On the Extent of Feedstock–System Interaction in Determining the Efficiency of Laser Powder Directed Energy Deposition

Abstract

Laser Powder Directed Energy Deposition (LP-DED) is an advanced additive manufacturing process that uses a focused laser beam to melt and fuse powder material onto a substrate. This technology enables the production of complex metal components with high precision and material efficiency. The properties of the powder feedstock are highly important and have been extensively studied in the literature. Powder size distribution and particle shape have been identified as key factors influencing the flowability, and it is imperative that nozzle designs take these into account for optimum material delivery. The laser–powder interaction, where the laser energy influences the melting behavior, as well as nozzle designs, have been highlighted in both historical and the more recent laser cladding literature. Finally, a comprehensive analysis of fluid dynamic simulations of the powder particles and their interaction with the nozzle design is provided.

1. Introduction

Additive manufacturing (AM) refers to a range of technologies that use layer-by-layer manufacturing techniques starting from a virtual model created using CAD software. A wide range of materials can be used in these processes, from polymers (with low melting points) to metals and ceramics (with high melting points). AM focuses on producing objects with complex shapes that cannot be made using traditional manufacturing methods. By adding material rather than removing it, the geometric complexity of the part produced does not add cost, offering unprecedented design freedom and production flexibility. Layered fabrication and increased complexity go hand in hand with a very high level of customization and the possibility of mass customization of the final product; these characteristics of AM, which cannot be compared to any other manufacturing technology, make it a great complement to conventional technologies [1,2,3,4,5,6,7,8].

Using the ASTM F42/ISO TC 261 [9] definition, the seven additive manufacturing families can be further divided into single-step and multi-step technologies. When considering metals and alloys, single-step technologies are those based on complete fusion of the initial feedstock and include powder bed fusion (PBF) and directed energy deposition (DED) technologies.

Laser Powder Directed Energy Deposition (LP-DED), already known since the early 1970s as laser cladding [10,11,12,13], is a versatile additive manufacturing technique with a wide range of applications across various industries. The most common DED technologies include Laser Engineered Net Shaping (LENS), Direct Metal Deposition (DMD), Laser-Aided Additive Manufacturing (LAAM), Wire Arc Additive Manufacturing (WAAM), and Electron Beam Freeform Fabrication (EB-FFF), among others [14].

DED involves the use of a laser to create a melt pool on a substrate into which metal powder is injected, allowing the fabrication of complex geometries, the repair of existing components, and the creation of multi-material structures [15]. Laser Powder Directed Energy Deposition (LP-DED) can be considered one of the most versatile DED technologies due to the fact that different powder batches can be used during the same 3D printing job (from two to six, depending on the considered machine) and laser–optic coupling can be tailored according to the minimum wall thickness to be addressed, from ~0.7 mm upwards. In LP-DED processes, deposition and fusion are localized to a single point and therefore do not cause significant shrinkage problems, as in LPBF processes. In addition, there are lower residual stresses in the finished parts because the solidification rates are slower with this technology and the localized interaction zone between the feedstock and the laser enables repairing procedures for the components.

Laser Powder Directed Energy Deposition (LP-DED) (Figure 1) is one of the most promising technologies in the field of additive manufacturing, particularly for the manufacture and repair of metallic components, as it enables the direct deposition of molten material through the use of continuous material flows and real-time melting. Once molten, the material is deposited layer by layer and rapidly solidifies to form the desired geometry. This feature makes the process particularly suitable for repairs, modifications, and the creation of complex geometries. An LP-DED machine (schematically illustrated in Figure 1) consists of a nozzle mounted on a multi-axis arm that deposits the molten material onto the substrate. This schematic provides a general overview of the LP-DED process, highlighting the laser–powder–substrate interaction that underpins the entire review. The material can be deposited from any angle, as the process utilizes machines with high degrees of freedom.

This manufacturing technology offers higher deposition rates and a wider processing window compared to powder bed-based AM technologies, making it more suitable for the production of large metallic components with medium geometric complexity [16,17,18].

Figure 2 illustrates the primary subsystems constituting a standard LP-DED apparatus. This schematic representation provides a structured overview of the functional components, such as the feedstock delivery system, energy source, and motion control unit, that work in synergy to ensure precision and quality in the deposition process. The main components include the following (Figure 2):

-

Material feed system: The material to be deposited, which may be in wire or powder form, is transported to the deposition nozzle. The feed can be automatic or manual, depending on the requirements of the production process.

-

Energy source: The energy source is critical to melting the material. Common options include high power lasers, electron beams, and electric arcs. The concentrated energy melts the material and creates a metallurgical bond between the deposited layers.

-

Movement system: An industrial robot or CNC system is used to position the deposition nozzle with high precision, enabling the construction of parts with complex geometries.

-

Monitoring and control: Advanced sensors monitor critical parameters, such as temperature, deposition speed, and fusion quality. Vision systems and automated software provide real-time control to ensure the quality of the finished part.

LP-DED is considered when large volumes and tailored composition (i.e., graded compositions of bimetallic parts) have to be fabricated with lower precision compared to laser powder bed fusion but with a higher build rate [20]. Applications of AM technology for metallic materials are continuously developing and impacting various industrial sectors, including oil and gas and mining [21,22,23]. In the aerospace industry, LP-DED is used for manufacturing large-scale components, such as nozzles and thrust chambers for liquid propulsion systems. This application benefits from the ability to produce large build volumes and complex geometries that are difficult to achieve with traditional manufacturing methods. The process also allows for the integration of bimetallic liners and closeout jackets, which are essential for the performance and durability of propulsion systems. This technology is also applied in the development of nuclear thermal propulsion chambers, where it offers advantages in terms of material selection and geometric flexibility [24,25,26,27,28].

Compared to other metal additive manufacturing technologies, LP-DED exhibits distinct advantages for the repair of high-value components. These offer a relatively small heat-affected zone (HAZ), excellent density and metallurgical bonding, minimal impact on the base component (e.g., reduced distortion and micro-cracking), and precise material deposition [29]. Additionally, a wide range of metals and alloys has been successfully tested for repair using LP-DED [30,31,32,33,34]. LP-DED is highly effective in the repair and refurbishment of high-quality components, such as gay cast iron. The process can restore up to 98.7% of the original tensile strength of damaged parts, making it ideal for applications in the automotive and agricultural sectors [35].

The interaction between the laser source and the powder stream, delivered through inert gas by specially designed nozzles, plays a key role in not only determining the final properties of the deposited materials but also in guiding the selection and optimization of process parameters and equipment configuration.

There is quite an extensive number of review papers dealing with different aspects of powder-based metal additive manufacturing technologies, ranging from laser beam shaping [36], laser–powder interaction [37,38], and thermophysical phenomena [39] to defect formation and numerical simulation [40], as well as powder recycling and reuse [37,41].

Interesting insights are given by the work of Dass et al. [3] concerning the state of the art of DED in general, with a focus on material design, as well, while the literature review by Guan and Zhao [42] addresses more specifically the modeling of LP-DED processes, also taking into account the physical interaction between the powder feedstock and the heat source.

Most of the available literature review effort is concentrated on the laser powder bed fusion (LPBF) technology, being the most widespread and applied metal AM technology. However, many aspects of laser–powder interaction, melt pool behavior, and microstructural evolution deeply studied for LPBF can be widened to LP-DED, thus taking advantage of the large heritage coming from laser cladding [43,44].

Unlike other literature reviews that consider these aspects in isolation, this work provides a comprehensive and integrated perspective. This review highlights how these factors are closely interrelated by combining consolidated knowledge from laser cladding with an in-depth analysis of powder characteristics, laser–powder interaction phenomena, and nozzle design in LP-DED systems. It highlights how their synergistic influence determines the stability and efficiency of the deposition process, the quality of the starting material, and, ultimately, the performance and integrity of the finished part. This holistic approach is a novel contribution of this work and provides valuable insights for both understanding process fundamentals and guiding future technological developments.

2. Metal Powders for LP-DED

2.1. Feedstock Properties

Restrictions on the metals and alloys that can be processed with AM in general and LP-DED in particular may be related to their physical and chemical properties. The use of highly reflective metals is not recommended, as excessive reflection can lead to undesirable phenomena, such as shrinkage and partial melting. In addition, intermetallic materials with low toughness are unsuitable for this type of processing as they do not allow for proper process control [3].

The properties of metal powders are critical to the success of additive manufacturing processes, as well as the effective reuse of powders in multiple processing cycles. Particle size, shape, distribution, and purity, together with physical properties, such as density and flowability, have a direct impact on the quality of the final product [45]. Careful selection of powders according to specific application requirements and process specifications is essential to achieve high-quality components with superior mechanical properties. Spherical particles generally exhibit better flowability than irregularly shaped particles, which can result in more uniform deposition and higher relative density in the final product. In the case of Al₂O₃ ceramics, the use of plasma spheroidized alumina powder (PSAP) resulted in higher flexural strength and relative density compared to irregular alumina powder (IAP) due to the consistent growth orientation of columnar crystals [46]. The composition of the powder influences the microstructure and mechanical properties of the deposited material. For example, the use of WC-12Co composite powder in DED resulted in coatings with high hardness and wear resistance due to the formation of dendritic carbides and herringbone planar crystals [47]. In the case of medium entropy CrCoNi alloys, variations in powder composition resulted in differences in crystallographic orientation and dislocation density, which in turn influenced mechanical properties, such as yield strength and elongation, at different levels of the deposited structure [48].

The interaction between the powder characteristics and the LP-DED process parameters, such as laser power and scanning speed, plays a crucial role in determining the final properties of the deposited material. The shape and size distribution of powder particles are critical in determining the flowability and packing density during the LP-DED process.

• Particle Size: Particle size plays a fundamental role in the additive manufacturing process. The optimal particle size varies depending on the type of process used and the specific final applications, but it generally ranges from 15 to 150 μm. Smaller particles offer a greater surface area, which increases energy absorption and radiation scattering [49], facilitating fusion between layers and thereby improving surface quality and the density of the final component. However, the use of excessively fine particles may lead to poor powder flow and promote agglomeration, causing difficulties during the deposition process. In contrast, larger particles may result in poor fusion between layers and lower density unless properly managed.
• Particle Size Distribution (PSD): A uniform particle distribution favors better cohesion between particles and more homogeneous powder deposition during the process. Additionally, a distribution that combines particles of different sizes, including both smaller and larger ones, improves flowability and reduces the risk of structural defects, such as porosity or lack of fusion. Optimizing the particle size distribution is therefore critical to achieve components with controlled density and superior quality. Significant variations in particle size distribution may require recalibration of printing parameters to prevent negative impacts on the mechanical properties of the produced components.
• Particle Shape: Spherical particles are generally preferred, as they allow for smooth flow and uniform distribution on the build platform. This morphology also promotes more efficient fusion between particles during heating. However, particles with irregular or acicular shapes, such as needle-like particles, can be used in specific applications to improve the final material’s density. While these particles may behave favorably under certain conditions, they tend to reduce powder flowability and complicate the deposition process.
• Chemical Composition: Impurities or contaminants resulting from powder production or handling can compromise the mechanical properties of the material, reducing resistance, durability, and corrosion resistance in the final component. Therefore, it is essential that metal powders are composed of pure materials and that their chemical composition is strictly controlled to meet the required standards for the final application. The choice of metal alloys should be made based on the technical specifications of the component to be produced. For example, in nickel-based superalloys, the inclusion of elements like Al, Ti, Nb, Ta, V, Cr, Fe, Co, and Mo, along with various trace elements, allows for the desired microstructure and mechanical properties to be achieved [50].
• Density: The apparent density directly influences powder flowability and deposition during the AM process. Powders with a higher apparent density tend to behave better during printing, promoting uniform deposition and reducing the formation of porosity in the final material [51,52]. Conversely, powders with low apparent density can cause fusion defects, compromising the surface quality of the component. Insufficient packing density can lead to high surface roughness and low relative density, negatively impacting the mechanical properties of the component [53]. Additionally, packing density affects the thermal conductivity of the powder bed [54] and the melt bath kinetics [55], as void formation in the powder bed alters heat flow. Therefore, controlling the apparent density is crucial to ensuring the success of the process.
• Flowability: Good powder flow ensures regular deposition and optimal fusion between layers, reducing the risk of defects, such as porosity or lack of fusion. Powder flowability is closely related to its particle shape, size, and surface, making it essential to manage these properties to achieve high-quality results in the AM process. Powders with poor flowability tend to agglomerate due to high resistance to movement, leading to the formation of non-uniform powder layers and low packing density [56]. Flowability and packing density are also influenced by particle shape, with spherical and smooth powders reducing particle friction, thus improving these parameters [57,58].
• Moisture Content: Moisture content is a critical factor. The presence of moisture can lead to oxidation of the metal particles, negatively affecting the melting process and the final component’s quality. Furthermore, moisture can promote the formation of unwanted compounds that compromise the mechanical properties of the material. Therefore, it is essential to store powders in low-humidity environments and treat them adequately to remove moisture traces before use.
• Reactivity and Thermal Stability: Some metals or alloys exhibit greater reactivity at high temperatures, which can lead to undesirable chemical or physical changes during the melting process. In particular, materials like titanium or nickel-based alloys require careful evaluation of their thermal stability to avoid the formation of unwanted phases that may compromise the final product’s characteristics.
• Oxidation Resistance: This issue can compromise the quality of metal powders, especially for metals like titanium, aluminum, and stainless steel. These materials are particularly sensitive to oxidative reactions at high temperatures, resulting in damage to the quality of the powders and the final product. For this reason, it is crucial to protect powders during the AM process and carefully control the printing atmosphere to prevent oxidation. For nickel powders used in AM processes, a significant issue is the oxygen content; the powders have a higher oxygen content than the cast ingot [59], leading to an increased risk of oxide inclusion formation [60], spattering, and balling [61] during the process. Excessive oxygen content can also reduce the ductility of nickel-based superalloy components produced via AM [62].

Characterizing powders is essential to ensure the quality of the final components. Therefore, it is critical to perform a thorough evaluation of the powders before use, following the methods established by the ASTM F3049 standard [63], with particular attention to recycled powders. During the recycling process, powders may undergo changes in particle size distribution, flowability, packing density, and chemical composition due to phenomena like spattering, oxidation, and the evaporation of binder elements [64].

2.2. Recycling of Metal Powders

Powder reuse plays an important role both in terms of environmental impact and in making the process economically advantageous, achieving a high level of material utilization given the significant cost of powders used in additive manufacturing (AM) processes.

Unmelted powder that is not used in the production cycle is recovered from the build chamber and used in subsequent cycles. It is important to determine if and how metal powders are affected by the process during reuse cycles. Powders can undergo changes in chemical composition and physical properties; if the chemical composition changes significantly, the components produced could fall outside of the specified requirements, while if the physical properties change, the machine parameters may no longer be optimal for the material being processed [65]. It is possible to run new cycles using unused powder from previous cycles taken from the build chamber and sieved to remove spatter and debris or further mixed with a percentage of new powder. Analyzing the variations that powders undergo during subsequent reuse cycles is fundamental to evaluating any changes in the mechanical performance of components produced with these powders [66].

The characteristics of the powders that may undergo changes during reuse are particle size distribution (PSD), particle morphology, flowability, apparent density, compressibility, thermal conductivity, and oxygen concentration. Due to its low thermal conductivity and apparent density, the thermal effusivity (a measure of the material’s ability to exchange thermal energy with its surroundings) of virgin powder is lower than that of recycled powder, indicating less heat transfer with the surrounding environment and, consequently, heat accumulation in the virgin powder particles. Virgin powder has higher permeability, which indicates how easily a fluid (usually gas) can flow between the powder particles [18], and it tends to move with more agglomerates, requiring more energy to flow, thus showing lower flowability compared to recycled powder.

In terms of flowability, previous research papers have clearly established a link between the physical properties of powders, their flowability, and the quality of the final components. For instance, Kiani et al. [67] found that particle morphology (e.g., AR50) affects how powders initiate flow from a static or compacted state, whereas D50 plays a key role in sustaining flow once it has started. Inconsistent powder flow in LP-DED systems like LENS^® can cause momentary shortages during deposition, resulting in parts with reduced density (increased porosity) and inferior mechanical properties. Supporting this, Freeman et al. [68] demonstrated that fluctuations in powder flow rate during LP-DED lead to irregular melt track dimensions, which in turn cause variations in the overall quality of the deposited parts. Kiani et al. [69] showed that compared to the recycled powders (8.1–8.5 g/s), the virgin powder had a greater flow rate of 9.2 g/s. Even though this drop in the flow rate value is minimal, prior research has demonstrated that even slight reductions in powder flowability can result in significant reductions in mechanical characteristics [31]. Therefore, it is critical to closely examine even slight variations in powder flowability. But, there was no discernible difference in flow rate for powders P1, P2, P3, and P4, suggesting that more reuse cycles have no appreciable effect on powder flowability.

Carrion et al. [70] stated that reusing powder seems to improve high-cycle fatigue resistance (HCF: high-cycle fatigue); this can be explained by the combination of higher flowability (less cohesion between particles) and lower compressibility (smaller voids between different powder particles), leading to a more uniform powder bed distribution and the formation of smaller pores within the component [70].

When powders are used in different cycles, they may become contaminated with impurities, foreign bodies, or interstitial elements (oxidation); consequently, elements can be introduced into the powders during pre-processing or post-processing stages. The most commonly used powder recycling methods are the following [71].

• Single Batch and Collective Aging: The single Batch method involves reusing a batch of powder in multiple processing cycles until an insufficient amount of powder remains to complete a print; sieving is performed between cycles to remove larger particles or agglomerates, but no virgin powder is added. The collective aging method is similar to the single batch method but on a larger scale; multiple powder batches are used to print with powder of the same age mixed for subsequent prints.
• Top up: This method involves placing virgin or sieved powder on top of the remaining powder in the chamber to maintain the necessary powder level for the subsequent print; this powder is not mixed with the powder already in the machine, resulting in a layered batch with sections with different processing histories.
• Refreshing: After printing, the unused powder is sieved and mixed with virgin powder, creating a homogeneous blend. This process can be performed after each print or after a certain number of prints, ensuring that there is always enough powder for subsequent prints.

Concerning the latter strategy, of particular interest is the work by Rousseau et al. [72], where batches of Ti-6Al-4V with different oxygen contents were characterized and compared. The high-oxygen powder was recycled 10 times over a two-year period, showing no significant oxygen pickup during repeated recycling or during the printing process. Instead, the primary change observed was a coarsening of the particle size distribution, where larger and more fragmented particles were seen compared to new powders. Parts made with recycled powder showed a significant loss of vanadium compared to those made with new powder; despite this chemical change, the overall tensile properties, including the ultimate tensile strength and ductility, were maintained within acceptable ranges for the Ti-6Al-4 V alloy (with some differences in the yield strength and UTS).

Table 1. Reuse of metal powders in LP-DED; R0 is for fresh powder, and R# indicates the number of reuses.

Material	N° of Reuses	Sieving	Mixed with Fresh Powder	Yield Strength (MPa)	UTS (MPa)	Elongation (%)	Ref.
316L SS	9	yes	no	R0: 425 ± 6	R0: 602 ± 2	R0: 55 ± 7	[73]
				R9: 445 ± 4	R9: 612 ± 2	R9: 44 ± 2
NASA HR-1	6	yes	no	R1: 455 ± 20	R1: 925 ± 24	R1: 34 ± 1	[74]
				R2: 459 ± 27	R2 961 ± 26	R2: 34 ± 2
				R3: 455 ± 33	R3: 942 ± 37	R3: 35 ± 2
				R4: 470 ± 26	R4 971 ± 25	R4: 34 ± 2
				R5: 479 ± 27	R5: 994 ± 16	R5: 35 ± 2
				R6: 470 ± 27	R6: 983 ± 24	R6: 35 ± 0
Ti-6Al-4V ELI	3	yes	no	NA	NA	NA	[75]
Ti-6Al-4V	10	yes	yes	R0 *: 839 ± 3	R0 *: 975 ± 4	R0 *: 22.0 ± 0.8	[72]
				R10 *: 934 ± 6	R10 *: 1054 ± 5	R10 *: 20.9 ± 0.5
316L SS	NA	yes	yes	R0: 469.6 ± 3.4 628.2 ± 6.6 R# **: 458.4 ± 29.9	R0: 628.2 ± 6.6 R# **: 651.6 ± 43.9	R0: 31.2 ± 2.2 R# **: 16.2 ± 2.1	[76]
AlSi10Mg ***	Up to 4	yes	no	R0: 180 ± 23	R0: 281 ± 15	R0: 9 ± 1	[69]
				R1: 174 ± 13	R1: 149 ± 50	R1: 5.0 ± 0.8
				R2: 189 ± 17	R2: 63 ± 25	R2: 6 ± 1
				R3: 195 ± 12	R3: 168 ± 16	R3: 5.7 ± 0.8
				R4: 193 ± 12	R4: 102 ± 43	R4: 6.0 ± 0.5
UNS S32750 (SAF 2507)	Up to 3	yes	no	NA	NA	NA	[77]
Stellite 21	Up to 3	yes	no	NA	NA	NA	[77]

* In reference [72], R = is intended as Ti64 fresh powder with medium oxygen content, while R10 is recycled Ti64 with high oxygen content. All of these values are obtained after heat treatment. ** Number of reuses not specified. *** Mechanical properties derived from figures.

presents an overview of the information available in the literature regarding the reuse of metal powders in AM processes with different technologies and materials. As shown in the table, powders can be reused for a variable number of prints, with sieving performed between each print to remove larger agglomerates.

As reported by Saboori et al. [76], the use of recycled powder mixed with a virgin one resulted in a very different mechanical behavior, with lower values of yield strength and ultimate tensile strength and a 50% reduction in the elongation compared to samples fabricated with the same process parameters and with virgin powder. The authors also found that the occurrence of large inclusions, such as oxides rich in Mn and Si with a diameter of 7–15 μm, is responsible for the significant variance in strain to fracture seen in the two distinct sets of samples made using recycled and new powder. These oxides were also present in samples fabricated with fresh powder, but with a lower frequency. As a result of powder reuse, Pereira et al. [77] found that the average micro-hardness of the deposited Stellite^® 21 and SAF 2507 super duplex stainless steel samples decreases as powder reuse increases. This decline is attributed to a rise in oxygen content, along with a decrease in carbon and manganese in the cobalt-based alloy, and the formation of oxides and austenite in the SDSS alloy.

Kiani et al. [69] showed that when AlSi10Mg is considered as powder feedstock for LP-DED, the ultimate tensile strength (UTS), which is the maximum stress the material can withstand, drops from 329 MPa in parts built from virgin powder to 303 MPa after just one cycle of powder reuse. Furthermore, along with the decrease in UTS, there is also a notable drop in the elongation to failure, the measure of how much the part can stretch before breaking. The elongation drops from 9.7% in the original build to 5% after one cycle of powder reuse, indicating that the parts become less ductile and more brittle after recycling. It is worth mentioning that while the ultimate tensile strength and the elongation to failure exhibited a marked decrease after the first recycling cycle, further cycles of recycling did not induce significant alterations in these properties. This finding indicates that the primary impact occurs during the initial reuse cycle, with subsequent cycles maintaining a consistent level of mechanical performance. The decline in mechanical properties, particularly in UTS and ductility, is likely to be associated with minor increases in lack-of-fusion porosity during the deposition process. It has been demonstrated that even a minimal diminution in powder flowability, occasioned by repeated recycling, has the potential to give rise to porosity issues. These, in turn, have been shown to have a deleterious effect on the strength of the final components.

It is worth mentioning the study by Shalnova et al. [75] comparing various powder mixtures containing 0%, 10%, 25%, and 50% recycled Ti-6Al-4V powder. The paper showed that the incorporation of recycled powder content up to 50% does not have a significant negative impact on the essential mechanical properties, such as strength, ductility, or impact toughness. Furthermore, the microstructural features in the as-built state remain consistent, with only slight modifications observed post-annealing [75].

3. Laser–Powder Interaction

The dynamic behavior of powder particles at the interface with the laser is a critical factor in determining the quality and integrity of the manufactured components in laser-based additive manufacturing. As discussed previously, the physical and chemical properties of metal powders, such as particle size, shape, composition, and flowability, as well as their evolution through multiple reuse cycles, critically influence not only the deposition process but also the way these particles interact with the laser beam. These interactions govern key phenomena, such as absorption, scattering, melting, and evaporation, all of which directly affect the energy distribution within the melt pool and, ultimately, the microstructural and mechanical properties of the final part [78]. Understanding the complex interplay between the laser and powder particles therefore requires an integrated approach in which powder properties and process parameters are not considered in isolation but as co-dependent factors shaping LP-DED process performance and output.

The trajectory, geometry, and physicochemical properties of the starting powder particles play a critical role in determining the quality and integrity of the final product. The melting and deposition process is inherently probabilistic; it is subject to the effects of laser power, powder size distribution, powder flow rate, material properties, and scanning speed [79].

The incident energy plays a major role in shaping the final microstructure of the part. This, in turn, is influenced by the laser power, deposition rate, and inert gas flow (shielding and carrier gas flow), as well as the efficiency of the metal powder flow [80]. It should also be noted that before reaching the substrate surface, the beam interacts with the gas/powder flow, resulting in a number of thermal events. For example, through absorption and scattering, the powder particles alter the distribution of laser intensity and attenuate the laser energy. As a result of absorbing laser energy, the particles experience temperature increases, melt, and even evaporate; convection and radiation are two ways in which the high-temperature particles exchange energy with their environment [81].

3.1. Powder Injection

During the interaction of the particle–carrier gas, the carrier gas provides the initial velocity of the powders and exerts a force upon them, driving them towards the molten pool. In the power–powder interaction, the powder absorbs part of the laser energy. This results in a subsequent rise in temperature and, concurrently, a portion of the energy absorbed by the powders is radiated. It has been established that the distribution of powder streams is a pivotal parameter in attaining the optimal conditions for LP-DED [82,83,84].

Powder injection can be divided into three concepts (Figure 3): discontinuous coaxial powder injection, which involves feeding three or more powder streams coaxially to the laser beam; continuous coaxial injection, which creates a cone of powder streams enclosing the laser beam; and off-axis injection, which involves feeding a single powder stream laterally into the laser beam. Because the clad track is dependent on the scan direction, the off-axis powder injection nozzle is only appropriate for 2D applications. The coaxial powder injection nozzle is used in 2D and 3D part reconditioning procedures because many of the parts (such airfoil tips) need tiny tracks. The main benefit of discontinuous coaxial powder injection is that it allows the deposition head to be tilted without changing the powder stream [85].

In recent years, several computational models of varying complexities and incorporating different physical phenomena have been proposed. These account for the powder particle stream [82,86,87], melt thermo-hydrodynamic particle-injecting models [88,89,90], and melt thermo-hydrodynamic homogenous models [91,92,93,94].

Aggarwal et al. [95] developed an integrated framework for directed energy deposition consisting of coupled coaxial gas–powder flow modeling, melt pool thermo-hydrodynamics, and particle impingement modeling in order to investigate particle–melt interaction. One interesting insight, due to the different interactions of the powder particles with the heat source, is that the temperature range is affected with very high variability.

3.2. Particle Incorporation

The incorporation of metal powder particles is influenced by several variables, such as processing parameters, substrate material and surface finish, and the powder composition itself. The deposition process is significantly affected by their properties, including specific heat capacity, diffusivity, thermal conductivity, and absorptivity, for the specific laser wavelength used [96,97].

The quality of the LP-DED process depends on the thermal state of the melt pool and how quickly the powder is incorporated into it. Ideally, the most efficient operation of LP-DED would involve rapid and complete melting of the incoming powder particles. However, as reported by Khairallah et al. [98], simulations using a static laser beam with a diameter of 600 mm, power of 200 W, and deposition rate of 4 g/min do not show that particles are submerged in the melt pool upon high-speed impact, like a stone thrown over a pond of water. In fact, the melt pool holds most of the particles in place, while the laser source completes the incorporation of the powder particles from top to bottom, with dwell time due to inefficient laser absorption coupled with laser shadowing of the incoming powder. A trend outlined by the simulations is that large particles (diameter greater than 140 microns) tend to submerge in place, while smaller particles float on top of the melt pool and are later radially melted by the laser source.

During floating, a thin layer of gas separating the particle and the surface is trapped at the end during melting, a theory also outlined by Zhao et al. [99], based on the assumption that the shell-like porosities resulting from the LP-DED fabrication of AlMgScZr alloys are due to a thin layer of air separating the particle and the surface of the melt pool.

If the temperature difference between the drop and the melt pool is greater than a threshold, the coalescence of liquid droplets in a pool indicates that a thin layer of air can form on the drop surface, separating the drop from the pool surface [100]. In an improperly configured LP-DED system, there may be an excessive temperature difference between the melt pool and the powder particles falling into it. This is because the laser beam above the melt pool may have previously overheated or underheated the powder particles. As a result, a thin layer of air may form on the powder particles, increasing the possibility of gas entrapment in the melt pool. Shell pores will form if the trapped gas cannot escape from the melt pool, such as, for example, due to rapid solidification [99]. Another possible cause is the formation of an oxide layer or even hydroxide in the case of reactive alloys (i.e., aluminium alloys), with partial decomposition during interaction with the laser source.

The catchment efficiency in LP-DED, or the percentage of incoming powder particles that are wasted and do not contribute to the build process, would begin to be affected by the incorporation time (defined as the time required for powder particles to be fully melted in the laser bead) as the deposition rate increased [97,98].

According to Prasad et al. [97], a sizable island of unmelted powder is seen floating and whirling over the melt pool, forcing incoming particles to bounce around. The ricochet effect may be worsened for highly oxidized metals because the oxide layer may further slow down complete coalescence and increase the part’s porosity. Therefore, it can be stated that the likelihood that the arriving particles may scatter, bounce, or clump together onto an unmelted powder surface increases with the length of time the particles stay floating on the surface (Figure 4) [97].

Molten particles can occasionally collide and merge in flight prior to impact due to the three-dimensional velocity field of the powder particles. During in-flight heating, the trajectory of the powder particles varies from the nozzle exit to the substrate surface; at this point, the energy density distribution of the laser beam (Gaussian in the simulation considered) causes the temperature of the particles to rise heterogeneously. Some of the powder particles can reach temperatures as high as the evaporation temperature. However, the magnitude of the impact velocity of the powder particles on the substrate surface is almost uniform. The maximum melt pool temperature can increase or decrease during impingement depending on the powder particle trajectory, the in-flight temperature rise, and the localized temperature of the impact zone in the melt pool. Therefore, it was observed during the simulations that a wave forms in the melt pool after impact, causing mixing and a drop in the maximum temperature.

The experimental study by Haley et al. [101] supports the work of Prasad [97]. The work, which was carried out by monitoring the interaction between the powder particles and the melt pool (the powder used was 316L stainless steel in an Optomec 750 LENS system), further demonstrated that the power particles create a ripple when incorporated into the melt pool and float above the surface of the melt pool before sinking below it (Figure 5). The clear morphological differences between particles that have liquefied in flight and then refrozen on the surface and particles that have frozen while floating confirm this surface residence time. The residence time of the particles on the surface of the melt pool prevents it from absorbing more particles.

Considering the necessity to build thin walls using LP-DED, Liu et al. [83] developed an analytical model to describe the effects of powder incorporation following experimental verification. Their conclusion was that the interaction between powder concentration distribution and power density distribution affects the manufacturing of thin wall parts in coaxial laser cladding, with fluctuations of the powder feed rate playing a pivotal role in degrading the integrity of the thin wall’s growth. In single-pass laser cladding, a rise in concentration distribution causes a decrease in the walls’ thickness and a growth rate increase.

3.3. Laser Attenuation

The laser attenuation generated by the interaction between the high-energy beam and the powder particles has been studied and modeled for laser cladding since the early 1990s [102,103,104]. The first attenuation models developed were based on the shadow cast by the powder particles on the laser beam. These models assumed that beam attenuation was proportional to the area occupied by the powder particles in each plane. Typical simplifications of shadow-based attenuation models were considered to be a valid theoretical framework, with these simplifications including neglecting the effect of beam divergence and considering constant particle size. Light scattering was also considered [102,105,106]. The conclusions drawn for this technique can be extended to directed energy deposition and help to understand the interaction between the feedstock to be entrained in the melt pool and the source of thermal energy responsible for melting and solidification.

The laser–powder coupling model [81] takes into account laser attenuation and concurrent particle heating due to the laser’s thermal energy. The first is the consequence of both particles grouping in powder streams and the subsequent absorption and scattering processes. On the other hand, the phenomenon of particle heating is attributed to the light–heat conversion effect when said particles are irradiated by the beam. Intensity attenuation due to a solid volume can be modeled using the Beer–Lambert equation, and the intensity loss (IL_i,j,k), as shown in Figure 6 [81], can be described with the following equation (Equation (1)):

$${IL}_{i,j,k}=\left\{\begin{array}{c}{I}_{i,j,k-1}\left[1-exp\left(-\pi r_p^2{Q}_{ext}{N}_{dcps}^k∆z\right)\right] for k\ge 1\\ 0 for k=0\end{array}\right.$$

(1)

where Q_ext represents the extinction coefficient of powder material and $N_{dcps}^k$ is the particle concentration of the volume element ΔV_k.

The total intensity loss can be calculated by accounting for the movement from point to point, and the final integration will lead to a formulation of the total intensity loss (TIL), outlined as follows (Equation (2)):

$${TPL}_k=\iint {\sum }_{n=1}^k{I}_{i,j,n-1}\left[1-exp\left(-\pi r_p^2{Q}_{ext}{N}_{dcps}^n∆z\right)\right]dxdy$$

(2)

Meanwhile, as the particles move across the laser beam, they experience transitory heat transfer events, such as conduction, convection, and radiation. The heat transport model of the laser–powder interaction can be described using the lumped parameter method (LPM) by accounting for the convection coefficient, the thermal conductivity, and the sphericity of powder particles. Considering the LPM and the energy balance, laser–powder interaction can be described as follows (Equations (3) and (4)):

$$m_p{c}_p{\displaystyle \frac{d{T}_p}{dt}}={\displaystyle \frac{1}{4}}{\delta }_I{S}_{p}A{I}_{att}-h{S}_p\left(T_p-T_0\right)-\epsilon \sigma S_p\left(T_p^4-T_0^4\right)-{\delta }_m{L}_m{\displaystyle \frac{d{m}_m}{dt}}$$

(3)

$${\delta }_I\left\{\begin{array}{c}0 Other \\ 1 Coupling zone,\end{array}\right. {\delta }_m=\left\{\begin{array}{c}0 T_p\ne T_m\\ 1 T_p=T_m\end{array}\right.$$

(4)

where A denotes the laser’s absorptivity, c_p is the powder-specific heat, T₀ and T_p refer to the surrounding temperature and the transient temperature of the particle, respectively, T_m stands for the melting temperature, S_p is the surface area of a single particle, σ and ε are the Stefan–Boltzmann constant and surface emissivity, separately, L_m is the latent heat of fusion, and m_m is the melted mass of the particle.

Huang et al. [107] further refined a numerical model that integrates the primary physical phenomena involved in the process. This includes the interaction between the attenuated laser beam, the thermally activated powder stream, and the semi-infinite substrate. The model incorporates the spatial distribution of both particle concentration and laser intensity across the computational domain. This model also includes powder entrainment efficiency considering both the powder spatial distribution and the melt pool shape variation. The result of laser attenuation in the case of Inconel 625 powder is shown in Figure 7.

Particle heating through laser radiation is one of the main issues for cladding quality and efficiency control, and it has been studied for cladding since the late 1980s to simulate and verify powder temperature profiles through experimental analysis [108,109,110,111].

The effect on LP-DED process efficiency has been further analyzed by Zhang et al. [112] and Tan et al. [113]. The former demonstrated that by incorporating a coupled electromagnetic wave heating (EWH) model, it is possible to simulate the laser–particle interactions. The authors found that the absorption of laser energy by powder particles reduces the effective heat transfer into the deposition layer, which directly affects the temperature fields during LP-DED. The reduced heat input results in clearly measurable changes in the thermal profile and the melt pool dimensions observed over multiple layers, allowing the local strain field of the deposited sample to be correlated with stress and localized strain.

Results from Tan et al. [113] showed that particles moving along different trajectories within the laser beam experience markedly different heating rates. In particular, an increase in the powder stream incident angle leads to a longer interaction distance between the laser beam and the powder particles; when the incident angle is increased from 50° to 65°, the temperature of the powder particles at the deposition surface increases significantly (up to 54.4%). Particles that fall close to the edge of the laser spot on the theoretical deposition surface have a final temperature that is 300% higher than particles that do not cross the central axis of the laser beam. In addition, the deposition quality can be improved at a lower laser power because the larger angle of powder incidence and the lower powder incidence velocity improve the laser–powder interaction and the temperature of the powder reaching the deposition surface. In terms of particle velocity, Tan et al. [113] showed that the speed of the powder particles significantly impacts the thermal energy they accumulate before deposition. Slower particles have more time to absorb laser energy, leading to a higher temperature when reaching the deposition surface. As a result, a reduction in particle speed from 9.3 m/s to 3.9 m/s results in a 144% increase in the particle’s temperature at the theoretical deposition surface; that is, lower particle speeds facilitate better melting and bonding [113].

Particle tracking and simulation of transport phenomena are additional key points to enhance the reliability of combining multiple materials during LP-DED fabrication (Figure 8) [114].

The theoretical and experimental investigation performed by Liu et al. [115] showed that laser power attenuation is an exponential function, and it is determined by the powder feed rate, the particle’s moving speed, spraying angles, waist positions and diameters of the laser beam and powder flow, the grain’s diameter, and the run of the laser beam through the powder flow. In addition, the attenuation of laser power increases with the powder feed rate or the passage of the laser beam through the powder stream.

3.4. Spatter Formation

The ejection of melt from the melt pool is known as scatter. The literature describes the origins and consequences of spatter production in powder bed fusion, laser welding, and cutting. When a section of the melt acquires sufficient momentum perpendicular to the melt surface to separate from the liquid, the spatter is formed; these events have been associated with vaporization, as well, as usual for laser powder bed fusion and laser welding [116,117]. For the latter, boiling and bursts in the melt pool due to vapor pressure gradients are also reported in the literature [97,111,118].

Different mechanisms for spatter formation have been proposed for several powder-based additive manufacturing techniques. In the case of LP-DED, the formation of a powder island in the melt pool as a result of the applied laser power not sufficiently melting the powder particles is the cause of spatter production [119]. Some of the metal droplets formed by the powder particles near the edge of the melt pool burst and either integrate into the melt pool or are ejected as spatter due to surface tension or boiling effects [97]. Recent high-speed imaging studies indicate that the two mechanisms by which spatter particles are generated in the powder bed fusion process are recoil pressure and vapor-driven entrainment, with the former being predominant [120].

In addition to wasting material, powder spattering lowers the quality of the part as a whole. For example, previous research has shown that keyhole apertures are created when molten particles are ejected from the deposition zone. Regardless of where they come from, the frequently oxidized spatter particles may also deposit elsewhere on the build surface, leading to unexpected morphological and chemical heterogeneity, as well as uneven surface morphology [79,121].

It is worth pointing out that the mechanisms outlined by Prasad et al. [97] were applicable to different steels (wear-resistant, as per [97], and 316L), and the formation of a similar powder island was seen with the same experimental setup for Rockit 401 and Satellite 71 when using a multi-jet coaxial nozzle. As the deposition rate of the process increases, the probability of powder grains ricocheting off of other powder grains in the melt pool increases. With 316L powder, the size of the island decreased with increasing laser power.

Recently, Niu et al. [122] demonstrated the impact of spatter formation on the prediction of surface roughness for LP-DED while suggesting that the spatter itself can be divided into two categories, namely, particles (i.e., hot/cold ejecta) and droplets (Figure 9). With regard to the former, the high viscosity of the molten metal powder and the impact of the metal vapor produced prevent a small proportion of the unmelted particles from entering the melt pool [123]. When these particles leave the melt pool, they form particle spatters. In contrast, a large number of particles manage to adhere to the surface of the melt pool after penetrating the metal vapor. However, due to surface tension, they are unable to penetrate the interior of the melt and are instead transported to the rear of the melt pool by the Marangoni flow. As the melt pool solidifies, these particles adhere to the deposition layer, increasing the surface roughness. At the same time, the violent fluctuations within the melt pool cause droplet spatter to form, resulting in the detachment of liquid metal droplets. Another source of spatter [97] is the powder grains that bounce off of the island, which are heated to partial or complete melting by the laser beam and land in the cladding zone or elsewhere.

As reported by Naesstroem et al. [119], the materials and properties of the build plate can also play a critical role in the formation of spatter from LP-DED fabrication. As iron oxides and steel exhibit similar melting temperatures, the presence of an oxide layer due to surface corrosion on the build plate may promote melt pool instability and enhance spatter generation during LP-DED processing. A significant oxide layer on the substrate caused spatter from the melt pool in another study using wear-resistant steels (Figure 10). There was no noticeable spatter when machining a substrate of the same material.

3.5. Effects of the Carrier/Shielding Gas Flow

The carrier gas flow oversees moving the powder particles from the powder feeders to the exit of the LP-DED nozzle, while the shaping or shielding gas is responsible for the efficient activity of the laser beam, as well as avoiding extreme oxidation events taking place during manufacturing. The gas flow is directly influenced by the characteristic design of the LP-DED nozzle, as the carrier gas flowing from the discrete powder nozzles will be very different from that of a coaxial powder nozzle. This will have an impact on the equipment’s behavior during fabrication, the powder catchment’s efficiency, and the overall performance of the technology [124].

The gas flow not only carries the powder but also influences how quickly and uniformly the particles accelerate as they leave the nozzle. Although the particles are propelled by the carrier gas, friction (drag) and collisions with the nozzle walls will tend to keep their velocity below gas velocity [125]. Gas flow conditions are also critical in controlling the diameter and concentration of the particle stream (Figure 11). Higher gas velocity increases the forces acting on the particles, which can result in a more accelerated but thinner stream. However, due to the fixed physical parameters of the nozzle, particles may experience multiple collisions, and these interactions further modify how well the particles deposit on the substrate.

Ferreira et al. [125] showed, in particular, that an increase in the axial and shaping gas flow rates has a measurable effect on the focal plane position of the powder stream; a 2 L/min increase in the axial or shaping gas flow rate can pull down the focus plane by approximately 1 to 1.5 mm. In contrast, an increase in the carrier gas flow shows an opposite behavior, where a 4 L/min increase can raise the convergence plane by about 1.5 mm. In terms of powder flow rate effects, Ferreira et al. [125] reported that with lower powder flow rates (e.g., 4 or 6 g/min), deposition tends to be more precise because the powder stream is tightly focused. However, when the powder flow rate increases to 12 g/min, the stream diameter can expand from 2 mm to 2.6 mm at the focal plane, primarily due to more frequent particle collisions, which lead to greater dispersion and less control over the deposition width [125].

The interplay between the gas flow and the particles means that there is an inherent balancing act; while a strong gas flow is necessary for transporting particles, it can also cause excessive dispersion if not properly controlled, ultimately affecting the efficiency of the particle cementation process [86,126].

4. Nozzle Configurations in LP-DED

The previous sections of this review paper emphasized the significance of laser-powder interaction in determining the efficiency of energy absorption and the dynamics of melting and deposition. However, this interaction does not occur in isolation, and it is strongly influenced by how the powder is delivered to the laser focal zone. In this context, the nozzle’s configuration becomes a key enabling factor, as it dictates the spatial and temporal characteristics of the powder stream, modulating its interaction with the laser beam. The design of the deposition head and the nozzle itself directly affects how the metal powder interacts with the laser beam, influencing powder catchment efficiency, melt pool stability, and, ultimately, the mechanical properties of the final build. The main nozzle configurations in LP-DED systems include coaxial (both continuous and discrete) and off-axis designs, as shown in Figure 12 [127,128]. Each configuration comes with its own set of advantages and limitations depending on the intended application, substrate geometry, and desired deposition quality. This section provides a detailed exploration of these nozzle types, supported by both numerical simulations and experimental studies that elucidate their influence on powder and gas dynamics.

4.1. Coaxial Nozzles: Design and Performance

Coaxial nozzles are most commonly used in LP-DED due to their ability to deliver metal powder symmetrically around the laser beam. This design ensures consistent deposition quality and high process efficiency, making it ideal for a wide range of applications. Coaxial deposition nozzles use a traditional powder design layout. It includes shielding gas arrangements to protect the laser optics from metal fumes and dust, paths for a concentrated powder stream, and shielding gas to protect the melt pool [129]. Coaxial nozzles are divided into continuous and discrete designs, each with unique powder flow characteristics influenced by geometric configurations and operating parameters. Discrete coaxial nozzles were developed as an evolution of off-axis nozzles; the latter is described in Section 4.2.

4.1.1. Continuous Coaxial Nozzles

Continuous coaxial nozzles (Figure 12b), widely utilized in Laser Powder Directed Energy Deposition (LP-DED), are characterized by their omnidirectional capabilities and their ability to deliver a consistent, homogeneous powder flow [128]. The fundamental design comprises two coaxial cones separated by a defined annular gap through which powder, carrier gas, and shielding gas are introduced. This annular outlet ensures that the powder stream encloses the laser beam in a conical shape, facilitating precise material deposition within the melt pool [130]. The nozzle design allows for adjustments in the cone angles, which can shift the focal spot of the powder stream along the Z-direction, aligning it precisely with the melt pool. This capability is essential for optimizing catchment efficiency [130]. The powder feeding of continuous coaxial nozzles is characterized by a consistent and uniform flow of powder particles into the melt pool. The powder, mixed with carrier gas, is directed through the annular channel surrounding the laser beam, forming a conical powder stream that converges at the focal point of the laser.

A prominent feature of continuous coaxial nozzles is the inclusion of a powder holding space or an accumulation chamber. This chamber, positioned between the metering station and the annular region, serves as a mixing zone for multiple feed materials, either blended uniformly or arranged to form a compositional gradient. This capability is particularly beneficial for the production of Functionally Graded Materials (FGMs), enabling the creation of components with spatially varying material properties [19,130,131].

Geometrical adjustability is another critical characteristic. Depending on the machine’s configuration, the cone angles can be modified to optimize the powder stream’s focus and adjust the interaction zone between the powder and the laser beam. This flexibility enhances deposition efficiency and precision. However, the tilting capability of continuous coaxial nozzles is limited due to gravitational effects on the powder cone. Experiments developed by Nasiri and Movahhedy [132] indicate satisfactory performance at tilt angles up to 20° [128,133]. Continuous coaxial nozzles can be further classified based on the nozzle tip’s configuration into inward and outward designs [128,134].

• Inward designs ensure powder convergence by bouncing particles off of the extended nozzle walls, enhancing focus at the melt pool. This configuration minimizes the ejecting effects of shielding gases, promoting stable powder distribution.
• Outward designs, often referred to as Extreme High-Speed Laser Material Deposition (EHLA) nozzles, melt the powder above the substrate, allowing for the deposition of liquid metal into the melt pool. This approach significantly increases feed rates, reduces heat input, minimizes distortions, and ensures a reduced heat-affected zone (HAZ). However, EHLA nozzles are less suitable for non-rotational parts and reflective materials like pure copper [135].

Numerical simulations, including Computational Fluid Dynamics (CFD) models, have demonstrated that inter-cone channels within these nozzles contribute to the formation of a stable, convergent powder stream, as reported in the study conducted by Arrizubieta et al. [136]. Conversely, nozzles lacking these channels tend to produce divergent, spray-like streams without a defined focal point. The inclusion of these channels is common but not universal, indicating variability in design practices [19]. Figure 13 schematically represents these inter-cone channels and their impact on powder flow dynamics [19,137]. The presence of inter-cone channels, even though they are common, is not universal. Ju et al. [138] identified key geometric parameters and carried out design optimization to achieve higher powder catchment efficiency with maximum depth of focus (the distance up to which the powder stream remains convergent after having converged at the focal plane). Through CFD analysis, the optimum structural dimension for the given laser power was obtained. It is noteworthy that the nozzle design featured in the study did not include any inter-cone channels.

Continuous coaxial nozzles offer several advantages in laser-based LP-DED, including high powder catchment efficiency, uniform material distribution, and directional independence, key for complex geometries and multi-axis operations [128,130]. The ability to adjust cone angles enables precise alignment of the powder stream’s focal point with the melt pool along the Z-axis, enhancing deposition consistency [130]. Nasiri and Movahhedy [132] demonstrated that optimizing nozzle geometry and gas flow can achieve catchment efficiencies up to 80% while also improving melt pool stability and minimizing oxidation. However, the performance of these nozzles remains highly sensitive to both geometric design and process parameters, requiring careful calibration for specific manufacturing scenarios [128,130,132].

4.1.2. Discrete Coaxial Nozzles

The discrete coaxial nozzle is a critical component in LP-DED, offering a structured approach to powder delivery. Unlike continuous coaxial nozzles, which provide an annular and uniform powder stream, discrete coaxial nozzles utilize multiple (typically three or four) individual powder jets, each delivered through independent and spatially arranged channels. These channels are oriented to form a converging powder stream at the focal point of the laser beam [130].

Key geometric parameters for the design of discrete coaxial nozzles, as identified by Wu et al. [139], are schematically shown in Figure 14, with a generic layout that is common in all variants. The geometry of a discrete coaxial nozzle significantly influences powder stream characteristics. Key design parameters include the injection radius (R_in), injection angle (ω), powder channel exit diameter (D_n), and standoff distance (S_f) [6]. The orientation of these channels determines the focal point of the powder flow relative to the laser beam and the substrate [19].

One of the defining features of discrete coaxial nozzles is their ability to accommodate a multidirectional deposition approach, making them particularly useful for complex geometries and multi-axis builds. This advantage is achieved by tilting the nozzles up to 180°, allowing for greater flexibility in powder delivery [128,133,140]. The main disadvantage of this nozzle design is that the powder flow is not uniform because of its working principles.

Very few studies [141] in the literature exist comparing continuous and discrete coaxial nozzles, making it difficult for the end user to select the one that best suits their application. A comparative analysis of continuous and discrete coaxial nozzle designs has demonstrated that continuous nozzles exhibit superior performance. Specifically, Zhong et al. [141] evaluated their effectiveness in the fabrication of Inconel 718 components, revealing that continuous coaxial nozzles produced deposition tracks with an 18% greater height, a higher material deposition rate, and a powder velocity of 8 m/s. In contrast, the discrete coaxial three-jet nozzle, while offering enhanced accessibility for complex three-dimensional applications due to the powder bouncing effect, exhibited lower powder capture efficiency compared to the continuous coaxial nozzle configuration [19,141]. Furthermore, variations in the powder stream density of discrete nozzles, depending on the deposition direction, were observed to influence the resulting deposit geometry [19]. Results from experiments by Zhong et al. [141] comparing the discrete (three-jet) and the continuous coaxial nozzle are summarized in Table 1 on page 12 of the original article [19]. In conclusion, regarding the particle intensity distribution (PID), both nozzle configurations display a comparable pattern characterized by a concentrated accumulation of particles at the focal center, followed by a rapid decrease in particle density along the radial axis. However, the continuous coaxial nozzle demonstrates superior powder capture efficiency, as evidenced by the larger clad area, in comparison to the discrete three-jet nozzle. This enhanced efficiency is attributed to the higher particle velocities achieved with the continuous coaxial design, which facilitate deeper penetration of particles into the melt pool. The increased velocity allows particles to overcome the melt pool’s surface tension, thereby reducing the likelihood of particle rebound.

In detail, discrete coaxial nozzles exhibit distinct differences in performance compared to their continuous counterparts.

• Powder Mass Concentration: The concentration of powder particles at the focal spot is inversely related to the nozzle exit diameter and injection radius. Larger values for these parameters increase the focal spot size, reducing powder concentration.
• Standoff Distance: This parameter negatively correlates with the injection angle and the nozzle exit’s diameter, meaning that an increase in these values shortens the standoff distance and shifts the focal plane closer to the nozzle exit [19].
• Particle Velocity and Catchment Efficiency: The catchment efficiency of discrete coaxial nozzles is generally lower than that of continuous nozzles due to the bouncing effect of powder particles. The lower velocity of powder streams in discrete nozzles (3.76 m/s) compared to continuous nozzles (11.28 m/s) further contributes to this inefficiency [141].
• Deposition Rate and Dilution: Discrete coaxial nozzles have a lower deposition rate (2.173 kg/h) compared to continuous coaxial nozzles (2.485 kg/h). However, they exhibit significantly reduced dilution, with a maximum of 129 μm compared to 506 μm in continuous nozzles [141].
• Multi-Material Capability: One of the primary advantages of discrete coaxial nozzles is their suitability for Functionally Graded Materials (FGMs). The independent control of multiple powder streams allows for localized material composition variations, a capability that is difficult to achieve with continuous nozzles [19].

Despite their advantages in multi-material deposition and multi-axis capability, discrete coaxial nozzles present several challenges [128,130]:

• Non-Uniform Powder Flow: Unlike continuous coaxial nozzles, discrete designs do not ensure even powder distribution, potentially leading to inconsistencies in the deposited track geometry.
• Lower Catchment Efficiency: Due to the discrete nature of the powder streams, some powder particles fail to enter the melt pool, increasing material waste.
• Dependency on Process Parameters: Achieving optimal deposition quality requires fine-tuning of standoff distance, injection angle, and nozzle orientation

4.2. Off-Axis/Lateral Feed Nozzles: Design and Performance

The off-axis powder feeding nozzle is primarily suitable for two-dimensional (2D) applications, as the deposited track is heavily dependent on the scan direction. However, its unidirectional nature and relatively low efficiency present certain limitations. These nozzles are commonly employed for coating cylindrical components, where the deposition technique follows a unidirectional approach [130,142].

The characteristics of the gas-powder stream, together with the size of the melt pool, have a significant influence on the shape of the final part, powder efficiency, surface roughness, and geometric accuracy. The gas-powder stream is determined by the design of the nozzle, including its vertical orientation, its vertical distance from the melt pool (standoff), and the velocities of the primary and secondary gas flows. Typically, off-axis nozzles are axisymmetric, with a circular cross-section and a converging–diverging internal geometry, allowing them to generate a high-velocity assisted gas jet while minimizing kinetic energy loss due to the absence of shock waves. Off-axis nozzles are designed with a slight tilt relative to the laser beam to mitigate choking effects and prevent blockages in the processing field. These nozzles are often used to overlay the outside diameter of cylindrical components and for applications where coaxial nozzles are impractical, such as in grooves or confined geometries. Although off-axis nozzles offer relatively high deposition rates, a notable drawback is the limited shape consistency of the deposited layers, which are only optimal in a single direction, resulting in limited deposition efficiency [130,143].

Compared to off-axis designs, coaxial nozzles, particularly the continuous coaxial configuration, exhibit significantly improved powder distribution characteristics. In coaxial designs, the powder stream is symmetrically arranged around the laser beam, enabling a focused conical powder envelope that promotes homogenous deposition and a stable melt pool. This geometry leads to higher powder capture efficiency and more consistent clad geometry. Off-axis nozzles, in contrast, deliver powder laterally, resulting in anisotropic powder streams and increased divergence as the powder approaches the melt pool. These inconsistencies lead to reduced catchment efficiency of approximately 20% versus over 30% for coaxial systems and higher sensitivity to scan direction and positioning relative to the melt pool, as shown by Silva et al. [144]. Consequently, coaxial systems, especially those with optimized inter-cone channel designs, are better suited for multi-directional, 3D builds requiring high deposition precision and uniformity. Li et al. [145] further explored the influence of scanning direction, demonstrating that powder catchment efficiency varies depending on feed positioning relative to the melt pool.

While nozzle geometry, particularly the advantages of coaxial configurations, plays a critical role in determining powder catchment efficiency and deposition stability, it must be considered in conjunction with the broader set of process parameters that govern the LP-DED system as a whole.

Typically, process parameters are balanced using the global energy density (GED) parameter, defined as P/vd, or, rather, the ratio between the laser power (P) and the combination of deposition speed (v) and spot size (d). Although this allows for obtaining a quick estimation of the switch between the keyhole mode and the lack of fusion to find a reliable range of processability parameters, the LP-DED system is much more complex. The parameters highlighted in the present paper are reported in the Ishikawa diagram in Figure 15 and represent a major contribution of the present paper to defining the current state of art of LP-DED systems and the associated variables.

5. Powder Distribution Behavior and Influence of Process Parameters on Nozzle Performance in LP-DED

In Laser Powder Directed Energy Deposition (LP-DED), the design and configuration of nozzles, whether coaxial or off-axis, play a critical role in determining the efficiency and quality of the process. Typically, the coaxial design facilitates a coordinated interaction between the powder stream and the laser beam, forming a stable powder cloud structure [146,147,148]. However, laser power reduction due to absorption and scattering by powder particles is a challenge in both configurations and depends on factors like the Gaussian dispersion of laser beam intensity, uniform spherical powder particle size, and steady-state powder flow characteristics at the nozzle outlet [115].

The performance of nozzles in LP-DED is significantly influenced by process parameters, particularly those affecting powder and gas flow dynamics. Various numerical simulations have been developed to analyze the dynamics of these flows, providing insights into the optimization of nozzle geometries. The geometry of coaxial nozzles plays a crucial role in determining the characteristics and stability of powder flow. Lin et al. studied different coaxial nozzle exit configurations and observed a strong correlation between nozzle design and powder flow stability [134]. Various parameters, including particle-wall interactions within the nozzle, influence the resultant powder distribution at the exit [149].

The results indicate that non-elastic interactions promote a more stable and well-controlled powder stream, leading to a more uniform and focused deposition area. In contrast, elastic rebounds cause powder divergence, reducing deposition efficiency. Moreover, it has been demonstrated that extending the powder conveying tube beyond 10 mm enhances powder stream convergence, leading to more uniform Gaussian distribution [150].

Pan and Liou further analyzed the effect of the powder delivery tube on output performance, showing that the width and outer diameter of the tube dictate the final powder stream structure [151]. Liu et al. [152] studied the evolution of powder distribution within the nozzle and found that recurrent collisions of powder particles against the inner walls significantly affect the trajectory and spread of the powder stream at the outlet [152]. They concluded that the powder diameter and the restitution coefficient are critical factors influencing the powder concentration’s distribution. Consequently, selecting a well-defined powder size range can improve the uniformity and predictability of the powder stream.

In addition to flow dynamics, powder material properties play a significant role in coating performance. Some of the most critical parameters are powder mass and density, which directly affect the size and position of the convergence zone (Figure 16) [153].

These findings suggest that materials with different densities require specific nozzle adjustments to maintain optimum powder flow convergence and deposition efficiency.

Another key factor influencing powder flow is particle size distribution. Numerical modeling has been used to evaluate the effect of different size distributions, and it has shown that larger particles have similar trajectories to smaller particles, thus changing the shape and position of the focal region, i.e., finer particles produce a narrower focal region compared to coarser particles (i.e., 20 μm vs. 80 μm) [149].

To consolidate the analysis presented in this section, the following comparison summarizes the main differences observed between coaxial and off-axis nozzles in terms of powder distribution behavior, deposition precision, and material utilization efficiency. The precision and efficiency of deposition are strongly influenced by the nozzle type. Coaxial nozzles provide a centralized, controlled powder flow aligned with the laser focal point, reducing thermal gradients and enhancing powder–laser coupling. As demonstrated in numerical and experimental studies (e.g., Zhong et al. [141]), continuous coaxial nozzles deliver higher powder velocities (11.28 m/s vs. 3.76 m/s in discrete nozzles), resulting in superior penetration into the melt pool and reduced particle rebound. This effect increases the deposition rate and reduces porosity. Off-axis nozzles, while simpler and more compact, suffer from asymmetric powder delivery and reduced powder convergence, compromising deposition uniformity. Moreover, the inability of off-axis systems to adapt to changing build directions limits their application to simpler geometries. In contrast, coaxial nozzles facilitate uniform builds across complex surfaces due to their omnidirectional powder delivery. These insights confirm that nozzle geometry is not only a passive component but a key driver in determining LP-DED build quality and material efficiency.

5.1. Numerical and Experimental Investigation of Gas-Powder Flow

The optimization of nozzle geometries in LP-DED processes has been driven primarily by numerical modeling, which provides critical insight into the behavior of powder and gas flows. These models are essential for understanding the complex dynamics within LP-DED nozzles and identifying design parameters that improve process efficiency and deposition quality. Numerical models typically treat the shielding and carrier gases as a continuous phase described by the Navier-Stokes equations. Reynolds averaged equations coupled with the k-ε turbulence model are used to describe the flow dynamics. The powder particles, on the other hand, are modeled as a discrete phase, and their trajectories are analyzed using a two-phase flow approach where the particles form the secondary phase. This two-phase modeling approach allows for accurate simulation of the interactions between gas flows, powder particles, and nozzle walls, which are critical for achieving uniform powder distribution and stable flow [150,152,154].

Among the nozzle configurations studied, coaxial nozzles have garnered the most attention due to their widespread use in LP-DED processes. The optimization of these nozzles focuses on numerically evaluating key parameters, such as powder concentration at the focal point, which directly influences laser attenuation and catchment efficiency [82,152]. Powder concentration is determined by the interplay between gas flow dynamics, powder properties, and nozzle geometry [155]. These factors collectively govern the uniformity and precision of the powder stream, making them central to nozzle design improvements.

Several studies have employed numerical and mathematical models to investigate the influence of nozzle geometry on powder stream characteristics in LP-DED processes. Pinkerton et al. [82,156] developed a model to predict powder concentration distribution, showing that parameters like the injection angle, exit diameter, and outlet passage width significantly affect focal point concentration. Yang [155], using a Gaussian distribution model, found that smaller exit diameters and reduced nozzle angles enhance powder focus ability and catchment efficiency. Lin [157] compared inward and outward coaxial designs, demonstrating that while the inward configuration yields a columnar stream with lower peak concentration, the outward design produces a more focused stream with higher concentration. Pan et al. [151] introduced a model incorporating non-spherical particle shapes, revealing the substantial effect of outlet width and outer diameter on powder stream formation. Liu et al. [152] advanced previous models by accounting for particle–wall and interparticle collisions, although their study focused on a single nozzle design. Li et al. [150] emphasized the benefits of continuous coaxial nozzles with ring slits, showing their effectiveness in generating concentrated focal spots and allowing for precise control over powder focus location.

5.1.1. Simulation Approaches and Findings

Simulation of powder-gas flow dynamics in P-DED plays a critical role in understanding nozzle performance and optimizing deposition efficiency. Various computational models, including Computational Fluid Dynamics (CFD) and Discrete Element Methods (DEM), have been used to analyze how nozzle geometry, gas flow rates, and powder characteristics influence powder flow behavior. The results of several studies provide a comprehensive understanding of how these parameters affect powder concentration, velocity distribution, and deposition quality.

Pan et al. [151] and Zekovic et al. [85] conducted seminal studies investigating the influence of nozzle geometry and gas flow parameters on powder flow behavior. Pan et al. focused on gravity-driven powder flow in coaxial nozzles using a two-stage modeling approach. First, they developed a particle–wall collision model to simulate the stochastic interactions of non-spherical powder particles with the nozzle walls. The output of this model was integrated into a Computational Fluid Dynamics (CFD) simulation using ANSYS Fluent 15.0 to evaluate the gas-powder flow below the nozzle. The simulations showed that the nozzle’s geometry, in particular the tilt angle and the passage width, had a profound effect on particle concentration and focus. For example, reducing the angle of the powder passages from 60° to 45° increased the concentration at the focus point by 30%, while narrower openings resulted in higher peak concentrations.

Kovalev et al. [158] expanded on these investigations by analyzing the gas-dispersed flows within a coaxial nozzle with a divided powder supply. The simulation results revealed that the focusing and dispersion of the powder jets depend significantly on the transport channel’s geometry and the interaction of the powder particles with the walls. Particles undergo multiple reflections within the transport channel, leading to variations in their exit angles and velocity distributions. This phenomenon was found to be a primary factor in determining jet stability and focus. As the powder exited the nozzle, its concentration distribution was computed to identify the optimal deposition zone, which was observed at a distance of approximately 25 mm from the nozzle exit, as shown in Figure 17. The simulation further demonstrated that a four-port nozzle facilitates an axially symmetric distribution of powder density, which plays a crucial role in ensuring consistent material deposition and minimizing powder wastage.

Nasiri and Movahhedy [132] introduced an innovative continuous coaxial nozzle design intended to mitigate the effects of gravity on powder distribution during Direct Metal Deposition. Their research focused on modifying the internal flow dynamics within the nozzle by dividing the powder flow into eight distinct streams before its entry into the nozzle chamber. These streams were directed through symmetrically positioned radial grooves embedded in the inner cone of the nozzle, effectively reducing the deviation of the powder’s focal point and enhancing the overall uniformity of the powder’s distribution. The simulations were conducted using the Euler-Lagrange approach, assuming a turbulent flow regime for the carrier gas and treating the powder particles as a discrete phase. The model incorporated the standard k-ε turbulence model and considered only the weight, drag, and inertia forces acting on the particles.

The numerical simulation approach of Nasiri and Movahhedy [132] demonstrated significant improvements in powder flow stability at different inclination angles. The results indicated that at nozzle tilt angles of 0°, 20°, 30°, and 45°, approximately 90%, 89%, 87%, and 86% of the powder mass was retained within a 2.5 mm focal region, respectively. Additional information can be found in Figure 15 on page 184 of the original article [132].

Additionally, Nasiri and Movahhedy [132] present simulation findings demonstrated that optimizing the carrier gas velocity and the inlet/outlet geometric parameters further enhanced the collimation of the powder stream. Unlike the study by Kovalev et al., which primarily concentrated on optimizing the internal surface finish and transport tube length to improve collimation, Nasiri and Movahhedy focused on modifying the internal flow architecture to counteract gravity effects directly. Their findings provide a substantial contribution to the development of more versatile and efficient nozzle designs for laser direct energy deposition applications [132]. Li et al. [159] further expanded on these investigations by focusing on coaxial discrete three-beam nozzles. Using an Euler-Lagrange model coupled with a k-ε turbulence approach, they simulated the internal and external powder stream characteristics. The study demonstrated that the geometry of the nozzle, particularly the length, diameter, and shrinkage angle of the powder passage, significantly influences the convergence and velocity distribution of the powder stream. Increasing the length of the passage from 15 mm to 100 mm led to a reduction in the focal spot diameter from 5.1 mm to 2.6 mm, exhibiting a logarithmic decrease in divergence. Similarly, reducing the passage diameter from 2.0 mm to 1.0 mm resulted in a more concentrated powder stream, with focal spot diameters shrinking from 5.9 mm to 2.4 mm, while the maximum particle velocity increased from 6.75 m/s to 16.4 m/s.

In Arrizubieta et al. [136], the passage of coaxial nozzles contains a shrinkage structure to improve the convergence of the powder stream, but there is a lack of discussion of the rationality of such a structure design. Contrary to expectations, as observed in Figure 18, the implementation of a tapered passage with a shrinkage angle of 2° did not enhance powder stream concentration. Instead, it reduced the maximum powder velocity from 12.9 m/s to 5.8 m/s and increased the focal spot diameter, suggesting that such geometrical modifications may lead to greater powder dispersion rather than improved focus [160].

Ferreira et al. [160] focused on the interaction of multiple gas flows in coaxial nozzles and the effect of nozzle geometry on powder flow behavior. Specifically, in their study, a combined computational and experimental analysis was performed on three different coaxial nozzle designs (A, B, and C), each characterized by different gas channel arrangements and spacings. Ferreira et al. [160] used CFD simulations in COMSOL Multiphysics to model the turbulent gas flow behavior. Specifically, a two-dimensional axisymmetric model was developed to analyze the gas flow characteristics inside of and below the nozzles. The study used a Reynolds-Averaged Navier-Stokes (RANS) turbulence model together with Transport of Diluted Species (TDS) to investigate the dispersion of powder particles [160]. The configuration is illustrated in Figure 3 of [160]. Their results highlighted the following aspects:

• Nozzle A: A highly collimated gas jet was observed, reaching velocities of up to 55 m/s.
• Nozzle B: A broader gas jet with a peak velocity of 9 m/s was reported.
• Nozzle C: The gas jet showed a more diffuse profile with peak velocities of approximately 4.2 m/s.

Yao et al. [161] conducted a comprehensive study of the impact of nozzle geometry on powder flow behavior in LP-DED. Their work focused on how variations in nozzle diameter, carrier gas flow rate, and the defocused amount influence powder spatial distribution, particle velocity, and energy absorption by the metal powder. The study employed a coaxial laser deposition system equipped with four symmetrically arranged nozzles.

To simulate the gas-powder interaction, ANSYS Fluent was used for CFD modeling of the gas flow, while EDEM software implemented DEM-based particle tracking. The two models were coupled to simulate particle trajectories under different nozzle configurations. Additionally, an electromagnetic wave heating model in COMSOL Multiphysics was applied to study laser energy absorption and temperature rise in powder particles.

Their results revealed that reducing the nozzle diameter led to higher powder concentration near the laser–particle interaction zone, but it also increased particle divergence angles beyond optimal conditions [161]. Furthermore, Yao et al. [161] explored the impact of defocused amounts on particle velocity and spatial distribution, demonstrating that a negative defocused amount increased particle convergence at the melt pool center.

Another study conducted by Pant et al. [162] investigates the dynamics of powder flow within a multi-channel coaxial nozzle employed in a Direct Metal Deposition system. The research integrates both numerical simulations and experimental validation to analyze how the nozzle geometry influences powder concentration, velocity distribution, and the overall efficiency of powder delivery. Pant et al. [162] distinguished their study by demonstrating that accurate turbulence modeling is critical for predicting powder transport in LP-DED systems, with their Eulerian-Lagrangian simulation revealing that the k-ω model outperforms the k-ε model in capturing near-nozzle particle dynamics and velocity distributions, particularly under low Reynolds number conditions.

5.1.2. Experimental Validation

Experimental validation plays a crucial role in confirming the accuracy of numerical models developed for understanding powder flow dynamics in LP-DED. Several research efforts have focused on validating computational predictions through optical measurement systems, high-speed imaging techniques, and powder collection experiments. To validate their simulations, Pan et al. [151] conducted experiments using a simplified optical measurement system. This system measured particle concentration below the nozzle under varying configurations. The experimental results confirmed the simulation predictions regarding the effect of nozzle angle and passage width on powder stream structure and focus. Building on these efforts, in the experimental setup of Zekovic et al. [85], high-resolution imaging techniques were employed to visualize powder flow. Their method, using structured laser light sheets, provided detailed cross-sectional views of the powder streams.

To validate the numerical model, Kovalev et al. [158] compared the calculated powder jet profiles with experimental data obtained through high-speed imaging. The high-speed visualization allowed for precise measurement of powder jet dispersion, convergence dynamics, and particle trajectory distributions. The model accurately predicted the jet convergence and the width of the waist region, aligning well with the experimental visualization of powder flow (Figure 19).

Similarly, Nasiri and Movahhedy [132] validated their findings through experimental measurements of powder concentration at different nozzle inclination angles. Their results show that even at a 45° inclination, over 86% of the powder remains within a 2.5 mm diameter focal region, indicating significant improvement in powder delivery efficiency compared to conventional coaxial nozzles. Furthermore, the novel nozzle design effectively minimized powder loss due to wall adhesion within the grooves. The experimental observations confirmed that the optimized gas velocity prevented powder settlement inside of the nozzle, ensuring a consistent and homogeneous powder stream at the outlet [132].

Li et al. [159] extended the validation efforts of their simulation results through high-speed imaging and charge-coupled device (CCD) photography, providing an in-depth analysis of the effect of nozzle geometry and process parameters on powder flow behavior. The measured velocities under varying parameters were compared with the simulated results, revealing a strong correlation between experimental data and simulation results. However, as the images provide only two-dimensional information, the velocity comparison is limited to two-dimensional vectors, which are lower than the maximum particle velocities predicted by the numerical simulations. Building on these findings, further investigations showed that varying the internal laser shielding gas flow rate (0 L/min, 10 L/min and 20 L/min) had a minimal effect on the convergence of the powder stream in coaxial discrete nozzles. Unlike continuous coaxial nozzles, where the shielding gas has a strong influence on the powder focus height, discrete nozzles allow the shielding gas to escape through powder passage gaps, thereby reducing its influence on the powder stream’s behavior (Figure 20).

For optimal LP-DED performance, the powder spot size must be matched to the laser spot and the melt pool dimensions to ensure high material utilization and deposition accuracy. Li et al. have formulated a basic principle for powder spot control, highlighting that increasing the passage length, decreasing the passage diameter, and optimizing the powder particle size distribution contribute to improved convergence. Figure 21a shows a schematic of the interaction of the powder with the laser and the melt pool in the LMD process. A critical observation was that powder collisions with the inner nozzle walls are a major cause of divergence at the nozzle outlet. Strategies to mitigate these effects include optimizing particle impact angles and controlling inelastic collision interactions within the nozzle. Figure 21b shows a schematic representation of particle motion in the nozzle passage.

In the studies conducted by Pan et al. [151] and Zekovic et al. [85], the powder stream was introduced parallel to the nozzle passage. This configuration resulted in a powder focus characterized by good convergence and uniformity. However, these studies rarely addressed the analysis of the powder focus size, as variations in powder or gas parameters were not found to significantly influence the convergence behavior of the powder stream. In practical applications, collisions between the powder particles and the passage walls are unavoidable. These interactions cause the powder flow to diverge at the nozzle outlet, ultimately forming powder spots of varying sizes [159].

As illustrated in Figure 22a, when the powder enters parallel to the passage, it is solely influenced by the gas flow along the passage. This results in the powder flow diameter at the nozzle outlet being consistent with the incident diameter. Under these conditions, the powder stream can produce an ideal Gaussian distribution at the focal point, and the spot diameter can be estimated using basic geometric formulas.

However, this theoretical approach does not accurately represent real-world scenarios [159]. Figure 22b represents a simplified depiction of a powder stream entering at random angles. If all powder particles initially enter the passage vertically, it is evident that the powder stream diverges at the nozzle outlet, resulting in a random powder spot. This spot typically follows a Gaussian distribution pattern [159].

Ferreira et al. [160] used a pitot tube system setup to measure localized gas velocities in the outer region of the nozzle outlet. Measurements were made using a 3 × 3 mm or 4 × 4 mm grid, capturing variations in both axial and radial directions. The data, interpolated using a bilinear function, showed that thinner gas channels increased gas velocity but also induced greater instability, as predicted by simulations.

Yao et al. [161] validated their CFD-DEM model with experiments on powder flow dynamics, confirming the relationships between particle velocity, the defocused amount, and the gas flow rate. The authors found that nozzle diameter has a significant effect on particle velocity, the divergence angle, and powder concentration in the interaction zone. Smaller nozzle diameters (1.8 mm) resulted in increased divergence angles (30°), which reduced powder delivery efficiency and increased dispersion losses. Conversely, an optimum nozzle diameter (2.4 mm) minimized divergence to 18°, resulting in better powder utilization and more uniform deposition [161]. Yao et al. [161] showed that increasing the carrier gas flow rate increased powder velocity but decreased particle temperature rise, while excessive shielding gas caused powder dispersion and efficiency loss; in addition, their study showed that a negative defocus amount (−1 mm) improved powder concentration at the laser center, increasing delivery efficiency and laser absorption.

For experimental validation, a CMOS-based image processing technique was employed by Pant et al. [162] to capture the powder stream’s behavior. The powder was illuminated using a double-pulse diode laser, and images were processed to extract the powder concentration and velocity. The material used was gas-atomized SS316L powder (average particle size of 45 µm), which was delivered via an Oerlikon Metco Twin 150 powder feeder using argon as the carrier gas and helium as the nozzle gas. A key outcome of the study was the identification of an optimal standoff distance (SOD) of 10–12 mm, corresponding to the highest powder concentration and the narrowest powder stream waist; the formation of bimodal powder distribution was observed prior to this distance, whereas a converged and focused stream was detected at the waist [162].

The study by Pant et al. [162] revealed that carrier gas flow rate (CGFR) significantly influences powder stream diameter. An increase in CGFR resulted in a more focused powder stream, while a higher nozzle gas flow rate (NGFR) caused wider powder distribution. This behavior is attributed to the interaction between the gas jet and the powder particles within the nozzle channels, where CGFR enhances powder focusing, whereas NGFR induces dispersion. For further details, see Figure 15 on page 8 of [162].

The powder velocity was measured both numerically and experimentally and showed a range of 2.0 to 6.2 m/s depending on the gas flow parameters. Trace length imaging velocity analysis confirmed a decrease in average particle velocity with increasing powder mass flow rate (MFR) due to momentum exchange between particles. It was also found that nozzle gas flow had a minimal effect on particle velocity, as the lower density of helium compared to argon resulted in negligible acceleration of powder particles [162]. The findings of Pant et al. [162] emphasize the critical role of nozzle design in determining powder flow characteristics in LP-DED processes.

According to all of the documents discussed in this paper,

Table 2. Comparison of the most relevant papers selected for this review.

Feedstock	Laser Beam Shape	Nozzle Type	Equipment Configuration	FEM Approach	Numerical Results	Experimental Results	Ref.
Stellite 6	NA	Coaxial	Precitec YC50	- Continuous gas phase: modeled using time-averaged Navier–Stokes equations - Dispersed phase: solved in Lagrangian frame by tracking particles through the computed flow field - Gas turbulent flow: the two-equation model or k–e model proposed by Launder and Spalding (CFD code, FLUENT, is used to solve the set of equations) - The Lagrangian approach of the discrete phase model (DPM) in FLUENT has been used to solve each particle’s dynamic behavior	Y	Y	[86]
316L SS	Top-hat	Coaxial discrete	Optomec 750 LENS		N	Y	[101]
NA	Gaussian	Coaxial	NA	- Simulation model: Simplified 1D analytical model for particle heating under laser irradiation - Approach: Solves energy balance with assumptions of uniform particle temperature, constant absorptivity, and no latent heat; accounts for drag, convection, and laser intensity variation	Y	Y	[115] *
316L SS	NA	Coaxial	HCX60 five-axis hybrid machine	NA	Y	Y	[139]
IN718	NA	Continuous coaxial	NA	2D axisymmetric Euler–Lagrange model in COMSOL 5.3a; two-phase flow (gas + particles); simulated velocity, density, diameter, focal position	Y	Y	[125]
316L SS	NA	NA	LENS 500 MTS HM		N	Y	[79]
IN718	NA	Coaxial	RFL-6000 fiber mounted on a KR50-R2100 six-axis robot—LCD-Z12-CW laser deposition head	Modified convolutional neural network (CNN)–long short-term memory (LSTM) model to extract spatio-temporal features like morphology fluctuations and spatter	Y	Y	[122]
Metco 42C SS	NA	Coaxial	Trumpf DMD505	Axisymmetric Navier–Stokes + discrete-trajectory model; WENO scheme on refined nonuniform mesh; simulates gas–particle flow and wall collisions in triple coaxial nozzle	Y	Y	[163]
Ti–6Al–4V	NA	Coaxial	1000 W laser (900–1070 nm)	Coupled electromagnetic wave heating (EWH) model used to simulate laser–particle interactions; laser modeled as high-frequency EM wave, powder heating computed via Maxwell’s equations and transient heat conduction; output used to modify a double-ellipsoid heat source in ABAQUS for thermal–mechanical analysis of residual distortion	Y	Y	[111] **
IN718	NA	Off-axis	IPG YLS-3000-CL laser (1070 nm)	NA	N	Y	[113]
Steel 24	NA	Coaxial	A 500 W CO2 laser on a CNC machine	NA	Y	Y	[83]
IN718	NA	NA	NA	NA	Y	Y	[93] ***
IN625	Gaussian	Coaxial	DMD—105D, DM3D, 1 kW diode laser	Coupled Eulerian–Lagrangian + VOF models; RANS k-ε for gas flow, discrete phase model for powder dynamics, OpenFOAM melt pool thermo-hydrodynamics using tracked particle data	Y	Y	[95]
NA	Gaussian	Off-axis + coaxial	NA	Analytical model to study the temperature distribution of the powder particles and their attenuation on the laser energy distribution	Y	N	[106] **
IN625	Gaussian	Off-axis	In-house developed LPF-AM setup	- Analytical model coupling Gaussian laser beam, powder stream, and semi-infinite substrate; simulates attenuated laser intensity, 3D melt pool geometry, and powder catchment efficiency - Powder stream concentration modeled with Gaussian distribution based on optical luminance (Mie theory) - Simulation in MATLAB: predicts substrate temperature field and up to 4% laser attenuation due to powder stream	Y	Y	[107] **
316L SS	From Super-Gaussian to top-hat (numerical)	Coaxial discrete	YLS-5000-S4 (5 kW) laser	- Laser–powder coupling model for DMD using homogeneous transformation (HT), Lambert–Beer law, and energy balance - Beam model: Super-Gaussian function and beam characteristic parameter identification method (BCPIM) for beam intensity reconstruction - Powder stream: DCPS model with four radially symmetrical nozzles; particle distribution influenced by scattering and absorption	Y	Y	[81]
IN718	NA	Coaxial discrete	Optomec LENS 450 Workstation	Analytical model developed in MATLAB to predict powder stream distribution beneath four-jet LP-DED nozzles - Inputs analyzed: powder flow rate, gas flow rate, and particle size - Output: Powder Deposition Efficiency (PDE) estimated from simulation results	Y	Y	[84]
Ti-6Al-4V	Gaussian/Ring (numerical)	Coaxial	NA	High-fidelity model based on the multiphysics ALE3D code, coupled with Cellular Automata Finite Element (CAFE) model for grain growth and DEM for powder flow	Y	Y	[98]
Steel	Top-hat	Coaxial continuous	NA	NA	N	Y	[97]
Ni60A	NA	Coaxial	Rofin-FL060 laser, V cladding head, ABB IRB4600 robotic arm	- Simulation model: CFD-based gas–solid two-phase flow model developed using ICEM CFD and ANSYS Fluent - Approach: Carrier gas modeled as continuous phase (Eulerian), powder as discrete phase (Lagrangian), turbulence solved via standard k–ε model	Y	Y	[164]
316L SS	NA	Coaxial	BAMPR + Magic 800 BeAM	NA	N	Y	[165]
Ti-6Al-4V	Gaussian	Coaxial	NA	- Simulation model: 2D time-dependent finite element model coupling heat transfer, fluid dynamics, and surface effects - Approach: Level-set method for interface tracking; laser modeled as Gaussian heat source; laminar flow assumed; effects of gravity, recoil pressure, and surface tension included	Y	Y	[166]
NA	NA	Coaxial discrete	NA	- Simulation model: 3D Discrete Element Method (DEM) using YADE to model particle–particle interactions - Approach: Simulated 10,000 spherical grains from a commercial discrete coaxial nozzle; elastic collisions modeled with frictional and normal forces; effects of carrier gas neglected	Y	N	[167]
H13 tool steel (atomized powder)	NA	Radially symmetrical nozzles (discrete coaxial + central coaxial for shielding gas)	MultiFab system with four radially symmetrical powder nozzles and a central coaxial shielding nozzle	3D CFD using FLUENT (discrete phase model), k-ε turbulence model, Lagrangian particle tracking	Y	Y	[85]
NA	NA	Coaxial nozzle (various outlet geometries)	Custom coaxial nozzles (Nozzle I–V); no carrier gas; FDM-fabricated ABS nozzles used for testing	3D stochastic Lagrangian model considering non-spherical particle–wall collisions	Y	Y	[151]
Technoclad 40 s-400 (Ni-Cr-based + WC), Hoganas 1540–00 (Ni-Cr-based), and 316l 20–53 (stainless steel)	NA	Coaxial nozzle with divided powder supply	DLD system with coaxial nozzle featuring separate powder feeding channels	3D CFD model using OpenFOAM; Eulerian–Lagrangian approach for powder particle tracking	Y	Y	[158]
NA	NA	Continuous coaxial nozzle with a novel design to mitigate gravity effects	Direct Metal Deposition (DMD) system utilizing the newly designed continuous coaxial nozzle	Computational Fluid Dynamics (CFD) simulations to analyze gas–powder flow dynamics and assess the impact of gravity on powder stream behavior - Simulation model: 3D Euler–Lagrange multiphase model implemented in ANSYS Fluent - Approach: Continuous phase solved with Navier–Stokes equations coupled with the k-ε turbulence model; discrete powder phase tracked via Lagrangian particle injection - Details: Complex nozzle geometry meshed in ANSYS ICEM CFD using hybrid structured/unstructured grid (>2.3M cells); particle trajectories computed within pre-solved gas flow field	Y	Y	[132]
NA	NA	Coaxial nozzle	Coaxial powder feeding system with internal carrier and shielding gas; laser cladding setup	- ANSYS Fluent 15.0 software (ANSYS Inc., Canonsburg, PA, USA) was used to simulate the powder stream characteristics - 3D gas–solid two-phase flow model; Euler–Lagrange approach for powder particles; k–ε turbulence model for carrier gas flow	Y	Y	[159]
Inconel 718	Gaussian	Continuous coaxial nozzle	Coaxial LMD system	CFD Model	Y	N	[136]
NA	NA	Coaxial nozzles	NA	- Simulation model: RANS-based turbulent CFD with k-ε model, coupled with Transport of Diluted Species and Particle Tracing modules in COMSOL Multiphysics - Approach: 2D axisymmetric simulation of gas flow in three nozzle designs; gas dynamics validated with experimental Pitot tube velocity measurements	Y	Y	[160]
NA	NA	Multi-channel coaxial nozzle	NA	- Simulation model: Two-way coupled CFD–DEM approach; gas flow solved with k-ε turbulence model in ANSYS Fluent, powder particles with EDEM; powder–gas interaction via UDF - Approach: 3D CFD-DEM model based on experimental four-nozzle LP-DED head; COMSOL used for coupled electromagnetic heating and thermal analysis of powder particles	Y	Y	[161]
SS316L	NA	Coaxial nozzle with multi-channels for powder delivery	In-house developed DMD system with a 1.2 kW continuous diode laser; Oerlikon Metco Twin 150 powder feeder used to deliver the powder using argon as an assist gas and helium as the inert (nozzle) gas	- Simulation Model: Eulerian–Lagrangian approach; gas flow modeled with k–ω turbulence model; powder particles treated as discrete phase in continuous gas phase - Approach: 3D numerical simulation using ANSYS ICEM CFD; coaxial nozzle with 20 channels for powder delivery; mesh with 298,474 tetrahedral elements; quarter-domain reduction for efficiency - Details: Compared k–ω and k–ε turbulence models; k–ω preferred for low-Reynolds, wall-bounded flows; validated with experimental image-based powder flow data	Y	Y	[162]

* Model accounts for particle–laser interaction only. ** Simulation model used to study the temperature distribution of the dust particles and interaction with the laser. *** The numerical and experimental results focus on microstructure prediction and its comparison with deposited tracks, highlighting the metallurgical aspects of the LP-DED process.

reports the main points and aims to give a comprehensive overview of the published research on the topic of this review paper.

6. Conclusions

The present paper addressed the topic of Laser Powder Directed Energy Deposition (LP-DED), highlighting the critical aspects of the equipment as well as the powder particles, the delivery system, and the deposition nozzle geometries and efficiencies. Through an in-depth analysis of the scientific literature addressing the main subsystems of LP-DED equipment, several key factors have been identified as crucial to process efficiency: powder properties, laser–powder interaction, nozzle design, gas flow dynamics, laser source characteristics, and melt pool behavior. Taking advantage of the laser cladding heritage and exploiting recent advancements in numerical simulations found in the literature, the following conclusions can be drawn:

▪

Powder characteristics must be consistent, and powder reuse strategies can be successfully implemented in LP-DED processes, thus leveraging the sustainability of the overall process.

▪

The efficiency of molten metal deposition is governed by process parameters, as well as the localized interactions among powder particles, carrier and shielding gas flows, and the laser beam. Consequently, real-time monitoring and control of these interactions are essential for optimizing process performance and achieving high-quality deposits.

▪

Optimized internal geometries improve powder convergence and enhance deposition precision.

▪

Surface treatments and material selection significantly influence powder jet stability, with smoother surfaces yielding better focusing.

▪

Gas flow control strategies offer potential for enhanced powder delivery efficiency and material utilization.

▪

Advanced turbulence modeling enables better predictive control of powder transport, contributing to refined nozzle configurations for industrial applications.

▪

These findings provide a robust foundation for continued advancements in LP-DED nozzle designs, paving the way for improved process control, reduced material waste, and enhanced build quality in metal additive manufacturing.