Process design for high throughput Laser Powder Bed Fusion

Gullane, Alex (2023) Process design for high throughput Laser Powder Bed Fusion. PhD thesis, University of Nottingham.

PDF (Revised document with minor corrections addressed) (Thesis - as examined) - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader
Available under Licence Creative Commons Attribution.
Download (9MB) | Preview


Laser Powder Bed Fusion (LPBF) suffers from modest build rates. A powder bed component typically requires many hours to produce; hence for LPBF to become time and cost effective, batch sizes are limited to especially low volumes. By reducing build times, components can be realised much faster and it is possible to increase the number of components for which LPBF remains economical, in turn broadening the scope of applications for which the technology is viable.

LPBF has received growing attention in industry for significant benefits when compared with traditional manufacturing methods, namely high geometrical design freedom and sub-millimetre local process control. The latter, however, is seldom exercised within the literature, whereby authors typically adopt a single set of laser parameters to consolidate a component. This one-size-fits-all approach solely satisfies one metric, most commonly mechanical performance, while failing to accommodate other beneficial metrics such as manufacturing productivity. By harnessing the local process control possible through LPBF, a given component can be optimised for multiple metrics simultaneously.

The work in the present thesis exploits these process design freedoms, by varying parameters within sub-volumes of components to achieve the optimal part for both service conditions and manufacturing productivity. This involves prioritising mechanical strength in areas of structural significance and high volumetric build rates in areas of low structural significance. In theory, a component with similar mechanical behaviour to that seen in standard LPBF parts can be built with reduced time and cost. In practice however, this method is yet to be investigated and the boundaries between sub-volumes are yet to be understood.

Two methods have been highlighted, in which build times can be significantly reduced. These are coarser layer thicknesses, and Hot Isostatic Pressing (HIPing) of shelled components. Used in isolation, each incurs a notable reduction in mechanical properties. However, these techniques are introduced to LPBF parts as sub-volumes, enabling the remaining volume of the components to be fabricated by standard laser processing parameters to maintain mechanical properties, while still benefitting from reduced build times.

An initial study demonstrates parts can be additively built using multiple layer thickness regions with similar ultimate tensile strength (1110 - 1135 MPa) and elastic moduli


to standard LPBF specimens. Varying penalties to ductility were observed, depending on layer thickness and interface design (elongation to failure reductions up to one-third in the most extreme case). New pore formations were discovered along the interface between sub- volumes that were understood to dominate failure.

In the following study, XCT revealed that sub-volume interface orientation has a great effect on porosity formation along the boundary, with interfaces perpendicular to the substrate experiencing the largest pore formations, while interfaces parallel to the substrate experienced no additional porosity. Realtime evidence was observed that these defects lead to fracture at the interface. By combining pre-test XCT and Focus Variation data of fracture surfaces, a novel 3D reconstruction technique has been demonstrated, enabling post-mortem evaluation of additively manufactured parts and tracking of pore deformation during subsequent mechanical testing.

The final study designs and tests new-class laser scan strategies, to enable the laser to raster back and forth between regions of varying parameters, changing parameters both instantaneously and using a ramped region, as well as using secondary rework passes. Rework passes were successful in halving the number of interfacial pores, however, the introduction of laser vectors that pass through sub-volumes continuously presented new pore formations at parameter increment planes. HIPing of semi-hollow specimens achieved high density material with only a small volume of < 10 μm pores and superior ductility. Both methods were effective in significantly increasing productivity, however, each presented notable issues in part quality when characterised for porosity using XCT and mechanical performance by way of fatigue testing.

Build times can be reduced as much as 31.2% and 34.6% using coarse layer or shelled regions respectively, while still dedicating at least a quarter of the component volume to optimal parameters to maintain part performance. It is also possible to increase annual profits by one-third by adopting this method – this may increase or decrease depending on the size and number of builds. By producing entirely shelled components with coarse layers and using a post-process HIP treatment to enclose the void, it is possible to reduce build times by up to 76% and increase annual profit by 259%.

Item Type: Thesis (University of Nottingham only) (PhD)
Supervisors: Clare, Adam
Hyde, Chris
Murray, James
Keywords: Additive manufacturing, Laser Powder Bed Fusion, Process, Process control, Productivity, Ti-6Al-4V, Layer thickness
Subjects: T Technology > TS Manufactures
Faculties/Schools: UK Campuses > Faculty of Engineering
Item ID: 73137
Depositing User: Gullane, Alex
Date Deposited: 31 Jul 2023 04:40
Last Modified: 31 Jul 2023 04:40

Actions (Archive Staff Only)

Edit View Edit View