X-Ray Diffraction Measurements for Inconel 718 Alloy Elements Created by Incremental Methods

In recent years, numerous interdisciplinary research groups have focused their efforts on the rapidly developing field of additive manufacturing (AM) (Pinkerton & Li, 2004; Wang et al., 2017), due to AM’s capability to fabricate parts with complex geometries that cannot be obtained by regular manufacturing methods. Unlike traditional subtractive manufacturing processes, such as machining, AM constructs parts directly in their intended shape by consolidating the material in the form of small volumes, layer by layer (Wang et al., 2017), following CAD models. The use of additive manufacturing expands the bounds of human creativity, enabling the design of parts that perform their intended functions to a better degree and extent than before. 3D printing methods that employ a laser energy source are finding applications in the production of relatively small parts, characterized by a high complexity of sophisticated geometry. These techniques are beginning to be implemented to produce components from materials commonly used in aerospace such as titanium alloys, aluminum or nickel superalloys (Wang et al., 2017), such as Inconel 718.

Due to the fact that additive parts may be characterized by the appearance of undesirable characteristics such as excessive surface roughness, porosity, the occurrence of thermal stresses as a result of the printing process itself (Knuth, 1991), and changes in microstructure, grain size and symmetry, as well as the distribution and size of various precipitates and phases, it is crucial to rigorously evaluate these materials and the parts, including strength and fatigue testing, before a given technology can be adopted for serial production. Residual stresses cause deformation and cracking of the part after machining and heat treatment, and the fatigue strength of the component deteriorates when the surface layer has relatively high tensile stresses (Shiomi et al., 2004). This is a particularly important in fields of industry and transportation where the fatigue strength of components is a crucial aspect for ensuring safety and safeguarding human lives – as is the case, for example, in air transportation.

From the point of view of fatigue testing, it is very important that the surface stresses on specimens dedicated to these tests need to be within the appropriate range as defined by the relevant standards before proceeding to their implementation. The study by Gadalinska et al. (2019) used diffraction methods to determine the level of surface stresses introduced in the process of turning and grinding in Inconel 718 samples to be used in fatigue tests, concluding that grinding introduces the highest absolute value of compressive stresses. Moreover, by analyzing the half-widths of the diffraction reflections, significant differences in their values were observed for the surface states after turning and after grinding, suggesting that the process of grinding may affect micro-stresses, dislocation density, and grain size in the surface layer of the investigated material. Malicki et al. (2016) also estimated the level of strengthening of the material from full width at half maximum of X-ray diffractions peaks and compared the hardness of the Inconel 718 alloy before and after heat treatment.

Ferreri et al. (2020) comparatively addressed the question of phase changes of γ, γ′, γ′′, δ and carbide MC phases for forged specimens and specimens produced using additive methods (Laser Powder Bed Fusion – LPBF) immediately after printing and after the Hot Isotastic Pressing (HIP) procedure for Inconel 718 alloy. Neutron diffraction was used as the main research method, verified by scanning electron microscopy. A significantly higher volume fraction of the δ phase and a significantly lower volume fraction of the strengthening γ phase were observed in the material after printing but by using the HIP procedure, than for the parts printed after the HIP procedure and forged. Ferreri et al. conclude that the LPBF process creates microstructural conditions that promote the transformation of γ′′ to δ during the heat treatment of AMS 5663. In addition, they found that the δ phase in the LPBF material is much smaller and is located both at the grain boundaries and inside the grains, in contrast to the forged material, where it is coarse-grained and intergranular. In the case of HIP-treated material, slightly smaller δ and MC particle sizes contributing to more dispersion strengthening were observed after printing. An interesting and surprising observation was that despite having lower γ′′ content, the LPBF material is stronger than both LPBF-HIP and wrought materials. Such behavior was rationalized by the finer grain structure in the LPBF material.

Measurement of the stresses produced by the printing process was carried out, among other things, in the work reported in the article by Shiomi et al. (2004), using the destructive method of applying a milling machine to remove layers of material of the printed part with a thickness of 0.1–0.2 mm. The next step of this measurement strategy was to measure the strain on a macroscopic scale. The measurement procedure was carried out until the entire printed part was removed. Information about the distribution of residual stresses was obtained, with tensile stresses in the surface and subsurface layers, and this effect was independent of the scanning speed of the printing process. Thanks to the applied heat treatment, Shiomi et al. managed to reduce the residual stresses by up to about 70%, while the layer-by-layer rescanning procedure reduced them by about 55%. Another proposed way to reduce the value of surface stresses was to preheat the base plate. This action made it possible to reduce surface stresses by 40%.

A comparative stress measurement methodology was employed by Mercelis and Kruth (2006) for SS316L stainless steel parts made by SLS and SLM methods. Two methods were used: the crack compliance method and the method using the X-ray diffraction phenomenon. The experimental methods verified model calculations. The experimental results in the paper confirm previous reports of tensile stresses just below the top surface, followed by a significant zone of compressive stresses. The authors report that the factors determining the magnitude of the stresses are the height of the element and the height and stiffness of the base plate. The strategy of the printing process will also be a factor affecting the stress values: the stresses will be higher for the direction perpendicular to the scanning direction, but lower for the direction in line with the scanning direction. If the island strategy is used, the value of the maximum stress will be lower while the stresses for mutually perpendicular directions will not show significant differences. Two strategies have been suggested to reduce residual stresses resulting from the printing process: heat treatment using a laser source and heating the base plate.

The microstructure and residual stresses for the Inconel 718 superalloy were also studied for the rapid laser-fed (LRFed) process by Liu et al. (2011). The microstructure of the tested printouts showed the appearance of columnar dendrites growing epitaxially along the direction of the print. After annealing at high temperature, the geometry of the dendrites changed – equiaxed grains were formed. Stresses were measured using the Vickers microindentation method. The authors concluded that the residual thermal stresses are distributed unevenly (they are high in the overlapping areas of two adjacent transitions and low in the inner areas of a single transition), and thus play a significant effect on recrystallization during solution annealing.

A very detailed analysis of the microstructure is reported by Amato et al. (2012), where the effective phase composition and formation of precipitates were analyzed for prints produced from Inconel 718 alloy by selective melting of powder in an argon or nitrogen atmosphere from pre-melted powder. First, for printed parts (here in the form of cylinders) not subjected to any additional treatment, the occurrence of columnar grains and arrays of γ′ phase (Ni₃Nb, body-centered tetragonal) with strong texture on the direction was observed. After application of the HIP process, the texture of the coagulation phase sharpened while the material partially recrystallized, adopting the FCC structure. After annealing, half recrystallization and the appearance of spheroidal γ′ phase precipitates occurred. The appearance of δ phase precipitates both before and after recrystallization was also observed. The hardness measurement showed that the hardest material was the one after the HIP process, followed by the annealed one, while the as-built material had the lowest hardness.

From an economic and environmental point of view, one of the most important aspects during the manufacture of components by incremental methods is the recyclability of the powder used in such manufacturing procedures. Ardila et al. (2014) addressed this issue with regard to powder made of Inconel 718 alloy, finding that the state of the powder does not undergo significant changes during reuse for the manufacture of components by incremental methods. The authors report that most of the particles remained spherical and, moreover, there was no significant increase in the defective morphology of the powders. In addition, the particle size distribution remained constant even after several manufacturing processes, except for a small number of particles aggregated to sizes of 50 to 100 microns. One very interesting observation was that although the prints were made in a chamber where oxygen (2000 ppm) was present, the composition of the powder did not change. As a result, even after 14 production iterations, the samples obtained from the printing process did not differ from each other in either the metallurgical or the mechanical context.

Researchers involved in the development of printing procedures are aware of the affinity of modern 3D manufacturing methods with welding processes, since both methods use an increase in process temperature to achieve the desired result, i.e., the formation of a weld/join between two parts or the formation of a part with any level of shape complexity. Significant temperature gradients cause significant changes in microstructure properties. As early as 1996, Huang et al. (1996) addressed the effect of homogenization temperature on microcracks in heat-affected zones (HAZ) for Inconel 718 superalloy after electron beam welding. It was observed that the tendency of microcracks in the heat-affected zone of the electron-beam-welded Inconel 718 casting (determined by total crack length – TCL measurements) first increases and then decreases as the homogenization temperature increases from 1037°C to 1163°C. The minimum TCL value was observed in the sample homogenized at 1066°C. Moreover, it was observed that the volume fraction of the δ-Ni₃Nb and Laves phase decreases with increasing homogenization temperature, but the volume fraction of MC carbide remains constant over the entire homogenization temperature range.

Radhakrishna and Prasad Rao (1997) sought to identify the morphology and composition of the Laves phase in areas near welds in Inconel 718 alloy. They found that the deleterious Laves phase tended to appear in interdendritic areas after the weld was made. This occurs as a result of microsegregation of alloying elements during weld solidification. The factor with the most significant effect on the morphology and composition of this phase was the cooling rate of the weld. High cooling rates proved to be much more favorable than low rates.

Blackwell (2005) compared microstructural and mechanical properties for Inconel 718 alloy prints manufactured by the LENS method in the as-built condition and after HIP treatment. The main objective of this study was to examine the effect of the HIP process on the properties of printed parts. The process was found to lead to a decrease in anisotropy in the printed element, but generated a significant increase in grains in the substrate.

Ghorbanpour et al. (2022) addressed the effects of printing parameters and heat treatment for functionally graded additionally manufactured elements obtained from the nickel superalloy under discussion by laser powder sintering on the microstructure and indirectly on fatigue crack growth. Laser power, scanning speed, unit layer thickness, laser inter-path distance, and beam distribution function were changed during the printing of the test specimens. In addition, the prints were carried out using different orientations of the print direction with respect to the geometry of the samples. After printing, the samples were subjected to various post-process thermal and thermal treatments associated with the HIP procedure. The properties of the samples were compared with analogous forged and printed parts, but for which no heat treatment was performed. It was observed that the tested samples differed significantly in terms of porosity, grain size, grain shape, content of strengthening and deleterious phases. To study the fatigue properties, the samples were subjected to three-point bending tests. The experiments showed that heat treatment could increase the damage tolerance of L-PBF Inconel 718 parts to the level of forged material. Interestingly, the effect of crack closure-induced roughness was found to be a function of the orientation of the structure, especially in the low stress-ratio regime. The study found that by using an appropriate post-treatment procedure, fatigue behavior can be improved to a level even higher than that of forged material. The results of the current study can be used as a benchmark for determining the most effective heat treatment, given the applications of functionally graded materials. In addition, Ghorbanpour et al. stated that although the results of their study are comprehensive with regard to fatigue crack growth at room temperature, further research is needed at higher temperatures.

Hönnige et al. (2021) studied the influence of printing parameters in the L-PBF process, such as temperature cycling, thickness of the layer applied at one time and post-process heat treatment. The main focus of this work was to study the effect of the choice of printing parameters on microstructure and residual stresses. As in many similar studies, the formation of dendritic grains in the microstructure of prints and the existence of Laves phases in interdendritic areas were observed. The neutralization of microstructural anisotropy was achieved by applying annealing thermal treatment while the Laves phase was thus partially dissolved in the basic γ-matrix. A change in morphology from coarse-grained to discrete particles was achieved by thermal annealing processes. After the annealing thermal treatment, the samples had lower compressive residual stresses in the mean and maximum values, which should be related to the evolution of the microstructure during the annealing thermal treatment. The microhardness values of L-PBF Inconel 718 parts are comparable with literature data.

Previous works has also investigated changes in microstructural features, phase formation and three-dimensional variation of the state of stress resulting as an effect of the implementation of printing with the use of the WAAM method, which allows the manufacture of parts of larger sizes, but characterized by a lower complexity of geometry than in the case of laser methods. One such work, describing in great detail the changes in microstructure and mechanical properties, is Hönnige et al. (2021). Diffraction measurements were used to determine the three-dimensional stress state under different manufacturing conditions. Inter-pass rolling was found to have a significant effect on reducing the values of deformation of the resulting part. It was shown that near the printing platform the values of stresses had a tensile character, while on the surface of the printed part the stresses assumed a compressive character. Another positive effect of using the inter-pass rolling procedure during printing was the alleviation of the disordered Laves phase, whose high concentration of niobium usually prevents it from being re-solved, and a reduction in grain size and crystallographic texture. Another advantage of inter-wall rolling, Hönnige et al. report, lies in the method’s potential to reduce the porosity of post-printed parts.

For many years, researchers have struggled with the problem of reliably measuring the absolute values of residual stresses of components with complex shapes. Unfortunately, this is often a methodological problem, and in the case of printing, the surface roughness also has a significant impact on the results of measurements by diffraction methods. Since diffraction methods are the only non-destructive methods that make it possible to obtain information about the absolute value of stresses, the present study strives to identify their values on samples produced in DMLM technology.

The main motivation behind the diffraction experiments we carried out on components made from Inconel 718 alloy using incremental technologies was to identify the level of stress occurring in the samples as a result of the additive manufacturing process.

In general, components produced using additive methods are characterized by surface tensile stresses, high porosity and significant roughness of the final component. All these features are undesirable phenomena and therefore require control and levelling. Diffraction methods are an excellent tool for obtaining information on the absolute values of residual stresses on the surface of final printed parts using the classic, standardized sin²ψ method. Research potential can also be seen in the application of methods such as the small angle X-ray scattering (SAXS) method for determining the porosity of printed parts and the reflectometry method, which in turn makes it possible to define the level of roughness.

This paper presents the results of diffraction measurements of the phase composition using the Rietveld methodology and the results of stress measurements using sin²ψ.

Measurement uncertainty is a non-negligible aspect of any experimental work. In the series of diffraction experiments reported herein, carried out to identify the level of stress in components made using incremental methods, relatively high levels of measurement uncertainty were observed, averaging around 137 MPa. Such a high level of measurement uncertainty probably stems from the complex nature of the material itself, which is a non-trivial material from the perspective of diffraction measurement: during diffraction reflection acquisition, overlapping of reflections from planes of different phases were observed. In addition, each of the specimens was characterized by significant roughness, which had a substantial effect on the experimentally undesired scattering of incident radiation. Unfortunately, despite using optimized measurement parameters and half-automatic measurement analyses that allowed for the identification stress levels in finite time for such a large number of unit measurements, effective separation of these diffraction reflections was unattainable.

To mitigate measurement errors, various methodological and technological solutions used in diffraction methodology were applied, such as the OMEGA measurement mode and focusing lenses on the incident beam path, which allow the full radiating power of the X-ray tube to be used and the influence of sample roughness to be negated. Moreover, the lenses facilitated the generation of a parallel beam, which substantially lowered the influence of the surface roughness. In addition, a fast semiconductor detector was used.

Despite the measurement uncertainty, however, consistent trends observed in the measurement results lend credence to the reliability of the experimental results obtained over the entire measurement cycle.

Stress measurements were performed on 10 specimens made using incremental methods from Inconel 718 alloy, across 92 different points on their surface. A total of 178 stress measurements were made using the sin²ψ method, for different measurement directions defined with respect to the geometry of the specimens. Preliminarily to the main measurement cycle, a phase analysis of the material was performed. It was found that detaching a specimen from the print platform could not be performed without changing the stress state, and the stress measurement parameters were optimized.

Summarizing the results of the obtained measurements, the following conclusions can be drawn:

The Inconel 718 alloy exhibits a multiphase structure predominantly characterized by a cubic FCC phase. However, it also contains precipitations in the form of niobium-nickel compounds with varying crystallographic structures.