.