rag_docs_final_review_tex_merged/tungsten_review.tex

% This LaTeX document needs to be compiled with XeLaTeX.
\documentclass[10pt]{article}
\usepackage[utf8]{inputenc}
\usepackage{ucharclasses}
\usepackage{amsmath}
\usepackage{amsfonts}
\usepackage{amssymb}
\usepackage[version=4]{mhchem}
\usepackage{stmaryrd}
\usepackage{hyperref}
\hypersetup{colorlinks=true, linkcolor=blue, filecolor=magenta, urlcolor=cyan,}
\urlstyle{same}
\usepackage{graphicx}
\usepackage[export]{adjustbox}
\graphicspath{ {./images/} }
\usepackage{polyglossia}
\usepackage{fontspec}
\setmainlanguage{english}
\setotherlanguages{romanian, hindi}
\newfontfamily\hindifont{Noto Serif Devanagari}
\newfontfamily\lgcfont{CMU Serif}
\setDefaultTransitions{\lgcfont}{}
\setTransitionsFor{Hindi}{\hindifont}{\lgcfont}

\title{Review }


\author{Alberico Talignani ${ }^{\mathrm{a}, 1}$, Raiyan Seede ${ }^{\mathrm{b}, 1}$, Austin Whitt ${ }^{\mathrm{b}}$, Shiqi Zheng ${ }^{\mathrm{a}}$, Jianchao $\mathrm{Ye}^{\mathrm{c}}$,\\
Ibrahim Karaman $^{\text {b,*, }}$, Michael M. Kirka ${ }^{\text {d,* }}$, Yutai Katoh ${ }^{\text {e, }}$, Y. Morris Wang ${ }^{\text {a,* }}$\\
a Department of Materials Science and Engineering, University of California, Los Angeles, CA 90049, USA\\
${ }^{\mathrm{b}}$ Department of Materials Science and Engineering, Texas A\&M University, College Station, TX 77843, USA\\
' Materials Science Division, Lawrence Livermore National Laboratory, Livermore, CA 94005, USA\\
${ }^{\mathrm{d}}$ Manufacturing Science Division, Oak Ridge National Laboratory, Oak Ridge, TN 37830, USA\\
${ }^{\mathrm{e}}$ Materials Science \& Technology Division, Oak Ridge National Laboratory, Oak Ridge, TN 37830, USA}
\date{}


%New command to display footnote whose markers will always be hidden
\let\svthefootnote\thefootnote
\newcommand\blfootnotetext[1]{%
  \let\thefootnote\relax\footnote{#1}%
  \addtocounter{footnote}{-1}%
  \let\thefootnote\svthefootnote%
}

%Overriding the \footnotetext command to hide the marker if its value is `0`
\let\svfootnotetext\footnotetext
\renewcommand\footnotetext[2][?]{%
  \if\relax#1\relax%
    \ifnum\value{footnote}=0\blfootnotetext{#2}\else\svfootnotetext{#2}\fi%
  \else%
    \if?#1\ifnum\value{footnote}=0\blfootnotetext{#2}\else\svfootnotetext{#2}\fi%
    \else\svfootnotetext[#1]{#2}\fi%
  \fi
}

\begin{document}
\maketitle
\section*{A review on additive manufacturing of refractory tungsten and tungsten alloys}


\section*{A R T I C L E I N F O}
\section*{Keywords:}
Tungsten

Tungsten alloys

Additive manufacturing

Laser powder-bed-fusion

Laser directed-energy-deposition

Electron beam powder-bed-fusion

\begin{abstract}
A B S T R A C T We review the progress of additive manufacturing effort on refractory metal tungsten and tungsten alloys. These materials are excellent candidates for a variety of high temperature applications but extremely challenging to fabricate via additive manufacturing due to a series of existing issues during the manufacturing. We outline these issues and discuss the current understanding and progress to tackle them. Laser powder-bed-fusion, laser directed-energy-deposition, and electron beam powder-bed-fusion are three common techniques that have been applied to additively manufacture pure tungsten. This overview discusses current observations and understanding on the issues associated with each of these techniques. We identify future research opportunities in additive manufacturing of refractory metals.
\end{abstract}

\section*{1. Introduction}
Due to their high density, excellent thermal conductivity and high temperature capabilities, high strength and hardness, and minimal sputtering yield and hydrogen interactions, refractory metal tungsten and tungsten alloys have a broad range of potential applications, including plasma facing components for fusion reactors [1,2], fusion targets [3], armor penetrators [4], and nuclear space power and propulsion [5]. For many of these applications, additive manufacturing (AM) offers unique geometrical design freedom and rapid prototyping capability, which is unparalleled by conventional manufacturing techniques. Moreover, AM offers additional potentials to fabricate functionally graded transitions from tungsten to various dissimilar materials. Due to its extremely high melting temperature (for pure tungsten, the melting temperature $\mathrm{T}_{m}=3422^{\circ} \mathrm{C}$ ) and brittle nature, tungsten is notoriously difficult to fabricate via either laser- or electron-beam-based AM techniques. Nevertheless, encouraging progress has been made in the past few years in AM tungsten and its alloys. Given the rising importance of refractory metals in various high temperature applications for harsh environments, this article aims at providing a timely overview of recent development in AM of refractory metals, in particular, tungsten and tungsten alloys.

To date, laser powder-bed-fusion (L-PBF) [sometimes also known as selective laser melting (SLM)], laser directed-energy-deposition (LDED), and electron beam powder-bed-fusion (EB-PBF) [or electron beam melting (EBM)] are the most common techniques to fabricate tungsten materials. The first two utilize laser energy to melt tungsten powder, and the latter electron energy. This review is arranged according to the materials made by the above three AM techniques, each of which has its own unique set of promises, challenges, and opportunities. As processing conditions determine the manufacturing defects, microstructure, and subsequent mechanical properties, Table 1 summarizes some key features and limitations of each AM technique, which help readers to better understand the microstructural origins of each type of materials and subsequent challenges involved in each approach. Notably, L-PBF offers substantially higher cooling rate and stronger temperature gradient compared to other two techniques, and thus may influence the residual stresses and cracking behavior. This review focuses on pure tungsten, as it is arguably one of the most challenging materials for AM. We contend, however, that many challenges and issues encountered in tungsten
\footnotetext{\begin{itemize}
  \item Corresponding authors.
\end{itemize}

E-mail addresses: \href{mailto:ikaraman@tamu.edu}{ikaraman@tamu.edu} (I. Karaman), \href{mailto:kirkamm@ornl.gov}{kirkamm@ornl.gov} (M.M. Kirka), \href{mailto:ymwang@ucla.edu}{ymwang@ucla.edu} (Y.M. Wang).

1 These authors contributed equally to this work.
}

Table 1

A summary of some key processing features [7] of three commonly used AM techniques for tungsten and tungsten alloys; i.e., laser powder-bed-fusion (L-PBF), laser directed-energy-deposition (L-DED), and electron beam powder-bed-fusion (EB-PBF).

\begin{center}
\begin{tabular}{llll}
\hline
 & L-PBF & L-DED & EB-PBF \\
\hline
Source power $(\mathrm{W})$ & $10^{2}-10^{3}$ & $10^{2}-10^{4}$ & $10^{2}-10^{3}$ \\
Beam size $(\mu \mathrm{m})$ & $30-200$ & $10^{2}-10^{3}$ & $10^{2}-10^{3}$ \\
Scanning speed $(\mathrm{mm} / \mathrm{s})$ & $10^{1}-10^{3}$ & $10-10^{2}$ & $10^{1}-10^{3}$ \\
Cooling rate $(\mathrm{K} / \mathrm{s})$ & $10^{5}-10^{7}$ & $10^{2}-10^{5}$ & $10^{3}-10^{4}$ \\
Temperature gradient (K/ & $10^{6}-10^{7}$ & $10^{4}-10^{6}$ & - \\
$\quad$ m) &  &  &  \\
Environment & Argon, & Argon & Vacuum, trace \\
 & nitrogen &  & helium \\
Material waste & High & Minimal & High \\
Pre-sintering & No & No & Yes \\
Spattering & Yes & No & No \\
\hline
\end{tabular}
\end{center}

manufacturing are likely applicable to a general class of refractory metals including high entropy alloys [6], which are prone to cracking during manufacturing. The review ends with our recommendations on the future opportunities for AM refractory metals, in the hope to spur future research in these interesting materials.

\section*{2. Laser powder-bed-fusion}
\subsection*{2.1. Method}
L-PBF is a well-known additive manufacturing technique for metals and alloys and sometimes ceramics. During this process, the powder is deposited layer-by-layer ( $\sim 20-150 \mu \mathrm{m}$ thick), and a laser beam (either continuous or pulsed wave) is applied to selectively melt the desired region. Some critical parameters that influence the build quality of materials include build layer thickness, laser power, scan speed, and hatch spacing. These parameters influence the volumetric energy density of processing conditions. Laser absorptivity is another important variable that determines the percentage of laser energy coupled into the powder layer. Notably, L-PBF is typically performed in an argon environment with oxygen levels ranging from tens to hundreds of ppm. Compared to L-DED and EB-PBF, the beam size of L-PBF is appreciably smaller, leading to a higher cooling rate and stronger temperature gradient, Table 1. Powder spattering and denudation phenomenon are common features of L-PBF processes, which cause processing defects/ pores that are difficult to eliminate $[8,9]$.

\subsection*{2.2. Cracking}
The ability of powerful lasers to melt essentially any types of metals makes L-PBF a natural choice to fabricate tungsten. However, cracking has been the biggest challenge in L-PBF W. None of the available literature has reported crack-free samples except when a femtosecond laser source was used [10]. As such, understanding the cracking mechanisms during L-PBF processes has been a central focus of recent studies. Generally speaking, two types of cracks have been observed in L-PBF W: longitudinal (with the crack direction parallel to the laser scanning direction) and branched or transverse cracks (with the crack inclined to the laser scan direction) [11-16].

The crack nucleation and propagation in L-PBF W are considered to be associated with the high ductile-to-brittle transition temperature (DBTT) $\left(\sim 200-400^{\circ} \mathrm{C}\right)$ of tungsten. The direct evidence supporting the above proposition is the appreciable time delay between the

\begin{center}
\includegraphics[max width=\textwidth]{2024_04_13_b812db9e8fc840c3b338g-02(4)}
\end{center}

(c) Frame $150, \mathrm{t}=3.00 \mathrm{~ms}$

\begin{center}
\includegraphics[max width=\textwidth]{2024_04_13_b812db9e8fc840c3b338g-02(3)}
\end{center}

Frame $329, \mathrm{t}=6.58 \mathrm{~ms}$

\begin{center}
\includegraphics[max width=\textwidth]{2024_04_13_b812db9e8fc840c3b338g-02(1)}
\end{center}

Frame $459, \mathrm{t}=9.18 \mathrm{~ms}$

\begin{center}
\includegraphics[max width=\textwidth]{2024_04_13_b812db9e8fc840c3b338g-02(2)}
\end{center}

Frame $638, \mathrm{t}=12.76 \mathrm{~ms}$

\begin{center}
\includegraphics[max width=\textwidth]{2024_04_13_b812db9e8fc840c3b338g-02}
\end{center}

Fig. 1. Scanning electron micrographs showing cracking behavior of tungsten bare plate during single-track experiments. (a) The laser power was set at $\mathrm{P}=300 \mathrm{~W}$ and speed $v=300 \mathrm{~mm} / \mathrm{s}$. Melt pool is marked with a dashed line. A branched crack is shown by the black arrow. (b) $\mathrm{P}=450 \mathrm{~W}$ and $\mathrm{v}=100 \mathrm{~mm} / \mathrm{s}$. The melt pool is well defined by the grains. Branched cracks are observed [12]. (c) Images taken from in-situ high speed camera experiments with $\mathrm{P}=300 \mathrm{~W}$ and $\mathrm{v}=100 \mathrm{~mm} / \mathrm{s}$. Cracks are highlighted inside red rectangles. The amount of time elapsed is shown in each image [15].\\
solidification and the appearance of cracks, as recorded by an in-situ high speed camera during single track experiments, Fig. 1 [12,15]. Another important observation is that cracks tend to propagate along high angle grain boundaries (HAGBs) [11-16], examples of which are shown in Fig. 2 [11]. This behavior can be attributed to the sensitivity of grain boundaries (GBs) in tungsten to impurities. Oxygen is a known impurity in tungsten powder and has been reported ranging from 30 to $370 \mathrm{ppm}[12,17]$. Several groups attributed the formation of cracks to aggregation of tungsten oxides during solidification [11,14,18,19]. However, cracks are not fully eliminated even when the oxygen level is very low [12], suggesting that impurities might not be the only factor influencing the cracking behavior. A systematic study of oxygen or other impurities such as hydrogen effects on the brittleness of L-PBF W remains missing. Tungsten powder size, shape, and distribution have also been reported to influence the cracking behavior of L-PBF materials [20]. However, a systematic study is needed in order to fully clarify this phenomenon.

To further understand the role played by residual stresses in cracking of L-PBF W, electron backscatter diffraction (EBSD) studies have been performed on cross-sections of printed tungsten samples [13, 16, 17, 19, 21-25]. A correlation between the density of HAGBs and cracks was observed. Although HAGBs are more prone to cracking than low-angle GBs (LAGBs), the formed cracks help to relieve intergranular stresses [13,19,23,24]. As shown in Fig. 3 [19], most cracks are observed along HAGBs, and perhaps even more importantly, the regions right next to the cracks have lower Kernel Average Misorientation (KAM) values compared to regions in which cracks are absent - evidence that most of the plastic deformation experienced by the material is concentrated near HAGBs [19]. Similar observations were made by another group, Fig. 4 [24], where cracks tend to appear near HAGBs (instead of LAGBs).

\subsection*{2.3. Strategies to mitigate cracks}
\subsection*{2.3.1. Alloying}
To suppress cracks in L-PBF W, incorporation of rare earth or other elements into pure tungsten has been explored. Researchers mixed pure tungsten powder with $1 \mathrm{wt} \%, 5 \mathrm{wt} \%$, and $10 \mathrm{wt} \%$ of Ta powder [23]. As shown in Fig. 5 [23], addition of $5 \mathrm{wt} \%$ Ta significantly decreased the grain size. However, no further grain refinement was observed with $10 \mathrm{wt} \% \mathrm{Ta}$. The refinement of grain size appears to reduce the cracks. The same approach was reported by another group [26,27], where alloying with Ta was found to reduce the average crack length per unit area by $30.7 \%$. One possible reason is that Ta has higher electron affinity to oxygen than $\mathrm{W}$ (the formation Gibbs free energy of $\mathrm{Ta}_{2} \mathrm{O}_{5}$ is $-1904 \mathrm{~kJ} / \mathrm{mol} \mathrm{vs}-761.5 \mathrm{~kJ} / \mathrm{mol}$ for $\mathrm{WO}_{3}$ ). Thus, Ta has the tendency of attracting oxygen and mitigating the impurity segregation along GBs. In a similar approach, $5 \mathrm{wt} \%$ of $\mathrm{Nb}$ was added to tungsten powder during L-PBF (the formation Gibbs free energies of $\mathrm{Nb}_{2} \mathrm{O}_{5}, \mathrm{NbO}_{2}$, and $\mathrm{NbO}$ are $-921 \mathrm{~kJ} / \mathrm{mol},-771 \mathrm{~kJ} / \mathrm{mol},-416 \mathrm{~kJ} / \mathrm{mol}$, respectively), and it was also found effective in suppressing cracks [24]. Although these alloying approaches achieved a certain degree of success, the underlying mechanisms have been poorly understood and require further investigations.

In a different study, $0.5 \mathrm{wt} \%$ of $\mathrm{ZrC}$ was added to tungsten. Grain refinement was observed and a reduction of crack density as high as $88.7 \%$ was achieved [25]. The beneficial effect of yttrium oxides $\left(\mathrm{Y}_{2} \mathrm{O}_{3}\right)$ has also been studied. In this case, no significant change of average grain size was noticed, crack reduction was still observed and attributed to tungsten grain shape changes [19]. A comparison between micro- and nano-sized $\mathrm{Y}_{2} \mathrm{O}_{3}$ was also carried out. The addition of nano- $\mathrm{Y}_{2} \mathrm{O}_{3}$ was found to reduce cracks due to the formation of a large fraction of LAGBs, whereas a reduction in hardness was seen with the addition of micro-sized $\mathrm{Y}_{2} \mathrm{O}_{3}$ [19]. In contrast, a separate single track experiment [28] found the addition of $\mathrm{Al}_{2} \mathrm{O}_{3}, \mathrm{Y}_{2} \mathrm{O}_{3}$, and $\mathrm{ZrO}_{2}$ to have no influence on suppressing cracks. The above results suggest that there is no consensus in the scientific community in terms of the choice of alloying element and other additives.

\subsection*{2.3.2. Remelting, scanning strategies, and substrate heating}
Other reported strategies to suppress cracks in L-PBF W include\\
\includegraphics[max width=\textwidth, center]{2024_04_13_b812db9e8fc840c3b338g-03}

Fig. 2. Electron backscatter diffraction (EBSD) images of two sides of a L-PBF W sample showing surface cracks. (a, c) As-printed W cross sections on different axes. Longitudinal, branched, and parallel build direction (BD) cracks are visible (red and blue arrows). (b, d) EBSD inverse pole figure (IPF) maps of (a, c). Scan tracks and scanning directions (SD) are visible, and 'ladder-shaped' grains and cracks along the grain boundaries are seen. The effect of rotation by 67 plus remelting between each layer is shown [11].\\
\includegraphics[max width=\textwidth, center]{2024_04_13_b812db9e8fc840c3b338g-04}

Fig. 3. Scanning electron micrographs (SEM), electron backscatter diffraction (EBSD), and Kernel average misorientation (KAM) data from the surface of a L-PBF W sample. Cracks are observed along high angle grain boundaries in L-PBF W. (a) SEM image of the L-PBF W sample. (b) EBSD inverse pole figure (IPF) map. (c) Image quality map. (d) KAM map [19].

scanning strategy adjustment, remelting, and substrate heating. These processes aim to alleviate or minimize the residual stresses formed in tungsten during printing, which are due to the high temperature gradients near the melt pool. Remelting plus rotating strategy has been studied in a series of tungsten builds [11]. It was found that rotating 67 between each layer randomized the grain orientation and shape, thus reducing the so-called "ladder-shaped" structure formed by grains without rotation (see Fig. 2). This process hindered the formation of cracks since the ladder-shaped grains provide crack-formation sites. While only one example is given, almost all works in the literature adopted the scan vector rotation strategy in between build layers most used 67 so as to minimize scan alignment in the same orientation, while others opted for either $45^{\circ}$ or $90^{\circ}[23,29]$. Remelting refers to the process of scanning a track more than once before recoating the sample with fresh powder. In conjunction with rotation strategies, remelting eliminated the columnar grains and helped to suppress longitudinal cracks [11]. As a result, the remelt sample was found to have smaller grain sizes and the average surface roughness was reduced [30]. Nonetheless, the combination of scan rotation and remelting was not sufficient to fully suppress cracks [11,30,31]. In addition, it was suggested that remelting may impact the density of the sample compared to a non-remelted reference [30]. This phenomenon is not well understood.

Substrate heating (up to $1000{ }^{\circ} \mathrm{C}$ ) is another strategy to suppress cracks in L-PBF W [29]. The purpose of substrate heating is two-fold: to reduce the temperature gradient (and thus residual stresses during L-PBF) and to bypass the DBTT of tungsten. Given that the DBTT is above room temperature $\left(200-400^{\circ} \mathrm{C}\right)$, embrittlement is likely to occur during solidification and cooling $[12,31,32]$. Theoretically, if the substrate is preheated above the DBTT, the screw dislocations in tungsten will have enough mobility and accommodate the plastic strain induced by the temperature gradient during melting and subsequent cooling [29]. Another important aspect is that maintaining an impurity-free environment during printing is crucial, as the DBTT of tungsten can be theoretically shifted by $200{ }^{\circ} \mathrm{C}$ between $10 \mathrm{ppm}$ and $50 \mathrm{ppm}$ of oxygen content [12]. Despite tremendous effort [29], substrate heating between 80 and $1000^{\circ} \mathrm{C}$ has not been enough to fully eliminate cracking. A detailed study of how to optimize the preheating conditions (e.g., temperature and cooling rate) is needed in the future in order to fully understand their influences on cracking behavior.\\
\includegraphics[max width=\textwidth, center]{2024_04_13_b812db9e8fc840c3b338g-05(1)}

Fig. 4. Electron microscopy was used to characterize surface cracks in Nb-alloyed W. (a) Scanning electron microscope (SEM) image of W-5 wt\% Nb alloy. Cracks are present and pointed by arrows. (b) Electron backscatter diffraction (EBSD) inverse pole figure (IPF) map of image (a), cracks are again indicated with arrows. (c) Inverse pole figure (IPF) grain boundary (GB) distribution map. Red lines represent high angle GBs. Cracks appear only on red lines. (d) GB distribution plot. Those images are adopted from [24].\\
\includegraphics[max width=\textwidth, center]{2024_04_13_b812db9e8fc840c3b338g-05}

Fig. 5. Electron backscatter diffraction (EBSD) images of three Ta-alloyed L-PBF W samples. From left to right, normal direction inverse pole figure (IPF) maps of cross-sections from pure W, W-5 wt\%Ta, and W-10 wt\% are shown. Grains become more refined between (a) and (b), while a smaller change is noticed in (c) [23].

\subsection*{2.4. Processing parameter windows}
Aside from cracks, achieving a high relative density is crucial in LPBF W. As summarized in Table $2[7,9,10,12-19,22-26,27]$, the relative density of tungsten obtained in the literature ranges from $\sim 80 \%$ to $\sim 99 \%$ [14]. To obtain good processing parameters for high density samples, laser power versus scan speed graphs have been used $[21,32$, 33]. In such a graph, different processing zones are marked according to a combination of laser power and speed, such as irregular crack region, regular crack region, balling region, warped region, and dense region $[21,32,33]$. Similar graphs have also been adopted to study cracking behavior in single-track experiments [12,32,33]. However, it has been demonstrated in the literature that such graphs are only useful for a specific type of L-PBF machine and with a fixed build layer thickness. An energy input diagram or normalized enthalpy diagram would be more useful to identify processing windows for tungsten and offer better

Table 2

A summary of laser processing parameters for L-PBF W and resultant mechanical properties in the literature.

\begin{center}
\begin{tabular}{|c|c|c|c|c|c|c|c|c|c|c|c|}
\hline
Ref. & Machine & \begin{tabular}{l}
Laser \\
Power \\
(W) \\
\end{tabular} & \begin{tabular}{l}
Energy \\
Density $^{1}$ \\
$\left(\mathrm{~J} / \mathbf{m m}^{3}\right)$ \\
\end{tabular} & \begin{tabular}{l}
Beam/ \\
Hatch \\
$(\mu m)$ \\
\end{tabular} & \begin{tabular}{l}
Layer \\
Thick. \\
$(\mu m)$ \\
\end{tabular} & \begin{tabular}{l}
Max Density \\
$(\%)$ \\
\end{tabular} & \begin{tabular}{l}
Hardness $^{4}$ \\
$(\mathrm{GPa})$ \\
\end{tabular} & \begin{tabular}{l}
Preheating \\
$\left({ }^{\circ} \mathrm{C}\right)$ \\
\end{tabular} & \begin{tabular}{l}
Powder \\
Size $(\mu m)$ \\
\end{tabular} & Alloying & Cracking \\
\hline
[31] & \begin{tabular}{l}
Renishaw \\
AM250 \\
\end{tabular} & \begin{tabular}{l}
200 \\
(pulsed) \\
\end{tabular} & - & $75 / 90$ & - & 82.90 & - & No & 19.4 & No & Yes \\
\hline
[23] & \begin{tabular}{l}
Custom \\
Built, DMP \\
320 \\
\end{tabular} & 300 & $150-900$ & 90 & - & $\sim 81-98.7$ & - & $0-400$ & - & \begin{tabular}{l}
No and \\
Yes (Ta) \\
\end{tabular} & \begin{tabular}{l}
Yes. Cracks are less \\
evident with alloying. \\
Both transverse and \\
longitudinal cracks are \\
observed. \\
\end{tabular} \\
\hline
$[21]$ & \begin{tabular}{l}
Renishaw \\
AM400 \\
\end{tabular} & \begin{tabular}{l}
$150-400$ \\
(pulsed) \\
\end{tabular} & $88-1185$ & \begin{tabular}{l}
$75 /$ \\
$75-150$ \\
\end{tabular} & 30 & $\sim 80-96$ & $\sim 2-3.79$ & No & 28 & No & Yes \\
\hline
$[11]$ & \begin{tabular}{l}
Renishaw \\
AM400 \\
\end{tabular} & \begin{tabular}{l}
400 \\
(pulsed) \\
\end{tabular} & 474 & $75 / 100$ & 30 & $92.5-96.5$ & - & No & 28 & No & \begin{tabular}{l}
Yes, cracks are longer \\
than $1 \mathrm{~mm}$ (along \\
HAGBs). \\
Transverse and \\
longitudinal cracks are \\
observed. \\
\end{tabular} \\
\hline
[13] & \begin{tabular}{l}
Renishaw \\
125 \\
\end{tabular} & \begin{tabular}{l}
200 \\
(pulsed) \\
\end{tabular} & $641-930$ & \begin{tabular}{l}
$43 /$ \\
$115-155$ \\
\end{tabular} & 50 & $94-98$ & - & No & $\sim 47$ & No & Yes \\
\hline
[19] & \begin{tabular}{l}
Renishaw \\
AM400 \\
\end{tabular} & \begin{tabular}{l}
250 \\
(pulsed) \\
\end{tabular} & 544-1587 & \begin{tabular}{l}
$70 /$ \\
$50-100$ \\
\end{tabular} & $20-35$ & $94.5-98.30$ & $\sim 3.63-4.21$ & 180 & \begin{tabular}{l}
$15-45 \mathrm{~W}$ \\
$15-53$ \\
$\mathrm{Y}_{2} \mathrm{O}_{3}$ \\
\end{tabular} & \begin{tabular}{l}
No and \\
Yes \\
$\left(\mathrm{Y}_{2} \mathrm{O}_{3}\right)$ \\
\end{tabular} & \begin{tabular}{l}
Yes, hundreds of \\
microns and in all \\
directions. \\
Oxides reduce \\
cracking. \\
\end{tabular} \\
\hline
$[30]$ & EOS M290 & $150-350$ & $\sim 94-875$ & 100 & 20 & 98.40 & $\sim 4.02-4.47$ & 180 & 15.8 & No & \begin{tabular}{l}
Yes, fewer cracks in \\
the bulk. \\
Remelting improves \\
cracking. \\
\end{tabular} \\
\hline
[17] & EOS M290 & $200-370$ & $250-1850$ & 50 & 20 & $97.72-98.50$ & $\sim 4.36-4.58$ & 50 & 16.24 & No & Yes \\
\hline
$[22]$ & \begin{tabular}{l}
SLM® \\
Solution \\
$125 \mathrm{HL}$ \\
\end{tabular} & $200-400$ & 198-905 & 70/105 & 30 & 98.51 & - & 200 & $5-25$ & No & Yes \\
\hline
[14] & \begin{tabular}{l}
EOSM100 \\
DMLS \\
\end{tabular} & $100-170$ & 125-1062 & \begin{tabular}{l}
$40 /$ \\
$40-70$ \\
\end{tabular} & 20 & 99.61 & $\sim 4.12$ & 80 & $10-25$ & No & \begin{tabular}{l}
Yes. Longitudinal: \\
straight, $30-100 \mu \mathrm{m}$. \\
Transverse: shorter \\
and S shaped along \\
GBs. \\
\end{tabular} \\
\hline
[18] & \begin{tabular}{l}
Custom \\
Built \\
\end{tabular} & $200-350$ & $500-1167$ & 50 & 20 & $87.8-89.4$ & $\sim 4.65$ & 200 & 14.41 & No & Yes \\
\hline
$[24]$ & EOS M280 & $250-370$ & - & $70-110$ & 30 & $93.3-98.0$ & $6.69-10.31$ & 200 & \begin{tabular}{l}
$5-25 \mathrm{~W}$ \\
$1-10 \mathrm{Nb}$ \\
\end{tabular} & Yes, Nb & \begin{tabular}{l}
Yes, along HAGBs. Nb \\
alloying partially \\
suppresses \\
cracks. \\
\end{tabular} \\
\hline
[29] & \begin{tabular}{l}
Aconity 3D \\
GmbH \\
\end{tabular} & $375-400$ & $196-446$ & $100 / 80$ & 40 & $\sim 94.7-98.5$ & - & $600-1000$ & $15-45$ & No & \begin{tabular}{l}
Yes (reduced at \\
$1000^{\circ} \mathrm{C}$ ). \\
\end{tabular} \\
\hline
$[26]$ & SLM 280 & 400 & - & 100 & - & - & - & - & \begin{tabular}{l}
$32 \mathrm{~W}$, \\
$18 \mathrm{Ta}$ \\
\end{tabular} & Yes, Ta & \begin{tabular}{l}
Crack density reduced \\
by alloying. \\
\end{tabular} \\
\hline
$[27]$ & \begin{tabular}{l}
Renishaw \\
AM400 \\
\end{tabular} & \begin{tabular}{l}
400 \\
(pulsed) \\
\end{tabular} & - & 100 & - & - & - & - & - & Yes, Ta & \begin{tabular}{l}
Yes, less with Ta \\
(along GBs). \\
\end{tabular} \\
\hline
$[16]$ & SLM 125HL & 400 & 238-1667 & \begin{tabular}{l}
$80 /$ \\
$100-120$ \\
\end{tabular} & 30 & $\sim 90-97$ & - & $\mathrm{HIP}^{3}$ & 32 & No & Yes, reduced with HIP. \\
\hline
\end{tabular}
\end{center}

${ }^{1}$ Energy Density $=\frac{P}{l \times v \times D_{\text {beam }}}$, where $\mathrm{P}$ is power, 1 is layer thickness, $\mathrm{D}$ is beam diameter and $\mathrm{v}$ is laser speed. Speed is calculated using point distance and exposure time for pseudo-pulsed laser machines.

2 These values are not calculated according to the equation in note 1 but are reported as found in the respective papers.

${ }^{3}$ HIP: Hot Isotactic Pressing. Note that HIP is not preheating. ${ }^{4}$ A compressive strength in the range of $900-1523$ MPa has been reported for L-PBF W [17,18,22,30,32].

machine-to-machine variation comparison [34-38]. Regretfully, such laser processing maps for pure tungsten do not exist yet.

\subsection*{2.5. Mechanical properties}
Due to the poor sample quality, few studies have been conducted on documenting the mechanical properties of L-PBF W. The available data are limited to hardness and compressive mechanical properties, Table 2. Depending on the processing parameters, a hardness value of 3.63-4.47 GPa [14, 17-19, 30, 32] was reported for L-PBF W, which was superior to samples made by conventional powder metallurgy and spark plasma sintering (the hardness values range between 3.14-3.92 GPa; i.e., $320-400 \mathrm{HV}$ [17]). Alloying with $5 \mathrm{wt} \% \mathrm{Nb}$ was found to elevate the hardness from $6.69 \mathrm{GPa}$ to $8.01 \mathrm{GPa}$ [24]. Similarly, the addition of nano-yttrium oxides increased the hardness to $\sim 4.51 \mathrm{GPa}(\sim 460 \mathrm{HV})$. The effect was attributed to dispersion strengthening. In the same study, it was also shown that introducing micro-yttrium oxides (instead of nano-sized oxides) resulted in lower hardness than that of conventionally manufactured tungsten. The reduction in hardness was rationalized by the agglomeration of micro-sized yttrium oxides, which weakened the material [19]. In terms of compressive properties, a wide range of compressive strength (900-1523 MPa) was reported [17,18,30,32,33], whereas no strength/ductility data were found in tension. With the ubiquitous existence of cracks, it is not surprising to see rather poor mechanical property data on L-PBF W.

\section*{3. Laser directed-energy-deposition}
\subsection*{3.1. Method}
L-DED is an AM technique in which metal powder is fed into a melt pool created by a laser. After the first layer is deposited the powder feeder moves upward and deposition of the second layer begins. L-DED is usually conducted under an argon atmosphere that utilizes an argon blower. L-DED is suitable for the manufacture of large parts relatively quickly and offers excellent design freedom due to the additional parameters involved in the process. Some unique features of L-DED are powder feed rate and the option to change input powder composition using multiple hoppers to manufacture composite materials or grade composition throughout AM parts. L-DED can also be used to repair parts due to its ability to accurately deposit material anywhere in the build chamber. These features allow L-DED to process functionally graded structures, which can mitigate the challenges associated with joining dissimilar materials such as $\mathrm{W}$ and ferritic-martensitic steels. Fig. 6 displays a schematic of the L-DED process [39] showing the laser power source, powder feeder, and build platform.

In this section, we will assess the current state-of-the-art in L-DED W and tungsten alloys and the effects of L-DED processing on their structure and properties. To understand these processes and how to successfully implement them, it is critical to recognize the processing challenges and defects associated with L-DED of tungsten and its alloys, and possible mitigation strategies.

\subsection*{3.2. L-DED tungsten and tungsten alloys}
\subsection*{3.2.1. Deposition of pure tungsten}
Several studies have demonstrated that tungsten can be printed with moderate success utilizing L-DED. However, these studies also reported difficulties in fully melting tungsten powder during deposition. Polygonal tungsten powder was printed on a reduced activation ferritic/ martensitic (RAFM) steel substrate in [40]. Single tracks were printed using the processing parameter combinations in the range of $\mathrm{P}=2000-$ $4000 \mathrm{~W}$ and $\mathrm{v}=200-600 \mathrm{~mm} / \mathrm{min}$ at a constant powder feed rate $(\dot{\mathrm{m}}=$ $29.3 \mathrm{~g} / \mathrm{min}$ ). Significant compositional mixing was observed between the steel substrate and powder, with tungsten content ranging from 12 to $55 \mathrm{wt} \%$ at various locations within the melt pools. Fig. 7 [40] displays single track melt pool cross sections at various parameter sets. Unmelted tungsten particles can be observed throughout the melt pools at each of the parameter sets displayed in Fig. 7. These unmelted tungsten particles were also observed in single laser clads of tungsten and tungsten-nickel alloys printed on a mild steel substrate studied in [41]. Mixing between substrate material and the deposited tungsten reportedly increased with laser power, which corresponded to a decrease in overall tungsten content and hardness of the single tracks [40]. This is due to increased laser penetration into the steel substrate, increasing the relative amounts of $\mathrm{Fe}$ and $\mathrm{Cr}$ in the melt pool.

Microstructural evaluation of multi-layer tungsten prints was also conducted [40]. Intermetallic precipitates were observed in the scanning electron micrograph (SEM) in Fig. 8a and were identified to be $\mathrm{Fe}_{7} \mathrm{~W}_{6}$ from transmission electron microscopy (TEM) analysis (Fig. 8b and c). Energy dispersive spectroscopy (EDS) analysis of the precipitates was consistent with the TEM observations of $\mathrm{Fe}_{7} \mathrm{~W}_{6}$. X-ray diffraction peaks of single- and double-layer prints on the RAFM steel substrate (Fig. 8e) identified the existence of W, $\mathrm{Fe}, \mathrm{Fe}_{7} \mathrm{~W}_{6}$, and $\mathrm{Fe}-\mathrm{Cr}$ phases. In a 9-layer print conducted at $3000 \mathrm{~W}$ and $3000 \mathrm{~mm} / \mathrm{min}$, the authors observed a significant compositional gradient from $25.23 \mathrm{wt} \% \mathrm{Fe}$, $2.73 \mathrm{wt} \% \mathrm{Cr}$, and $72.04 \mathrm{wt} \% \mathrm{~W}$ near the substrate to $3.91 \mathrm{wt} \% \mathrm{Fe}$, $0.28 \mathrm{wt} \% \mathrm{Cr}$, and $95.82 \mathrm{wt} \% \mathrm{~W}$ at the top of the build, as can be seen in Fig. 9 [40]. These observations show that although tungsten content increases with each additional layer, $\mathrm{Fe}$ and $\mathrm{Cr}$ from the substrate continue to diffuse into the upper layers of the matrix after 9 deposited layers of tungsten. SEM analysis of this multi-layer build indicates the existence of unmelted W-rich particles, dendrite structures, and microcracking within the tungsten particles. Overall, this study highlighted the need for designing a compositionally graded transition from steels to tungsten to avoid producing the undesirable intermetallic phases.

Cracking and porosity were observed within tungsten single tracks, single layers, and multi-layer deposits as can be seen in Fig. 9j \& k [40]. Cracks were speculated to be due to liquation cracking and residual stresses from the rapid heating and cooling cycles. Single layer cracking occurred at the top of the deposit and propagated toward the substrate causing $\mathrm{W}$ particles along the crack path to internally fracture and others to de-bond with the matrix. Spherical porosity also occurred within single layers of deposited tungsten, likely due to gas trapped within the melt pool during solidification. Pores are a typical defect observed in materials manufactured via L-DED and can be mitigated by manipulating process parameters [42]. Multi-layer tungsten deposits contained tungsten particles with a large number of microcracks near the top of the deposits, as can be seen in Fig. 9d \& g [40]. These microcracks did not appear as frequently in the lower layers of the build. This may be due to the higher degree of remelting experienced at the bottom of the sample, relative to the top.

Many commercially available L-DED machines are not capable of achieving the large laser powers utilized in reference [40]. A study [43] attempted to print $12 \mathrm{~mm}$ high thin-walled vertical tungsten tubes using L-DED using polyhedral $99.7 \%$ purity tungsten powder at lower values of laser power. They printed 35 samples with parameters ranging between $\mathrm{P}=600-1000 \mathrm{~W}, \mathrm{v}=50-350 \mathrm{~mm} / \mathrm{min}, \dot{\mathrm{m}}=5-25 \mathrm{~g} / \mathrm{min}$, and a laser diameter $=750 \mu \mathrm{m}$ to determine optimal parameters for printing the material. With a fixed layer thickness of $100 \mu \mathrm{m}$ and a constant but undisclosed hatch spacing, they reported 7 of the 35 parameter sets reached the targeted build height with the rest of the specimens either

\begin{center}
\includegraphics[max width=\textwidth]{2024_04_13_b812db9e8fc840c3b338g-07}
\end{center}

Fig. 6. A schematic of the directed-energy-deposition process showing the laser source, powder feeder or sprayer, and the build platform [39].

\begin{center}
\includegraphics[max width=\textwidth]{2024_04_13_b812db9e8fc840c3b338g-08}
\end{center}

Fig. 7. Single tracks fabricated at various parameter sets (a-f), displaying different melt pool dimensions as well as varying levels of unmelted W particles [40].

exceeding or falling short of the target range, as can be seen in Fig. 10 [43]. However, no analysis of microstructural homogeneity or degree of melting the tungsten particles was conducted in the study.

\subsection*{3.2.2. Deposition of tungsten alloys}
3.2.2.1. Tungsten-nickel deposition. Due to the various applications of tungsten alloys, many studies attempted to print tungsten alloys. One group successfully printed a $60 \mathrm{~W}-40 \mathrm{Ni}$ collimation component (Fig. 11a) using L-DED with the following parameters [41]: $2000 \mathrm{~W}$ laser power, $300 \mathrm{~mm} / \mathrm{min}$ scan speed, and $8 \mathrm{~g} / \mathrm{min}$ feed rate. Single laser clads of $60 \mathrm{~W}-40 \mathrm{Ni}$ printed on a mild steel substrate were observed to contain unmelted tungsten particles and dendritic structures that were speculated to be $\mathrm{Ni}_{4} \mathrm{~W}$ and $\mathrm{NiW}_{2}$ intermetallic phases.

Another group similarly observed unmelted tungsten particles in a LDED W-15Ni alloy along with a $\gamma$-Ni phase containing $15 \mathrm{wt} \% \mathrm{~W}$ [44]. They reported a layered microstructure with unmelted tungsten particles dominating regions of the deposit that were only subjected to initial melting, and $\mathrm{W}$ dendrite structures in regions that were subjected to remelting by the subsequent layer (Fig. 11d and e). These W-15Ni specimens had tensile strengths of $\sim 500 \mathrm{MPa}$ and were prone to brittle fracture ( $\sim 3 \%$ strain to fracture) at room temperature.

3.2.2.2. Tungsten-nickel-iron deposition. Unmelted W-rich particles embedded in an FCC Ni-Fe matrix were observed in L-DED manufactured $90 \mathrm{~W}-7 \mathrm{Ni}-3 \mathrm{Fe}[45,46]$, similar to what was observed in $\mathrm{W}$ and W-Ni deposits. An alternating microstructure with layers dense in unmelted tungsten and partially melted tungsten particles was also observed [45,46], similar to those observations in W-15Ni, Fig. 11 [41, 44]. Tensile testing of $90 \mathrm{~W}-7 \mathrm{Ni}-3 \mathrm{Fe}$ samples resulted in high strength (1037 MPa) and low ductility (3.5\% elongation). These as-printed materials displayed significantly higher ultimate tensile strength (UTS) and lower ductility than those of traditionally manufactured $90 \mathrm{~W}-7 \mathrm{Ni}-3 \mathrm{Fe}$ via liquid phase sintering (LPS), Fig. 12a [45]. Additionally, large periodic variations in microhardness ( $\sim 76 \mathrm{HV}$ ) were observed, Fig. 12b. These variations are attributed to a periodic sublayer change where $\mathrm{W}$-particle dense regions are observed above and below the regions with lower relative amounts of W particles. L-DED materials have an average hardness of $\sim 415 \mathrm{HV}$, higher than specimens made by LPS. This is due to higher amounts of hard W-particles embedded in the matrix in L-DED specimens. Fig. 13 displays fracture surfaces of $90 \mathrm{~W}-7 \mathrm{Ni}-3 \mathrm{Fe}$ where large pores were observed, indicating that the porosity acts as fracture initiation sites [45]. Lower ductility during tensile testing in the as-deposited specimens compared to traditional LPS is attributed to the residual porosity observed at these fracture surfaces. Tungsten particle cleavage and tearing of the ductile Ni-Fe matrix were also observed at the fracture surfaces.

\subsection*{3.3. Challenges in L-DED tungsten}
Common challenges associated with L-DED $\mathrm{W}$ on ReducedActivation-Ferritic-Martensitic (RAFM) steels are illustrated in Fig. 14. There are many critical parameters that affect melt pool morphology during L-DED, including laser power (P), scan speed (v), hatch spacing (h), laser focus (f), substrate temperature, powder size distribution and morphology, and powder feed rate ( $\dot{m}$ ). These variables are critical in

\begin{center}
\includegraphics[max width=\textwidth]{2024_04_13_b812db9e8fc840c3b338g-09(1)}
\end{center}

(e)

\begin{center}
\includegraphics[max width=\textwidth]{2024_04_13_b812db9e8fc840c3b338g-09}
\end{center}

Fig. 8. Phase analysis of pure W printed on a RAFM steel substrate using (a) scanning electron microscopy (SEM), (b, c) transmission electron microscopy (TEM), (d) energy dispersive spectroscopy (EDS), and (e) x-ray diffraction (XRD) [40].

achieving successful prints in L-DED. Only a limited number of these variables have been explored in L-DED W and tungsten alloys. Challenges such as attaining targeted build heights and mitigating porosity can be resolved by optimizing parameters such as $\mathrm{P}, \mathrm{v}, \mathrm{h}$, and $\dot{\mathrm{m}}$ as well as improving feedstock quality [42, 47-49]. Additionally, residual stress-induced cracking has been shown to be mitigated by substrate preheating [50-53] which, to our knowledge, has not been attempted on L-DED W. Substrate preheating may also result in increased melting of tungsten particles that would remain unmelted if printed on a room temperature substrate. Mitigation of intermetallic particle formation may be achieved by introducing filler alloys between the steel base plate and tungsten, circumventing regions in the alloys' phase diagrams in which detrimental phases are stable. These strategies may improve the feasibility of additively manufacturing tungsten via directed energy deposition.

\section*{4. Electron beam melting}
\subsection*{4.1. Method}
The EBM or EB-PBF process belongs to the powder bed family of additive manufacturing technologies. Similar to L-PBF, the heat source is selectively moved across the powder bed to melt the regions of interest in a layer-by-layer process. Electron beams, compared to lasers, are high in energy density exceeding several kilowatts focused into spot sizes of several hundred microns in diameter when melting. EB-PBF occurs under controlled vacuum conditions to both maintain the quality of the electron beam spot size and offset fluctuation in pressure associated with

\begin{center}
\includegraphics[max width=\textwidth]{2024_04_13_b812db9e8fc840c3b338g-10}
\end{center}

Fig. 9. Scanning electron microscope (SEM) and energy dispersive spectroscopy (EDS) analysis of a 9-layer W sample printed at $3000 \mathrm{~W}$ and $3000 \mathrm{~mm} / \mathrm{min}$. (a, b) Low magnification cross-sectional SEM images of the specimen, indicating where EDS analysis was conducted. (c) The results of EDS analysis conducted on the areas displayed in (b). (d-f) Higher magnification images of the top, middle, and bottom of the specimen, respectively. (g-i) High magnification images displaying W particles, dendrite structures, and microcracks [40]. (j, k) Low and high magnification images displaying cracks and porosity in W deposited on RAFM steel [40].

vaporization of the metal in the liquid state as the beam is melting. Additionally, in EB-PBF the powder bed is heated to elevated temperatures through defocusing the electron beam and rapidly rastering it across the powder bed surface to allow for the powder particles to loosely sinter to one another and conduct the negative charge of the electron beam away. If the negative electrical charge of the imparted electrons is not conducted away, the powder bed will build-up a negative charge which results in the repulsion of the powder particles from one another, i.e., a "smoking" event. In the instance of tungsten, the powder bed is heated to between 1000 and $1400{ }^{\circ} \mathrm{C}$ [54]. As a result of heating the powder bed, materials processed through EB-PBF often have lower levels of residual stress than corresponding materials processed through L-PBF [54]. One of the advantages of the EB-PBF process over that of L-PBF is the ability to rapidly manipulate the electron beam heat source over the entirety of the build area to locally control thermal conditions of the material. This has been shown as beneficial for controlling the microstructure [55] as well as stress states of the material to suppress defects such as cracks in non-weldable materials [56].\\
(a)

\begin{center}
\includegraphics[max width=\textwidth]{2024_04_13_b812db9e8fc840c3b338g-11(2)}
\end{center}

$10 \mathrm{~mm}$

(b)

\section*{Proper build domain}
\begin{center}
\includegraphics[max width=\textwidth]{2024_04_13_b812db9e8fc840c3b338g-11(1)}
\end{center}

$10 \mathrm{~mm}$\\
Over build domain

\begin{center}
\includegraphics[max width=\textwidth]{2024_04_13_b812db9e8fc840c3b338g-11}
\end{center}

$10 \mathrm{~mm}$

(c)\\
\includegraphics[max width=\textwidth, center]{2024_04_13_b812db9e8fc840c3b338g-11(3)}

Fig. 10. Tungsten specimen fabricated at various parameter sets displaying heights that are either below, matching, or exceeding the targeted build height [43]. (a) Images of specimens that are under built, over built, or matching the targeted height. (b) The measured height increment for each layer in samples printed at different speeds plotted against laser power. (c) Cubes printed at various parameters in the under build, over build, or target printing regimes.

\subsection*{4.2. Porosity}
Various levels of success in the processing of tungsten through EBM (EBM W) have been reported. Key to the processing of tungsten is the ability to obtain materials that approach the theoretical density of pure tungsten at $19.3 \mathrm{~g} / \mathrm{cc}$. In literature, four states for EBM $W$ based on density and porosity of the material have been identified, as depicted in Fig. 15 [57].

In the figure, volumetric energy densities ranging from 208 to $3840 \mathrm{~J} / \mathrm{mm}^{3}$ with a substrate temperature of approximately $850{ }^{\circ} \mathrm{C}$ were used. These states are (a) limited fusion, (b) insufficient fusion, (c) proper fusion, and (d) excessive fusion. Limited fusion is characterized by resultant relative densities of $<70 \%$ with excessive balling of the tungsten observed within the melt layers. Insufficient fusion exhibits relative densities between $70 \%$ and $90 \%$, however, significant interconnected porosity exists within the material including chimney porosity [58]. Proper fusion is the optimal processing state where densities greater than $90 \%$ are achievable and interconnected porosity is mitigated through full melting and wetting of the tungsten. The last state, excessive fusion, can be defined as fully dense material that exhibits swelling due to too much energy being imparted into the material. Similar trends were also identified in work that varied the linear energy used to melt tungsten from 333 to $5000 \mathrm{~J} / \mathrm{m}$ with a powder bed temperature of $1000{ }^{\circ} \mathrm{C}$ [59].

Various levels of porosity have been reported in the literature, with many studies achieving success for high density tungsten. Densities as high as $\mathbf{9 9 . 5 \%}$ have been measured; nevertheless, microcracking was found $[57,59]$. SLM and EBM W have also been compared and it was found that comparable densities could be obtained using either technique, although the build temperature greatly influenced the porosity level and defect levels [60]. Lastly, nondestructive techniques such as in-situ near-infrared (IR) defect detection have been used to report defect-free, highly dense samples (>99\%) [61,62].

\subsection*{4.3. Cracking}
The occurrence of cracking in EBM $\mathrm{W}$ is a problem akin to that observed in SLM processed tungsten with the debate ongoing for the specific mechanism(s) by which tungsten cracks during processing. It has been suggested that cracks occur in the solid state as a result of significant inelastic deformation along grains neighboring GBs [61]. This was supported through electron backscatter diffraction (EBSD) analysis of the areas surrounding cracks that revealed localized orientation gradients near the edges of cracks. Representative EBSD micrographs showing the cracking in tungsten are shown in Fig. 16 [61]. This is consistent with other studies that observed the cracking phenomena through high-speed in-situ videos of the SLM process with a heated powder bed temperature range above and below the DBTT range of tungsten. This was attributed to the development of significant von Mises stresses when the tungsten cycled below DBTT [12]. It was theorized that thermal stress generated from thermal gradients during SLM processing of tungsten can only be compensated by crack formation along low-strength GBs, particularly those with impurities [19,63].

From the reported studies, the influence of the build substrate temperature and the overall build temperature has a clear influence on the cracking in AM W. A report [29] utilized a SLM system with substrate preheating and showed that increasing build temperature from $200{ }^{\circ} \mathrm{C}$ to $1000^{\circ} \mathrm{C}$ significantly reduced cracking in AM tungsten, though it did\\
\includegraphics[max width=\textwidth, center]{2024_04_13_b812db9e8fc840c3b338g-12(2)}

Fig. 11. Laser directed-energy-deposition (LDED) of W-Ni alloys. (a) W60-Ni40 collimation component printed at $2000 \mathrm{~W}$ laser power, $300 \mathrm{~mm} / \mathrm{min}$ scan speed, and $8 \mathrm{~g} / \mathrm{min}$ feed rate [41]. (b, c) Low and high magnification SEM micrographs showing the microstructure of a single laser clad of W60-Ni40 printed on a mild steel substrate [41]. (d) Scanning electron micrographs of a W-15Ni L-DED part showing a layered microstructure containing unmelted $\mathrm{W}$ particles in regions subjected to initial laser melting [44], and (e) dendritic W structures in remelted regions of the deposit [44]. (a)

\begin{center}
\includegraphics[max width=\textwidth]{2024_04_13_b812db9e8fc840c3b338g-12}
\end{center}

(b)

\begin{center}
\includegraphics[max width=\textwidth]{2024_04_13_b812db9e8fc840c3b338g-12(1)}
\end{center}

Fig. 12. Mechanical properties of additively manufactured W-Ni-Fe alloys. (a) Engineering stress-strain curves for $90 \mathrm{~W}-7 \mathrm{Ni}-3 \mathrm{Fe}$ specimens manufactured by laser metal deposition (LMD, alternatively directed-energy-deposition) and liquid phase sintering (LPS). (b) Hardness values along the build direction for 90W-7Ni-3Fe specimens manufactured by LMD and LPS [45].

not eliminate cracking entirely. Multiple studies explored the role of substrate heating: samples have been built with a powder bed temperature of approximately $850{ }^{\circ} \mathrm{C}$, and observed minor levels of cracking [57]. In a comparison between SLM and EBM, significant cracking was found in SLM, with no cracking in EBM. The lack of cracking in EBM W was attributed to a combination of build plate temperature of $1000{ }^{\circ} \mathrm{C}$ and addition of a support structure to raise the tungsten samples off the build plate $[19,60]$. Theoretically, if the substrate is heated above the\\
\includegraphics[max width=\textwidth, center]{2024_04_13_b812db9e8fc840c3b338g-13(1)}

Fig. 13. Tensile fracture surfaces of $90 \mathrm{~W}-7 \mathrm{Ni}-3 \mathrm{Fe}$ fabricated via laser directed-energy-deposition (L-DED). Features such as (a) porosity, (b-d) W particle cleavage, and (d) matrix failure are indicated with white arrows [45].

\begin{center}
\includegraphics[max width=\textwidth]{2024_04_13_b812db9e8fc840c3b338g-13}
\end{center}

Fig. 14. Graphical representation of challenges typically encountered during directed energy deposition of tungsten on RAFM steel. These challenges include achieving targeted build heights, cracking and porosity, difficulty of melting W particles, formation of intermetallics and dendrites, and layered melted/remelted structures within the build.

DBTT of tungsten, dislocations should be significantly more mobile, thus reducing the chance of cracking. Higher substrate temperatures allow for lower temperature gradients during cooling, reducing the stresses generated on the tungsten components. Similarly, by raising the printed components of the build plate by using a support structure, heat is not allowed to dissipate quickly, which in turn reduces the stresses on the components. The use of different metals, such as steel and titanium, as build substrates was also investigated [61]. Titanium build plates have been used due to the high degree of solubility the elements have in one another in an effort to create a metallurgical bond at the interface of part and build plate. Cracking in tungsten was observed to be sensitive to the build preheat temperature. Crack density was drastically lower when build surface temperatures of $1500{ }^{\circ} \mathrm{C}$ were used, compared to $1100{ }^{\circ} \mathrm{C}$.\\
Additional studies leveraged an ever-higher surface preheat temperature of $1800^{\circ} \mathrm{C}$ to demonstrate the ability to successfully suppress crack formation in EBM W [62]. Mitigation techniques for suppressing cracking in tungsten aside from the processing science include alloying tungsten with elements such as tantalum. However, for nuclear fusion application tungsten-tantalum alloys are generally considered to be problematic due to tantalum's degradation into the undesirable isotopes during nuclear exposure [64].

\subsection*{4.4. Microstructure}
Rather distinct to the pure refractory material systems processed through AM such as tungsten, as well as some common BCC or HCP type\\
(a) Limited fusion

\begin{center}
\includegraphics[max width=\textwidth]{2024_04_13_b812db9e8fc840c3b338g-14(2)}
\end{center}

(b) Insufficient fusion

\begin{center}
\includegraphics[max width=\textwidth]{2024_04_13_b812db9e8fc840c3b338g-14(1)}
\end{center}

(c) Proper fusion

\begin{center}
\includegraphics[max width=\textwidth]{2024_04_13_b812db9e8fc840c3b338g-14}
\end{center}

(d) Excessive fusion

\begin{center}
\includegraphics[max width=\textwidth]{2024_04_13_b812db9e8fc840c3b338g-14(3)}
\end{center}

\includegraphics[max width=\textwidth, center]{2024_04_13_b812db9e8fc840c3b338g-14(4)}\\
\includegraphics[max width=\textwidth, center]{2024_04_13_b812db9e8fc840c3b338g-14(5)}

Fig. 15. Observed states of pure tungsten fabricated through EBM using different parameters. (a) Limited fusion, (b) insufficient fusion, (c) proper fusion, and (d) excessive fusion.

The images are adopted from [57].

\begin{center}
\includegraphics[max width=\textwidth]{2024_04_13_b812db9e8fc840c3b338g-15(1)}
\end{center}

Fig. 16. Scanning electron microscope (SEM) and electron backscatter diffraction (EBSD) data of boundary cracking in EBM pure tungsten. (a) EBSD showing surrounding texture, and (b) an optical image [61].

alloys such as Ti-6V-4Al, are the anomalous textures that form as a result of the AM process [65,66]. Shown in Fig. 17 [62] are representative EBSD micrographs depicting this anomalous texture. While many in the literature for EBM of tungsten and other refractory metals have observed the phenomena, its significance in relation to the processing science of the materials has only been briefly mentioned.

Columnar grain structures aligned parallel to the build direction resulting from epitaxial growth of EBM W are reported in literature. Controlling the $\{001\}$ and $\{111\}$ fibrous texture of pure tungsten via $67^{\circ}$ interlayer rotation is also explored. Similar observations regarding the effect of texture on yield strength anisotropy have also been reported for tantalum $[57,60,61,67]$. The ability to obtain a mixed $\{001\}$ and $\{111\}$ fiber texture with the possibility of material having either strong $\{001\}$ or $\{111\}$ build direction fiber textures was also identified. In a similar study of EBM of molybdenum, the role played by area energy density to melt the material on the texture was also discussed [61,63]. This phenomenon was hypothesized to be associated with sensitivities of the melt pool shape to the electron beam energy density coupled with\\
\includegraphics[max width=\textwidth, center]{2024_04_13_b812db9e8fc840c3b338g-15}

Fig. 17. Electron backscatter diffraction (EBSD) images showing mixed $\{001\}$ and $\{111\}$ fibers in pure EBM tungsten. (a) Cross-sectional inverse pole figure (IPF) map, and (b) build direction IPF map [62]. The build direction is vertical for both images.\\
formation of networks of LAGBs driven by thermal stresses from solidification.

\subsection*{4.5. Performance of EBM Tungsten}
Analysis of the performance of EBM $\mathrm{W}$ for thermomechanical behavior is currently limited. The bend strength of an EBM W during three-point bending test was measured to be $340 \mathrm{MPa}$, which is significantly lower than reference wrought tungsten [60]. This has been partially attributed to the porosity of the EBM W samples. The strength of EBM W parallel to the build direction was also evaluated, with fracture being observed to occur along the fibrous GBs via a combination of decohesion and transgranular failure [57,60]. Lastly, the hardness of EBM W as well as its surface deterrence to ITER-like plasma heat load exposures at steady state $\left(10 \mathrm{MW} / \mathrm{m}^{2}\right.$ ) and transient (105 pulses with $0.14 \mathrm{GW} / \mathrm{m} 2$ ) was also investigated [59]. Ultimately, it was found that the EBM W performed similarly to baseline wrought recrystallized tungsten product as surface deterrence to plasma heat load exposures.

\section*{5. Summary, outlook, and recommendations}
Although tungsten and tungsten alloys are notoriously difficult to print due to their high melting temperatures, high thermal conductivities, and brittleness, encouraging progress has been made in the last decade to additively manufacture this unique class of materials. Cracking has been and remains to be one of the dominantly challenging issues in the field. Nevertheless, advance has been made to overcome this issue. For example, EBM has shown promises to manufacture crack free samples. Another important issue, which has not been investigated to a large extent, is the pore formation and control mechanisms under keyhole mode processing conditions. In addition to cracks, porosities inevitably influence the mechanical properties of additively manufactured tungsten and tungsten alloys. Up to date, limited mechanical property data are available (especially those related to elevated temperature properties) that are of critical relevance to practical applications. With the rapid progress of additive manufacturing techniques and processing conditions, we expect to witness a rising amount of data in this direction. In addition, meticulous microstructure control and new alloy design strategies are expected in near future for this class of high temperature alloys. We further contend that many issues encountered during refractory metals additive manufacturing are likely applicable to numerous fracture-prone metals such as multi-principal alloys. Strategies are needed to enable us to "print these unprintable" alloys.

\subsection*{5.1. Recommendations}
Substrate heating has been proved to be an effective strategy to overcome cracking in the EBM process. Similar approaches have not been demonstrated successfully for L-PBF or L-DED. This could be partially due to the higher oxygen contents in the laser processes. Further studies are needed in this direction to make the laser AM processes feasible to manufacture crack-free W components. Microstructure control such as grain shape manipulation has been reported to be effective in reducing the residual stresses and thus cracks in brittle materials. This approach has not been well studied for tungsten and tungsten alloys yet, or any other refractory alloys. Inoculation via the addition of nanoparticles could be another rewarding strategy to overcome the cracking issue for tungsten. In addition to experimental endeavors, computer modeling of thermal history, microstructure, and resultant residual stresses is likely to further advance this field. The above recommended research directions are likely applicable to all three AM techniques reviewed in this work.

Revision note: while this paper was under review, a parallel overview paper was published [68].

\section*{CRediT authorship contribution statement}
Talignani Alberico: Writing - original draft, Investigation, Conceptualization. Seede Raiyan: Writing - original draft, Investigation, Conceptualization. Whitt Austin: Writing - original draft, Investigation. Zheng Shiqi: Investigation, Conceptualization. Katoh Yutai: Writing - review \& editing, Supervision, Project administration, Investigation, Funding acquisition, Conceptualization. Wang Y. Morris: Writing - review \& editing, Writing - original draft, Supervision, Project administration, Investigation, Funding acquisition, Conceptualization. Kirka Michael M: Writing - review \& editing, Writing - original draft, Supervision, Investigation, Conceptualization. Ye Jianchao: Supervision, Investigation. Karaman Ibrahim: Writing - review \& editing, Supervision, Investigation, Funding acquisition, Conceptualization.

\section*{Declaration of Competing Interest}
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

\section*{Data availability}
No data was used for the research described in the article.

\section*{Acknowledgments}
This research was sponsored by the US Department of Energy, Office of Fusion Energy Sciences and Advanced Research Projects AgencyEnergy (ARPA-E) under contract DE-AC05-00OR22725 with UTBattelle LLC. The work at LLNL was performed under the auspices of the US Department of Energy under contract no. DE-AC52-07NA27344.

\section*{References}
[1] M. Rieth, S.L. Dudarev, S.M.G. de Vicente, J. Aktaa, T. Ahlgren, S. Autusch, D.E. J. Armstrong, M. Balden, N. Baluc, M.F. Barthe, W.W. Basuki, M. Battabyal, C. S. Becquart, D. Blagoeva, H. Boldyryeva, J. Brinkmann, M. Celino, L. Ciupinski, J. B. Correia, A. De Backer, C. Domain, E. Gaganidze, C. Garcia-Rosales, J. Gibson, M. R. Gilbert, S. Giusepponi, B. Gludovatzj, H. Greuner, K. Heinola, T. Hoschen, A. Hoffmann, N. Holstein, F. Koch, W. Krauss, H. Li, S. Lindig, J. Linke, C. Linsmeier, P. Lopez-Ruiz, H. Maier, J. Matejicek, T.P. Mishra, M. Muhammed, A. Munoz, M. Muzyk, K. Nordlund, D. Nguyen-Manh, J. Opschoor, N. Ordas, T. Palacios, G. Pintsuk, R. Pippan, J. Reiser, J. Riesch, S.G. Roberts, L. Romaner, M. Rosinski, M. Sanchez, W. Schulmeyer, H. Traxler, A. Urena, J.G. van der Laan, L. Veleva, S. Wahlberg, M. Walter, T. Weber, T. Weitkamp, S. Wurster, M.A. Yar, J. H. You, A. Zivelonghi, Recent progress in research on tungsten materials for nuclear fusion applications in Europe, J. Nucl. Mater. 432 (1-3) (2013) 482-500.

[2] Y. Katoh, L.L. Snead, L.M. Garrison, X. Hu, T. Koyanagi, C.M. Parish, P. D. Edmondson, M. Fukuda, T. Hwang, T. Tanaka, A. Hasegawa, Response of unalloyed tungsten to mixed spectrum neutrons, J. Nucl. Mater. 520 (2019) 193-207.

[3] A.M. Engwall, S.J. Shin, J. Bae, Y.M. Wang, Enhanced properties of tungsten films by high-power impulse magnetron sputtering, Surf. Coat. Technol. 363 (2019) 191-197.

[4] Q. Wei, K.T. Ramesh, B.E. Schuster, L.J. Kecskes, R.J. Dowding, Nanoengineering opens a new era for tungsten as well, JOM 58 (9) (2006) 40-44.

[5] S.J. Zinkle, L.J. Ott, T. Ingersoll, R.J. Ellis, M.L. Grossbeck, Overview of materials technologies for space nuclear power and propulsion, AIP Conference Proceedings, 2002: pp. 1063-1073.

[6] C.J. Han, Q.H. Fang, Y.S. Shi, S.B. Tor, C.K. Chua, K. Zhou, Recent advances on high-entropy alloys for 3D printing, Adv. Mater. 32 (26) (2020).

[7] T. Mukherjee, T. Debroy, A digital twin for rapid qualification of 3D printed metallic components, Appl. Mater. Today 14 (2019) 59-65.

[8] S.A. Khairallah, A.A. Martin, J.R.I. Lee, G. Guss, N.P. Calta, J.A. Hammons, M. H. Nielsen, K. Chaput, E. Schwalbach, M.N. Shah, M.G. Chapman, T.M. Willey, A. M. Rubenchik, A.T. Anderson, Y.M. Wang, M.J. Matthews, W.E. King, Controlling interdependent meso-nanosecond dynamics and defect generation inmetal 3D printing, Science 368 (6491) (2020) 660-665.

[9] M.J. Matthews, G. Guss, S.A. Khairallah, A.M. Rubenchik, P.J. Depond, W.E. King, Denudation of metal powder layers in laser powder bed fusion processes, Acta Mater. 114 (2016) 33-42.

[10] S. Bai, J. Liu, P. Yang, M. Zhai, H. Huang, L.-M. Yang, Femtosecond fiber laser additive manufacturing of tungsten, Laser 3D Manufacturing III, San Francisco, 2016, p. $97380 \mathrm{U}$.

[11] D.-Z. Wang, K.-L. Li, C.-F. Yu, J. Ma, W. Liu, Z.-J. Shen, Cracking Behavior in Additively Manufactured Pure Tungsten, Acta Metall. Sin. 32 (1) (2019) 127-135.

[12] B. Vrancken, R.K. Ganeriwala, M.J. Matthews, Analysis of laser-induced microcracking in tungsten under additive manufacturing conditions: Experiment and simulation, Acta Mater. 194 (2020) 464-472.

[13] A.T. Sidambe, Y. Tian, P.B. Prangnell, P. Fox, Effect of processing parameters on the densification, microstructure and crystallographic texture during the laser powder bed fusion of pure tungsten, Int. J. Refract. Met. Hard Mater. 78 (2019) 254-263.

[14] P. Rebesan, M. Bonesso, C. Gennari, R. Dima, A. Pepato, M. Vedani, Tungsten Fabricated by Laser Powder Bed Fusion, Berg. Huttenmann. Mon. 166 (5) (2021) 263-269.

[15] B. Vrancken, W.E. King, M.J. Matthews, In-situ characterization of tungsten microcracking in Selective Laser Melting, Procedia CIRP 74 (2018) 107-110.

[16] J. Chen, K. Li, Y. Wang, L. Xing, C. Yu, H. Liu, J. Ma, W. Liu, Z. Shen, The effect of hot isostatic pressing on thermal conductivity of additively manufactured pure tungsten, Int. J. Refract. Met. Hard Mater. 87 (2020), 105135.

[17] C. Tan, K. Zhou, W. Ma, B. Attard, P. Zhang, T. Kuang, Selective laser melting of high-performance pure tungsten: parameter design, densification behavior and mechanical properties, Sci. Technol. Adv. Mater. 19 (1) (2018) 370-380.

[18] M. Guo, D. Gu, L. Xi, H. Zhang, J. Zhang, J. Yang, R. Wang, Selective laser melting additive manufacturing of pure tungsten: Role of volumetric energy density on densification, microstructure and mechanical properties, Int. J. Refract. Met. Hard Mater. 84 (2019), 105025.

[19] Z. Hu, Y. Zhao, K. Guan, Z. Wang, Z. Ma, Pure tungsten and oxide dispersion strengthened tungsten manufactured by selective laser melting: Microstructure and cracking mechanism, Addit. Manuf. 36 (2020), 101579.

[20] A.C. Field, L.N. Carter, N.J.E. Adkins, M.M. Attallah, M.J. Gorley, M. Strangwood, The effect of powder characteristics on build quality of high-purity tungsten produced via laser powder bed fusion (LPBF), Metall. Mater. Trans. A 51 (3) (2020) 1367-1378.

[21] D. Wang, C. Yu, X. Zhou, J. Ma, W. Liu, Z. Shen, Dense pure tungsten fabricated by selective laser melting, NATO Adv. Sci. Inst. Ser. E Appl. Sci. 7 (4) (2017) 430.

[22] X. Ren, H. Liu, F. Lu, L. Huang, X. Yi, Effects of processing parameters on the densification, microstructure and mechanical properties of pure tungsten fabricated by optimized selective laser melting: From single and multiple scan tracks to bulk parts, Int. J. Refract. Met. Hard Mater. 96 (2021), 105490.

[23] A. Iveković, N. Omidvari, B. Vrancken, K. Lietaert, L. Thijs, K. Vanmeensel, J. Vleugels, J.-P. Kruth, Selective laser melting of tungsten and tungsten alloys, Int. J. Refract. Met. Hard Mater. 72 (2018) 27-32.

[24] J. Xue, Z. Feng, J. Tang, C. Tang, Z. Zhao, Selective laser melting additive manufacturing of tungsten with niobium alloying: Microstructure and suppression mechanism of microcracks, J. Alloy. Compd. 874 (2021), 159879.

[25] K.L. Li, D.Z. Wang, L.L. Xing, Y.F. Wang, C.F. Yu, J.H. Chen, T. Zhang, J. Ma, W. Liu, Z.J. Shen, Crack suppression in additively manufactured tungsten by introducing secondary-phase nanoparticles into the matrix, Int. J. Refract. Met. Hard Mater. 79 (2019) 158-163.

[26] K. Li, Y. Li, W. Chen, C. Zhao, Y. Yuan, L. Cheng, T.W. Morgan, W. Liu, Z. Shen, Effect of Ta addition on the fuzz formation of additively manufactured W-based materials, Nucl. Fusion 60 (6) (2020), 064004.

[27] K. Li, G. Ma, L. Xing, Y. Wang, C. Yu, J. Chen, J. Ma, G. Wu, W. Liu, Z. Shen, X. Huang, Crack suppression via in-situ oxidation in additively manufactured W-Ta alloy, Mater. Lett. 263 (2020), 127212.

[28] B. Vrancken, R.K. Ganeriwala, A.A. Martin, M.J. Matthews, Microcrack mitigation during laser scanning of tungsten via preheating and alloying strategies, Addit. Manuf. 46 (2021), 102158.

[29] Av Müller, G. Schlick, R. Neu, C. Anstätt, T. Klimkait, J. Lee, B. Pascher, M. Schmitt, C. Seidel, Additive manufacturing of pure tungsten by means of selective laser beam melting with substrate preheating temperatures up to 1000 ०C, Nucl. Eng. Des. /Fusion 19 (2019) 184-188.

[30] Z. Xiong, P. Zhang, C. Tan, D. Dong, W. Ma, K. Yu, Selective laser melting and remelting of pure tungsten, Adv. Eng. Mater. 22 (3) (2020), 1901352.

[31] X. Zhou, X. Liu, D. Zhang, Z. Shen, W. Liu, Balling phenomena in selective laser melted tungsten, J. Mater. Process. Technol. 222 (2015) 33-42.

[32] S. Wen, C. Wang, Y. Zhou, L. Duan, Q. Wei, S. Yang, Y. Shi, High-density tungsten fabricated by selective laser melting: Densification, microstructure, mechanical and thermal performance, Opt. Laser Technol. 116 (2019) 128-138.

[33] B. Zhang, R. Seede, L. Xue, K.C. Atli, C. Zhang, A. Whitt, I. Karaman, R. Arroyave, A. Elwany, An efficient framework for printability assessment in Laser Powder Bed Fusion metal additive manufacturing, Addit. Manuf. 46 (2021), 102018.

[34] Y.M. Wang, T. Voisin, J.T. McKeown, J. Ye, N.P. Calta, Z. Li, Z. Zeng, Y. Zhang, W. Chen, T.T. Roehling, R.T. Ott, M.K. Santala, P.J. Depond, M.J. Matthews, A. V. Hamza, T. Zhu, Additively manufactured hierarchical stainless steels with high strength and ductility, Nat. Mater. 17 (1) (2018) 63-71.

[35] T. Voisin, J.B. Forien, A. Perron, S. Aubry, N. Bertin, A. Samanta, A. Baker, Y. M. Wang, New insights on cellular structures strengthening mechanisms and thermal stability of an austenitic stainless steel fabricated by laser powder-bedfusion, Acta Mater. 203 (2021), 116476.

[36] A.M. Rubenchik, W.E. King, S.S. Wu, Scaling laws for the additive manufacturing, J. Mater. Process. Technol. 257 (2018) 234-243.

[37] J. Trapp, A.M. Rubenchik, G. Guss, M.J. Matthews, In situ absorptivity measurements of metallic powders during laser powder-bed fusion additive manufacturing, Appl. Mater. Today 9 (2017) 341-349.

[38] J. Ye, S.A. Khairallah, A.M. Rubenchik, M.F. Crumb, G. Guss, J. Belak, M. J. Matthews, Energy coupling mechanisms and scaling behavior associated with laser powder bed fusion additive manufacturing, Adv. Eng. Mater. 21 (7) (2019), 1900185.

[39] T. Majumdar, N. Eisenstein, J.E. Frith, S.C. Cox, N. Birbilis, Additive Manufacturing of Titanium Alloys for Orthopedic Applications: A Materials Science Viewpoint, Adv. Eng. Mater. 20 (2018), 1800172.

[40] J. Xie, H. Lu, J. Lu, X. Song, S. Wu, J. Lei, Additive manufacturing of tungsten using directed energy deposition for potential nuclear fusion application, Surf. Coat. Technol. 409 (2021), 126884.

[41] M. Zhong, W. Liu, G. Ning, L. Yang, Y. Chen, Laser direct manufacturing of tungsten nickel collimation component, J. Mater. Process. Technol. 147 (2004) 167-173.

[42] T. DebRoy, H.L. Wei, J.S. Zuback, T. Mukherjee, J.W. Elmer, J.O. Milewski, A. M. Beese, A. Wilson-Heid, A. De, W. Zhang, Additive manufacturing of metallic components-process, structure and properties, Prog. Mater. Sci. 92 (2018) $112-224$.

[43] W. Jeong, Y.-S. Kwon, D. Kim, Three-dimensional printing of tungsten structures by directed energy deposition, Mater. Manuf. Process. 34 (2019) 986-992.

[44] G. Wang, X. Sun, M. Huang, Y. Qin, Y. Yao, S. Yang, Influence of processing parameters on the microstructure and tensile property of $85 \mathrm{~W}-15 \mathrm{Ni}$ produced by laser direct deposition, Int. J. Refract. Met. Hard Mater. 82 (2019) 227-233.

[45] Y.P. Wang, S.Y. Ma, X.S. Yang, Y.Z. Zhou, X. Liu, J.F. Li, J.J. Zhang, C. Li, X. Y. Wang, G.M. Le, Y. Zhang, Microstructure and strengthening mechanisms of 90W-7Ni-3Fe alloys prepared using laser melting deposition, J. Alloy. Compd. 838 (2020), 155545.

[46] C. Li, Y. Wang, S. Ma, X. Yang, J. Li, Y. Zhou, X. Liu, J. Tang, X. Wang, G. Le, Densification, microstructural evolutions of 90W-7Ni-3Fe tungsten heavy alloys during laser melting deposition process, Int. J. Refract. Met. Hard Mater. 91 (2020), 105254.

[47] B. Bax, R. Rajput, R. Kellet, M. Reisacher, Systematic evaluation of process parameter maps for laser cladding and directed energy deposition, Addit. Manuf. 21 (2018) 487-494.

[48] J.C. Haley, B. Zheng, U.S. Bertoli, A.D. Dupuy, J.M. Schoenung, E.J. Lavernia, Working distance passive stability in laser directed energy deposition additive manufacturing, Mater. Des. 161 (2019) 86-94.

[49] P.N. Sibisi, A.P.I. Popoola, N.K. Arthur, S.L. Pityana, Review on direct metal laser deposition manufacturing technology for the Ti-6Al-4V alloy, Int. J. Adv. Manuf. Technol. 107 (2020) 1163-1178.

[50] V. Fallah, M. Alimardani, S.F. Corbin, A. Khajepour, Impact of localized surface preheating on the microstructure and crack formation in laser direct deposition of Stellite 1 on AISI 4340 steel, Appl. Surf. Sci. 257 (2010) 1716-1723.

[51] W. Meng, Y. Xiaohui, W. Zhang, F. Junfei, G. Lijie, M. Qunshuang, C. Bing, Additive manufacturing of a functionally graded material from Inconel 625 to Ti6Al4V by laser synchronous preheating, J. Mater. Process. Technol. 275 (2020), 116368.

[52] L. Zhou, S. Chen, M. Wei, J. Liang, C. Liu, M. Wang, Microstructure and properties of $24 \mathrm{CrNiMoY}$ alloy steel prepared by direct laser deposited under different preheating temperatures, Mater. Charact. 158 (2019), 109931.

[53] A. Sadhu, A. Choudhary, S. Sarkar, A.M. Nair, P. Nayak, S.D. Pawar, G. Muvvala, S. K. Pal, A.K. Nath, A study on the influence of substrate pre-heating on mitigation of cracks in direct metal laser deposition of NiCrSiBC-60\% WC ceramic coating on Inconel 718, Surf. Coat. Technol. 389 (2020), 125646.

[54] I. Gibbons, D. Rosen, B. Stucker, Additive Manufacturing Technologies, Springer,, 2014.

[55] M.M. Kirka, K.A. Unocic, N. Raghavan, F. Medina, R.R. Dehoff, S.S. Babu, Microstructure Development in Electron Beam-Melted Inconel 718 and Associated Tensile Properties, JOM 68 (3) (2016) 1012-1020.

[56] Y.S. Lee, M.M. Kirka, J. Ferguson, V.C. Paquit, Correlations of cracking with scan strategy and build geometry in electron beam powder bed additive manufacturing, Addit. Manuf. 32 (2020).

[57] G. Yang, P. Yang, K. Yang, N. Liu, L. Jia, J. Wang, H.P. Tang, Effect of processing parameters on the density, microstructure and strength of pure tungsten fabricated by selective electron beam melting, Int. J. Refract. Met. Hard Mater. 84 (2019), 105040.

[58] Z.C. Cordero, R.B. Dinwiddie, D. Immel, R.R. Dehoff, Nucleation and growth of chimney pores during electron-beam additive manufacturing, J. Mater. Sci. 52 (6) (2017) 3429-3435.

[59] D. Dorow-Gerspach, A. Kirchner, T. Loewenhoff, G. Pintsuk, T. Weißgärber, M. Wirtz, Additive manufacturing of high density pure tungsten by electron beam melting, Nucl. Mater. Energy 28 (2021), 101046.

[60] J. Wright, Additive Manufacturing of Tungsten via Selective Laser Melting and Electron Beam Melting, University of Sheffield,, 2019.

[61] E.A.I. Ellis, M.A. Sprayberry, C. Ledford, J.P. Hankwitz, M.M. Kirka, C.D. Rock, T. J. Horn, Y. Katoh, R.R. Dehoff, Processing of tungsten through electron beam melting, J. Nucl. Mater. 555 (2021), 153041.

[62] C. Ledford, P. Fernandez-Zelaia, J.O. Rojas, Q. Campbell, A.M. Rossy, M.M. Kirka, Electron Beam Powder Bed Fusion Additive Manufacturing of Refractory Metals, 2021 Proceedings of the Tenth International Conference on Tungsten, Refractory \& Hardmaterials 2021, pp. 159-165.

[63] P. Fernandez-Zelaia, C. Ledford, E.A.I. Ellis, Q. Campbell, A.M. Rossy, D. N. Leonarda, M.M. Kirkaa, Crystallographic texture evolution in electron beam melting additive manufacturing of pure Molybdenum, Mater. Des. 207 (2021), 109809.

[64] M.R. Gilbert, J.C. Sublet, Neutron-induced transmutation effects in W and W-alloys in a fusion environment, Nucl. Fusion 51 (4) (2011), 043005.

[65] M. Higashi, T. Ozaki, Selective laser melting of pure molybdenum: Evolution of defect and crystallographic texture with process parameters, Mater. Des. 191 (2020), 108588.

[66] O. Andreau, I. Koutiri, P. Peyre, J.-D. Penot, N. Saintier, E. Pessard, T. De Terris, C. Dupuy, T. Baudin, Texture control of 316L parts by modulation of the melt pool morphology in selective laser melting, J. Mater. Process. Technol. 264 (2019) 21-31.\\
[67] L. Thijs, M.L.M. Sistiaga, R. Wauthle, Q. Xie, J.-P. Kruth, J. Van Humbeeck, Strong morphological and crystallographic texture and resulting yield strength anisotropy in selective laser melted tantalum, Acta Mater. 61 (12) (2013) 4657-4668.

[68] P. Morcos, A. Elwany, I. Karaman, R. Arróyave, Additive manufacturing of pure tungsten and tungsten-based alloys, J. Mater. Sci. 57 (21) (2022).

\begin{itemize}
  \item 
\end{itemize}


\end{document}