*3.2. Formation of the SLM Titanium and Copper-Coated Diamond/Copper Composites* 3.2.1. SLM Manufacturing of Titanium-Coated Diamond/Copper Composites

The rectangular contour samples were printed in the single-track experiment, and their morphology was characterized and correlated with the process parameters. Due to scanning a single-track layer-by-layer along the Z-axis to form each rectangular contour sample, the width of a single wall was the width of a single weld pool. Samples with different process parameters were observed via OM to estimate the range of process parameters in the cubic experiment. The SLM laser parameters selected for the single-track formation test of the titanium-coated diamond/copper combined materials were as follows. The laser power was improved from 140 to 160 and 180 W, and the scanning rate was improved from 200 to 300 and 400 mm/s at a fixed powder layer thickness of 0.025 mm. It was important to observe the continuity of the molten pool, as any discontinuity would produce defects, for example, porosity, delamination and surface roughness. Figure 3a,b exhibited that the superior process parameters of the rectangular contour samples with 3 and 5 vol.% titanium-coated diamond/copper composites comprised a laser power of 140 W and a scanning rate of 200 mm/s, that with 1 vol.% titanium-coated diamond/copper composites comprised a laser power of 180 W and a scanning rate of 200 mm/s. In the single-track experiment, when the layer thickness was below the median powder particle size (0.025 mm), a significant amount of powder particles larger than the layer thickness

would rub the previous layer in the melting region as the scraper moved, leading to failure of the print formation. Therefore, a layer thickness of 0.025 mm was optimal. *Micromachines* **2022**, *13*, x FOR PEER REVIEW 6 of 16

**Figure 2.** (**a**) The morphologies of the titanium-coated diamond particles and the corresponding EDS element mappings; (**b**) The thickness of the titanium-coating layer on the diamond particle surface and the corresponding EDS element mappings; (**c**) The morphologies of the copper-coated diamond particles and the corresponding EDS element mappings; (**d**) The thickness of the coppercoating layer on the diamond particle surface and the corresponding EDS element mappings; The XRD pattern of (**e**) the titanium-coated and (**f**) the copper-coated diamond particles. **Figure 2.** (**a**) The morphologies of the titanium-coated diamond particles and the corresponding EDS element mappings; (**b**) The thickness of the titanium-coating layer on the diamond particle surface and the corresponding EDS element mappings; (**c**) The morphologies of the copper-coated diamond particles and the corresponding EDS element mappings; (**d**) The thickness of the copper-coating layer on the diamond particle surface and the corresponding EDS element mappings; The XRD pattern of (**e**) the titanium-coated and (**f**) the copper-coated diamond particles.

their morphology was characterized and correlated with the process parameters. Due to scanning a single-track layer-by-layer along the Z-axis to form each rectangular contour sample, the width of a single wall was the width of a single weld pool. Samples with

*3.2. Formation of the SLM Titanium and Copper-Coated Diamond/Copper Composites*  3.2.1. SLM Manufacturing of Titanium-Coated Diamond/Copper Composites

**Figure 3.** (**a**) Rectangular contour samples of the 1, 3 and 5 vol.% titanium-coated diamond/copper composites; (**b**) the corresponding processing window of the laser power and scanning rate; (**c**) The SLM manufactured morphology: top and front view of the 1, 3 and 5 vol.% titanium-coated diamond/copper composite, respectively. **Figure 3.** (**a**) Rectangular contour samples of the 1, 3 and 5 vol.% titanium-coated diamond/copper composites; (**b**) the corresponding processing window of the laser power and scanning rate; (**c**) The SLM manufactured morphology: top and front view of the 1, 3 and 5 vol.% titanium-coated diamond/copper composite, respectively.

different process parameters were observed via OM to estimate the range of process parameters in the cubic experiment. The SLM laser parameters selected for the single-track formation test of the titanium-coated diamond/copper combined materials were as follows. The laser power was improved from 140 to 160 and 180 W, and the scanning rate was improved from 200 to 300 and 400 mm/s at a fixed powder layer thickness of 0.025 mm. It was important to observe the continuity of the molten pool, as any discontinuity would produce defects, for example, porosity, delamination and surface roughness. Figure 3a,b exhibited that the superior process parameters of the rectangular contour samples with 3 and 5 vol.% titanium-coated diamond/copper composites comprised a laser power of 140 W and a scanning rate of 200 mm/s, that with 1 vol.% titanium-coated diamond/copper composites comprised a laser power of 180 W and a scanning rate of 200 mm/s. In the single-track experiment, when the layer thickness was below the median powder particle size (0.025 mm), a significant amount of powder particles larger than the layer thickness would rub the previous layer in the melting region as the scraper moved, leading to failure

of the print formation. Therefore, a layer thickness of 0.025 mm was optimal.

Compared with the single-track experiment involving only three process parameters (layer thickness, scanning rate and laser power), the cubic experiment further considered the hatch distance and the scanning strategy. The hatch distance (*h*) was calculated as fol-Compared with the single-track experiment involving only three process parameters (layer thickness, scanning rate and laser power), the cubic experiment further considered the hatch distance and the scanning strategy. The hatch distance (*h*) was calculated as follows:

$$h = (1 - Hr) \times w \tag{2}$$

where *Hr* and *w* indicate the overlap rate and melt pool width, respectively. The higher the overlap rate, the higher the density and the lower the porosity. The highest densities and lowest porosities were obtained for each parameter group when the melt pool overlap rate reached 60%. The center of the next scan track overlapped the edge of the adjacent single-track melt pool and was partially remelted to achieve good metallurgical bonding between the pools and reduce the porosity. However, when the overlap was greater than 60%, the porosity increased slightly. If the laser energy density increased further upon a rise in the laser power or a decrease in the hatch distance, the porosity increased slightly when the density reached its peak value, that is, when the porosity was at its lowest. Existing research confirmed these phenomena [37–39].

lows:

The data of the single-track formation test of the 1, 3 and 5 vol.% titanium-coated diamond/copper combined materials could be obtained through the ImageJ software, the melt pool width was found to be approximately 250 µm, and the h was calculated as approximately 100 µm. Figure 3c shows the unpolished samples of the SLM-printed 1, 3 and 5 vol.% titanium-coated diamond/copper combined materials. The 1 and 3 vol.% composites had higher relative density than that of the 5 vol.% composite, the relative density of 1, 3 and 5 vol.% titanium-coated diamond/copper composites were 96%, 90% and 81%, respectively. A higher relative density is related to better specimen performance [40]. Hence, 1 and 3 vol.% coated diamond/copper composites featuring high relative density were chosen for the investigation below.

Additionally, with the emergence of a metal vapor above the molten pool, a recoil pressure was induced onto the surface of pool. Meanwhile with the quick development of a thermal gradient inside the liquid, a flow of molten metal was produced from the hot region to the coldest one, referred to as the Marangoni role. Both roles brought particle ejection, known as recoil-induced ejection, as described in Figure 4 [41,42]. Besides, a denudation area was built on the sides of the molten track, where we could sweep powders under the flow of the metal vapor and the gas shield. This kind of particle ejections was known as entrainment particle (Figure 4) [43]. Metal vapor-induced entrainment caused powderspattering behavior [44]. To be intuitive, micrometric particles changed the surface tension, molten pool rheology, and metal vapor density according to declaration [45–47]. Also, it was believed that the imbalance of the molten pool, hot spatter collision during flight and entrained particles gathered on the solidified area were among the major driving forces for big spatter production. A pronounced spattering situation was found while mixing 5 vol.% coated diamond with copper by comparing with a pure copper bed condition, thus determining the negative role of the coated diamond in the molten pool. The hot particles partially fused with the solid materials after falling back into the powder bed, cooled down after ejection [29]. *Micromachines* **2022**, *13*, x FOR PEER REVIEW 9 of 16

**Figure 4.** Spatter ejection phenomenon in SLM of the coated diamond/copper combined materials. **Figure 4.** Spatter ejection phenomenon in SLM of the coated diamond/copper combined materials.

The preparation of dense copper/diamond composite materials with a high TC re-

many researchers adopted electroless plating for coating the copper layer on diamond powders [49]. The superior SLM process parameters of the rectangular contour samples with 1 vol.% copper-coated diamond/copper composites comprised a laser power of 180 W and a scanning rate of 200 mm/s (Figure 5a), that with 3 vol.% copper-coated diamond/copper composites comprised a laser power of 160 W and a scanning rate of 100 mm/s (Figure 5c,d). According to the quality of the melted rails, the processing window was separated into four different zones (Figure 5d,e), that is, a weak sintering zone (52.8%; A), an unstable melting zone (22.2%; B), a continuous track zone (2.8%; C), and an overmelting zone (22.2%; D), in reference to their different line-energy densities (LEDs, J/m). The less continuous track area suggested a formation difficulty under the constrained process parameters. It was not easy to form a scanning track due to the insufficient LED within zone A. The powder failed to form a stable width track within zone B, and a large amount of unmelted powder adhered to the surface due to insufficient energy. In zone C, the melt flow in the molten pool stabilized and had a sufficient penetration depth into the previous layer, thus obtaining a relatively smooth trajectory when the input LED was 1600 J/m. When the LED was too high (>1700 J/m) in zone D, a track featuring a width of ap-

3.2.2. The Formation of SLM Copper-Coated Diamond/Copper Composites
