Currently, knowledge on the performance of many additively manufactured (AM) materials is lacking (1), (2). Selective laser melting, or selective laser sintering, is an additive manufacturing (AM) technique that produces final parts from process parameters that describe the process and instructions that describe the desired geometry (2). These process parameters and instructions describe how a part was produced and hopefully provide insight into the mechanical properties of a final part. However, for a given process, the final microstructure and properties are not known (1). Part of the issue is that for a constant process and instructions, the final properties are more inconsistent than traditional manufacturing methods (3). Much of this inconsistency has been attributed to structure, especially the mesostructural features porosity and surface roughness (4)–(6). However, the microstructure is also different from traditional manufacturing methods (2) and the exact variation in the microstructure of AM materials is not well understood. Other variations from traditional manufacturing process have also been identified. These variations include deviations from the nominal alloy composition (3), (5) and a higher concentration of oxides (1). AM also introduces residual stresses in final parts, which can drive crack growth (7). However, these residual stresses can be reduced with a heat treatment.
This project is concerned with a dataset received from Sandia National Labs (4), (5). The AM parts being considered are microtensile specimens manufactured by an unspecified process, with the nominal composition of the 17-4PH stainless steel alloy, and heat treated by the H900 process. The cross-section of the specimens is approximately 1mm by 4mm not considering the surface roughness or dimensional inaccuracy of the AM process. The below figure shows the real microtensile specimens (5).
The specimens were created by the same process, but have very different mechanical properties. Digital image correlation was used to determine strain during tensile testing. Load was measured, but the stress was calculated by the minimum cross-sectional area. The yield strength can vary from 778-1014 MPa. And the strain to failure can vary from 4.4 – 14.0%. This is shown in the below figure below. The modulus can vary from 164-201 GPa, but is not shown in the figure below.
The dataset also includes structural data in the form of micro computed tomography (µCT) images. These images can scan through a material and reveal the 3D mesostructure. These 3D images have cubic voxels, where one side of the voxel is somewhere between 7-10µm, depending on the specimen. The raw 3D images can be sliced and viewed as shown in the following figure.
From the 3D images it is possible to identify the pores in the specimens using segmentation techniques. Sandia performed a segmentation by thresholding each of the images by a handpicked value. This allowed the pores to be identified as shown in the following image.
Currently, knowledge on the performance of many additively manufactured (AM) materials is lacking 1, 2. One way to address this issue is to develop models that link the process, structure, and properties of AM materials. Developing models is a challenge because the structure of AM parts varies on two separate length-scales from traditional manufacturing methods. The problem that will be addressed directly is that the mechanical properties of stainless steel micro-tensile specimens are very different, although they were produced by an identical AM process.
This project will study the structure on the meso length scale, where the surface roughness, pores, and matrix can be distinguished as separate phases. However, this method does not account for variation in the microstructure of the matrix. Therefore, we also propose to study the nano length scale properties of the matrix for different specimens. The fundamental questions being addressed by this project are:
1. What is the mesostructural variation between AM microtensile specimens?
2. How does the mesostructural variation impact performance/macroscale properties of the specimens?
(1) Y. Huang, M. C. Leu, J. Mazumder, and A. Donmez, “Additive manufacturing: current state, future potential, gaps and needs, and recommendations,” J. Manuf. Sci. Eng., vol. 137, no. 1, p. 14001, 2015.
(2) W. E. Frazier, “Metal Additive Manufacturing: A Review,” J. Mater. Eng. Perform., vol. 23, no. 6, pp. 1917–1928, Jun. 2014.
(3) W. E. Luecke and J. A. Slotwinski, “Mechanical Properties of Austenitic Stainless Steel Made by Additive Manufacturing,” J. Res. Natl. Inst. Stand. Technol., vol. 119, p. 398, Oct. 2014.
(4) B. L. Boyce et al., “Extreme-Value Statistics Reveal Rare Failure-Critical Defects in Additive Manufacturing: Extreme-Value Statistics Reveal Rare Failure- Critical Defects…,” Adv. Eng. Mater., vol. 19, no. 8, p. 1700102, Aug. 2017.
(5) B. C. Salzbrenner et al., “High-throughput stochastic tensile performance of additively manufactured stainless steel,” J. Mater. Process. Technol., vol. 241, pp. 1–12, Mar. 2017.
(6) A. Bauereiß, T. Scharowsky, and C. Körner, “Defect generation and propagation mechanism during additive manufacturing by selective beam melting,” J. Mater. Process. Technol., vol. 214, no. 11, pp. 2522–2528, Nov. 2014.
(7) S. Leuders et al., “On the mechanical behaviour of titanium alloy TiAl6V4 manufactured by selective laser melting: Fatigue resistance and crack growth performance,” Int. J. Fatigue, vol. 48, pp. 300–307, Mar. 2013.