Simultaneous strength and ductility enhancements of high thermal conductive Ag7.5Cu alloy by selective laser melting

Abstract

High electrical and thermal conductive metals (HETCM) play a key role in smart electronics, green energy, modern communications and healthcare, however, typical HETCM (e.g., Ag, Au, Cu) usually have relatively low mechanical strength, hindering further applications. Selective laser melting (SLM) is a potentially transformative manufacturing technology that is expected to address the issue. Ag is the metal with the highest thermal conductivity, which induces microscale grain refinement, but also leads to high internal stresses by SLM. Here, we select Ag7.5Cu alloy as an example to demonstrate that multi-scale (micro/meso/macro) synergies can take advantage of high thermal conductivity and internal stresses to effectively strengthen Ag alloy. The mimicry of metal-hardened structures (e.g., large-angle boundary) is extended to the mesoscale by controlling the laser energy density and laser scanning strategy to manipulate the macroscale internal stress intensity and mesoscale internal stress direction, respectively, to form mesoscale large-angle "grains", resulting in multiple mutual perpendicular shear bands during fracture. The presented approach achieved a significant enhancement of yield strength (+ 145%) and ductility (+ 28%) without post-treatment. The results not only break the strength-ductility trade-off of conventional SLM alloys, but also demonstrate a multi-scale synergistic enhancement strategy that exploits high thermal conductivity and internal stresses.

Figure 1

High Q Ag7.5Cu alloy forms a variety of microstructures under different AM parameters (see Fig. S2, Supplementary Information for details): (a) An EBSD-IPF map of directional columnar grains. The major axis of the grain is directed to the laser heat source. (b) A SEM image of dendrites in unmelted spherical powders and molten regions. (c) An EBSD-IPF map of equiaxed grain.

Figure 2

Schematic diagram of multi-scale synergistic reinforcement mechanism.

Figure 3

(a) Scheil-Gulliver solidification curves of Ti8.5Cu, Ti7.5Cu, Ag8.5Cu and Ag7.5Cu. (b) Thermal conductivity of Ti8.5Cu and Ag7.5Cu during solidification. (c) Cooling curves of Ti8.5Cu and Ag7.5Cu during solidification. All the three figures (a–c) are obtained from JMatPro 7.0.0 software. (d) An EBSD-IPF map of ultrafine equiaxed grains. (e) Grain size distribution based on the EBSD-IPF map.

Figure 4

Effect of process parameters on morphology, volume density, distortion and defects (See Fig. S3, Supplementary Information for details). (a) Effect of process parameters on volume density. In general, LED is proportional to the volume density. According to LED (Eq. 1), LED is directly proportional to the laser power, and inversely proportional to scan speed, layer thickness and hatch distance. (b) XRD spectra of Ag alloy powders and representative sample sample-α, -β, -γ. (c) Effect of LED on total distortion by FEA (the figure is obtained from Simufact.welding 6.0.0 software). An OM image of excessive LED results in the formation of (d) cracks and (e) a SEM image of micro-pores.

Figure 5

(a) OM images of single-layer morphology. (b) 3D view of meso-scale columnar grains induced by three scan types of small domains: X-axis progressive scan (type A), Y-axis progressive scan (type B) and cross-scan (type C). (c) Thermal field simulation in forming process. FEA of peak temperature distribution (d) and total distortion (f) by bi-offset scanning strategy. (e) Morphology of sample-α observed by OM corresponding to the peak temperature distribution by FEA. (c), (d) and (f) are obtained from Simufact.welding 6.0.0 software.

Figure 6

Schematic diagram of multi-scale (macro: (a), meso: (b), micro: (c)) synergistic reinforcement mechanism. All figures are obtained from Simufact.welding 6.0.0 software.

Figure 7

SEM image of fracture surface morphology of (a) sample-α with synergy multi-scale effects, (b) sample-β with high volume density and (c) sample-γ with low volume density and (d) casting sample. (e) Stress–Strain diagrams of representative samples produced by metal AM and casting. (f) Summary of the elongation versus stress for silver alloy (examples of unspecified components are Ag7.5Cu alloys)^–.

Figure 8

FEA of the peak temperature distribution and equivalent plastic strain distribution of (a) meso-scale columnar “grains”, (b) equiaxed “grains” and (c) equiaxed “grains” with precipitates introduced. (d) localized region of excessive stress emerged by residual stress accumulation in single direction. (e) localized region of high stress at the meso-scale “grain” boundary. (f) precipitate domains. All figures are obtained from Simufact.welding 6.0.0 software.

Figure 9

(a) An OM image showing morphology of sample-β with the highest density in the ZY-plane: continuous laser processing causes the thermal field to gradually increase in the Z-axis direction. (b) A SEM image showing “columnar grain morphology” and the corresponding EDS maps. (c) A SEM image showing “density & homogeneous morphology” and the corresponding EDS maps. (d) An OM image showing morphology of sample-α with a fully meso-scale “columnar grain”. (e) A SEM image showing “columnar grain morphology” and (f) the corresponding EDS analysis.

Figure 10

(a) 3D view of three scan types of small domains: X-axis progressive scan (type A), Y-axis progressive scan (type B) and cross-scan (type C). (b) Molten pool morphology of sample-α observed by OM. (c) EBSD mappings and (d) EDS line-scan of different scan types of small domains.

Figure 11

Schematic of (a) process parameters, (b) single-wall and (c) cube tests. (d–h) Schematic of bi-offset scanning strategy. (d) Angular-offset scanning in the XY-plane: The directions of progressive scanning are perpendicular to each other between adjacent square regions with 1.0 mm sides. (e) Position-offset scanning on the next layer: After being offset by $\sqrt{2}$/2 mm in the 45° direction on the XY plane, the scanning is performed in the same manner as the previous layer. (f) Three scan types form smaller square domains with 0.5 mm sides after layer-by-layer deposition. (g) Schematic of three scan types of small domains: X-axis progressive scan (type A), Y-axis progressive scan (type B) and cross-scan (type C). (h) 3D view of meso-scale columnar “grains” induced by the scanning strategy.

Figure 12

(a) Specimen dimension for metal tensile test. (b) Schematic of the bi-offset scanning strategy. (c) a SEM image of fracture surface morphology. (d) an OM images of molten pool topography. (e) The relationship between the meso-scale “grain” and the directions of the tensile test. (f) an OM image of printed tensile test specimen.