Page 96 - MSAM-2-3

Materials Science in Additive Manufacturing Validation of a novel ML model for AM-PSP

engineer net shape (LENS), laser powder fusion (LPF), laser- time into the melt pool, which leads to in-situ alloying .
[16]
aided direct-metal deposition (LADMD), laser-based multi- However, like PBF processes, secondary processes such as
direction metal deposition (LBMDMD), and laser-aided machining and heat treatment are required for DED parts.
manufacturing process (LAMP) . A summary of comparisons between the different metal
[16]
Figures 4 and 5 show the typical powder DED machine AM techniques is presented in Table 2 below:
and wire DED machine structure, respectively.
1.3. Titanium metal AM processing
During the DED process, material feed rate and energy
power can be adjusted to achieve the desired microstructural Metal AM processes have a significant impact on the
feature. In addition, variation to process parameters and microstructure and performance of the titanium alloys. In
multi-nozzle feedstock with different alloy systems can L-PBF and DED processes, elongated prior β grain boundaries
lead to a unique advantage for the DED process to fabricate grow in the same direction as the build orientation due to
functionally graded components. Due to the unique the cyclic thermal history [16,42,43] . However, microstructure
advantage of the material delivery system, the DED process and microhardness vary depending on build orientation and
can be used to repair and clad valuable parts that cannot build height. The prior β grain size in DED parts is larger
be processed by other AM processes. Furthermore, the than in L-PBF, while the microhardness is higher in L-PBF
[44]
DED process provides a higher deposition rate and a wider parts . In DED titanium, columnar, equiaxed, and a mix of
process volume compared to other AM processes. During columnar and equiaxed microstructure are observed, while
the DED process, different powders can be fed at the same in L-PBF, the lamellar colony, columnar, and Widmanstätten
are usually observed [19,45-48] .

Microstructural evolution in the AM processes and heat
treatment could influence the machinability of titanium
components. In general, finer microstructures lead to
higher hardness and machining force. In L-PBF parts, heat
treatment could lead to coarser microstructures, especially
when the heat treatment temperature approaches or
surpasses the β transus. This increases the average α laths
size significantly and reduces yield stress and ultimate tensile
stress . In L-PBF parts, the near-surface microstructure
[32]
is typically fine, and its grain orientation limits crystal
dislocation, leading to higher slip resistance, and increased
machining force . In DED titanium components,
[49]
subsurface deformation depth decreases with increasing
machining speed. At a low cutting speed, the built-up edge
phenomenon is observed at the cutting tooltip .
[50]
Figure 4. Powder-fed directed energy deposition.
As AM-processed titanium alloys are known to have
significant property differences, it is critical to build
unique process-structure-property linkages (PSP) for a
better understanding of the influence that varying grain
morphology has on material performance. Previous research
simulated grain structure evolution in the as-deposited state
for L-PBF using computational fluid dynamics and phase-
field framework . Several researchers have developed
[51]
analytical and finite-element models only to investigate
residual stress and melt pool profiles without consideration
for machining [52,53] . There is a growing number of reported
efforts to develop PSP linkages to capture the underlying
interactions between AM parameters and structure
properties [54,55] . Machine learning (ML) models explore
the relationship between structure and property based on
existing data without constructing AM process-physics-
Figure 5. Wire-fed directed energy deposition. based assumptions and numerical models [54,56,57] . Additional

Volume 2 Issue 3 (2023) 4 https://doi.org/10.36922/msam.0999