Computational modeling of selective laser melting process

Selective laser melting (SLM) is a metal-based additive manufacturing (AM) process which can produce printed parts with high density and good mechanical properties. However, the properties of the printed part depend on several processing parameters including laser power, scanning speed, layer thickn...

Full description

Saved in:
Bibliographic Details
Main Author: Le, Kim Quy
Other Authors: Du Hejun
Format: Thesis-Doctor of Philosophy
Language:English
Published: Nanyang Technological University 2020
Subjects:
Online Access:https://hdl.handle.net/10356/136867
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
id sg-ntu-dr.10356-136867
record_format dspace
institution Nanyang Technological University
building NTU Library
continent Asia
country Singapore
Singapore
content_provider NTU Library
collection DR-NTU
language English
topic Engineering::Mechanical engineering::Fluid mechanics
Engineering::Mathematics and analysis::Simulations
spellingShingle Engineering::Mechanical engineering::Fluid mechanics
Engineering::Mathematics and analysis::Simulations
Le, Kim Quy
Computational modeling of selective laser melting process
description Selective laser melting (SLM) is a metal-based additive manufacturing (AM) process which can produce printed parts with high density and good mechanical properties. However, the properties of the printed part depend on several processing parameters including laser power, scanning speed, layer thickness, powder distribution, and hatch spacing. Moreover, SLM is a multi-physics process, including heat transfer, fluid flows, multi-phase interactions, phase-change process, laser-material interaction, and chemical reactions. In order to model and simulate the SLM process correctly, several methodologies have to be adopted. Discrete element method (DEM), which is a finite-difference method widely used for simulating granular systems, has been applied to describe the Newtonian interactions of the particles (or powders) in SLM, while computational fluid dynamics (CFD) has been adopted to model the thermodynamics of the melting process. In this work, open-source software packages including LIGGGHTS® and OpenFOAM® were chosen to simulate the SLM process. LIGGGHTS® is a DEM solver and OpenFOAM® is a CFD solver. The heat transfer, fluid flows, laser-material interaction, phase-change, and multi-phase interactions were all performed in OpenFOAM®. A model has been developed to simulate the deposition of powder, the melting process, and the solidification of the melt track. Moreover, multi-physics in SLM process have been imitated, including heat transfer, fluid dynamics, phase-change, multi-phase interaction, and the interaction of the laser beam with the powder bed. Different forces, such as surface tension, recoil pressure, and Marangoni force, have been presented. Two models for the laser beam source has been developed – as a simplified heat source and one that mimics the multiple reflections by employing a ray-tracing model. Both models have been validated by comparing the melt pool depths and widths of the simulation to experiments. The laser beam, when simulated as a simplified heat source, was applied on the top surface of the powder bed. The simplified heat source model was employed to examine the levelness of the melt tracks produced with different scanning patterns. The study found that the sequential scanning pattern leads to a non-flat printed surface, while the boustrophedon pattern produces a flat printed surface. Multiple reflections of the laser beam have been developed by employing the ray-tracing model to simulate the keyhole-mode melting phenomenon. The melt pool can be in conduction-, transient-, or keyhole-mode. In the study, the melt pool is considered to be in conduction-mode when the ratio of the melt pool depth to the melt pool width is less than 0.5 and keyhole-mode when the ratio is above 1.0. The region between the conduction-mode and the keyhole-mode is known as the transient-mode. Although the simplified heat source model was able to simulate the conduction-mode melting with results close to the experiments, it is inadequate for keyhole-mode melting which requires the representation of multiple reflections of the laser beam to be included. As such, the ray-tracing model was developed to simulate the keyhole-mode and the melt pool dynamics of the keyhole-mode melt track. Furthermore, the model offers a channel to predict the keyhole-induced porosities in the solidified melt track. Hence in this study, two types of keyhole-mode melting were investigated, namely medium-deep keyhole and well-deep keyhole. When the ratio between melt pool depth and melt pool width was greater than a reference value of 1.5, it can be considered as the well-deep keyhole. The study found that the melt pool dynamics of the well-deep keyhole is unsteady as compared to the medium-deep keyhole. Different modes of fluid flow, such as downward flow, bottom backward flow, clockwise flow and top forward flow are noticed in the well-deep keyhole melt pool. On the other hand, the melt pool dynamics of the medium-deep keyhole is more stable with two main flows: the downward flow and backward flow. In addition, the ray-tracing model was also employed to investigate the melt track morphology of the overhanging structure. The study found that the overhanging melt track often experiences discontinuity due to wetting effect, low surface tension gradient. In addition, short lifetime of the melt pool also leads to the discontinuity, where the lifetime of a melt pool was estimated from the time when the material was melted until it was fully solidified. However, by applying high laser power while maintaining low energy density alleviates the discontinuities and at the same time achieves high geometrical accuracy of the overhanging printed part.
author2 Du Hejun
author_facet Du Hejun
Le, Kim Quy
format Thesis-Doctor of Philosophy
author Le, Kim Quy
author_sort Le, Kim Quy
title Computational modeling of selective laser melting process
title_short Computational modeling of selective laser melting process
title_full Computational modeling of selective laser melting process
title_fullStr Computational modeling of selective laser melting process
title_full_unstemmed Computational modeling of selective laser melting process
title_sort computational modeling of selective laser melting process
publisher Nanyang Technological University
publishDate 2020
url https://hdl.handle.net/10356/136867
_version_ 1761782023183138816
spelling sg-ntu-dr.10356-1368672023-03-11T18:01:23Z Computational modeling of selective laser melting process Le, Kim Quy Du Hejun School of Mechanical and Aerospace Engineering Singapore Centre for 3D Printing MHDU@ntu.edu.sg Engineering::Mechanical engineering::Fluid mechanics Engineering::Mathematics and analysis::Simulations Selective laser melting (SLM) is a metal-based additive manufacturing (AM) process which can produce printed parts with high density and good mechanical properties. However, the properties of the printed part depend on several processing parameters including laser power, scanning speed, layer thickness, powder distribution, and hatch spacing. Moreover, SLM is a multi-physics process, including heat transfer, fluid flows, multi-phase interactions, phase-change process, laser-material interaction, and chemical reactions. In order to model and simulate the SLM process correctly, several methodologies have to be adopted. Discrete element method (DEM), which is a finite-difference method widely used for simulating granular systems, has been applied to describe the Newtonian interactions of the particles (or powders) in SLM, while computational fluid dynamics (CFD) has been adopted to model the thermodynamics of the melting process. In this work, open-source software packages including LIGGGHTS® and OpenFOAM® were chosen to simulate the SLM process. LIGGGHTS® is a DEM solver and OpenFOAM® is a CFD solver. The heat transfer, fluid flows, laser-material interaction, phase-change, and multi-phase interactions were all performed in OpenFOAM®. A model has been developed to simulate the deposition of powder, the melting process, and the solidification of the melt track. Moreover, multi-physics in SLM process have been imitated, including heat transfer, fluid dynamics, phase-change, multi-phase interaction, and the interaction of the laser beam with the powder bed. Different forces, such as surface tension, recoil pressure, and Marangoni force, have been presented. Two models for the laser beam source has been developed – as a simplified heat source and one that mimics the multiple reflections by employing a ray-tracing model. Both models have been validated by comparing the melt pool depths and widths of the simulation to experiments. The laser beam, when simulated as a simplified heat source, was applied on the top surface of the powder bed. The simplified heat source model was employed to examine the levelness of the melt tracks produced with different scanning patterns. The study found that the sequential scanning pattern leads to a non-flat printed surface, while the boustrophedon pattern produces a flat printed surface. Multiple reflections of the laser beam have been developed by employing the ray-tracing model to simulate the keyhole-mode melting phenomenon. The melt pool can be in conduction-, transient-, or keyhole-mode. In the study, the melt pool is considered to be in conduction-mode when the ratio of the melt pool depth to the melt pool width is less than 0.5 and keyhole-mode when the ratio is above 1.0. The region between the conduction-mode and the keyhole-mode is known as the transient-mode. Although the simplified heat source model was able to simulate the conduction-mode melting with results close to the experiments, it is inadequate for keyhole-mode melting which requires the representation of multiple reflections of the laser beam to be included. As such, the ray-tracing model was developed to simulate the keyhole-mode and the melt pool dynamics of the keyhole-mode melt track. Furthermore, the model offers a channel to predict the keyhole-induced porosities in the solidified melt track. Hence in this study, two types of keyhole-mode melting were investigated, namely medium-deep keyhole and well-deep keyhole. When the ratio between melt pool depth and melt pool width was greater than a reference value of 1.5, it can be considered as the well-deep keyhole. The study found that the melt pool dynamics of the well-deep keyhole is unsteady as compared to the medium-deep keyhole. Different modes of fluid flow, such as downward flow, bottom backward flow, clockwise flow and top forward flow are noticed in the well-deep keyhole melt pool. On the other hand, the melt pool dynamics of the medium-deep keyhole is more stable with two main flows: the downward flow and backward flow. In addition, the ray-tracing model was also employed to investigate the melt track morphology of the overhanging structure. The study found that the overhanging melt track often experiences discontinuity due to wetting effect, low surface tension gradient. In addition, short lifetime of the melt pool also leads to the discontinuity, where the lifetime of a melt pool was estimated from the time when the material was melted until it was fully solidified. However, by applying high laser power while maintaining low energy density alleviates the discontinuities and at the same time achieves high geometrical accuracy of the overhanging printed part. Doctor of Philosophy 2020-02-03T06:24:56Z 2020-02-03T06:24:56Z 2020 Thesis-Doctor of Philosophy Le, K. Q. (2020). Computational modeling of selective laser melting process. Doctoral thesis, Nanyang Technological University, Singapore. https://hdl.handle.net/10356/136867 10.32657/10356/136867 en This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License (CC BY-NC 4.0). application/pdf Nanyang Technological University