The nickel-based superalloy, Haynes 230 (H230), is widely used in high-temperature applications, e.g., heat exchangers, because of its excellent high-temperature mechanical properties and corrosion resistance. As of today, H230 is not yet in common use for 3D printing, i.e., metal additive manufacturing (AM), primarily because of its hot cracking tendency under fast solidification. The ability to additively fabricate components in H230 attracts many applications that require the additional advantages leveraged by adopting AM, e.g., higher design complexity and faster prototyping. In this study, we fabricated nearly fully dense H230 in a laser powder bed fusion (L-PBF) process through parameter optimization. The efforts revealed the optimal process space which could guide future fabrication of H230 in various metal powder bed fusion processes. The metallurgical analysis identified the cracking problem, which was resolved by increasing the pre-heat temperature from 80 °C to 200 °C. A finite element simulation suggested that the pre-heat temperature has limited impacts on the maximum stress experienced by each location during solidification. Additionally, the crack morphology and the microstructural features imply that solidification and liquation cracking are the more probable mechanisms. Both the room temperature tensile test and the creep tests under two conditions, (a) 760 °C and 100 MPa and (b) 816 °C and 121 MPa, confirmed that the AM H230 has properties comparable to its wrought counterpart. The fractography showed that the heat treatment (anneal at 1200 °C for 2 h, followed by water quench) balances the strength and the ductility, while the printing defects did not appreciably accelerate part failure.

Additive manufacturing (AM) is gaining attention because of its ability to design complex geometries. Direct energy deposition (DED), one of the AM processes, is widely used nowadays for its high deposition rate. When using DED process in manufacturing or repairing, it is important to know the melt pool dimensions as a function of processing parameters to obtain high deposition rate and avoid defects such as lack of fusion. In this study, we used the random forest (RF) algorithm to the predict melt pool dimensions and compared the results against existing physics based lumped model by Doumanidis et al. [1]. The results show that RF model works well to predict the DED melt pool dimensions, where energy density and material volume deposited govern the dimensions. Further, we tested the ubiquitous semi-ellipsoidal shape assumption for DED cross section against the circular shape, and found semi-ellipsoidal shape to be fair when deposition process is stable and free of defects. Overall, this study highlights the applicability of machine learning algorithms for small AM datasets. Keywords: RegressionAnalysis,DirectedEnergyDeposition,Ti—6Al—4V,MeltPoolShape

Hydride-dehydride (HDH) Ti-6Al-4V powders with non-spherical particle morphology are typically not used in laser-beam powder bed fusion (LB-PBF). Here, HDH powders with two size distributions of 50—120 µm (fine) and 75—175 µm (coarse) are compared for flowability, packing density, and resultant density of the LB-PBF manufactured parts. It is shown that a suitable laser power-velocity-hatch spacing combination can result in part production with a relative density of > 99.5\% in LB-PBF of HDH Ti-6Al-4V powder. Size, morphology, and spatial distribution of pores are analyzed in 2D. The boundaries of the lack-of-fusion and keyhole porosity formation regimes are assessed and results showed parts with a relative density of > 99.5\% could be LPBF processed at a build rate of 1.5—2 times of the nominal production rates in LPBF machines. The synchrotron x-ray high-speed imaging reveals the laser-powder interaction and potential porosity formation mechanism associated with HDH powder. It is found that lower powder packing density of coarse powder and high keyhole fluctuation result in higher fractions of porosity within builds during the LB-PBF process.

Hydride-dehydride (HDH) Ti-6Al-4V alloy with particle size distribution of 50—120 µm is laser powder bed fusion (L-PBF) processed using optimum processing parameters and a near-fully dense structure with a density of 99.9 \% is achieved. Microstructural observations and phase analyses indicate formation of columnar βgrains with acicular α/α′phases in as-built condition. The roughness of the as-fabricated samples is significant with an average roughness of Ra = 15.71 $\pm$3.96 µm and a root mean square roughness of Rrms = 108.4 $\pm$24.9 µm, however, both values are reduced to Ra = 0.19 $\pm$0.04 µm and Rrms = 4.9 $\pm$0.6 µm after mechanical grinding. Mechanical tests are carried out on as-fabricated specimens followed by stress relief treatment. All samples are tested to failure in fatigue, under fully-reversed tension-compression conditions of R = −1. The as-built samples failed from the surface with crack initiation mainly at micro-notches, whereas after mechanically grinding, crack initiation changed to subsurface defects such as pores. Minimizing surface roughness by mechanically grinding eliminates surface micro-notches which improves fatigue strength in the high cycle fatigue region. Fatigue notch factor calculations showed that the effect of surface roughness was significantly lower when HDH powder is used compared to standard spherical powder. X-ray diffraction analysis revealed an in-plane compressive stress, micro-strain and grain refinement on the surface of the mechanically ground samples. Fractography observations (macroscale) revealed a fully brittle fracture in the first stage of crack growth with a transition to a dominantly ductile fracture in the third stage of crack growth. On the other hand, at the micro scale, even the brittle fracture regions showed evidence of ductile fracture within the α′martensite laths.

The efficacy of an elasto-viscoplastic fast Fourier transform (EVPFFT) code was assessed based on blind predictions of micromechanical fields in a sample of Inconel 625 produced with additive manufacturing (AM) and experimentally characterized with high-energy X-ray diffraction microscopy during an in situ tensile test. The blind predictions were made in the context of Challenge 4 in the AFRL AM Modeling Challenge Series, which required predictions of grain-averaged elastic strain tensors for 28 unique target (Challenge) grains at six target stress states given a 3D microstructural image, initial elastic strains of Challenge grains, and macroscopic stress—strain response. Among all submissions, the EVPFFT-based submission presented in this work achieved the lowest total error in comparison with experimental results and received the award for Top Performer. A post-Challenge investigation by the authors revealed that predictions could be further improved, by over 25\% compared to the Challenge-submission model, through several model modifications that required no additional information beyond what was initially provided for the Challenge. These modifications included a material parameter optimization scheme to improve model bias and the incorporation of the initial strain field through both superposition and eigenstrain methods. For the first time with respect to EVPFFT modeling, an ellipsoidal-grain-shape Eshelby approximation was tested and shown to improve predictive capability compared to previously used spherical-grain-shape assumptions. Lessons learned for predicting full-field micromechanical response using the EVPFFT modeling method are discussed.

Although the deformation behavior of high-entropy alloys (HEAs) has been extensively studied at the macroscale, many important properties have yet to be explored for these alloys at the microscale, thus hampering accurate prediction of damage and failure processes. Synchrotron high-energy diffraction microscopy (HEDM) and fast-Fourier transform-based crystal plasticity modeling was conducted to investigate the three-dimensional (3D) grain-resolved micromechanical response for approximately 1,900 constituent grains within a single-phase FCC HEA up to 1% applied strain. The evolution of grain-resolved elastic strains, lattice reorientations, and maximum resolved shear stresses (mRSS) were evaluated to quantify elastic, yield, and fully plastic behavior. Overall, the initial critical resolved shear stress (CRSS), determined via in situ HEDM and companion modeling, was found to be > 20% higher than estimated using the classical polycrystalline Taylor factor (M = 3.06). However, a descriptive parameter based on the average grain-resolved Taylor factor (M¯) was found to show excellent agreement with plastic yielding behavior observed within HEDM datasets. Noticeable deviations in HEDM lattice reorientations compared to both EVP-FFT simulations and classical predictions for FCC polycrystals were discovered, highlighting the complexity in correlating local lattice reorientations, Taylor, and Schmid factors with plastic response for this material at the grain-scale. Therefore, it is anticipated that the overall trends and parameter identification of 3D grain-resolved properties in this study can serve as an important foundation for continued mesoscale investigation on both well-established and newly developed Cantor-like HEAs.

The simulation and modeling of part-level distortion and residual stress in diverse metal additive manufacturing (AM) geometries has great potential to enable the rapid adoption of this technology in engineering design. Moreover, the use of additive manufacturing component libraries (CLs) offer a computationally efficient means of quantifying these part-level defects resultant from AM processing. We report on how the individual simulations of simple shapes, potential entries in a CL, can be superimposed to provide an indication of distortion and residual stresses in complex geometries. Laser powder bed fusion AM was used to construct test geometries of varied shapes and their combinations in the form of CLs in an effort to characterize location-dependent and feature-dependent distortion distributions. Blue light scanning was used to experimentally measure 3D distortions in order to investigate the interaction between the component shapes and local boundary conditions. Overall, part-level distortions were highly dependent on test component geometry, local boundary conditions, and shape combination. Commercial finite element software was used to verify experimental trends and to make predictions of distortion. The use of CLs resulted in over 20 times savings in computational cost while reproducing overall trends in distortion for test geometry assemblies. Therefore, it is anticipated that the use of CLs for L-PBF AM geometries has demonstrated potential to facilitate efficient simulations of full component AM assemblies, thereby reducing the need for costly trial-and-error-type experimental analysis.

Using a large-scale, experimentally captured 3D microstructure data set, we implement the generative adversarial network (GAN) framework to learn and generate 3D microstructures of solid oxide fuel cell electrodes.The generated microstructures are visually, statistically, and topologically realistic, with distributions of microstructural parameters, including volume fraction, particle size, surface area, tortuosity, and triple-phase boundary density, being highly similar to those of the original microstructure.These results are compared and contrasted with those from an established, grain-based generation algorithm (DREAM.3D). Importantly, simulations of electrochemical performance, using a locally resolved finite element model, demonstrate that the GAN-generated microstructures closely match the performance distribution of the original, while DREAM.3D leads to significant differences. The ability of the generative machine learning model to recreate microstructures with high fidelity suggests that the essence of complex microstructures may be captured and represented in a compact and manipulatable form.