Melting modes in laser powder bed fusion additive manufacturing
MetadataShow full item record
Laser powder bed fusion (LPBF) is a metal AM technology that has one of the highest industrial uptake at the moment in the aerospace, automotive, and biomedical sectors. LPBF enjoys such popularity as it enables the manufacturing of near-net-shape geometrically complex metal parts. LPBF allows for optimised designs to be explored for manufacturing, such as topology optimised or loading field-driven designs for product lightweighting and customization, while also reducing environmental impact through energy reduction and low carbon dioxide emissions, helping the transition towards sustainable manufacturing. The manufacturing of products using LPBF is almost entirely digitally controlled, from a computer-aided design model to layer-by-layer customization of process parameters, to monitoring and controlling the process while parts are being manufactured. The digitization of metal AM opens up exciting new avenues in areas of design, process planning, process monitoring, and process control. This dissertation focuses primarily on process planning for LPBF. Process planning involves developing a theoretical understanding of the effects of the numerous process parameters that could be digitally controlled in LPBF on the final product outcomes. Within LPBF process planning, it is highly challenging to understand and model the complex laser-material interaction phenomena in LPBF, often resulting in marginally stable process parameters or resulting in a high number of experiments in the development of process parameters required to meet part quality metrics. This dissertation focuses on process physics modelling and simulation at the mesoscale to develop a theoretical understanding of the impact of LPBF process parameters on outcomes such as porous defects, surface topography, and residual stresses. For this purpose, normalized processing diagrams have been developed to visualize the three melting modes (conduction, transition, and keyhole mode) observed in LPBF. The normalized processing diagrams obtained in this dissertation, for the first time in LPBF, are shown to be independent of material, LPBF system, and processing parameters such as powder layer thickness within the datasets presented herein. Additionally, a temperature prediction model has been developed to predict the thresholds between the conduction, transition, and keyhole melting modes. The efficacy of these predicted thresholds has been evaluated experimentally for low reflectivity (titanium and ferrous) alloys and high reflectivity (aluminium) alloys. For low reflectivity alloys, a vaporisation depth greater than 0.5 and 0.8 times the beam spot radius corresponds to the thresholds between conduction to transition mode and transition to keyhole mode respectively. For high reflectivity alloys, surface vaporisation and a vaporisation depth greater than 0.5 times the beam spot radius used corresponds to the thresholds between conduction to transition mode and transition to keyhole mode respectively. Simulations using the normalized processing diagrams and the temperature prediction model are then used to develop a fundamental understanding of porous defects, surface topography, and residual stresses during LPBF of an aluminium alloy (AlSi10Mg) and two titanium alloys (Ti-6Al-4V and Ti-6242Si). For high reflectivity materials such as aluminium alloys, when considering density optimization, divergent beams with resulting focal diameters >100 μm help to obtain a conduction mode microstructure leading to parts with densities of over 99.98%. When working with a focused beam, stabilizing melt pool and spatter dynamics in the transition melting mode by using an appropriate laser power and velocity combination can help in minimizing defects and obtaining densities close to 99.98%, similar to conduction mode densities, albeit with a narrower process parameter window for success. Additionally, a melt pool aspect ratio (ratio of depth to width) of ≈0.4 is observed to be the threshold between conduction and transition/keyhole mode melt pools, which differs from the conventionally assumed melt pool aspect ratio of 0.5. This dissertation thereby provides a novel method to obtain high-quality aluminium alloy parts with defocused and focused beams in LPBF. Such findings can be expanded to other high reflectivity alloys for LPBF. For low reflectivity alloys, when considering density optimization during LPBF of Ti-6242Si, the use of processing diagrams alongside X-ray computer tomography and imaging show that Ti-6242Si has a broad process window with parts above 99.90% density observed in conduction, transition, and keyhole melting modes of LPBF. While the highest density parts (up to 99.98%) are observed in the transition melting mode for Ti-6242Si, transition and keyhole mode LPBF of Ti-6242Si could also lead to macroscopic cracking perpendicular to the build direction, which is attributed primarily to the higher residual stresses during solidification. Furthermore, when considering surface topography, a combination of statistical approaches, simulations, and experiments show that LPBF processing parameters that lie in the keyhole melting mode with lower beam velocity settings and conservative laser powers lead to surface roughness, Sa, values of lesser than 10 μm, which is significantly lower the roughness values obtained for conduction and transition mode LPBF process parameters for Ti-6Al-4V and Ti-6242Si. This significant reduction in surface roughness is due to a negligible contribution from partially melted powder particles in the keyhole melting mode border. Lastly, the fundamental understanding of LPBF developed in this dissertation was leveraged towards biomedical, military, and defence applications. The North American industry has shown a cautious approach to the adoption of LPBF, due to high initial investment costs, the iterative R&D nature of part production, and emerging certification needs. Successful industry adoption of metal additive manufacturing relies on understanding the complex interactions between design, materials, and process to ensure high product quality and reliability. This dissertation would help lower the risk of LPBF technology adoption by virtue of offering a better understanding of the physics behind the laser-material interaction in the process and reducing the need for extensive empirical approaches toward part quality-driven process parameter development.
Cite this version of the work
Sagar Patel (2021). Melting modes in laser powder bed fusion additive manufacturing. UWSpace. http://hdl.handle.net/10012/17217