Computational fluid dynamics modelling of powder-based laser additive manufacturing processes

Abstract

Powder-based laser additive manufacturing (LAM) applies a laser beam to melt injected or pre-placed powder. A three-dimensional (3D) object can be fabricated layer-by-layer. Many investigations on LAM processes have been carried out. Numerical modelling is an effective tool to understand the complex physical phenomenon in multiple scales. Efforts have been made on developing numerical models for various LAM processes. However, in the existing thermal-fluid models for powder-based LAM processes, the powder stream or layer is usually simplified. In L-DED processes with a high powder feed rate, the impact of powder injection on the melt pool dynamics cannot be neglected. It requires an improved powder stream model to represent the powder injection more realistically. Powder filler narrow gap laser welding (NGLW) process is a novel welding technique that applied powder to fill up the groove layer-by-layer in an additive manufacturing manner. It shares similar the mechanism to off-axial L-DED process. However, no related study has been reported. As for the recently developed multiple materials laser powder bed fusion (MMLPBF) process, no numerical model was available to describe the complex physical behaviour of this new additive manufacturing technology. In this research, a series of investigations have been carried out around the powder-based LAM processes. An in-situ, coaxial melt pool measurement technique for coaxial L-DED was developed. This study successfully determined the width and length of the melt pool in real-time. With the obtained melt pool width, a process was conducted to predict the cross-sectional outline of the deposited track. The melt pool dimensions obtained through the in-process measurement can be applied to validate the developed theoretical models. A computational fluid dynamics (CFD) model was developed to investigate the L-DED processes with relatively high powder feed rates. It modelled the powder source with a discrete phase method (DPM). The impact of the injected powder stream on the melt pool dynamics was analysed. In addition, a case study was carried out to investigate the powder filler NGLW process. The influence of the main processing parameters on the formation of macro defects was investigated qualitatively. An analytical model was coupled with the developed numerical model to predict the primary dendrite arm spacing (PDAS) in the deposited track. Stainless steel blocks of 8.5 mm in thickness were successfully welded with four filling passes. The microstructure was mapped for the joint cross-section. Moreover, a CFD model was developed for a multiple materials laser powder bed fusion (LPBF) process. The mixture of the dissimilar metals was considered. The temperature distribution in the melt pool and the morphology and components distribution of melting multi-material powder bed were characterised. Powder-scale inhomogeneous melting behaviour of dissimilar metallic powders was predicted. The single-track morphology on mixed dissimilar powder bed was analysed. Moreover, adjacent-track melting was simulated. The research has resulted in the submission of 4 papers for journal publications.

Date of Award	1 Aug 2020
Original language	English
Awarding Institution	The University of Manchester
Supervisor	Lin Li (Supervisor), Zhu Liu (Supervisor) & Wei Guo (Supervisor)