Abstract
In this research, the microstructure and mechanical properties of Fe–1.8Ni–0.5Mo–2Cu–0.45C powder metallurgy (PM) steels with 0–0.3% vanadium were investigated by optical microscopy, scanning electron microscopy, transverse rupture, hardness and tribological tests. It is revealed that the volume fraction of polygonal ferrite in the sintered PM steels increased with the increasing of V addition, and more fine V(C,N) particles were formed during the sintering process. The enhancement of transverse rupture strength of the sintered and heat-treated PM steels with V addition was due to the grain refinement and precipitation strengthening effect. The friction coefficient the heat-treated steels decreased with an increase in V addition, whereas the wear loss increased.
Introduction
Fe-based powder metallurgy (PM) steels are widely used in the automotive industry due to its high productivity and flexible alloy composition design. Fe–Ni–Mo–Cu–C alloy system exhibits excellent integrated mechanical properties for sintered steels and is used to compete with wrought grades, and the FLA-08N2M-C2-620SH (ISO 5755:2012(E) is one of the PM steels widely used in the practical production [1,2]. However, for products with high mechanical performance requirements, such as heavy-duty gear, the strength and toughness of the sintered steel cannot meet the performance characteristics [3]. The addition of microalloying elements provides an effective way to improve both the strength and toughness of steel, which results in precipitation strengthening and grain refinement [4–10]. To achieve these beneficial properties, steels containing microalloying elements are usually subjected to the controlled hot working process, which is different in PM than in conventional wrought steel processing utilising forging and rolling [11].
The microalloying elements suitable for PM steel should be able to form precipitates during sintering or cooling and refine grain. The choice of microalloying element should be to consider the solubility of the carbide or nitride that is precipitated. Titanium has a high propensity for forming oxides and sulphides, the lowest solubility of its carbide or nitride, and is generally ineffective as a precipitation strengthener in medium- to high-carbon steels. V(C,N) is more soluble in high-carbon steels and less sensitive to carbon level compared with Nb(C,N), which shows a reduced solubility in steel at higher carbon levels [11,12]. Vanadium can combine preferentially with nitrogen to generate nitrides in situ during sintering and form precipitates during cooling. This is advantageous for Fe-based PM materials, which typically contain higher carbon and use nitrogen-containing sintering atmospheres. Moreover, V(C,N) can effectively limit the growth of austenite grains during the heat treatment, and obtain the martensite with higher strength and toughness after quenching. Therefore, vanadium is an ideal microalloying element for Fe-based PM materials. Studies have shown that vanadium and silicon can enhance both the strength and the hardenability in Fe-based sintered steels [13].
As the content of carbon in the FLA-08N2M-C2-620SH is relatively high, which will led to a poor toughness, the content of carbon was reduced to 0.45% (wt-%, similarly hereinafter), the vanadium was added to in the form of Ferrovanadium powder in the studied PM steels, and the effect of vanadium micro-addition on microstructure and mechanical properties of Fe–1.8Ni–0.5Mo–2Cu–0.45C PM steels in the sintered and heat-treated conditions were evaluated in the paper.
Materials and experimental procedure
The chemical composition of the 1300 WA pre-alloyed Fe powder.
The density of sintered specimens was measured by the water displacement method. Specimens for microstructural characterisation were prepared by the standard metallographic methods. Both the sintered and heat-treated specimens were etched by 4% nital and the etching times were carefully controlled. The microstructure of samples was observed by optical microscope, scanning electron microscopy (SEM, Hitachi S-4800) and the Energy dispersive spectrometry was used to determine the phase compositions. The volume fractions of phases were calculated by Image-pro plus software, and the grain size was measured using the linear intercept method. Transverse rupture was carried out using the universal testing machine (CSS-2202) according to the standard ISO 3325: 1996. The hardness and microhardness of the sintered PM steels were evaluated using a Vickers hardness tester with the load of 5 and 0.2 kg, respectively, and 10 measurements were conducted for each sample to ensure satisfactory statistical reliability. The wear resistance of the heat-treated samples was evaluated by a pin-on-disc tester at ambient conditions, and GCr15 steel ball was used as a stationary slider. The normal load of 10 N, sliding speeds of 0.117 m s−1 and the friction time of 40 min were applied as test parameters for each test specimens. The abrasion resistance was evaluated by measuring mass loss after test using electronic balance (accurate to 0.0001 g) and the dynamic friction coefficient recorded by the tester.
Results and discussion
Microstructure of sintered specimens
Figure 1 shows the optical micrograph of the sintered PM specimens. The microstructure of the sintered specimens mainly consisted of bainite, accompanied with a small amount of polygonal ferrite and porosity, and no pearlite was observed in the sintered PM specimens. The grain size and the volume fraction of polygonal ferrite and bainite in the sintered PM steels were shown in Table 2. The bainite in the sintered V0 specimen was in the form of a coarse acicular ferrite and lath structure with streak cementite distributed parallel between ferrite laths, which was upper bainite [14]. Compared with the sintered V0 specimen, the microstructures of the specimens with V addition were refined, and more of the bainite from upper bainite to granular bainite. The volume fraction of polygonal ferrite increased with increasing V content, while the amount of porosity, the size and the volume fraction of the bainite decreased. Compared with that of sintered V0 specimen, the volume fraction of polygonal ferrite of sintered V3 specimen increased from 22 to 41%, while the grain size reduced from 31.5 to 19.7 µm. The optical microstructure of the sintered PM specimens: (a) V0, (b) V1, (c) V2 and (d) V3. UB: upper bainite; PF: polygonal ferrite; K: porosity; GB: granular bainite. The grain size and volume fraction of polygonal ferrite and bainite in the sintered PM specimens.
Figure 2 presents the optical micromorphology of bainite in V0 and V3 specimens with higher magnification. For the bainite in V0 specimens, the lath ferrite nucleated at the austenite grain boundary, as indicated by black arrows, was shown in Figure 2(a); the growth of ferrite led to the C-rich zone on both sides of the ferrite laths as the C atoms are expelled from the ferrite. Owing to the slower diffusion rate of C atoms in austenite than in ferrite, and the presence of Mo could significantly restrain the transformation of austenite to proeutectoid ferrite and pearlite, but exhibit no obvious retardation to the formation of bainite [15–17], the discontinuous streak cementites would precipitate at the C- rich zone parallel to the ferrite laths, and the coarse upper bainite could be obtained from the sintered V0 specimen during the slow cooling process [18]. However, V(C,N) can precipitate during cooling process from sintering temperature when V atoms dissolved into the PM steel [12], which will promote the formation of polygonal ferrite at the prior austenite grain and isolate austenite into islands of C-rich zone; thus, the volume fraction of polygonal ferrite increased with increasing V content, while the grain size decreased. The formation of polygonal ferrite decorates the prior austenite grains, disabling them from their role as nucleation sites for lath ferrite to form upper bainite [19]. Considering the slow cooling rate (6–8°C min−1) of the sintered PM steels, the C-rich austenite islands transformed into equiaxed ferrite and granular cementite during cooling, and formed granular bainite. The granular bainite forming temperature was higher than that of upper bainite transition range, which enabled the growth of equiaxial ferrite and granular carbide precipitate in the austenite islands [18,20]. The presence of fine polygonal ferrite and granular bainite can provide excellent ductility and toughness for the sintered PM specimens with V additions. Optical morphology of the bainite in the sintered PM specimens. (a) V0 and (b) V3. SC: streak cementite; PF: polygonal ferrite; GB: granular bainite.
In fact, vanadium nitride (VN) and vanadium carbide (VC) can be generated from in situ reaction of V with nitrogen gas and graphite during the sintering process. The V atoms in the solid solution could also combine with carbon and nitrogen atoms to form carbonitride. There are three reactions which might occur during the sintering process [9,21]
Figure 3 displays the SEM image of the Ferrovanadium particle in the sintered V3 specimen and the corresponding EDS analysis results for spot 1, 2 and 3. The Ferrovanadium particle still remained in its original place and exhibited a core-rim structure. Compared with that of spot 3, the EDS analysis results of spot 1 indicate that the grey spherical regions with a size of about 0.5–1 µm at the core area were the remains of the undissolved Ferrovanadium particle, where the content of C atom was lower while the content of V atom was higher, indicating that the C atom can diffuse into the Ferrovanadium particle, and the V atom in the centre of Ferrovanadium particle had not diffused completely into the near pre-alloy iron powders’ particles. From the EDS analysis result of spot 2, the black rim area was found to be mainly composed of V, N and C elements (the sum of atom per cent of V, N and C element was 84.11 at.-%, and that of N and C element was 26.16 at.-%), indicating that the V(C,N) will be formed during the sintering process. Previous studies have revealed that VN can be formed at a temperature of 674°C by the reaction between Ferrovanadium and nitrogen gas for the higher affinity between vanadium and nitrogen [22,23]; thus, the diffused V atoms would prefer combining with N atoms rather than with C atoms in the sintering atmosphere. Furthermore, the above reaction (2) could occur at a lower sintering temperature, resulting in the content of N atom in the spot 2 being higher than that of C atom, and the amount of VN would be higher than that of VC. SEM micrograph of FeV particle in the sintered V3 specimen (a), and the corresponding EDS analysis results for spot 1 (b), spot 2 (c) and spot 3 (d).
Figure 4 shows the EDS surface scanning analysis results of V and N elements distribution in the sintered V2 specimen. Except for the areas of the undissolved Ferrovanadium particles remained after the sintering process, the distribution of V elements was fairly homogeneous in the specimen, while relatively more N atoms distributed in the outer layer of the particles. The V atom would diffuse into the Fe matrix from the Ferrovanadium particles during the sintering process, and the typical sintering temperatures (1120–1260°C) are sufficient to get in situ synthesised V(C,N) particles in solution. The slow cooling from the sintering temperature will allow for the formation of fine V(C,N) particles’ precipitates in the matrix [11,12,14,24]; thus, the distribution of V elements is fairly homogeneous in the specimen. These precipitated V(C,N) particles can play the role of strengthening phase in the PM steel, resulting in increased strength and grain refinement. As the holding time of sintering temperature was not enough, the in situ formed V(C,N) could not dissolve completely into the pre-alloy iron powders’ particles; thus, relatively more N atoms distributed in the outer layer of the particles. The V(C,N) particles will precipitate during cooling, which results in precipitation strengthening and grain refinement. The V(C,N) particles that precipitate at a ferrite/austenite interface can provide a coherent, low-energy interfaces with ferrite, thus reducing the interfacial energy related to ferrite nucleation [25]. This explained the increased volume fraction of polygonal ferrite with the increase of V content. EDS surface scanning analysis results of V and N elements distribution in the sintered V2 specimen, FeV: Ferrovanadium.
The thermodynamic data of the solubility product K for VN and VC (generally defined as V4C3 or VC0.75) in austenite can be expressed as the following based on Turkdogan [26]:
The solubility product K versus temperature for VN and VC0.75 in austenite.

Mechanical properties of sintered specimens
Table 3 presents the density, apparent hardness and microhardness of the as-sintering specimens, and the transverse rupture strength of the PM specimens is shown in Figure 6. Micro-addition of V had little effect on the sintering density, and the apparent hardness of the sintered specimens increased with the increase of V addition. Compared with that of a V-free specimen of V0, the transverse rupture strength of the sintered specimens with V additions enhanced, and the transverse rupture strength of the V2 specimen reached a maximum of 827.4 MPa. The increase in both strength and hardness of sintered specimens with V addition is probably caused by the grain refinement and the precipitation strengthening effect induced by V(C,N) particles. Microhardness of bainite increases with V addition, while there was no effect for the polygonal ferrite. For Fe–Ni–Mo–Cu–C system PM steels, the addition of 2%Cu can saturate the ferrite, and it is already at its maximum reinforcement effect of the solution; therefore, the addition of V could not result in a further significant increase for the microhardness of ferrite [13]. Even so, the morphology of the bainite of the specimens with V addition changed to a denser granular bainite cluster, and there was an increase in the microhardness of bainite, then, the apparent hardness of the sintered specimens increased with the increase of V addition. Compared with V2 specimen, it is worth noting that upon further increase in V addition to 0.3%, the transverse rupture strength of V3 specimen was found to decrease. Studies have demonstrated that large hard particles can raise locally the stress of matrix, which induces generation and propagation of cracks in steel [29]. The great amount of in situ synthesised V(C,N) presented in the outer layer of iron particles reduced the toughness of the steel, and Furthermore, undissolved Ferrovanadium particles also have a negative impact on the toughness of the sintered steel. Transverse rupture strength of the sintered PM specimens. The density, apparent hardness and microhardness of the sintered PM specimens.
Figure 7 shows the SEM micrographs of fracture surfaces of V0 and V2 specimen after transverse rupture tests. Both the dimples and cleavage surface are observed from the fracture morphology, which reveals a mixed ductile–brittle fracture mode. Moreover, the fracture surface of V2 sample exhibited more dimples than that of V0 sample, which are correlated with the refiner grain and the granular bainite. SEM micrographs of the fracture surface of the sintered V0 specimen (a) and the V3 specimen (b).
Figure 8 presents the stress–strain curve obtained from transverse rupture tests of the sintering specimens. It can be seen that the V-free specimen exhibited the lowest transverse rupture strength but the highest strain. The area surrounded by the stress–strain curve and the abscissa can be regarded as the fracture absorbing energy. It can be calculated from the curves that V2 and V3 specimens exhibited the highest and lowest fracture absorbing energy, respectively. The great amount of in situ synthesised V(C,N) presented in the outer layer of iron particles reduced the toughness of the sintered steel when the addition of V was more than 0.2%. Stress–strain curve for transverse rupture tests of the sintered PM specimens.
Microstructure and mechanical properties of heat-treated specimens
Figure 9 shows the microstructures of the quenched V0 and V3 specimens. The microstructures of all the specimens were composed of mixed lath martensite and acicular martensites with a small amount of residual austenite. Compared with that in V0 specimen, V3 specimen had much finer grain and less porosity, and a smaller amount of acicular martensite and residual austenite. The residual austenite contents measured by image software in quenched V0–V4 specimens are 8.73, 6.59, 5.44 and 4.26%, respectively. It has been reported that the initial structure of plate/lath-like carbides dissolves more rapidly than globular carbides during austenitising [30]. V(C,N) was not dissolved in the matrix under the heat treatment temperature of 930°C from the results of Figure 5, therefore, the content of C dissolved in austenite of V3 specimen should be smaller, and C element distribute less evenly compared with that of V0 specimen, which is favourable for obtaining less amount of acicular martensite and residual austenite after quenching. The presence of V(C,N) and fine granular bainite retard the growth of austentite grain prior to quenching, and thus refine the quenched microstructure. Optical microstructures of the quenched PM specimens: (a, c) V0 specimen, (b, d) V3 specimen. RA: residual austenite; AM: acicular martensite.
The hardness and the transverse rupture strength of the heat-treated PM specimens are shown in Figure 10. Compared with the sintered specimens, both the hardness and transverse rupture strength of the heat-treated specimens were improved significantly due to the formation of martensites. However, the hardness of the heat-treated specimens decreased slightly with the increase of V addition. The effect of V addition on the transverse rupture strength was the same as that of sintered specimens; the transverse rupture strength of the V2 specimen reached a maximum of 1156.4 and 120.9 MPa more than that of V0 specimen.
Transverse rupture strength and apparent hardness of the heat-treated PM specimens.
The effect of V addition on the hardness of the heat-treated specimens is complicated, which is related to the microstructure determined by the austenitising temperature [11]. The austenitising temperature of the heat treatment was not high enough to dissolve all the V(C,N) precipitates, and then the amount of carbon dissolved in the austentite was lowered. The content of C dissolves in martensite will be decreased with the increase of V addition because of a great amount of undissolved carbonitride in the heat-treated samples, thus, the hardness of the heat-treated specimens decreased slightly because the hardness of the martensite matrix essentially depends on the amount of the solution carbon. The reason for the transverse rupture strength of the heat-treated specimens increased with the increase of V addition lies in the much finer grain and precipitation strengthening effect induced by the precipitates. However, further increasing the V addition as that in the V3 specimen, the great amount of undissolved V(C,N) particles in the heat-treated microstructure, which are susceptible to microcrack formation, will reduce the transverse rupture strength.
Wear resistance of the heat-treated specimens
The dynamic friction coefficient of the heat-treated PM specimens is shown in Figure 11. Compared with that of V0 specimen, the dynamic friction coefficient of the heat-treated PM samples with V addition fluctuated slightly, indicating that the specimens with V additions had a higher frictional stability. The averaged value of the friction coefficient of heat-treated PM V0–V4 is 0.316, 0.289, 0.265 and 0.262, respectively, indicating that the friction coefficient of the heat-treated specimens decreased with the increasing of V addition, which can be explained by the finer grain and less porosity. The dynamic friction coefficient of the heat-treated PM specimens.
Figure 12 shows the effect of vanadium addition on wear loss of the heat-treated specimens. It can be seen that the wear loss of the heat treated specimens increased with an increase in V addition. According to the classical theory, the wear volume (V) proportional to the normal load (N), the sliding distance (S) and the inverse of the hardness (H) [31] is
Effect of vanadium addition on wear loss of the heat-treated specimens.

Conclusions
Compared with that of the V-free sintered PM steels, the microstructures of the sintered PM steels with V addition were refined, and more of bainite presented in the form of granular bainite, and fine V(C,N) particles were formed during the sintering process. The volume fraction of polygonal ferrite increased with increasing V content, while the size and the volume fraction of bainite decreased. The apparent hardness and the transverse rupture strength of the sintered specimens increased with the increase of V addition, but upon further increasing the V addition to 0.3%, the transverse rupture strength decreased. The increase in both strength and hardness of sintered specimens with V addition is probably caused by the grain refinement and the precipitation strengthening effect induced by V(C,N) particles. The microstructures of the quenched PM steels were composed of mixed lath martensite and acicular martensites with a small amount of residual austenite. The V addition refined the grain, and decreased the amount of acicular martensite and residual austenite. The hardness of the heat-treated specimens decreased slightly with the increase of V addition for the reduced amount of acicular martensite. The transverse rupture strength of the heat-treated specimen with 0.2% V addition reached a maximum of 1156.4 and 120.9 MPa more than that without V addition. The heat-treated PM samples with V addition had a higher frictional stability. The friction coefficient the heat-treated steels decreased with an increase in V addition, whereas the wear loss increased.
Footnotes
Disclosure statement
No potential conflict of interest was reported by the authors.
