Objective Heat-source models play a key role in the numerical simulation of the selective laser melting process. To examine the applicability of the double ellipsoid and Gaussian surface heat-source models on simulation, the temperature field of a single-pass selective laser melted 18Ni300 metal powder was calculated using the two heat-source models and compared with experimental data.

Methods The ANSYS APDL commercial software was used to establish a three-dimensional finite element model and to simulate the temperature field of the selective laser melting process. The double ellipsoid and Gaussian surface heat-source models were defined and applied to an 18Ni300 metal powder. Single-pass selective laser melted 18Ni300 samples were fabricated using EOS M290 by applying different laser power levels and scanning speeds. The sample widths were measured using scanning electron microscopy to determine the width of molten pools. In the calculated temperature fields, the sizes of molten pools of the single-pass selective laser melted 18Ni300 were analyzed and compared with experimental results.

Results and Discussions The shapes of the temperature field distribution calculated using the double ellipsoid and the Gaussian surface heat-source models are all ellipsoidal after applying a 210 W laser power and a 1000 mm/s scanning speed. However, their width, length, and trail calculated using the double ellipsoid heat-source model are larger than those calculated using the Gaussian surface heat-source model (Fig.3). This is due to the difference in the energy distribution of the two heat-source models. In the Gaussian surface heat-source model, the energy is distributed on the surface of the powder layer, whereas in the double ellipsoid heat-source model, the energy is distributed deep into the powder layer. Moreover, the maximum temperature calculated using the Gaussian surface heat-source model is much higher than that calculated using the double ellipsoid heat-source model. This indicates that the diffusion speed of heat through the powder layer is faster than that through the surrounding environment, resulting in a lower maximum calculated temperature of the molten pool surface using the double ellipsoid heat-source model.

Along with the laser scanning, the peak temperature calculated using the double ellipsoid heat-source model is lower than that calculated using the Gaussian surface heat-source model and fluctuates significantly with a rising trend. In contrast, the peak temperature calculated using the Gaussian surface heat-source model is higher and fluctuates slightly with small variations (Fig.4).

The sizes of the molten pools calculated by applying a 210 W laser power and a 1000 mm/s scanning speed using the two heat-source models show that the width of the molten pool calculated using the Gaussian surface heat-source model is larger than that calculated using the double ellipsoid heat-source model. In contrast, the length and depth of the molten pool calculated using the double ellipsoid heat-source model are larger than those calculated using the Gaussian surface heat-source model (Fig.5).

However, under different processing parameters, the calculated sizes of the molten pools using the two models exhibit different trends. Although the calculated width and depth of the molten pool using the two heat-source models decrease as the scanning speed increases (for the same laser power), the decreasing trend of the results calculated using the Gaussian surface heat-source model is slower than that calculated using the double ellipsoid heat-source model(Fig.6). It is confirmed that the depth of the molten pool calculated using the double ellipsoid heat-source model is larger than that calculated using the Gaussian surface heat-source model. The width of the molten pool is not only related to the laser power and scanning speed, but also to the diffusion of the laser energy in the powder. Since the Gaussian surface heat-source model does not take into account the energy diffusion in the powder, the width of the molten pool calculated using the Gaussian surface heat-source model is smaller than that calculated using the double ellipsoid heat-source model when the scanning speed is low, but it is larger when the scanning speed is high. The scanning speed has an exponential rather than linear effect on the width and depth of the molten pool. As a result, the width and depth of the molten pools are not the same under the same laser energy density. The width and depth of the molten pools obtained using high laser power and high scanning speed are larger than those obtained using low laser power and low scanning speed, although the same laser energy density was used.

The experimental results confirm that the simulation results obtained using the double ellipsoid heat-source model are more accurate and reliable than those obtained using the Gaussian surface heat-source model in the simulation of the 18Ni300 selective laser melting process(Fig.8).

Conclusions The double ellipsoid heat-source model takes into account the laser energy diffusion in the powder, and its calculated-results accuracy is better than that of the Gaussian surface heat-source model in the simulation of the 18Ni300 selective laser melting process. Thus,the double ellipsoid heat-source can describe the temperature field evolution and moten pool size during the selective laser melting process better than the Gaussian surface heat-source model. The width of the molten pool is not only related to the laser power and scanning speed, but also to the laser energy diffusion in the powder. The effect of the scanning speed on the width and depth of the molten pool is nonlinear. Under the same laser energy density, the depth and width of the molten pool obtained using high laser power and high scanning speed are larger. The width calculated using the double ellipsoid heat-source model is in good agreement with the experimental data, thus providing a basis for the optimization of the selective laser melting process.