A Sensory Material Approach for Reducing Variability in Additively Manufactured Metal Parts

Abstract

Despite the recent growth in interest for metal additive manufacturing (AM) in the biomedical and aerospace industries, variability in the performance, composition, and microstructure of AM parts remains a major impediment to its widespread adoption. The underlying physical mechanisms, which cause variability, as well as the scale and nature of variability are not well understood, and current methods are ineffective at capturing these details. Here, a Nickel-Titanium alloy is used as a sensory material in order to quantitatively, and rather rapidly, observe compositional and/or microstructural variability in selective laser melting manufactured parts; thereby providing a means to evaluate the role of process parameters on the variability. We perform detailed microstructural investigations using transmission electron microscopy at various locations to reveal the origins of microstructural variability in this sensory material. This approach helped reveal how reducing the distance between adjacent laser scans below a critical value greatly reduces both the in-sample and sample-to-sample variability. Microstructural investigations revealed that when the laser scan distance is wide, there is an inhomogeneity in subgrain size, precipitate distribution, and dislocation density in the microstructure, responsible for the observed variability. These results provide an important first step towards understanding the nature of variability in additively manufactured parts.

Figure 1

Schematic illustrations of the Selective Laser Melting (SLM) process, (a) showing the consolidation of a fully solid part in a layer-by-layer manner, and the consolidation of powder at the melt pool. The schematic in (b) shows the pattern of the laser as it rasters in a hatch pattern. The long axis of the hatches is rotated by 90° with every new layer.

Figure 2

SEM images of the top surfaces of 3-D printed NiTi cubes using a distance between laser scans of (a) 35 μm and (b) 120 µm. The edges of the final laser track are highlighted on the right side of the image. In the center of the right image, two adjacent tracks are highlighted with their right edges estimated based on the width of the final track, showing the overlap region between them.

Figure 3

Results of DSC experiments for all NiTi samples 3-D printed using laser scan distances of 35 μm and 120 μm. The raw DSC curves are shown in (a) for multiple samples (10 repetitions with the same processing parameters). The transformation peaks are highlighted in orange, and the average width is indicated next to each peak. The graph in (b) shows the variability in measured transformation temperatures. The four columns correspond to the start (M_s) and finish (M_f) of the martensitic transformation during cooling, and the start (A_s) and finish (A_f) of the reverse transformation to austenite during heating. The height of the bars indicates the range of measurements, and the standard deviation is indicated beside each.

Figure 4

DSC curves for NiTi cubes 3-D printed with various process parameter values for laser power, speed, and scan distance. The curves are grouped by peak width.

Figure 5

Optical microscopy images of the cross sections cut in the X-Y plane from the NiTi samples 3-D printed using laser scan distances of (a) 35 μm and the (b) 120 μm. The color contrast indicates differences in grain orientation. The small black spots in (b) are not porosity, but artifacts from the etching process.

Figure 6

SEM images of the X-Z surfaces etched to reveal the cross sections of the NiTi samples 3-D printed using laser scan distances of (a) 35 μm and the (b) 120 μm. The visible portions of the laser tracks under investigation have been highlighted in green. The location of Focused Ion Beam (FIB)-prepared TEM specimens is highlighted in red.

Figure 7

TEM images of the AM fabricated NiTi samples showing subgrain structure and precipitates decorating the subgrains in the (a) 35 µm hatch distance sample taken from the center of the laser track, the 120 µm hatch distance sample taken from (b) the center of the laser track, and (c) the edge of the laser track.

Figure 8

TEM images of the AM fabricated NiTi samples: (a) 35 µm hatch distance sample taken from the center of the laser track, the 120 µm hatch distance sample taken from (b) the center of the laser track, and (c) the edge of the laser track. Specimens have been tilted to show maximum contrast of the dislocations.

Figure 9

TEM image of the AM fabricated NiTi sample using laser scan distance of 35 μm showing small <5 nm precipitates along with two larger Ti₂Ni precipitates.