The experimental tests for the powder A were conducted by varying the plasma power in the range 9–17 kW and keeping constant the other parameters (gas flow rate and powder feeding rate), according to Table 1.
Table 1
Plasma process parameters for SS316L < 325 mesh (A powder)
Test | Power (kW) | Ar (slm | He (slm) | Ar carrier (slm) | Pressure Test (bar) | Feeding rate (g/min) |
A1 | 9 | 40 | 10 | 1 | 0,91 | 3 |
A2 | 12,4 | 40 | 10 | 1 | 0,91 | 3 |
A3 | 17 | 40 | 10 | 1 | 0,93 | 3 |
Figure 2 shows the SEM images of the samples A raw (a) and after plasma treatment (b-d). The raw powder presents irregular shaped particles, with a quite broad dimensional distribution centred at 44 micron. Alongside aggregates of higher diameters, up to over 60 micron, very small particles of few microns are present.
The SEM images show that already at 9 KW of power, a large number of spheres are visible; the spheroidization rises at higher power up to 12 KW. Nevertheless, at the maximum power (17 KW) it is evident the formation of huge quantity of nanoparticles, probably due to the large evaporation of the material.
This evidence can be explained considering that during the process the injected particles absorb energy by the plasma and if the energy is enough, they melt; once out of the plume, the particles start to cool down and due to the action of surface tension forces, their shape becomes spherical. If the absorbed energy exceed that required for melting, the particles evaporate; in this case the rapid quenching of the reactor determines their recondensation in the form of nanoparticles [17].
The XRD patterns of the samples before and after the plasma treatment are reported in Fig. 3.
The raw A powder mainly exhibits fully austenitic structures (face-centered cubic structure) with peaks at 2θ value of 43.6°, 50.8°, 74.8° corresponding to (111), (200), (200) reflections,respectively. The ferrite phase (body-centered cubic structure) is present in small quantity at 2θ values of 44.6°, 64.8° due to (110) and (220) reflections.
Also the processed powders show austenitic structure as main phase, alongside an appreciable amount of ferrite. The transformation between austenite and ferrite takes place during the cooling phase (rapid quenching) of the particles. The maximum conversion is evident in the powder treated at 12.4 kW (test A2 in Table 1).
Other diffraction peaks of minor intensity are also observed at 2θ of 30°, 35°, 38°, 57° and 63°, attributable to oxides formed during the annealing of the commercial steel or more generally due to the oxidation of the products (mainly Fe2O3, Fe3O4 or FeCr2O4 [20–21]).
Generally, the oxides formation, which can result during the cooling phase, is attributed to the presence of nanoparticles, which tend to oxidise [22]. Indeed, the powder processed at 17 kW, with the highest nanoparticles content, exhibits diffraction peaks corresponding to oxides structures. A further purification treatment of the powder by a sonication step at room temperature in liquid solvent, greatly improved the purity of the material, by removing the most part of the nanometric deposits, as shown in Fig. 4.
As further confirmation of the effectiveness of the purification step from nanoparticles, the XRD pattern of the A3 powder is free from the oxide signals (Fig. 5).
As matter of fact, the diffraction peaks at 30°, 35°,38°, 57° and 63° of 2θ, which occur in presence of Fe3O4 or FeCr2O4 oxides and observed in the XRD pattern of A3 powder before the purification, is greatly reduced after the treatment.
A second tests set was conducted on SS316L < 100 mesh (149 micron) powder, called B. This sample, according to the SEM image (Fig. 6), shows irregular shaped particles and a rather wide distribution of dimensions, with aggregates major of 100 micron along with small particles around 10 micron.
Such large distribution of dimensions, unlike the previous analysed powder (SS316L < 325 mesh), suggested the utilization of the DC thermal plasma at higher plasma powers.
Table 2 shows the process parameters used for the test of the powder B.
Table 2
Plasma process parameters for SS316L < 100 mesh (B powders)
Test | Power (kW) | Ar (slm) | He (slm) | Ar carrier (slm) | Pressure Test (bar) | Feeding rate (g/min) |
B1 | 12,5 | 40 | 10 | 1 | 90 | 5 |
B2 | 17 | 40 | 10 | 2 | 92 | 5 |
B3 | 21 | 40 | 10 | 1,5 | 91 | 5 |
B4 | 21 | 40 | 10 | 1 | 87 | 5 |
B5 | 21,5 | 40 | 15 | 1 | 105 | 5 |
The SEM images of the samples B treated under different plasma powers (range 12.5–21.5 kW) are reported in Fig. 7.
Spheroidization process begins at the lowest power, with the formation of spheres of limited size (below 30 microns), but the larger particles appear to be only slightly touched by the flame (rounding). The spheroidization degree increases at higher power, already involving most of the material at 21 kW. By further increasing the power, next to the spheres, it is generally observed the formation of an appreciable quantity of nanoparticles, probably due to the vaporization of the smaller size material. This phenomenon is evident above all on the fraction of material deposited on the cold walls of the reactor. For this reason, higher plasma powers do not seem to be really advantageous for the process.
The XRD patterns of the samples B, before and after the plasma treatment, are depicted in Fig. 8.
The powders B, after plasma treatment, exhibit a XRD spectrum almost identical to the pristine powder. The main phase of the raw powder is austenite, that turns to be the most abundant phase in the products. The ferritic phase is nearly absent and it is appreciably found only in the spectrum of the powder treated at 12.5 kW (test B1), the lowest power used in this set of tests. Similarly, as in the case of powder A, it is worth mentioning the formation of oxides: likewise, the set A, the oxide signals grow at higher power and consequently with the production of nanoparticulate material, more susceptible to oxidation.
Based on the SEM and the XRD analyses, we selected two set of process parameters, one for each powder, which were considered to produce the best materials: indeed, the best results were achieved for B5 test (Table 2) and A2 test (Table 1). Such process conditions were used to duplicate both the tests and to produce a larger quantity of materials. By the way, the test coded as B5rev and A2rev replies B5 and A2, respectively. As a first result, the tests turned to demonstrate a good reproducibility.
The SEM pictures were used and further processed to measure the particles dimensions and particularly the degree of spheroidization. Image processing and calculation was carried out with the image editor Zen Core 3.6. The circularity calculated according to the following formula was assumed as the main reference parameter of the shape factor:
Circ = 4πA/P2
where P and A are the particle perimeter and area, respectively. The value varies from zero to 1, with a perfectly spherical particle having a circularity equal to 1.
The tests show that the average circularity value is always higher than 0.8, indicating an extensive spheroidization of the powders, whatever the raw powder processed. Itagaki et al. [17] also obtained powders with high sphericity and uniform size distribution by performing a DC arc plasma spheroidization under appropriate processing conditions, but nanoparticle-modified spherical particles were obtained already at 17 kW starting from raw powder of averaged diameter of 38 µm. In our conditions, the spheroidization process was conducted on two set of commercial powders of wider averaged diameter (till to 149 µm) with encouraging results both in terms of sphericity and nanoparticles formation; furthermore, these results are particularly representative due to the relevant dimensions of our experimental set-up, which could be easily scaled up.
The flowability of metal powders is not an inherent property – it depends not only on the physical properties (shape, particle size, humidity, etc.), but also on the stress state, the equipment used and the handling method [23–25]. That’s why the term “flowability” describe a complex behaviour of powder, when it is mobilized or subjected to stress and its evaluation is highly dependent on the equipment and the way in which it is tested. A uniform and comprehensive way of describing the flow of metal powders does not yet exist. Therefore, it is necessary to consider the possibilities of experimental determination techniques, to compare the quality of equipment and the measured results.
Numerous possibilities exist to test powder flowability. Among them, widely used and standardized methods are the Hall flowmeter funnel (ASTM B213) [26] and the Carney funnel (ASTM B964) as summarized in ASTM F3049-14, but no one index describing the flowability of powder streams is universally applicable and flowability tests have to reflect the state of the powder in the process under consideration.
In the case of our steel powder, we used the POWDERFLOW kit by Carpenter Additive, which allows to perform four different flowability tests and, particularly, the Hall Flow and the Carney Flow. The obtained values are reported in Table 3
Table 3
Hall Flow Results (HFR) on steel samples A and B (raw powders and after plasma treatment)
TEST (Tamb = 15°C, Moisture = 75%) | A raw powder | A2 | B raw powder | B5 | B5 (after purification) | B5 (32–40 µm fraction) | B5 (40–63 µm fraction) |
HFR (s/50g) | nd | 32,26 | 38,93 | 31,69 | 29,54 | 24,61 | 23,58 |
A-type powder revealed a cohesive behaviour during the test, obstructing the exit hole both of the Hall and Carney funnels and therefore not providing any flowability value. The same powder after plasma treatment (A2rev) shows a good flowability even if with average dimensions below 40 micron.
With respect to B-type powder, we registered an improvement of flowability, passing from a value of about 38.93s/50g to 31.69s/50g. The further solvent purification, eliminating the nanoparticles that act as "bridge" for the larger particles, improves the flowability: in fact, purified B5 has an HRF of 29.54 s/50g. The sieved B5 fractions also show good results compared to the unsieved powder: B5 (32–40 µm fraction) and B5 (40–63 µm fraction) have 24.61 and 23.58 s/50g of HRF, respectively. It’s worth noticing that the effect of sieving on increasing flowability is greater than that of purification.
The recorded values are comparable with reference values of commercial powders, as reported in the corresponding technical datasheets. Table 4 shows a selection of commercial SS316L powders.
Table 4
Flowability values for commercial SS316L powders
COMMERCIAL POWDER | SUPPLIER | PRODUCTION PROCESS | PARTICLE SIZE DISTRIBUTION (µm) | REFERENCE FLOWABILITY VALUE (s/50g) |
316L | HOGANAS AB | Gas Atomization | 20–53 | 18,1 |
316L | HOGANAS AB | Water Atomization | 25–63 | 37,4 |
MARS 316L | MIMETE METAL POWDERS | Gas Atomization | 15–45 | ≤ 30 |