*2.2. Powder Material*

In this work, four kinds of powders were prepared in the fabrication of the 90W-7Ni-3Fe powder. Figure 2 illustrates the powder size distribution of the different powders used in this work: pure tungsten powder (D10 = 4.425 μm, D50 = 7.211 μm, and D90 = 11.729 μm), Fe–Ni alloy powder (D10 = 11.420 μm, D50 = 16.478 μm, and D90 = 23.620 μm), nickel powder (D10 = 10.794 μm, D50 = 16.492 μm, and D90 = 24.961 μm), and sub-micro nickel powder (D10 = 0.932 μm, D50 = 1.968 μm, and D90 = 4.002 μm). Considering that the powders had different densities, the average particle size of the tungsten powder was smaller than that of the nickel powder and Fe–Ni alloy, so a relatively uniform powder layer distribution could be obtained. Submicron nickel powder was used to adjust the nickel content in the 90W-7Ni-3Fe alloy. Powders were mixed in a general powder mixer (AM300S-H) under an argon atmosphere was used for 0.5 h to obtain 90W-7Ni-3Fe powder. Figure 3 depicts the morphology of the mixed 90W-7Ni-3Fe powder. It can be observed that the shape of the powder particles was not changed and kept good sphericity. Thus, the 90W-7Ni-3Fe powder had good fluidity, and a uniform and suitable a powder layer could be obtained. Furthermore, fine powder particles adhered to larger particles without falling off (Figure 3b), and the adhesion contributed to the powder's spread and uniformity.

**Figure 2.** The distribution of Four kinds of powder sizes. (**a**) Pure tungsten powder; (**b**) nickel powder; (**c**) Fe-Ni alloy powder; (**d**) submicron nickel powder.

**Figure 3.** Morphology of the 90W-7Ni-3Fe powder. (**a**) Low magnification; (**b**) high magnification.

#### *2.3. Process Parameters and Conditions*

The process parameters used in this work are listed in Table 1. The parameters included four levels of laser power, scan speed, and hatching distance. The powder layer thickness was kept at 30 μm. Various combinations of process parameters were obtained according to the Taguchi experimental design method (Table 2). The scan strategy used in this experiment and building direction is illustrated in Figure 4a. The scan strategy was rotated by 67◦ between adjacent layers in order to reduce residual stress during the process of SLM. The 90W-7Ni-3Fe alloy samples with dimensions of 10 mm × 10 mm × 5 mm were prepared (Figure 4b). Based on consideration of the combined effect of the process parameters, the volumetric energy density (*VED*) is as follows:

$$VED = \frac{LP}{SS \times HD \times LT'} \tag{1}$$

where *LP* is laser power, *SS* is scan speed, *HD* is hatching distance, and *LT* is layer thickness.

**Table 1.** Process parameters used in this work.


**Table 2.** Process parameter combinations and the value of the volumetric energy density (VED).


**Figure 4.** (**a**) Scan strategy used in this experiment; (**b**) 90W-7Ni-3Fe samples fabricated by the designed process parameters.

#### *2.4. Characterization and Test*

All the samples were removed from the substrate by wire electrical discharge machining (WEDM). Then, the samples were cleaned with an ultrasonic cleaning machine and dried. After that, the dried samples were ground gradually with sandpaper of di fferent grits, 280#, 400#, 600#, 800#, 1000#, 1200#, 1500# and 2000#. Subsequently, the samples were polished using a diamond suspension of 2.5 and 0.5.

In this work, the theoretical density ρ*T* of 90W-7Ni-3Fe can be calculated as [26]

$$\frac{100}{\rho\_T} = \sum \frac{W\_i}{\rho\_i} \,\,\,\,\tag{2}$$

where *Wi* and ρ*i* are the mass fraction and the theoretical density of the *i*th alloy element, respectively. The theoretical density of W, Ni, and Fe are 19.3 g/cm3, 8.9 g/cm3, and 7.9 g/cm3, respectively [27].

According to the Archimedes method, the actual density can be calculated as follows:

$$
\rho\_{\rm a} = \frac{M\_0}{M\_2 - M\_1} \times \rho\_{0\prime} \tag{3}
$$

where *M*0 is the mass of sample in the air, *M*1 and *M*2 are the indications of the balance before and after the sample is placed in the beaker containing water, and ρ0 is the density of water.

Relative density ρ*RD* is calculated as follows:

$$
\rho\_{\rm RD} = \frac{\rho\_{\rm a}}{\rho\_{\rm T}} \times 100\%\_{\rm r} \tag{4}
$$

The transverse and vertical morphology of the samples (Figure 5a) were observed under an optical microscope (OM, Nikon MA 200) in order to analyze the defects' characteristics and distribution. The transverse section of the sample was etched with a mixture solution of 10 g KOH: 10 g K3 [Fe(CN)6]:100 mL H2O to ensure a clear microstructure. The microstructure was also characterized by an optical microscope (OM, Nikon MA 200). For a better analysis of the microstructure, a scanning electron microscope (SEM, S-4800, Hitachi, Tokyo, Japan) was used, and an energy-dispersive spectrometer (EDS) was adopted for analysis of the element distribution. The phase composition of the sample was tested by an X-ray di ffractometer (XRD, Advanced D8, Bruker, Billerica, MA, USA) with a Cu K α radiation at 40 KV and 40 mA in a 2θ range of 30◦–90◦ by using a step size of 0.02◦. Tensile property tests were performed at room temperature using an Instron550R at a constant tensile rate of 1 mm/min. The tensile specimen is illustrated in Figure 5b. The microhardness of the transverse morphology was measured by using a digital microhardness measurement system (MH5) with a load of 0.3 kg and a dwell time of 15 s. Five points were taken at an interval of 0.5 mm along the building direction for all selected samples fabricated under di fferent VEDs.

**Figure 5.** (**a**) Illustration diagram of sample direction; (**b**) dimensions of tensile sample.

## **3. Results and Discussion**
