Research

Additive Manufacturing (AM)

During the past decade, increasing attention has been drawn to additive manufacturing (AM), also known as 3D printing, a unique manufacturing process that allows for parts of various complexity to be built layer-by-layer with relatively more ease. Direct Laser Deposition (DLD) and Laser-Powder Bed Fusion (L-PBF) are two common laser-based AM techniques for metals fabrication.

1. Direct Laser Deposition (DLD)

2. Laser-Powder Bed Fusion (L-PBF)

AM Process-Structure-Property-Performance Relationships

The inter-related relationships among process parameters, thermal history, microstructure, and fatigue behavior of AM components are presented schematically here. As seen, utilized process and design parameters affect the thermal history (i.e. cooling rate, thermal gradients, and cyclic reheating) of the AM part. The thermal history during fabrication governs solidification, and consequently, all the resultant microstructural details such as: grain size, morphology, and orientation; defect size, type, and distribution; residual stress, etc. Accordingly, these microstructural features dictate the structural properties, and especially the fatigue performance, of fabricated parts.

Building Orientation and Mechanical Properties

The orientation in which AM parts are built (i.e. build orientation) may greatly affect defect directionalities (i.e. aspect ratio in shape), and thus, generates and dictates their anisotropic structural response, especially in tensile strength, elongation to failure, and fatigue resistance. In addition, the anisotropy may also be resulting from changes in the thermal history during fabrication (i.e. cooling rate and cyclic re-heating from subsequent layers), which affect microstructural details (i.e. grain size, phase fraction, defect size, type and distribution, etc.).

In general, as-built AM parts inherently consist of anisotropic microstructure due to an uneven thermal history and directional heat transfer that the parts experience during fabrication. Microstructural features, including grain size, grain morphology, and crystallographic orientation, affect the fatigue performance and failure mechanism of the part, especially pertaining to crack initiation and short crack growth. Therefore, the effects of microstructural features, as driven by the thermal history during the AM process, need to be considered when investigating the anisotropic behavior of AM parts.

In the absence of voids and inclusions, slip bands usually drive crack initiation in metallic materials. In general, finer microstructures provide better crack initiation resistance than coarser microstructures due to a higher density of slip bands – when crack initiation occurs in slip bands within grains. Additionally, fatigue crack initiation of textured materials is controlled by the orientation of the active slip system(s) with respect to the loading direction(s) – i.e. maximum shear stress. For instance, higher fatigue strength has been reported for Ti-6Al-4V when the maximum resolved shear direction is perpendicular to basal planes, where the easiest and most common slip systems, basal slip, reside in this alloy. Crystallographic orientations of the adjacent grains may also act as a barrier for short crack growth. The crack path deflection across a grain boundary is strongly influenced by the relationship between the grain orientations of neighboring grains. Therefore, high-angle grain boundaries act as an effective barrier to transgranular short crack growth. Crack growth may be retarded/arrested when none of the available slip systems are oriented closely. Grain size and morphology can also influence intergranular fatigue crack growth, leading to anisotropy in fatigue performance of AM materials. Typically, coarser grains can provide better crack growth resistance due to their larger grain boundaries, causing larger crack deflections. In addition, anisotropic grain growth, leading to an elongated grain morphology, may affect crack growth for different loading directions. Elongated grains (i.e. columnar) typically form during the AM process in the direction of solidification, which tends to be near-parallel with the building direction. In cases where loading is perpendicular to the building direction (i.e. the elongated direction of grains), cracks typically grow parallel to the building direction, as shown here, and therefore, they experience less deflection in the path, leading to a lower crack growth resistance. On the other hand, a higher crack growth resistance can be expected when the crack growth is perpendicular to the building direction, as shown here; such cracks experience a more tortuous and deflective crack path.

It has been found that 17-4 PH SS, fabricated via an L-PBF method in the vertical and horizontal orientations, reveal anisotropy in tensile strength, elongation to failure, and fatigue resistance, as shown here. It may be seen that horizontally-built specimens in their as-built condition exhibit higher monotonic tensile strength, fatigue resistance, and elongation to failure relative to the L-PBF parts built vertically.

It is expected that post manufacturing heat treatments, such as solution/homogenizing annealing, will remove the aforementioned microstructural directionality or heterogeneities imposed by part build orientation or directional solidification during the AM process. Inspection of L-PBF 17-4 PH SS microstructure has revealed no difference in grain size, grain morphology, or crystallographic orientation within vertically- and horizontally-built samples after heat treatment (solution annealing for 30 min at 1040 C and peak-aging for 1 h at 482 C). Although the microstructures were almost homogenized after heat treatment, the L-PBF 17-4 PH SS specimens still displayed anisotropy in tensile and fatigue strengths as well as elongation to failure, as shown here. This figure shows engineering stress–strain curves and fully-reversed (R = 1) strain–life fatigue experimental data for L-PBF 17-4 PH SS in heat treated condition. These results suggest that the observed structural anisotropy in AM parts may be more influenced by defects rather than microstructure.

Voids & X-Ray Tomography

A rotating 3D volumetric rendering from X-ray CT scan of vertical and horizontal L-PBF 17-4 PH SS specimens showing the void distribution within the gage section. The detected voids are color-coded based on volume (mm3). Blue: <1E-3; Green: 1E-3 – 2.5E-3; Orange: 2.5E-3 – 4.5E-3; and Red: >4.5E-3.

The building orientation of parts during L-PBF was found to have a significant influence on their eventual fatigue properties. Higher fatigue strength of horizontal specimens was mainly attributed to the orientation of deposited layers with respect to loading axis. Defects that formed between layers of vertical specimens were much more detrimental as they provided higher stress concentrations since a larger area was exposed to loading.

A Note on Process Optimization

Cognizant of the voids/defects present within the AM parts investigated, one may question whether optimizing AM process parameters based on density, as measured using Archimedes’ principle, is a sufficient means for improving mechanical properties, especially fatigue resistance of parts. Results indicate that maximizing density, measured per part by employing Archimedes’ principle, is not an ineffective means for improving the fatigue strength of AM part, as the existence of slit-shaped defects, or un-melted regions, cannot be easily prevented or measured.

These slit-shaped flaws, which can cover a broad region with small volume, may not be detected by density measurements that utilize Archimedes’ principle. This may be an issue since these un-melted regions appear to be the most detrimental type of flaw with respect to fatigue. Aside from their relatively large size, their irregular shape. Thus, process parameter optimization based on density alone, as measured by Archimedes’ principle, may not be a sufficiently-accurate criterion for achieving enhanced mechanical properties – or at the least fatigue strength. This does not mean that Archimedes’ principle should not be used; indeed, Archimedes’ method can be taken into account for a priori optimization of process parameters, but not beyond – especially when fatigue is targeted as an output property.

In order to improve fatigue strength of AM parts, measuring porosity using X-ray computed tomography (CT) can be more beneficial, because it more exhaustively characterizes the level of porosity within a material. Using this method, defects’ size/distribution, shape (e.g. volume to area ratio), and spacing, which are more representative of the defect statistics, and also the most influential ones affecting fatigue behavior, can be quantified and utilized in the optimization process. A novel systematic, ‘multi-objective’ process optimization method for obtaining fully-dense materials with target mechanical properties is required.

Building Orientation and Surface Roughness

In order to improve fatigue strength of AM parts, measuring porosity using X-ray computed tomography (CT) can be more beneficial, because it more exhaustively characterizes the level of porosity within a material. Using this method, defects’ size/distribution, shape (e.g. volume to area ratio), and spacing, which are more representative of the defect statistics, and also the most influential ones affecting fatigue behavior, can be quantified and utilized in the optimization process. A novel systematic, ‘multi-objective’ process optimization method for obtaining fully-dense materials with target mechanical properties is required. Anisotropic, distinct roughness can also exist along the surface of an AM part. For instance, for a part fabricated via L-PBF while oriented in a vertical incline, the overhanging side (i.e. facing downward toward the build-plate) is found to possess a higher surface roughness relative to the contracting surface (i.e. upward facing side), as shown in this figure. This figure presents the X-ray CT image of a 45° orientated Inconel 718 specimen, fabricated via an L-PBF method. Higher surface roughness of the overhanging side is attributed to the more direct contact of this face with the powder bed during manufacture and in this giving rise to melt pool thermal/fluidic edge effects. As with defects and microstructure, the surface roughness varies with respect to position within the part . It has been found that more near-surface voids form along the downward-facing side of a part fabricated at an incline, due to melt pool thermal/fluidic edge effects while in contact with powder bed during manufacture, e.g., capillary action, heat build-up, and more.

Specimen Design and Surface Machining

The X-ray CT scans, taken from the gage section of an as-built L-PBF Inconel 718 specimen, revealed the presence of large voids along its perimeter, as can be seen below. Direct contact of the part’s surface with the powder bed during manufacture may have given rise to melt pool thermal/fluidic edge effects (i.e. instabilities), leading to near-surface voids along the edges of the part. In addition, the HIP process cannot remove open voids (i.e. surface-connected voids), because these type of voids act as an extension of the specimen’s surface . Therefore, for as-built L-PBF parts, the probability of the voids being near the surfaces is higher than machined ones. These observations suggest that the thickness of material removed during machining may play a significant role on the fatigue behavior of post-machined AM parts. In other words, depending on the thickness of the outer layer that is trimmed away during machining, the voids may be removed or brought to the surface. As a result, fatigue behavior of machined specimens may be different based on the specimen design, as shown below.

This figure schematically shows different specimen designs, including fabricating the near net shape specimen as well as cylindrical rod. As seen, the thickness of the outer layer that needs to be trimmed away during machining is thinner for the near net shape specimen as compared to the rod one. Accordingly, by removing a thin layer from the surface of a near net shape fabricated specimen, the effect of surface machining on fatigue life may not be as pronounced. These findings suggest that the specimen design procedure may need to be standardized for AM materials in order to obtain a better understanding of their fatigue behavior as it pertains to the more robust engineering of parts fabricated via AM. However, if the process/design parameters are optimized in the way that there are not any near surface voids in the as-build specimens, the thickness of the removed layer during the post-manufacturing machining process should not greatly affect the fatigue behavior.

Sources

Thompson, SM., Bian, L., Shamsaei, N., & Yadollahi, A. (2015). “An overview of Direct Laser Deposition for additive manufacturing; Part I: Transport phenomena, modeling and diagnostics.” Additive Manufacturing, 8, 36-62.

Shamsaei, N., Yadollahi, A., Bian, L., & Thompson, SM. (2015). “An overview of Direct Laser Deposition for additive manufacturing; Part II: Mechanical behavior, process parameter optimization and control.” Additive Manufacturing, 8, 12-35.

Yadollahi, A., Shamsaei, N., Thompson, SM., Bian, L., & Elwany, A. (2017) “Effects of Building Orientation and Heat Treatment on Fatigue Behavior of Selective Laser Melted 17-4 PH Stainless Steel.” International Journal of Fatigue, 94, Part 2, 218–235.

Yadollahi, A., Shamsaei, N., (2017) Additive Manufacturing of Fatigue Resistant Materials: Challenges and Opportunities.” International Journal of Fatigue, 98, 14–31.