Abstract

Selective laser melting is one of the powder bed fusion processes which fabricates a part through layer-wised method. Due to the ability to build a customized and complex part, selective laser melting process has been broadly studied in academic and applied in industry. However, rapidly changed thermal cycles and extremely high-temperature gradients among the melt pool induce a periodically changed thermal stress in solidified layers and finally result in a distorted part. Therefore, the temperature distribution in the melt pool and the size and shape of the melt pool directly determine the mechanical and geometrical property of final part. As experimental trial-and-error method takes a huge amount of cost, different numerical methods have been adopted to estimate the transient temperature and thermal stress distribution in the melt pool and powder bed. The most existing research utilizes the moving Gaussian point heat source to model the profile of the melt pool, which consumes a significant amount of computational cost and cannot be used to implement the part-level simulation. This research proposes a new line heat source to replace the moving point heat source. Some efforts are applied to reduce the computational cost. Specifically, a relatively large step size is used for the line heat source to reduce the number of time steps. In addition, a mesh refinement scheme is adopted to reduce the number of cells in each time step by refining the mesh close to the heat source and coarsening the mesh far away from it. On the other hand, efforts are implemented to increase the accuracy of the simulation result. Temperature-dependent material properties are considered in this FE framework. In addition, material transition among powder, liquid, and solid are incorporated in the developed FE framework. In this study, temperature simulation of one scanning track based on self-developed FE code is applied for Stainless Steel 316L. The simulation results show that the temperature distribution and history of melt pool within line heat source are comparable to that of the moving Gaussian point heat source. While the simulation time is reduced by more than two times depending on the length of line heat input. Therefore, this FE model can be used to numerically investigate the process parameters and help to control the quality of the final part.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call