Metal powder feedstock evaluation and management for powder bed fusion: a review of literature, standards, and practical guidelines

Abstract

Metal powders are key to metal additive manufacturing technologies such as powder bed fusion. These powder feedstocks experience a range of forces and physical phenomena both during the powder bed fusion process and additional post-processing stages that can alter their composition and material properties. To evaluate such effects, these powders need to be characterized, tested, and analyzed at critical stages of their lifecycles. This paper provides a review of the current state of the art for powder evaluation methods and their applicability for powder bed fusion production. Methods are categorized by the properties they evaluate, either particulate or bulk properties. Industry standards are identified for each method if applicable and the advantages and disadvantages of each are defined. Effects of these properties on the flowability and spreadability of powders are synthesized and practical management guidelines are defined. This paper aims at providing an overview of powder evaluation for powder bed fusion, practical considerations for the development of powder test and evaluation programs, and provide insights for future research undertakings in the field.

1 Introduction

The use of metal powders in modern production began in the 1920s under the general name of powder metallurgy (PM), also known as press-and-sinter [1]. Since then, various types of production methods have been industrialized for the production of components from metal powders, namely powder metal injection moulding (MIM), hot isostatic pressing (HIP), and more recently additive manufacturing (AM) technologies such as powder bed fusion (PBF), directed energy deposition (DED), and binder jetting technologies. Each of these production methods has specific requirements for the metal powder feedstock, but some requirements are common between them. The powder is the beginning of the physical product lifecycle for these production technologies and effective controls and evaluation methods are required to achieve high-quality components.

In metal AM, various metal alloy powders can be used including alloys consisting of aluminum, steel, nickel, titanium, copper, and even precious metals such as platinum and gold [2]. This is one of the advantages of using technologies such as PBF. Additionally, there are novel powders currently being developed for PBF to unlock the specific capabilities of the PBF technologies as well as to fulfill the stringent requirements of critical industries such as medical, aerospace, nuclear, military, and oil and gas. Careful consideration is required when developing test and qualification programs to ensure the selected powder evaluation methods cater for the production processes, the specific powder metal alloy, the powder lifecycle, and the environments to which the powder is exposed.

Publications and research trends for powder evaluation and management for PBF production have increased in the past five years. The following Boolean search was conducted on Web of Sciences to identify relevant research publications:

ALL=(((“powder bed fusion”) OR (PBF) OR (SLM) OR ( “selective laser melting”)) AND (( “powder analysis”) OR ( “powder management”) OR ( “powder characterisation”) OR ( “powder characterization”) OR ( “powder evaluation”) OR ( “powder testing) OR ( “feedstock management”) OR ( “feedstock characterization”) OR ( “feedstock analysis”) OR ( “feedstock characterisation”) OR ( “feedstock evaluation”) OR ( “feedstock testing”)))

This search returned 63 relevant publications, additional searches were also conducted using other search engines such as Google Scholar. Figure 1 presents the research trends for research focusing on this topic based on the literary search conducted on 7 May 2023, highlighting the steady increase in research and research interest in recent years. Some of the notable works in this field include the works by Slotwinski et al. [3], Spierings et al. [4], Sutton et al. [5], Strondl et al. [6], Tan et al. [7], and Leung et al. [8].

Fig. 1

Research trends in powder evaluation and management

2 Background

Powder and granule physics is applied in many fields such as soil science, food production, the pharmaceutical industry, and manufacturing [9]. Powders are complex materials. Understanding and evaluating powder materials is a challenge largely due to their nature being that of an assembly of multiple individual particles, each of which exhibits its own properties and contributes to the properties of the powder bulk. The properties of the bulk powder can differ drastically based on how the individual powder particles are assembled or packed as well as the number of particles that make up the bulk.

Many of the powder evaluation methods currently used were originally developed for specific production processes other than PBF [5, 10, 11]. Examples include the Hall flow rate and the tap density methods which were developed for the traditional PM processes that typically involve loading powders through a cavity into a die or mould before compacting them into a green part. Although such methods were not developed specifically for metal PBF, they can still offer great benefits for evaluating powders for this process, but their application must be carefully considered.

2.1 Powder production and terminology

Metal powders can be produced via two fundamental methods, a mechanical or chemical method. Many of the production methods are limited by the amount of powder they can produce and their associated costs. Salak et al. [12] provided a thorough review of the various mechanical and chemical methods available for powder production. Some of the more common methods for metal powder production include atomization, milling, electrolysis, and mechanical alloying [13]. The atomization methods are the most popular and in some cases, versatile [13, 14]. The basic principle of the atomization method involves passing a stream of molten metal through a dispersion mechanism which atomizes the stream into small particles which then solidify into powder. Typical mechanisms of dispersion include rotating discs or centrifuges, liquid jets such as water or oil, pressurized gases, and plasma torches [14]. The gas and plasma atomization methods are popular for the production of powders for PBF. Of these two atomization methods, gas atomization is more widely used largely due to its high throughput and ability to produce fine and spherical powder particles [15]. Plasma atomized powders are typically highly spherical and exhibit narrow particle size distributions (PSD) but come at a higher cost [5, 16]. Figure 2 presents two scanning electron microscope (SEM) images of powder produced via these two atomization processes reported by Ahsan [16] and Meier [17], respectively, highlighting the improvement in particle sphericity for the plasma atomized powder particles. General powder production requirements for AM are specified in SAE AMS7002 [18]. Such requirements can be leveraged for powder procurement specifications.

Fig. 2

Micrographs of a gas [16] and b plasma [17] atomized powders (not to scale). Reproduced with permission from [16, 17]

Terminology is important when referring to powders, specifically when referring to powders in different process conditions or lifecycle phases. ASTM B243 and ISO 3252 define standard terminology for PM [19, 20]. Interestingly, neither has a definition for batch, with ASTM B243 defining a lot as “a specified quantity of product manufactured under traceable, controlled conditions as agreed between producer and user” [19]. The term “heat” is typically used to refer to the material that the powder producer uses to produce the powder. Powder lots typically are a blend of powder from different heats. In the case of PBF where the powder is reused, rejuvenated, or recycled, the powder deviates from the lot from which it comes to various degrees. Such powder requires a new identifier that can be used to define the process condition while still upholding traceability to the original powder lot. The term “batch” is defined as a “quantity of feedstock with uniform properties and composition” [21]. The term batch can refer to any quantity of powder, typically this term is used when the quantity of powder has deviated from the controlled condition of the original powder lot or as a generic term to refer to a quantity of powder [22]. Additional AM and PBF terminology is defined in ISO/ASTM 52900 including definitions for used powder, mixtures, beds, and blends [21]. SAE AMS7031 defines a list of powder terminology specifically for PBF production [23].

2.2 Overview of physical phenomena

It can be said that interparticle forces are the driving factor for many of the bulk powder properties. Understanding these forces and how they affect bulk powder properties provides insights for improving the quality of the formed powder beds during PBF production. Various factors contribute to interparticle relationships, which result in cohesive forces throughout the powder. Capillary forces are one such factor. In wetted particles, liquid bridges form between adjacent particles. This causes the liquid to be drawn into zones where the interparticle gap can be seen as the smallest. Capillary forces acting on the powders contribute to high moisture contents, which increase cohesiveness and, thus, can reduce flow [24]. Moisture pick-up can also lead to liquid bridges forming, contributing to agglomeration [25].

Van der Waals forces also influence the interaction between particles. These are the dominant interaction force between particles in a powder. These forces are only noticeable when the powder particles come into close contact with one another within a separation distance of around a single molecule [26]. These forces attract neighboring particles contributing to agglomeration. The higher the Van der Waals force, the more force is needed to allow these powders to flow. Electrostatic forces are also present which arise from powder handling. Charged particles may adhere more firmly to metallic surfaces. These forces cause the powder to be cohesive in nature depending on the extent and strength of these forces [26]. The powder particle sizes have a great effect on the tensile strength that exists between particles due to attractive forces, such strength is generally weaker for coarser particles resulting in less cohesion in the powder bulk [9]. Visser [26] and McGlinchey [9] provided a review of the various interparticle forces that exist and their effect on the bulk powder properties including:

• Liquid bridges (due to liquid)

• Capillary forces (due to liquid)

• van der Waals forces (molecular forces)

• Valence forces (molecular forces)

• Electrostatic forces

• Magnetic forces

• Interlocking forces (due to physical irregularities)

Once the powder layer has been applied by the recoating mechanism, the fusion source, be it a laser or electron beam, fuses the powder as defined by the scan strategy. When powders are reused in subsequent production runs, this powder contains contaminants due to this interaction between the fusion source, powder, and the environment. This fusion process involves complex physical phenomena and can be described as a chaotic system. Figure 3 depicts the types of interactions that occur between a particle and a light, or laser, source.

Fig. 3

Scattering geometries of particles. Reproduced with permission from [27]

Boley et al. [28] investigated this “chaos” by performing analyses on an idealized hexagonal close-packed powder bed. Figure 4 illustrates the complex interaction between the laser beam and the powder bed and particles.

Fig. 4

Illustration of typical a laser scattering, and b ray trajectories. Reproduced with permission from [28] © The Optical Society

Boley et al. [28] noted that the powder particle sizes and their distribution affect the laser absorptivity. Figure 5 demonstrates this effect with the image on the left showing that powders with a mix of large and smaller particles reduce the laser absorptivity, whereas powders with more homogeneous particle sizes and narrower PSD increase laser absorptivity. This demonstrates the importance of understanding the packing behavior of powders in the formed powder bed. This effect was demonstrated for powders with bimodal and Gaussian distributions [28].

Fig. 5

Absorptivity, $\alpha$, for bimodal stainless steel powder. Reproduced with permission from [28] © The Optical Society

Two results of this chaotic interaction between the fusion source and the powder of particular interest are the condensate and melt pool spatter that forms. Figure 6 presents an illustration from Sutton et al. [29] that depicts how the spatter and condensate are formed during the fusion process. Sutton et al. [29] identified recoil pressure or vapor entrainment as the cause of the melt pool spatter. The condensate or soot is vaporized material that can disrupt the laser beam as well as contaminate the powder bulk [29]. Another consideration is the radiation pressure. As the laser beam is essentially an electromagnetic wave, it carries momentum and a force. The correct radiation pressure must be considered as if the force is too high, particles will be forced away from the melt pool resulting in inadequate melting and fusing [30].

Fig. 6

Illustration of spatter and condensate formation. Reproduced with permission from [29]

2.3 Powder recoating mechanisms and powder bed defect formation

During the PBF process, the powder is spread across the build by two fundamental and dynamic processes, first the dosing process and then the recoating process. Ideally, this results in a level and homogeneous layer of powder to be fused by the fusion source. These processes differ significantly by machine model and make. Some PBF systems use an integrated dosing and recoating system, while others use two separate systems. Such machine model-specific mechanisms pose challenges for developing standardized test methods for powders that provide information that is representative of the production mechanisms that these powders experience.

Three common recoating mechanisms are depicted in Fig. 7. For the first mechanism, the blade recoater, various types of blades can be used for the blade-type recoater including brush, rubber, and solid metal blades each providing its own advantages and disadvantages [4]. The brush blades are typically made from carbon fiber bristles and the more rigid blades are typically made from a metal alloy or rubber composite. The second recoating mechanism differs significantly from the others in that the powder feedstock is fed from above by gravity. This mechanism typically contains two rollers at the base of the recoater. The third recoating mechanism uses a roller-type blade which rotates in the direction of recoating to apply the powder layers. These three mechanisms are highlighted as they, or variants thereof, are used by more than one machine OEM, other mechanisms include non-contact type mechanisms amongst others [31].

Fig. 7

Powder recoating mechanisms

All PBF recoating mechanisms stress the bulk powder to varying degrees. The investigations by Krantz et al. [32] highlighted that powder flowability is dependent on the stress state of the powder and that both static and dynamic evaluation methods are needed to completely understand the flowability of the powder for specific applications and conditions. Spierings et al. [4] noted that PBF recoating mechanisms typically aerate the powders when recoating and that these powders tend to exhibit high free surfaces when recoating. Figure 8 depicts an illustration of the recoating process and the forces involved. This figure has been adapted from Weinberg [33] with the information presented by Van der Schueren and Kruth [34].

Fig. 8

Illustration of blade-type recoating forces and defects

The spreadability, also referred to as rakeability [35], of a powder is a novel concept that describes how well a powder spreads. This can be determined by imitating the PBF recoating process when a recoater blade coats the base plate with powder in successive layers. Currently, there exists no standard for directly evaluating spreadability although both ISO and ASTM standards are currently under development [36, 37]. There are, however, metrics that can be used to evaluate how well a powder spreads. Such metrics include the percentage of powder coverage over a bed, surface topography (ST) and homogeneity, and powder bed density (PBD), which is a specific type of bulk density. ST and PBD are closely related, with ST being more affected by the recoating mechanism. The term surface homogeneity is used in this paper to refer to the degree to which the formed powder bed surface is free from defects such as voids, surface texture, and levelness deviations. A high-quality powder bed is one which is densely packed with powder, has a level surface, has no voids, has a smooth surface, and does not contain excessive particle segregation. It shall be noted that various other aspects are important such as moisture, homogeneity of chemistry, and foreign object debris (FOD) amongst others. Figure 9 illustrates some roller-type recoating defects based on simulations by Shaheen et al. [38]. In addition, Scime [39] presented typical powder bed defects as seen in Fig. 10. Powder bed defects and potential causes as reported by Foster et al. [40] and Scime et al. [39, 41] include:

• Streaking—Causes include damaged recoater blade, incorrect recoating speed, large particulates, and foreign objects in the powder.

• Recoater Hopping, Powder Flicking, and Clumping—Causes include overdosing, recoater collisions with parts, moisture, and incorrect recoating speed.

• Incomplete Spreading and Voids—Causes include underdosing, damaged recoater blade, and incorrect recoating speed.

Fig. 9

Roller-type recoating defects [38]

Fig. 10

Powder bed defects. Reproduced with permission from [39]

3 Sampling and sample preparation

Adequate sampling methods are important for obtaining samples that are representative of the particle properties and of the bulk from which they are obtained. Careful consideration needs to be made when sampling powders both to ensure the sample is representative of the bulk and also to not perform needlessly expensive and time-consuming analyses [42]. Sampling can be performed statically or dynamically. Standard sampling methods for metal powders are defined in ASTM B215 [43], ISO 3954 [44], and MPIF Standard 01 [45]. ISO 14488 [46] defines the procedure for sampling powders specifically for the determination of particulate properties such as particle size and shape. ASTM B215 [43] defines two methods for extracting samples from the bulk, one for sampling powders being transferred from storage tanks or blenders, dynamic sampling, and another for sampling from powder packaged in containers, static sampling. Both these methods may be applicable for PBF production depending on the powder lifecycle implemented and the quality requirements for the particular industry. The spin riffling sampling method is reported to have the lowest standard deviation [47]. Other sampling methods include cutting and scoop sampling [14, 42]. Devices used to standardize the sampling and splitting processes include samplers such as keystone samplers, also referred to as sampling thieves, and devices for homogenizing and dividing samples such as blenders, splitters, and rifflers [22, 43]. It is important to sample powders that have been homogenized or take samples at multiple locations throughout the bulk to reduce the effects of powder segregation. A combination of both homogenization and multiple samples from different locations in the bulk is preferred. Allen [42] provides a detailed explanation with examples of how to determine the number of samples required as well as the weight of each to achieve a predefined statistical confidence interval.

For PBF production sampling powder before, in situ, or after the build may be beneficial. The powder dosing system, collection system, powder bed, soot trap, depowdering system, sieving system, wet separator, and storage containers or vessels are examples of locations where powder may be sampled for various reasons including process characterization, qualification, quality control (QC), or nonconformance investigations. The powder at these various locations within the PBF production process chain could be exposed to different thermal, environmental, chemical, or physical conditions. The static sampling methods are applicable for such sampling but may require custom or non-standard sampling devices. For powders that have been reconditioned for reuse, a 100 per cent sampling program will likely be required per ASTM B215 [43] as it is uncommon to recondition a batch of powder that is subsequently split into more than 5 containers.

Powder samples need to be further prepared for certain analyses such as PSD, size analysis. and certain chemical analyses. ISO 14887 [48] defines a standard procedure for the dispersion of powders in liquid media. ASTM B821 [49] and B859 [50] provide guidance for liquid dispersion and deagglomeration of metal powders respectively. Such procedures are required for PSD test methods such as laser diffraction, light scattering, gravity sedimentation, and size analysis methods amongst others, while MPIF standard 67 [51] defines the appropriate procedures for sample preparation specifically for chemical analysis. ASTM E3 [52] for the preparation of metallographic specimens provides guidance for mounting, grinding, polishing, and etching metallographic specimens which is useful for preparing powder specimens for microscopy and other evaluation methods such as nanoindentation and certain chemical analysis methods. Additional guidance for metal powder sample preparation specifically for AM is provided in application notes by Vedel-Smith et al. [53].

4 Evaluating metal powders

Characteristics of metal powders can be defined in terms of their particulate properties as well as their bulk properties. General requirements for the characterization and evaluation of powders for PBF are defined by ISO/ASTM 52907 [54], ASTM F3049 [55], and the MPIF collection of standards for metal AM powder characterization [56].

Sutton et al. [5] provided a thorough review of metal powder for PBF focusing on the particulate properties. Sutton et al. [5] noted that the particulate properties contribute greatly to the properties of the bulk powders. The bulk properties should be evaluated in addition to the particulate properties. Evaluation of the bulk properties of metal powders provides valuable insights into process-specific aspects and the selection of the bulk property test methods should be conducted based on the specific process being investigated and the powder processing mechanisms involved. The effects of powder properties on the produced PBF material have been investigated and reviewed in previous publications including those by Tan et al. [7], Strondl et al. [6], Dietrich et al. [57], and Vock et al. [35]. QC limits shall be determined based on data obtained from qualification programs and industry standards and specifications. The efficacy of the evaluation method in evaluating the relevant metrics needs to be determined, especially if non-standard methods are to be used.

Figure 11 presents a breakdown of the different powder properties. This graphic was developed from the various literary sources reviewed in this paper. The properties depicted on the right-hand side of the tree diagram can be further classified into specific parameters and metrics, although these depend on the evaluation method being applied. It shall be noted that certain powder bulk properties such as tap density, angle of repose (AoR), and PSD typically involve a dynamic aspect during their evaluation although are measured in their static state or represent the static bulk properties. Additional properties not included in this diagram include thermal, electrical, magnetic, and environmental properties. Such properties are beyond the scope of this paper, although the reader should take note of these and assess whether such properties are applicable to their specific application.

Fig. 11

Breakdown of powder properties

4.1 Powder particulate properties

Powder particles differ greatly for the various metal alloys and powder production methods. It is often challenging to correlate particulate properties with the bulk from which they are sampled with confidence. Evaluation of powder particulate properties is performed on small samples, typically 2 grams or less. This requires the application of stringent and proven powder sampling to achieve accurate results [58]. Powders in such small quantities, especially fine powders, are also highly susceptible to contamination which can negatively affect the evaluation results. Methods for evaluating powder particles are typically applicable for all PM applications, as opposed to bulk powder property methods which should be correlated with the mechanisms of the application production technology.

Preparing powders for particulate property evaluation is performed by different methods. Common techniques include dispersion in various liquid or gaseous media, mounting in epoxy resins, fixing to adhesive tape, dispersion over glass using compressed air, and mounting in plastic straws [59, 60]. Some of these can too be used for evaluation of certain bulk properties although the small sample sizes do not represent the bulk powder well. It is important that the preparation technique and mounting approach does not deteriorate or adversely affect the powder itself. Equipment safety must also be considered when selecting an appropriate mounting approach to reduce the risk of powder spills which can cause damage to the equipment and instrumentation or contaminate subsequent analyses.

Particulate property evaluation methods are identified in Table 1 along with the properties they evaluate, applicable standards, and advantages and disadvantages. The are many methods and approaches for evaluating the chemical and elemental composition of metal powders and metal alloys in general. The ones listed in Table 1 are some of the common methods. Hardness is the main property evaluated by performing indentation tests, although various other properties such as stiffness and toughness can also be evaluated [27]. For additional methods and information for evaluating powder particles, the reader is referred to Neikov et al. [14], ASM Handbook Volume 7 [58] and 24 [13], Higashitani et al. [27], and Schulze [61].

Table 1 Powder particulate evaluation methods

Evaluation of metal powder chemical composition is often performed by ICP methods such as atomic or optical emission spectrometry [58]. Such methods are effective at characterizing the main alloying elements but are often ineffective at characterizing gas elements such as O, H, and N [58]. The inert gas fusion method is effective at characterizing the content of these elements which greatly affect the mechanical performance of structural metal alloys, such as titanium. ISO/ASTM 52907 provides a list of combustion and fusion-based chemical evaluation methods for various metal alloy powders [54]. Inert gas fusion also offers value in that it can be used for powder QC. Both ferrous and nonferrous powders can become contaminated when processed incorrectly. Metal powders can pick up moisture, oxygen, and other elements from the environment. The application of inert gas fusion to characterize these gas elementals at different stages of the powder lifecycle allows for the determination of powder quality. It is common practice to set QC limits on interstitial elements below those defined in the relevant material specification to allow for small variations in the powder chemistry during powder processing. Additional methods for characterizing and evaluating powder particle chemistry include XPS, EDS, and XRF. The reader is referred to the ASM Handbook Volume 7 [58] and Sutton et al. [5] for a review of such methods. Dalley et al. [64] showed how computer-controlled SEM-EDS can be used to characterize metal powders and their contaminants. They applied heavy liquid separation to the samples to extract and concentrate the contaminants prior to analysis.

The use of titration methods such as Karl Fischer titration has been investigated for metal PBF powders as an additional QC approach. The Karl Fischer titration method can be used to characterize the moisture content of metal powders, a common form of contamination when powders are openly exposed to the production environment. Mellin et al. [63, 65] investigated the application of this method and its implications for metal PBF. Reduction methods such as the hydrogen loss method can be used to evaluate the oxygen content of certain metal powder alloys [66]. Such methods offer value from a QC point of view as they are relatively simple and inexpensive, but should not be used to characterize oxygen content in powders as they lack the accuracy of combustion and fusion-based methods.

Microscopy offers great versatility for the evaluation of powder particulate properties. In certain cases, it can also be used for investigating bulk properties such as PSD as well. In addition to the evaluation of particulate properties, microscopy offers the ability to inspect powders for contamination, agglomeration, and other practical production QC aspects [67]. Such QC activities that can be performed with microscopy, specifically optical microscopy (OM), are inspection and evaluation of powder discoloration as an indicator for excessive thermal exposure during the PBF process due to repeated powder reuses [68]. Additionally, microscopy provides capabilities to investigate attrition and particle damage, which may indicate inadequate powder processing. Powder particle damage can be caused by a range of factors including fast recoating speeds, aggressive blending or mixing, and the various powder recovery techniques to name a few. There are various types of microscopes, the main types being OM and electron-based microscopes such as SEM and transmission electron microscopes (TEM). SEM offers additional versatility when compared to traditional optical microscopes in that various add-ons can be added to the microscope for different analyses. Such add-ons include detectors for X-rays and electrons for performing EDS and electron backscatter diffraction (EBSD) analyses amongst others. EDS offers the capability to perform element mapping to identify if powder contamination has occurred [54]. The application of metallographic practices and etching powder samples allows for the evaluation of particle porosity, microstructures, and crystal structures with the use of microscopy. Very few methods allow for such analysis at the particle level, XRD being one of the additional methods for the evaluation of powder particle crystal structures [69].

Particle morphology is a key factor affecting powder flowability, bulk density, and PBF-specific aspects such as spreadability. Morphology can be evaluated qualitatively or quantitatively. Qualitative and quantitative analysis and descriptions of powder particle morphologies are defined in standards such as ASTM F1877 [70] and ISO 9276-6 [71]. Powder particle morphologies are described similarly to that of metallographic grains. Particle morphology can be evaluated using various types of microscopes, computed tomography (CT), and dynamic image analysis techniques. Quantitative analysis for particle morphology can be performed by calculating particle size diameters, such as Feret, Martin, or project area diameters amongst others. Figure 12 depicts such particle sizes from McGlinchey [9].

Fig. 12

Types of particle diameters. Reproduced with permissions from [9]

Particle shape can be evaluated by calculating various shape factors. Image analysis with tools such as the open-source software ImageJ [72] can be leveraged for performing such evaluations and allows for calculating particle shape factors, sizes, as well as their distributions throughout the sample. ASTM F1877 [70] defines some standard shape factors and the formulae for calculating them including particle aspect ratio, elongation, roundness, equivalent circle diameter, and form factor. Powder particle morphology has a great effect on the flowability of metal powders and the quality of the resulting powder beds formed in the PBF process [73]. Spherical and near-spherical powders result in greater apparent densities (AD) when compared to powders of less spherical and irregular shapes [74]. The powder PSD should also not be too narrow as smaller particles are required to fill in the spaces between larger particles resulting in improvements in the bulk density [74]. There are a host of additional particle shape metrics and methods for calculating them. The reader is referred to Rodriguez et al. [75] for an extensive review of such shape metrics, their histories, and the different calculation methods. Performing particle morphological evaluations, as well as PSD evaluations, provides greater insights into the powder characteristics as opposed to just PSD alone [76]. SEM is often used for evaluating particle morphology as it can provide a greater depth of focus as opposed to OM allowing micrographs of powder particles that appear 3D in nature [62]. OM can be used if powders are dispersed on a flat surface or mounted using metallographic practices and is a relatively inexpensive method compared to the other particle morphological evaluation methods. Image analysis can be performed on the acquired micrographs obtained from performing microscopy. Such analyses are often included in the microscope operating software packages. Murphy et al. [77] provide a procedure for performing such image analyses of powder particles for PBF applications. Charbonneau et al. [78] studied the application of image analysis for evaluating particle shape. They investigated various standard shape descriptors and their sensitivity to morphologies commonly observed in metal AM. Their results indicate that roundness, sphericity, and irregularity shape descriptors provide capabilities for characterizing common particle morphologies seen in metal AM [78].

Skeletal density, also referred to as true density, is the density of the individual powder particles due to their internal pores and voids. Skeletal density is directly related to the density of the produced material in PBF [3]. It has been shown that a low powder skeletal density with a high internal particle porosity increases the porosity in the produced PBF material [79]. Gas pycnometry is commonly used to evaluate skeletal density. Recent research has demonstrated the effectiveness of CT analysis for evaluating both the skeletal density and particle porosity [60]. In addition, gas permeability and the adsorption methods can be used to evaluate powder particle morphological properties. The gas permeability method involves passing gas through a powder bed under regulated pressure and typically in a standard tube or vessel, after which the gas flow rate is measured [74]. The measured permeability can be used to infer information about bulk density or calculate the PBD and average particle size from the envelope-specific surface area [80]. Adsorption methods such as physical adsorption per ASTM B922 can be performed to evaluate the surface area of the powder particles. This surface area includes both the external surface and accessible internal surfaces from defects such as cracks and provides an approximation of the powder particle surface texture [81]. Powders with low particle sizes and high surface area typically result in improvements to both the PBD and density of the PBF-produced material [82].

The application of evaluation methods such as AFM and indentation testing, specifically nanoindentation, provides information for calculating a wide range of mechanical and material properties of individual powder particles. For instance, elastic modulus, hardness, yield stress, and fracture toughness can be calculated from the data obtained from AFM and nanoindentation tests [83, 84]. AFM can be used for evaluating the frictional forces and adhesion amongst other properties of powder particles that greatly affect the flow of powder bulks [85]. Gorji et al. [86] demonstrated the use of AFM and nanoindentation for characterizing metal powders in the virgin and reused conditions. They noted a decrease in the hardness and effective modulus and an increase in the powder particle surface roughness of the reused 316 L stainless steel powder. Lu et al. [87] reported an increase in hardness and elastic modulus of reused 316 L powder using the nanoindentation method. They hypothesize this increase is due to the secondary dendritic arm spacing and residual stress build-up due to repeated thermal exposure [87]. Lu et al. [87] also reported an increase in magnetization intensity and magnetic field strength using a multifunctional vibrating sample magnetometer to evaluate the magnetic hysteresis loop of the 316 L powders.

4.2 Powder bulk properties

Powder bulk properties are the properties that powders exhibit in their bulk form as opposed to individual particles. Table 2 provides a review of applicable methods for determining such bulk properties along with applicable standards and their advantages and disadvantages.

Table 2 Powder bulk evaluation methods

A common bulk powder property that is evaluated for PBF is the PSD. Snow [92] noted that it is one of the most important properties for PBF applications. PSD has been shown to have a significant effect on the resultant mechanical properties of the PBF-produced material [93]. This property aims at describing the size distribution of powder particles and is typically represented in terms of volume or count-based distributions [92]. PSD is commonly represented as a cumulative distribution graph of the particle diameters or by defining the resulting D₁₀, D₅₀, and D₉₀ values. The PSD range is also used whereby the range is calculated by subtracting the D₁₀ value from the D₉₀ value. ASTM E1617 defines the standard practice and templates for reporting PSD data [94]. There are many methods for evaluating powder PSD as defined in Table 2. Ramakrishnan [62] and ASTM E2651 [95] provided extensive reviews of such methods including additional methods for particle sizing and size distributions not included in this review. In addition, ASTM E2651 provides guidance on sampling, sample dispersion, and the limitations of the various PSD evaluation methods [95]. PSD can be measured both statically by methods such as microscopy or CT with subsequent image analysis and dynamically by methods such as dynamic light scattering. It is important to note that various PSD evaluation methods such as laser diffraction assume the particles are spherical. Verification of particle sphericity shall be performed prior to such testing.

Bulk density is defined by ASTM B243 as “the mass per unit volume of a powder under nonstandard conditions” [19]. The bulk density is a measure of specifically untapped powder and includes the interparticle void volume. It is important to differentiate between bulk and AD, whereby AD is a specific type of bulk density and is determined by a specific method such as ASTM B212 using a funnel of standard dimensions and a density cup of standard volume [19]. Bulk density specific to the PBF process is typically termed PBD and is a measure of the volumetric density formed when the powder is layered in the powder bed [96]. The bulk density is very sensitive to handling and preparation, hence repeatability is often a challenge and therefore it is important to specify the preparation method in detail [58]. The bulk density is affected by factors such as particle size, morphology, cohesiveness and moisture. Therefore, it is important to consider these factors when performing this experiment for repeatability purposes and how such environments correlate with the production environment where the powders are being used (humidity of the powders if kept in an open environment, the temperature on the day, etc). Another type of density that is measured for bulk powder is tap density. ASTM B527 is one of the standards that define the method for evaluating tap density [97]. Generally, cohesive powders stick together, and when allowed to pack several air gaps are present, therefore, decreasing the bulk density of the powder. Using an automatic tapper, the particles are allowed to rearrange and reduce the number of air pockets increasing the density. Such evaluation methods make use of a graduated cylinder for measuring the volume of tapped powder. These readings are susceptible to operator bias and are less accurate than mass measurements taken using an analytical balance [98]. Increases in the bulk density due to tapping are beneficial for PBF although the typical recoating mechanisms do not apply similar tapping forces. Of the types of recoating mechanisms, the gravity-fed mechanisms have the closest correlation with the tap density method although further research is needed to quantify such relationships.

The use of standardized funnels can also be used to evaluate powder flow rate and AoR. In the case of PBF, the powder flow rate is typically determined with a Hall flowmeter with some less flowable powders requiring a Carney flowmeter due to PBF powders typically exhibiting low cohesion and being of spherical morphology [96]. Other funnel geometries include the Gustavsson funnel. The flow rate is highly influenced by properties such as particle sizes and cohesiveness. The flow rate is the measure of the time it takes a standard weight or volume of powder to flow through a funnel of standard geometry. The standard method for mass flow rate recommends using 50 g of powder for the flow rate experiment, yet it may be useful to consider using a set volume when comparing different powders. If the powders have different densities, it stands to reason that one powder will be heavier than another. Gravity-induced forces may be the reason one powder flows faster than another. However, it is important to note that if a vast difference in flow rate time is recorded, this is likely not just the result of gravity, yet the effect of other factors such as particle size and cohesiveness. A faster flow rate relates to a less cohesive powder with better flowability properties. If a powder fails to flow through either of the funnels its result is deemed “Does not Flow”, which generally is a product of high cohesive forces. For non-free-flowing powders, the quantification of flow rate typically provides little value and exhibits low repeatability. For powders that do not flow through either a Hall or Carney funnel, the Scott volumeter or Arnold meter can be used for calculating AD. Petersen et al. [99] compared the Hall flowmeter and the Arnold density meter, noting the Hall flowmeter AD results are more sensitive to changes in particle surface texture, whereas the Arnold density meter AD results better simulate die filling, are more sensitive to changes in PSD, and apply to both free and non-free-flowing powders. The use of the Arnold density meter may, therefore, be better suited for evaluating bulk density for gravity-fed PBF processes. Kroeger et al. [100] developed an empirical model for mass flow rate and AD of typical powder metal alloys used for PBF applications as evaluated using either a Hall or Carney funnel. Figure 13 depicts their results, showing the fitted minimum and maximum curves. Beverloo’s law was applied for fitting cohesionless flow and Johanson’s model was used for fitting cohesive flow at different cohesion parameters [101, 102]. Such models provide insight for developing powder material specifications.

Fig. 13

Model of minimum and maximum flow rates and apparent densities for common PBF metal powder alloys. Reproduced with permission from [100]

The AoR is defined as the angle of inclination of the free surface to the horizontal of a bulk solid heap, which is a criterion for the flowability of a powder. AoR can also be determined using funnel methods such as the Hall and Carney funnels. There is, however, a range of additional methods for evaluating AoR [14]. Few standards exist for evaluating AoR; ISO 4324 [103] provides a method for evaluating AoR for powder and granules although there exists no standard method for metal powders for PBF applications specifically. AoR can be evaluated simply by taking pictures of the powder pile formed by passing powder through a standard funnel onto an upturned AD cup. It is best practice to take at least three pictures or measurements located at 120 degrees around the base of the powder pile. Such experiments provide insights into the cohesive forces and flowability of the powder. A low AoR value, typically 35 degrees or less, indicates a less cohesive powder, which in turn indicates better flowability. Geldart et al. [104] demonstrate the usefulness of AoR evaluation and note when evaluated using standardized procedures AoR provides a good indication of powder flowability that is comparable to that of shear testers. The results presented by Sun et al. [105] suggest AoR is a better method for characterizing decreases in powder flowability due to repeated reuses as opposed to the Hall flow rate method. Geldart et al. [104] are of the opinion that AoR is a useful, robust, and reliable method specifically for QC and powder monitoring but note it is not the most appropriate method for very cohesive powders. Related to the AoR are the properties of avalanching such as avalanche angle, often referred to as dynamic AoR, surface fractal, and volume expansion ratio amongst others [106]. Avalanche properties are evaluated by performing the rotating drum test method. This method is an inherently dynamic method and according to Spierings et al. [4] is well suited to the PBF process. Powders typically exhibit one of the four types of flow behavior when rotated in a drum, rolling, slumping, slipping, or cataracting flow behavior, with rolling flow behavior indicating the highest flowability [107]. The drum can be rotated at different speeds. At high speeds, a steady-state behavior is achieved whereas at lower speeds discrete behavior is achieved [106]. Having the ability to alter the speed of the drum provides the analyst with the opportunity to conduct tests at similar speeds to that of the recoater in the PBF process. This is demonstrated by Amado et al. [106] for polymer PBF systems. Whiting et al. [108] studied the repeatability and sensitivity of the rotating drum method and suggest testing at rotational speeds of 4 RPM or greater. Currently, there is no publicly available standard for this method although standard procedures are provided by instrument manufacturers. The results by Krantz et al. [32] indicate that powder AoR and avalanche angle results are closely correlated and provide interchangeable results.

Shear testing is an approach used to evaluate the dynamic properties of bulk powders under specific conditions. There are three types of shear testers that are typically used, the Jenike, Schulze, and Freeman testers [109,110,111]. Some of which can be used for the evaluation of additional properties such as compressibility, densities, and permeability. The additional dynamic testing capabilities of certain devices such as the Freeman FT4 tester, referred to as a rheometer, are particularly beneficial for assessing dynamic flow properties under low powder stress levels similar to those experienced in most PBF processes [90]. Rheometers can measure basic flow energy (BFE), flow rate and consolidation indices, conditioned bulk density (CBD), specific energy (SE), and aerated energy amongst other properties [112]. Low energies such as specific energy indicate better flowability as less energy is required to displace the powder. Automated testing through the use of such rheometers has the benefit of less human interaction when compared to other methods such as the funnel methods which results in increased repeatability [112]. The powders require conditioning before performing such tests to reduce the stress history and improve repeatability [113]. Powders can be evaluated in various states such as consolidated, aerated, or fluidized states [112]. Clayton et al. [114] demonstrated the use of the Freeman tester for evaluating different AM powders at different phases of their lifecycles. Freeman [112] performed a comparative study of a powder rheometer and shear cell tester for powders of different conditions, noting how the rheometer can differentiate non-cohesive powders better than the shear cell tester. Spierings et al. [4] stated that the Schulze ring shear testing approach is not well suited for powders for PBF as it assesses powders under a compressive load, which is not typically seen in PBF processes. Shear forces in PBF are generally exhibited in low magnitudes at the base of the recoating device due to the compressive forces of the weight of the powder bulk as illustrated in Fig. 8. Lyckfeldt [115] investigated the use of the Freeman FT4 tester for steel powders for PBF; he highlights the value of the Freeman FT4 tester for QC and for confirming processing performance.

Various information and bulk properties can be determined from shear testing. This information is determined from graphically constructed Mohr circles and yield loci, and for the case of the Freeman FT4 shear cell tester includes [111]:

• Major principle stress (MPS)

• Unconfined yield strength (UYS)

• Flow function (FF or ff_c)

• Angle of internal friction (AIF)

• Effective angle of friction

• Cohesion

• Wall friction angle

Some shear cell testers and rheometers can perform evaluations under inert atmospheres and at elevated temperatures [116, 117]. This allows for the imitation of the PBF environments and provides results that are more closely related to the PBF processes. For a thorough review of powder shear and flow testing, the reader is referred to the works by Schwedes [118] and Zegzulka et al. [119].

Additional methods for the evaluation of bulk powder include the evaluation of the thermal properties and optical properties such as laser absorptivity. The evaluation of the thermal conductivity of metal powder bulks has been investigated by Gong et al. [120]. Cooke and Slotwinski [22] and Sih et al. [121] provided a review of methods for characterizing the thermal properties of powders as well as their relation to AM, noting the difference between steady-state and transient methods. Pujula et al. [122] measured the thermal conductivity of metal powders using a differential scanning calorimeter (DSC). The Guarded-Hot-Plate apparatus is used for the evaluation of steady-state heat flux and thermal transmission properties as defined by ASTM C177 [123]. Slotwinski and Moylan [91] noted that guidance is needed for applying this method to metal powders. Cooke and Slotwinski [22] and Tan et al. [7] provided a review of additional methods for the evaluation of the thermal properties of metal powders. In general, increased PBD results in better thermal conductivity [124]. Laser absorption can be described as the thermal interaction between the powder feedstock and the laser beam, and is affected by various properties including the packing density and particle arrangement [125]. Particle sizes also affect the laser absorptivity as the literature notes that the irradiation area decreases with increasing particle sizes, and this limits the absorbed irradiance on the powder bed. Therefore, there is a negative correlation between the particle size and the laser absorptivity [125]. Tan et al. [7] noted that the measurement of thermal absorptivity due to the interaction between a laser source and the powder bed is a complicated and expensive undertaking. The use of simulation and numerical methods for evaluating laser absorptivity is a subject of current research. Yao et al. [126] demonstrated the use of such computational methods for evaluating laser absorptivity for different powder PSDs and packing configurations. The application of calorimetry such as the method proposed by Rubenchik et al. [127] offers good agreement with direct numerical calculations.

5 Evaluation of powder flowability and spreadability

In addition to the standard evaluation methods for bulk powders, as reviewed in the previous section, there are various metrics for classifying and comparing bulk powders. Terms such as flowability and spreadability are often used when describing metal powders for PBF applications, these terms are difficult to quantify, and often multiple bulk powder properties are evaluated to characterize them as composite properties.

Various principles are relevant to the flowability of powders when used in bulk. Two such important principles are agglomeration and cohesion. Agglomeration can be regarded as the formation of clusters of particles within a set volume which has a larger density than their surrounding particles. These can form due to forces such as Van der Waals forces and capillary forces amongst other reasons. They can be detrimental to production processes, as the presence of agglomerations can lead to pore formation which can ultimately lead to the formation of defects in the produced component. They may, however, be produced on purpose to aid flowability [128]. Walton and Mumford [129] described various particle agglomerates and their structures. They also explained the influence agglomeration has on flowability and cohesiveness. Cohesion can be described as the degree to which powder particles “stick” together or are attracted to one another and is evaluated as the “resistance of a powder to shear at zero compressive normal loads” [130]. Although cohesive forces are typically small, they can result in powder agglomeration, this is especially relevant to fine particles [130]. Various factors influence the cohesiveness of powders, and this, in turn, affects other properties such as flowability or spreadability. Powder cohesion is inversely proportional to powder flowability [85]. Moisture, particle size, and morphology are major factors affecting the cohesiveness of powders. Smaller particles have higher interparticle forces resulting in increased cohesiveness [58]. Abdullah and Geldart [98] regarded the Hauser ratio (HR) as a useful measure of powder cohesion [131]. HR has also been used as a metric for characterizing changes in the electrostatic charge of powders [132].

Powder flowability is a key aspect of understanding the dynamic behavior of powders. It is not an inherent material property but rather an “umbrella” term for describing the complex behavior of powders in motion [35, 114]. Flowability should be evaluated in relation to the application of the powder, such as the PBF recoating mechanism or powder dosing mechanism. Powders used for PBF typically exhibit good flowability [91]. Terms such as spreadability or rakeability are often used when investigating the ability of powders to be spread across the PBF powder bed [35, 96, 133]. Spreadability is closely related to flowability but can also include aspects of layer homogeneity, packing or bulk density, and other quality metrics such as cohesion. Powder flowability tends to increase with an increase in sphericity, increase in size, and decrease in particle surface roughness. Figure 14 visually depicts how the powder particle morphologies affect flowability, specifically in terms of AoR [134]. Correlating flowability with the spreadability of powders in the powder bed is required to determine the powder’s ability to achieve conformant powder layers. Powders having a wide PSD, being spherical in shape, and containing few fines results in improved spreadability and powder layer density. Particle size and size distribution are often found to have the greatest effect on the flowability of the powder bulk [135]. Wide PSDs result in improved packing density or powder bed cohesion but also can reduce flowability and increase the risk of segregation. Haferkamp et al. [136] investigated the effect of particle shape on powder flowability and other PBF spreadability metrics. They found that powders with a circularity¹ of greater than 0.8 exhibit improvements in flowability but do not result in significant improvements to the density of the produced material [136]. Young et al. [137] characterized PBF powder PSDs and avalanche angles and demonstrated their effects on the PBF process. They noted a decrease in the volume of spatter particles with an increase in the percentage of fine particles in the powder PSD [137].

Fig. 14

Relationship between particle morphology and AoR [134]

Common metrics and bulk properties used for characterizing powder flowability are tabulated in Table 3 including AoR, Carr’s Compressibility Index (CI), HR, and the flowability or Flow Function (ff_c). An additional metric for characterizing powders is the gas fluidization approach proposed by Geldart [138] in his seminal work on the topic. Such metrics provide information for performing statistical process control (SPC) on powders at various phases of their lifecycle. It shall, however, be noted that there is little consensus or standardization on flowability classifications and metrics. Table 3 is compiled from the seminal publications by Carr [139], Jenike [140], and McGlinchey [9], and should be used only as a guide for classifying powder flowability. Santomaso et al. [141] provide an investigation of density ratios and their effects on powder flowability. The HR provides a good indication of flowability and relates the tapped density to the bulk density. Guidance for PBF by Karapatis [142] suggests using powder with an HR of less than 1.25 is required to ensure reliable flowability. Additional metrics have been proposed by Karapatis et al. [143] and Spierings et al. [144] for optimizing the D₁₀, D₅₀, and D₉₀ values of the powder PSD in relation to powder flowability and the PBF layer thickness. Spierings et al. [144] proposed that the layer thickness to D₉₀ ratio should be near 1.5 to allow for effective powder layering and the D₉₀ to D₁₀ ratio should be near 5 to allow for sufficient fine particles. There is a caveat in that the D₁₀ value should be greater than 6 microns to reduce excessive powder agglomeration and hinder powder flowability [144]. Brika et al. [145] proposed an AM suitability metric for comparing powders from different suppliers for the LPBF process. Marchetti and Hulme-Smith [146] performed a regression analysis to evaluate the relationships between eight different flowability metrics for steel powders. Figure 15 presents the correlation matrix they developed, abbreviations are defined in Sect. 4.2 under rheometry and shear testing [146]. They found specific energy (SE), HR, compressibility (CPS), and AoR to be inter-correlated [146]. Caution shall be taken when evaluating powders on multiple metrics. Krantz et al. [32] showed how the development and use of composite metrics and indices can provide misleading and in certain cases contradictory results.

Table 3 Powder flowability metrics and classification

Fig. 15

Correlation of flowability metrics showing Pearson coefficients (white = positive, black = negative) [146]

By relating all the powder bulk properties, we can look at a more practical parameter to determine the suitability of a powder for PBF processes. The spreadability, as mentioned earlier, looks at how well the powder spreads across the build plate to form a powder bed and is greatly affected by the powder flowability. Figure 16 depicts a cause-and-effect diagram developed based on the synthesized results from the literature reviewed in this publication. Green items provide improvements to spreadability and red items can hinder effective spreadability. Causes marked with an asterisk indicate they contain caveats and interrelated causes such as the effect of flowability on PBD are not indicated in this figure.

Fig. 16

General causes of improvements to powder spreadability

The powder bulk properties will have a significant influence on the spreadability. A cohesive powder with agglomerates will clump up and result in inadequate spreading. The PSD also affects spreadability. The layer height of the recoater blade should be bigger than the particle sizes of the majority of powders. If not, only a limited amount of powder can fit under the recoating mechanism and hindering powder flow and spreadability [89]. It is evident that spreadability is affected by the cohesive nature of the powder, the PSD, and the AD among other factors. From these, the cohesive nature and spreadability can be evaluated indirectly by conducting multiple experiments, such as the flow rate, AoR, and AD tests. To effectively determine a powder’s spreadability all of these factors should be considered. From this, the suitability of a powder for PBF can be determined. Snow et al. [11] demonstrated how AoR can be used as a simple and inexpensive method and metric for quantifying powder spreadability.

Spreadability can also be quantified by determining the PBD and other properties such as powder bed ST. A common approach to evaluate spreadability directly is to use a spreading rig with a base of a known area and recoating mechanism. These devices range from simple manual spreadability apparatuses to more complex and automated devices. The development and use of spreadability devices is a topic of current research in the field with notable works by Snow et al. [11], Le et al. [147], Parker [88], Hulme-Smith et al. [148], and Mitterlehner et al. [149], depicted in Fig. 17. There is currently a lack of standardization for the direct evaluation of spreadability using such devices [150].

Fig. 17

Spreadability testing device. Reproduced with permission [149]

Spreadability devices typically spread powders over the build plate using a recoating mechanism, after which spreadability metrics can be determined by applying image analysis and taking density measurements [133, 148]. The percentage coverage is determined by spreading the exact amount of powder needed to cover the build plate at a specific height and then capturing an image of the coverage of the build plate. Image processing methods such as color thresholding are then performed to convert the image to black and white pixels and segregate the voids, after which the percentage coverage, which is related to ST, can be calculated [89]. The powder layer density (PLD) is determined through the same rig in a similar fashion. Powder in excess is spread across the build plate. The remaining powder on the build plate is weighed, and the volume is known (area of the build plate and layer height chosen) [89]. Evaluation of PBD requires a rig that has capabilities for adjusting the layer height consecutively. Such capabilities allow for the evaluation of PBD at different stages of the build and incorporate the effects of the powder weight on the resulting PBD. An additional approach is proposed by Jacob et al. [151] who designed and tested a capsule-type specimen that can be built in situ for characterizing PBD at various locations within the PBF build. This work refers to the original investigation by Karapatis et al. [143]. Jacob et al. [96] conducted further investigations specifically on the factors that influence powder spreadability for PBF with a rigid recoater blade. Mussatto et al. [134] used a digital microscope to characterize the ST, including height profiles, void volumes, and the segregation of powders spread across the build plate under different spreading conditions including layer thickness, recoater speed, and powder suppliers. Wischeropp et al. [152] performed spreadability experiments to determine the actual layer height and packing density during LPBF. Their results suggest the actual layer height can be as much as five and a half times greater than layer thickness. Spatter and the denudation effect are reported as explanations for this discrepancy [152]. Figure 18 illustrates how the fine powder particles are deposited earlier towards the beginning of the recoating process due to granular convention, also known as the Brazil Nut effect [134]. For a review of spreadability metrics and parameters for different LPBF recoater types, the reader is referred to the work by Sehhat and Mahdianikhotbesara [153]. Additional in situ approaches include the use of image analysis for characterizing the powder layer homogeneity. Such approaches are available as in-process monitoring capabilities for certain commercial PBF systems. Neveu et al. [154] investigated the use of image analysis performed in situ for the evaluation of powder bed surface homogeneity.

Fig. 18

Powder segregation during recoating [134]

In addition to conducting physical testing, research has been conducted on the use of simulation and numerical modeling for investigating powder flowability and PBF spreadability, largely through the application of the discrete element method (DEM). Parteli and Pöschel [155] developed such a model for investigating the roller-type recoating mechanism as depicted in Fig. 7. They found that increasing the recoating speed and broadening the powder PSD resulted in increased surface roughness of the resulting powder bed. Gürtler et al. [156] applied a numerical method for investigating the effects that different powder distributions have on produced material properties. Their findings highlight that powders with small particles result in better powder beds and such quality powder beds lead to reduced porosity in the produced material. Chen et al. [157] utilized DEM to investigate powder spreadability for PBF by considering contact and cohesion forces. They found that a reduction in friction between the recoater and powder particles results in improved powder fluidity and powder bed quality. Additionally, they found that for powders with particle sizes greater than 21.8 $\mu m$ the fluidity of the powder is improved [157]. A recent study by Xu and Nan [158] using DEM simulations found that spreadability is not necessarily directly related to flowability and that the shear action of the recoater blade on the powder is the main contributing factor to spreadability. For a thorough review of powder spreadability in various AM modalities, the reader is referred to the paper by Miao et al. [159].

6 Metal powder processing and management

Processing of metal powders has been performed since the inception of PM. Processing can be performed for various reasons including transporting powders, altering or improving their properties, and homogenizing powder bulks among other conditioning processes [27]. Two risks that are prominent when processing metal powders are the risk of contamination and the risk of combustion. Certain metal alloys and powders with high percentages of fine particles are susceptible to elemental pick-up from their environment. Processing of metal powders requires a highly controlled environment and is typically performed in a closed-loop system.

6.1 Powder processing and reuse

For PBF production, typical powder processing phases are the processes performed for reclaiming and reusing powders. Figure 19 presents a depiction of a typical powder lifecycle during PBF production. Reclaimed powders require varying forms of reconditioning or rejuvenation before they can be used in subsequent production runs. This phase consists of different subprocesses including powder removal from the build chamber, sieving, blending and/or mixing, and in certain situations, drying. Packaging and transportation can occur between these subprocesses as well depending on the production setup, which adds additional risk.

Fig. 19

Powder lifecycle during PBF production

A common reuse strategy implemented for PBF production is the continuous refreshing or replenishing approach where any reclaimed powder is reconditioned and blended with virgin powder to top up the batch for subsequent production. Figure 20 depicts the continuous refreshing approach along with a collective aging approach for powder reuse as presented by Lutter-Günther et al. [160]. Other strategies include no reuse whereby only virgin powder is used, such an approach is uneconomical in an industrial environment. Harkin et al. [161] investigated powder reuse of Ti–6Al–4V grade 23 powder using the continuous refreshing or top-up approach up to 9 cycles. They noted a gradual increase in oxygen content, no significant changes to the PSD, and improvements to the Hall flow rate and both HR and CI values with repeated reuse [161]. Similar results have been reported by Delacroix et al. [68] for 316 L powder. Delacroix et al. [68] implemented a reuse strategy that involved continuous use without refreshing with virgin powder and did report a slight increase in particle size. Tang et al. [162] and Shanbhag and Vlasea [163] investigated Ti–6Al–4V powder reuse during EBPBF production, both reporting an increase in oxygen content, with Tang et al. [162] reporting increases in Hall flow rate and decreases in tap density with repeated reuses.

Fig. 20

Powder reuse strategies. Reproduced with permission from [160]

There is a need for powder reuse and recycling standards for PBF [150]. Recently, two standards have been published. SAE AMS7031 defines requirements for powder reuse and identifies additional reuse strategies for non-closed-loop systems [23]. In addition, ASTM F3456 provides guidance for powder traceability throughout its production lifecycle by defining a reuse schema [164].

Sieving, blending, and mixing are common processes performed for reconditioning powders for reuse. Thermal treatment processes are also used for certain applications. Sieving is performed to remove powder agglomerates and defective particles that form during the PBF process. Sieving is typically performed with a sieve mesh size that correlates to the upper D₉₀ value of the qualified powder PSD or above. Other approaches include correlating the sieving mesh size with the PBF layer thickness to reduce the probability of particles larger than the layer thickness that could result in powder bed surface inhomogeneities such as recoater streaking. Sieving of virgin powder before use is also performed as a QC mechanism. It is important to differentiate between blending and mixing. Mixing is a process where a powder batch is processed to homogenize the powder particles throughout the batch [14]. Blending refers to the combining of multiple powder batches that consist of powder of different conditions or properties [14], an example being the blending of reclaimed powder with virgin powder for rejuvenation and reuse in PBF production. Mixing can also be performed to reduce powder particle size segregation due to settling or lengthy storage periods and improve the resulting powder bed density [7]. Olakanmi [165] demonstrated the importance of adequate powder mixing and how mixing reduces powder agglomeration and improves the PBD, this in turn improves the produce material density and reduces porosity. Mixing can be applied to aerate metal powders or to eliminate the powder’s stress history due to previous processing steps or storage, ultimately resulting in more repeatable spreading.

Thermal powder treatment processes have been applied in PM since its inception. Such thermal treatment processes include drying and annealing [14]. Li et al. [166] and Weingarten et al. [167] demonstrated how powder drying can result in improvements in the density of the produced material and parts. It is well known that dry powders or powders with a low moisture content also exhibit higher flowability and have a lower risk of oxidation during the PBF process. Some PBF machines can perform powder drying in situ using a heated platform. Other in situ drying approaches involve the use of the laser at low power to dry the powders in the PBF machine [167]. Drying under a vacuum or inert environment offers additional benefits, especially for fine and reactive metal powders. Cordova et al. [25] investigated the use of vacuum drying and their results show the benefits of vacuum drying on powder flowability for PBF applications as depicted in Fig. 21.

Fig. 21

Hall flow rate for different metal powders of different moisture conditions. Reproduced with permission from [25]

Powder annealing, although not well researched specifically for PBF applications, is often applied during powder production and for traditional PM applications [58]. Powder annealing can be applied to reduce the carbon, nitrogen, and oxygen content of powders [168]. These thermal treatments can also alter the microstructure of the powder particles. Such annealing processes may prove beneficial as a rejuvenation step during powder reuse for PBF applications.

6.2 Powder management and safety

Management of powder feedstock within the production facility is an important task. Standard operating procedures (SOP) are required to ensure such materials are processed, handled, and controlled correctly. Environmental controls are required in the production facility and at all locations where powders are processed, stored, or evaluated. Keeping temperature and humidity within these locations constant is important to reduce the risk of contamination due to aspects such as condensation. Hall et al. [169] investigated different powder storage and handling approaches for common PBF metal powder alloys. Factors investigated included storage container material, container atmosphere, atmosphere control approaches, facility environmental control approaches, and the effects of condensation and handling large powder batches. Their general recommendations include the use of metal storage containers backfilled with inert gas and temperature and humidity-controlled processing facilities [169]. The use of inert processing environments not only reduces environmental contamination but also reduces the risk of powder combustion.

Powder spills are a prominent concern for operator health and safety, fire hazards, and material contamination. Powder spills can go unnoticed and accumulation of powder on work surfaces and floors can spread through the facility. Particle sizes of metal powders used for PBF production are small enough to be inhaled and can penetrate deep into the lungs [170]. The workplace needs to be designed with this in mind. Figure 22 illustrates the importance of flooring selection and how the color of facility flooring can be used to improve the visibility of powder when spilt and reduce the risk of spills going unnoticed [171].

Fig. 22

Visibility of Ti–6Al–4V powder on floor tiles of different colors. Reproduced with permission from [171]

Powder disposal is another important aspect when working with powders. The powders need to be disposed of in the correct manner as defined in their accompanying material safety data sheets (MSDS). Incorrect disposal could lead to cross-contamination as well as adverse health effects if certain metal alloy powders come in contact with skin or are inhaled. A recent study conducted in industry demonstrated the use of resin for producing a solid passivated block from the waste powder condensate to allow for safe transportation and disposal [172]. Laboratory precautions are also important when working with powder, such as the correct PPE and environmental protection and control.

7 Challenges

There are currently no standardized evaluation methods developed specifically for the PBF process. Research has been conducted to assess the applicability of currently available methods to the PBF process to varying degrees. Certain methods show greater applicability than others, although it shall be noted that most currently available methods provide good capabilities for QC and powder monitoring even if they are not well suited to aspects of PBF production. The selection of evaluation methods shall account for repeatability, reproducibility, and measurement uncertainties. Cooke et al. [22] noted that standardizing methods to evaluate chemical composition, thermal properties, and morphology for PBF metal powders may pose challenges. Clayton et al. [114] highlighted the need for multiple powder evaluation methods to characterize process behavior. The majority of evaluation methods are conducted at room temperature and in air. The PBF processes are typically performed under inert atmospheres, and the electron beam PBF processes are performed at elevated temperatures. Mimicking these environmental conditions offline during powder testing and evaluation can provide additional insights into application-specific powder behavior and properties.

It has been found that the properties and composition of metal powders can differ significantly between suppliers and production lots. This results in the need to perform powder evaluation in-house and on a frequent basis. Such activities add additional costs and result in more expensive final products. The frequency and cost of these activities are exacerbated when powder reconditioning and reuse are performed, especially when complex strategies such as collective storage and aging are implemented. Powder reuse poses a range of additional challenges. Traceability and documentation of powder batches become increasingly difficult the more times the powder is reconditioned and reused and is compounded when powders lots and batches are blended. Standards have recently been published defining reuse schemas and requirements, although there are few best practice solutions to this problem [164].

The use of fine powders provides benefits and challenges for PBF. Fine powders can result in highly dense powder beds but are challenging to work with and process. Fine sieve mesh sizes are expensive, damage easily, and powders tend to clump up when used. Sieving is often applied to remove nonconformant powder particles such as agglomerates and particles fused with satellites, although this method does not remove such particles that are smaller than the sieve mesh size. This can result in poor powder flowability, high concentrations of soot, and irregular powder beds amongst other issues. Typically powder bulks with smaller particle sizes exhibit higher surface areas, as a result, such powders are more susceptible to contamination from their environment [124]. As highlighted by Tan et al. [7], further research is needed to fully understand the effects of powder particle surface area on flowability and the properties of the produced material. Evaluation of the PSD of powders is a commonly performed method. The evaluation of particle shape and morphology is less common, at least quantitatively. Recently, research has been conducted to investigate the effects of particle morphologies on spreadability as well as approaches for such evaluations. It is theorized that the evaluation of both PSD and particle shape distributions will provide a greater understanding of such effects and allow for improved powder control. Research into the production and use of powders with multi-model PSDs specifically designed for the PBF recoating mechanisms is an area of interest. Such research is needed to reduce the inverse effects that exist between powder flowability and PBD due to different PSDs [7]. Vock et al. [35] highlighted such contradictory effects as a current challenge.

Terminology and standardization of spreadability evaluation methods and metrics are needed. There has been a fair amount of research on this topic in recent years with various proposals, although there is little consensus on best practice approaches or terminology for describing critical spreadability and PBF-specific density aspects [7].

Occupational and environmental health and safety when handling, processing, and disposing of metal powders are not unique to PBF. Precautionary measures such as the use of PPE are common in the industry although there is little research on the health and environmental effects of metal powder exposure [170]. Recent research has investigated disposal approaches for metal powders. Disposal of metal powders is a costly expense, and this is substantial when scaling up to industrial production. Further research is needed to investigate powder disposal methods as well as methods for recycling nonconformant powders.

8 Conclusion

Powder evaluation and management are important aspects of PBF production both in an academic and industrial setting. Careful thought is needed when selecting the methods and equipment to be used for these activities. Different aspects of the powders, both particulate and bulk, need to be evaluated to provide a holistic understanding of the powder composition and properties. Powder evaluation need not be an expensive undertaking, through the use of simple funnel method apparatuses and microscopy a wide range of both particulate and bulk powder properties can be evaluated. Further research is needed to fully understand the effects that different powder properties have on the PBF recoating mechanisms, specifically spreadability, and the produced material and components. Standardization is needed for PBF-specific powder terminology and direct spreadability evaluation methods.

Powder evaluation methods and cause-and-effects of important PBF powder characteristics are reviewed and synthesized in this paper for powder bed fusion production. This information provides value for the development and conducting of efficient and effective powder testing and evaluation programs. Investigating the needs of the specific PBF application in tandem with the different evaluation methods is required for ensuring powders are adequately used and managed during production.