Abstract
The field of metal additive manufacturing (AM) has expanded in the past 10 years with several methods categorized as either fusion-based or solid-state. One of the dominant fusion-based AM methods, termed selective laser melting (SLM), is a powder-bed process that is a current topic of many studies focusing on the process parameters and how they affect the overall build quality. For this research, the alloy of specific interest is Inconel 718 (IN718). This nickel-based superalloy is widely used in high-temperature aerospace applications, such as nozzles, injectors, and other engine components where SLM offers the ability to increase the production efficiency of these complex and precise parts. Utilizing SLM as the production method for IN718 components can potentially allow for fewer manufacturing steps, less assembly, and less material waste, leading to lower costs and lead times. This process also allows for new design complexities that are not possible with traditional manufacturing, such as latticing and internal features. These capabilities have driven immense amounts of research focused on understanding the intricacies of how the SLM process affects IN718 builds.
SLM is a manufacturing method that has a large number of process parameters functioning in parallel, all of which uniquely affect the outcome of the parts produced. Key parameters include laser power, scan speed, layer thickness, and hatch spacing. Due to the broad parameter space and lack of standards, defects are commonly observed. Defects can include porosity, cracking, residual stress, and part distortion, which affect part lifetime and application. One parameter that has received minimal attention is the inter-layer time interval (ILTI). The ILTI is the amount of time it takes for the laser to return to the exact position during the building of consecutive layers, thus relating the role of heat accumulation from layer to layer. The ILTI is not a process input in SLM, since it cannot be programmed into the machine, however it can be influenced by sample geometry, and the number of parts within a build. The broad range of variations makes the impact and understanding of the ILTI complex and difficult to predict. In this research study, the impact of ILTI on a parts overall build quality was investigated. From this research data, the melt pool characteristics, microstructure, texture, and amount of porosity all showed variation, while changes in the composition and hardness were determined to be statistically insignificant. The experiments were designed such that the ILTI varied from 126 to 13 seconds, with the most prominent differences observed from the shortest time interval.
At the shorter intervals, the solidified melt pool structures changed in two aspects. The first is the degree of melt pool staggering. Melt pool staggering is referring to the horizontal offset that melt pools have with respect to a melt pool in the previous layer. The melt pools exhibited variations in the staggered geometry along with a different shapes based on the ILTIs. Analysis of the results suggest that variations in the melt pool shape are due to changes in the laser beam characteristics from a conduction mode to a transition mode observed in laser welding, along with the keyhole mode. Secondly, the radii of curvature for the melt pools varied within a range of 91 to 131 μm for an ILTI of 126 seconds to a range of 50 to 70 μm at an ILTI of 13 seconds. The lower radius of curvature indicated that the depth-to-width ratio of the melt pool is increasing, consistent with the literature regarding a shift to the transition mode of welding. To investigate the validity of a shift in the laser beam mode, the Eagar-Tsai model for temperature fields produced by a traveling distributed heat source was calculated. The model supported the melt pool shape for the longer ILTIs, but could not accurately predict the melt pool shape for the shortest ILTI (13 seconds). Since the Eagar-Tsai model is limited to application of the conduction mode of welding, the results suggest that at shorter ILTIs, the behavior of the melt pools is more consistent with the transition mode of operation. This change of energy/penetration beam mode could also explain the why the amount of sample porosity increased as the ILTIs decreased. The average area density of porosity, as measured by x-ray computed tomography (XCT) and image analysis, ranged from 0.15 mm-2 to 0.29 mm-2 as a function of decreasing ILTI.
Optical and electron microscopy along with electron backscatter diffraction (EBSD) analysis showed that the grain structure is well organized in the 13 second ILTI regions. This grain structure consists of 30 μm wide grains spaced 120 μm apart, extending up the centerlines of the melt pools, with larger grains taking up the spaces between. Both grain types showed evidence of growth through multiple build layers, as opposed to the nucleation of new grains at the melt pool interface, which was observed in the longer ILTI regions. The results suggest that the epitaxial growth through multiple build layers is due to the increased depth of penetration, minimal melt pool staggering, and higher temperatures. Consistent with changes in the melt pool alignment and geometry, the crystallographic texture of the samples was found to strengthen in the reduced time interval regions. Analysis of the EBSD data showed that the two distinct grain types dominate the texture. The stronger of the two is a Goss texture, {110}<001>, oriented in the build direction, found in the wider to the two grain types. The weaker component was a cubic texture, {100}<001>, oriented in the build direction, found in the other of the two grain types. There was a strengthening of texture observed, where longer ILTIs had a maximum multiple of uniform density (MUD) value of roughly 6, and the 13 second ILTI had a maximum MUD value of roughly 9. The change in the MUD values were attributed to a reduction in melt pool staggering and increased remelt of the previous layer, which both affect the directionality of the thermal gradients and solidification.
From this research, it is clear that the inter-layer time interval is a process input that will affect the overall quality of a SLM deposit in terms of grain morphology, texture, and porosity. For this reason, it should be a topic for further investigation including a more detailed evaluation of the mechanical properties, corrosion performance, fatigue response, and precipitate phases.
