Effects of the Gaussian-based heat source models on the finite element simulated temperature history of additive manufacturing components

The overall performance and lifetime of additively manufactured parts are strongly influenced by the local thermal history of the material. A realistic thermal process simulation allows prediction, interpretation and optimization of local microstructural evolutions, residual stress formations, defect formations and the resulting material properties. A pivotal point in thermal simulation is the adequate modeling of the local energy deposition imposed on the material by the heat source. Most commonly, a Gaussian-based model describes the heat distribution, based on measurements of the laser intensity profile. However, due to the complex nature of heat source–material interaction and its dependency on material, temperature and process parameters, it is unclear whether the heat flux also follows a Gaussian distribution. To study this assumption, four common heat source models, two-dimensional Gaussian, double ellipsoid, conical Gaussian and hemispherical, were investigated. First, their functional relations and thermal outcomes were implemented via FE simulations of a bead-on-plate model, enabling study of process parameters, material properties and geometrical aspects as process variables. Next, the models were validated for two AM cases using electric arc and laser beam heat sources, and the thermal histories were compared. The results reveal that using the Gaussian-based heat distribution can cause the temperature maxima to trail the heat source centre. This is particularly related to the rate of heat conduction from the melt pool to the surrounding base material and the heating rate. Since these phenomena depend on material properties and process parameters, considering both enables more accurate optimization of the AM process.