This research investigated the effect of changing the primary powder particle size distributions and layer thicknesses on part density. It also presented regression models for the process, which allowed the prediction for the density of fabricated parts. The research also conducted the first build location study, which assessed the influence of building in different locations on the build platform on the part density. However, not all related aspects have been considered in this research as they are out of the scope of this research. Other processing parameters need to be included in future work.
The database and the findings of this work would be highly beneficial for the practitioners. Repeating the work for other materials would definitely increase the understanding of the process and result in a greater understanding of the laser-material interaction and comparison studies. Moreover, for an enhanced understanding of process parameter and their interactions, it would be better to conduct the same study using a continuous laser beam. A continuous beam would be beneficial as it would allow certain parameters to be studied in isolation, to further understand their interactions on final fabricated parts. For instance, the point distance factor will no longer be applicable, as the laser would fire continuously. The combination of point distance, exposure time and jump speed as scan speed will be replaced by one parameter, solely scan speed which is the speed of Golva. Furthermore, the effects of ignition and intermittently stopping the laser beam will be mitigated as the number of times the laser is turned on and off will be reduced.
One of the main reasons for process instability from one material to another is the powder absorptivity of the heat source. Further research using a superior heat source such as a laser beam with different wavelength should help increasing the use and adaptation of powder bed fusion process into manufacturing and would allow the inclusion of additional materials, which could be used in fabricating part using PBF processes.
The volumetric energy density (VED) has been used to evaluate the ideal input energy zone for fabricating a material. As shown during this research, using VED alone as a single parameter to determine and compare the applied energy is not appropriate. The
individually, to ensure a comprehensive understanding. As future work, another energy indicator factor needs to be found or at least a modification for the current VED should be established, which is more representative and applicable to be used. Spot size, overlapping rate and penetration depth (welding layers) are factors that need to be included in an applied energy indicator, especially if it is known that each material behaves differently. Small spot size penetrates vertically into the layers but does not cover a wider region horizontally for adequate overlapping between adjacent melt pools, and vice versa. The rate of overlapping controls the melting area and depends on the applied energy. The overlapping between adjacent melt pools is determined by the value of hatching distance and resulting melt pool width. For a pulsed laser system, the overlapping between points in a melt track is determined by the point distance. The spot size influences both values while it is not included in the common formula that is used to calculate the VED.
Spatter, which is ejected particles during laser beam scanning and melt-pool formation, has a huge impact on quality of fabricated parts. It is a complex phenomenon that has not been comprehensively studied yet. An intensive study that investigates spatters associated with different materials and process parameters is a must. During the current study, it was observed that spatter is material dependent. Titanium alloy (Ti-6Al-4V ELI) produces fewer spatters in comparison to Stainless Steel (316L-SS) for similar VED and PSD. The amount and size of 316L-SS spatter is more and larger, which is more challenging. This means that the phenomenon is material dependent and requires a deeper investigation to understand what affects the spatter and how to control/minimise spatter. It is believed that the direction of gas flow inside the chamber and flow rate must be optimised for the best spatter control. They should be adjustable according to the material being process and fusion process parameters. In short, machine design is critical to control spatter. Ejected particles do not only impact the quality of fabricated parts but also increase the cost of the process; the more the spatter, the more the waste powder. The quantity of waste powder would be huge in mass production systems of AM. A dynamic optical system that has been embedded in Renishaw’s new machine may improve the part quality and allow for more reproducibility in any location on the build platform. In addition, the speed of the ejected particles for different material needs to be studied.
To achieve a robust and comprehensive prediction model for the laser powder bed fusion process, more process parameters, material properties, powder characterisation and the physical phenomena associated with the process need to be considered. As the models developed in the current study included the process parameters and their interaction, including power-powder interaction and physical properties for a material should increase the accuracy of the model. The model may be linked with simulation for the purpose of validation. Also, an intelligent software that is able to reduce the build time and improve the productivity would be highly beneficial. For example, software that enables slicing in different layer thicknesses can help increase productivity. The different slicing can be done according to the importance of a part section. Supports for instance, are not necessarily to be melt with every layer. If a part is sliced to have a layer thickness of 30µm, supports can be fused every other layer (i.e. at layer thickness of 60µm). Another example, when a part has a section that is not required to be full dense with other sections of the part that need to be full dense, a faster melting energy/scan can be calculated and suggested by software to speed up the build process in this section. Moreover, the software should be smart enough to identify any instability in the process, distinguish the cause and correct the process parameters accordingly in a closed loop system. There are a variety of sensors are available which can be used to modify current machines (especially laser machines), such as thermal imaging, spectrometers, optical imaging, etc., to detect process instabilities. Nowadays, there are some machines in the market with in-process monitoring capabilities.
However, the difficulty would be capturing a high volume of data and processing it in short time to correct the process automatically.