Understanding the hot isostatic pressing effectiveness of laser powder bed fusion Ti-6Al-4V by in-situ X-ray imaging and diffraction experiments.

In the present study, in-situ observation of Hot Isostatic Pressure (HIP) procedure of laser powder bed fusion manufactured Ti-6Al-4V parts was performed to quantitatively estimate the densification rate of the material and the influence of the defect initial size and shape on such rate. The observations were performed in-situ using the Ultrafast Tomography Paris-Edinburgh Cell and the combination of fast phase-contrast synchrotron X-ray tomography and energy dispersive diffraction. With this strategy, we could quantify how the effectiveness of HIP depends on the characteristics of a defect. Smaller defects showed a higher densification rate, while the defect shape did not have significant effect on such rate.

Synthetic Data Generation for Automatic Segmentation of X-ray Computed Tomography Reconstructions of Complex Microstructures.

The greatest challenge when using deep convolutional neural networks (DCNNs) for automatic segmentation of microstructural X-ray computed tomography (XCT) data is the acquisition of sufficient and relevant data to train the working network. Traditionally, these have been attained by manually annotating a few slices for 2D DCNNs. However, complex multiphase microstructures would presumably be better segmented with 3D networks. However, manual segmentation labeling for 3D problems is prohibitive. In this work, we introduce a method for generating synthetic XCT data for a challenging six-phase Al-Si alloy composite reinforced with ceramic fibers and particles. Moreover, we propose certain data augmentations (brightness, contrast, noise, and blur), a special in-house designed deep convolutional neural network (Triple UNet), and a multi-view forwarding strategy to promote generalized learning from synthetic data and therefore achieve successful segmentations. We obtain an overall Dice score of 0.77. Lastly, we prove the detrimental effects of artifacts in the XCT data on achieving accurate segmentations when synthetic data are employed for training the DCNNs. The methods presented in this work are applicable to other materials and imaging techniques as well. Successful segmentation coupled with neural networks trained with synthetic data will accelerate scientific output.

Different Approaches for Manufacturing Ti-6Al-4V Alloy with Triply Periodic Minimal Surface Sheet-Based Structures by Electron Beam Melting.

Targeting biomedical applications, Triply Periodic Minimal Surface (TPMS) gyroid sheet-based structures were successfully manufactured for the first time by Electron Beam Melting in two different production Themes, i.e., inputting a zero (Wafer Theme) and a 200 µm (Melt Theme) wall thickness. Initial assumption was that in both cases, EBM manufacturing should yield the structures with similar mechanical properties as in a Wafer-mode, as wall thickness is determined by the minimal beam spot size of ca 200 µm. Their surface morphology, geometry, and mechanical properties were investigated by means of electron microscopy (SEM), X-ray Computed Tomography (XCT), and uniaxial tests (both compression and tension). Application of different manufacturing Themes resulted in specimens with different wall thicknesses while quasi-elastic gradients for different Themes was found to be of 1.5 GPa, similar to the elastic modulus of human cortical bone tissue. The specific energy absorption at 50% strain was also similar for the two types of structures. Finite element simulations were also conducted to qualitatively analyze the deformation process and the stress distribution under mechanical load. Simulations demonstrated that in the elastic regime wall, regions oriented parallel to the load are primarily affected by deformation. We could conclude that gyroids manufactured in Wafer and Melt Themes are equally effective in mimicking mechanical properties of the bones.

X-ray Computed Tomography Procedures to Quantitatively Characterize the Morphological Features of Triply Periodic Minimal Surface Structures.

Additively manufactured (AM) metallic sheet-based Triply Periodic Minimal Surface Structures (TPMSS) meet several requirements in both bio-medical and engineering fields: Tunable mechanical properties, low sensitivity to manufacturing defects, mechanical stability, and high energy absorption. However, they also present some challenges related to quality control, which can prevent their successful application. In fact, the optimization of the AM process is impossible without considering structural characteristics as manufacturing accuracy, internal defects, as well as surface topography and roughness. In this study, the quantitative non-destructive analysis of TPMSS manufactured from Ti-6Al-4V alloy by electron beam melting was performed by means of X-ray computed tomography (XCT). Several advanced image analysis workflows are presented to evaluate the effect of build orientation on wall thicknesses distribution, wall degradation, and surface roughness reduction due to the chemical etching of TPMSS. It is shown that the manufacturing accuracy differs for the structural elements printed parallel and orthogonal to the manufactured layers. Different strategies for chemical etching show different powder removal capabilities and both lead to the loss of material and hence the gradient of the wall thickness. This affects the mechanical performance under compression by reduction of the yield stress. The positive effect of the chemical etching is the reduction of the surface roughness, which can potentially improve the fatigue properties of the components. Finally, XCT was used to correlate the amount of retained powder with the pore size of the functionally graded TPMSS, which can further improve the manufacturing process.

Advances in Laser Additive Manufacturing of Ti-Nb Alloys: From Nanostructured Powders to Bulk Objects.

The additive manufacturing of low elastic modulus alloys that have a certain level of porosity for biomedical needs is a growing area of research. Here, we show the results of manufacturing of porous and dense samples by a laser powder bed fusion (LPBF) of Ti-Nb alloy, using two distinctive fusion strategies. The nanostructured Ti-Nb alloy powders were produced by mechanical alloying and have a nanostructured state with nanosized grains up to 90 nm. The manufactured porous samples have pronounced open porosity and advanced roughness, contrary to dense samples with a relatively smooth surface profile. The structure of both types of samples after LPBF is formed by uniaxial grains having micro- and nanosized features. The inner structure of the porous samples is comprised of an open interconnected system of pores. The volume fraction of isolated porosity is 2 vol. % and the total porosity is 20 vol. %. Cell viability was assessed in vitro for 3 and 7 days using the MG63 cell line. With longer culture periods, cells showed an increased cell density over the entire surface of a porous Ti-Nb sample. Both types of samples are not cytotoxic and could be used for further in vivo studies.

Pandora's Box-Influence of Contour Parameters on Roughness and Subsurface Residual Stresses in Laser Powder Bed Fusion of Ti-6Al-4V.

The contour scan strategies in laser powder bed fusion (LPBF) of Ti-6Al-4V were studied at the coupon level. These scan strategies determined the surface qualities and subsurface residual stresses. The correlations to these properties were identified for an optimization of the LPBF processing. The surface roughness and the residual stresses in build direction were linked: combining high laser power and high scan velocities with at least two contour lines substantially reduced the surface roughness, expressed by the arithmetic mean height, from values as high as 30 µm to 13 µm, while the residual stresses rose from ~340 to about 800 MPa. At this stress level, manufactured rocket fuel injector components evidenced macroscopic cracking. A scan strategy completing the contour region at 100 W and 1050 mm/s is recommended as a compromise between residual stresses (625 MPa) and surface quality (14.2 µm). The LPBF builds were monitored with an in-line twin-photodiode-based melt pool monitoring (MPM) system, which revealed a correlation between the intensity quotient I2/I1, the surface roughness, and the residual stresses. Thus, this MPM system can provide a predictive estimate of the surface quality of the samples and resulting residual stresses in the material generated during LPBF.