Abstract
For highly loaded precision parts required in large numbers, the powder metallurgy route by pressing and sintering is attractive both economically and ecologically. To obtain the required mechanical properties, heat treatment is necessary. For this purpose, sinter hardening, that is, gas quenching of the parts immediately after leaving the high temperature zone of the furnace, is a highly cost-effective variant. Compared to standard oil quenching, however, the cooling rates are significantly lower, which means that alloying of the sintered steels plays a major role. In the present study, hybrid alloyed steels were investigated, based on Cr prealloyed powder with the addition of Ni or Mn elemental powder grades, as well as a fine Mn–Si masteralloy powder. The sinter hardening behaviour was investigated as a function of the alloy element and the carbon content, respectively, with moderate cooling in the range of 0.7 K/s (linearised) being applied. It showed that in this cooling regime, less Mn is required for sinter hardening than Ni or the masteralloy. However, the impact energy is significantly lowered by Mn addition, as a consequence of intergranular failure, while both Ni alloying and addition of the Mn–Si masteralloy offer an attractive combination of hardness and toughness. This underlines the importance of the alloying route, in particular when using Mn.
Introduction
The classical powder metallurgy (PM) production route of uniaxial die compaction plus sintering has been established for the manufacturing of complex-shaped ferrous parts with high geometrical precision and in large numbers. In addition to the geometrical properties, also the mechanical properties have been improved in the last decades, by suitable alloying techniques as well as by optimised production processes, which also include heat treatment.1–3
Most of the parts thus produced have been consumed by the automotive industry, the parts being installed in particular in internal combustion engines and transmissions. With the increasing trend to alternative, in particular electrical, drivetrain systems, which require such precision parts not at all or at least in lower numbers, 4 there is considerable pressure on the PM industry to find new markets for their products. This also involves further improvement of the property profiles as well as of the economy. In the latter case, traditional sintered steels have been at a disadvantage compared to wrought structural steels since PM has been using alloy elements as Cu, Ni and Mo. These elements offer low oxygen affinity but are more expensive than the elements used in wrought steels, such as Cr, Mn and Si, and also present more volatile prices. 5 Therefore, using the latter – cheaper – alloys has been a focus of PM research for a long time, in particular regarding Cr and Mn. For the former, prealloyed steel grades are commercially available,6,7 while for Mn, the prealloying route is feasible in principle. 8 It is, however, less attractive because of the strong hardening effect of Mn on ferrite, which makes Fe–Mn prealloy powders less compactible than, for example, Fe–Cr or Fe–Mo grades. Here, the advantage of PM compared to ingot metallurgy can be used: it offers numerous alloying routes, prealloying, blended elemental, masteralloy, diffusion bonding or also combinations, that is, hybrid alloying (e.g. Geroldinger et al. 9 ). In particular, the masteralloy route has made significant progress since the first approaches in the 1970s,10–13 in terms of yield, particle morphology and compositions. 14
As stated above, heat treatment is a viable way to improve the mechanical properties also of sintered steels. For low alloy grades, oil quenching and tempering can be successfully applied, but it increases cost. For PM parts, which usually contain open porosity, 15 also removal of the oil from the pore network is tricky and critical from the environmental viewpoint. Here, PM offers the variant of ‘sinter hardening’. In this case, the parts are quenched with cold inert gas immediately after leaving the high-temperature zone of the furnace. This results in clean, shiny parts for which just tempering is required, but no washing operations. However, the cooling rates attained by sinter hardening are significantly lower than for oil quenching, in industrial production typically in the range of 2.5–3 K/s. This implies that steel grades with better hardenability are needed, which usually means higher alloy element content than required for oil hardening grades.16–20
As an alternative to adding more and more alloy elements, also combining suitable elements can be done. Special combinations have shown to be particularly effective for sinter hardening, such as, for example, Mo and Cu, and here, combining not only different alloy elements but also different alloying techniques proved to be successful (e.g. Karamchedu et al. 20 ). One commercially available powder grade combines prealloyed Mo with diffusion bonded Cu and is widely used in the parts manufacturing industry. In the present study, it was investigated how a Cr prealloyed base powder can be combined with Mn added either as elemental powder or through a suitable masteralloy (which also contains some Si). In parallel, elemental Ni was added as a reference. Moderate cooling conditions were applied to simulate heavy loading in sinter hardening furnaces.
Materials and methods
The starting powders used were Astaloy CrA (Höganäs AB, Fe-1.8% Cr, water atomised) as base powder and electrolytic Mn (Poudmet, < 10 µm) as well as carbonyl Ni (Norilsk, < 10 µm) as additives. Furthermore, a fine Mn–Si masteralloy powder produced by ultra-high pressure water atomisation (Atomising Systems Ltd, Sheffield, UK, see Table 1) was employed. Natural graphite UF4 96/97 (Kropfmühl) was added as carbon carrier, to an amount resulting in the defined nominal carbon content in the green compacts (which considered also the C content introduced through the masteralloy). In one test series, the alloy element content was varied, keeping the nominal C content constant (see Tables 2 and 3), and in the second series, the reverse approach was chosen (see Table 4); in this latter case, also the carbon-free variant was investigated.
Chemical composition (wt-%) and mean particle size (µm) of the masteralloy used.
Chemical composition data obtained by XRF, C and O by LECO.
Composition of the mixtures used in study 1 – varying alloy element content. a
Carbon content: nominal = admixed graphite; for MA: + carbon from the masteralloy.
Elemental composition of the mixtures prepared using masteralloys. a
Carbon content: nominal = admixed graphite; for MA: + carbon from the masteralloy.
Composition of the mixtures used in study 2 – varying carbon content. a
Carbon content: nominal = admixed graphite; for MA: + carbon from the masteralloy.
The powders were dry mixed for 60 minutes in batches of 300 g in a Turbula mixer (Willy A. Bachofen AG, Switzerland), and the mixes were compacted at 700 MPa in a pressing tool with a floating die. Charpy impact specimens 55 × 10 × 10 mm (ISO 5754, https://www.iso.org/standard/85821.html) were produced. Die wall lubrication was afforded using Multical sizing fluid as a lubricant, which eliminated the need for a separate dewaxing step.
The green compacts were sintered in an electrically SiC rod heated pusher furnace with gas-tight superalloy tube retort of 65 mm inner diameter. Sintering was done by push-in-push-out, that is, the boats containing the specimens were pushed into the high temperature zone of the furnace, set at 1250 °C and operated in N2-10% H2 atmosphere (flow rate 2 Nl/min; both gases of 99.999% purity). After the desired isothermal soaking time of 1 hour they were pushed into the water-jacketed exit zone of the furnace, and the gas flow was changed to plain N2 and set at the higher rate of 5 Nl/min. There, the specimens cooled following Newton's law of cooling; the linearised cooling rate was about 0.7 K/s. This relatively low cooling rate was selected to identify those compositions that are particularly tolerant regarding changes in the cooling rate. These thus can be regarded as robust also in industrial production, in particular in the case of heavy furnace loading. Because of the fairly slow cooling rate in the low temperature range, tempering or stress relieving was not regarded as necessary.
The as-sintered specimens were characterised by standard techniques (three samples were measured for each test). Green density was calculated from mass and dimensions, and the sintered density was determined by water displacement (Archimedes method) after impregnation. The dimensional change was determined by measuring the length of the bars before and after sintering, that is, perpendicular to the pressing direction. Charpy impact tests (unnotched) and dynamic Young's modulus measurements were done, the latter on an ICME resonance system. Cross sections of the samples were examined by metallography, and hardness measurements were done in the sections (here it should be noted that the apparent hardness HV30 was measured, which includes the effect of porosity). The oxygen content was measured through hot fusion analysis using a LECO TC 400 analyzer and the combined carbon content through combustion analysis in a LECO CS-230 analyzer. The electrical conductivity was measured using the four-point method, and the magnetic properties were determined on a Foerster Koerzimat.
Experimental results and discussion – Part 1: Variation of the alloy element content
The dimensional and mechanical properties, as well as the carbon and oxygen content of the sintered samples, are shown in Figure 1 (the exact data are presented in Table S1 in the Supplemental Material).
Properties of steels with 0.6% admixed carbon, compacted at 700 MPa, sintered 60 minutes at 1250 °C in N2-10% H2. (a) Green density; (b) sintered density; (c) dimensional change; (d) apparent hardness; (e) dynamic Young's modulus; (f) impact energy (unnotched); (g) electrical conductivity; (h) coercive force; (i) magnetic saturation; (j) combined carbon content; and (k) total oxygen content.
Dimensional properties
The effect of the additives on the green density is rather low; Ni addition slightly increases it, while Mn and the MA lower it. Here, also the theoretical density should be considered, which is slightly increased by Ni but slightly lowered by Mn and also the masteralloy. However, since the density of all alloy elements is not too far off that of iron and their content is low, their effect on the theoretical density is marginal. The sintered density shows similar effects, which are, however, more pronounced for Mn and rather similar to those for the green density in the case of Ni and MA. This agrees also well with the dimensional change: the Ni alloyed grades show marked shrinkage, more than the reference steel and the more, the higher the Ni content is. This densifying effect of Ni is well known from previous studies21,22; of course, the lower resulting porosity positively affects the mechanical properties.
Mn, in contrast, promotes expansion during sintering, the more, the higher the Mn content is, and thus lowers the sintered density (and, consequently, to some extent also the mechanical properties). Also, this effect, which can be dubbed ‘Mn swelling’, is well known from the literature (see e.g. Danninger et al. 23 ). The reason is the unique homogenisation of Mn during sintering through gas phase transport,24,25 which is a ‘one-way’ process, the vapour pressure of Mn being several orders of magnitude higher than that of Fe. Through the interconnected porosity, Mn vapour is transported to the Fe surfaces and alloyed there, but the reverse – alloying of Mn by Fe vapour – is not possible. Mn diffuses into the Fe particles, increasing their volume and this inducing macroscopic swelling. The Mn–Si masteralloy does not show this undesirable effect since in this system, homogenisation is primarily attained through transient liquid phase, the vapour pressure of Mn above the masteralloy being significantly lower than above elemental Mn. With all materials studied here, however, the dimensional changes are small, < 1% linear, and thus can be expected not to adversely affect the dimensional precision of the sintered parts.
Mechanical properties and microstructures
Young's modulus
As evident from Figure 1(e), Young's modulus is well correlated to the density, as described, for example, by Leheup and Moon, 26 Cytermann 27 and Haynes and Egediege. 28 The lowest values are, therefore, recorded with the highest content of admixed Mn, while the effect of Mn introduced through the masteralloy is small. The slightly lower-than-expected values for the Ni alloyed steels, despite their high density, can be attributed to local areas of retained austenite, see Figure 2(b).
Metallographic sections of PM steels with 0.6% admixed carbon, compacted at 700 MPa, sintered for 60 minutes at 1250 °C in N2-10% H2; Nital etched. (a) CrA-0.6C, 172 HV30; (b) CrA-3Ni-0.6C, 362 HV30; (c) CrA-2Mn-0.6C, 398 HV30; (d) CrA-3Mn-0.6C, 407 HV30; (e) CrA-3MA-0.6C, 228 HV30 and (f) CrA-4MA-0.6C, 381 HV30.
Hardness
The apparent hardness shows a pronounced effect of the additives. The plain Cr steel is relatively soft as-sintered, about 170 HV30. All additives increase the hardness, but the effect is not too pronounced at 1% Mn added, for Ni up to 2% and for the MA even up to 3% (which, however, means just 1.2% Mn). At the sintering conditions applied here, sinter hardening, meaning an apparent hardness > 350 HV30, is attained for 3% Ni, 4% MA and for Mn both at 2 and 3 mass%, which underlines the strong effect of Mn on the hardenability of steel, as indicated also by its high Grossmann factor,1,29 which is significantly higher than that of Ni.
Microstructures
The hardness data are also supported by the microstructures. In Figure 2, some typical images are shown. At the cooling rate applied here, the plain Cr steel resulted in a fully pearlitic microstructure, which agrees well with the low hardness and moderate Grossmann factor (see e.g. Krauss 29 ). In the case of 3% Ni, a heterogeneous material is attained, in addition to martensite also some retained austenite being visible, as well as pearlitic areas in the core of the largest base powder particles. This indicates the slow bulk diffusion of Ni even at 1250 °C. Nevertheless, the combination of Cr and Ni seems to improve the hardenability in a synergistic way. The steels with 2% and 3% Mn, in contrast, are almost fully martensitic, with only a few pearlitic or bainitic areas. This confirms both the effectivity of the gas phase transport for homogenisation as well as the fact that, although Mn does not diffuse faster than Ni in austenite, 30 less Mn than Ni is required to transform also the core of the larger base powder particles into martensite. Also here, the combination of (prealloyed) Cr with Mn is beneficial towards (sinter) hardenability, even more than Cr + Ni, as indicated by the high hardness. The steel prepared using the Mn–Si masteralloy, finally, shows a well-homogenised microstructure. Also here, the accelerated homogenisation, in this case by transient liquid phase, was shown to be highly effective for transporting the alloy elements, but without the adverse swelling effect. With 3% MA (which corresponds to 1.2% Mn), however, the microstructure is still pearlitic with some bainite, while at 4% MA (= 1.6% Mn), a fully martensitic structure is attained, with accordingly higher hardness. This shows that for the present cooling conditions, the critical Mn content above which sinter hardening occurs is between 1.2% and 1.6% Mn.
Impact energy
Hardness, however, is not the only criterion for sintered steels; sufficient toughness is required, too. Here, the (unnotched) Charpy impact energy is well suited for measuring the toughness, this property being particularly sensitive both to the ductility of the matrix and the strength of the sintering contacts.
When comparing the impact energy data with the hardness, it is not surprising that higher hardness, as attained by alloying, corresponds to lower impact energy. This effect is least pronounced for Ni; with 1% Ni, the impact energy is even higher than for the reference material, despite the higher hardness. The well-known positive effect of Ni on the toughness of steels has been described also for sintered steels, 31 for these materials also the heterogeneous microstructure being regarded as positive. For the MA alloyed steels, the gain in hardness results in a more pronounced drop of the impact energy than for the Ni steels, but the effect is still tolerable: for 4% MA, 380 HV30 is combined with about 23 J·cm−2, which is an attractive combination.
The Mn alloyed steels, however, are an exception here; the loss of toughness is significantly higher than with the other steel grades, despite the only small hardness difference. For 2% Mn, about 400 HV30 are combined with 13 J·cm−2, and for 3% Mn, which results in virtually the same hardness, the impact energy drops to an unacceptably low level of just 6 J·cm−2. It must therefore be concluded that the positive effect of Mn on hardness and hardenability is offset by a negative effect on the toughness, at least if Mn is added as elemental powder.
Fractography
The reason for these pronounced differences can be found by fractographic studies, as shown in Figure 3: the reference material as well as the Ni alloyed steel show typical ductile fracture, for the plain Cr steel, the dimples being slightly coarser than for the Ni alloyed grade, which agrees with the higher impact energy. The MA grade shows a mix of ductile fracture and transgranular cleavage, while in the steel alloyed with 3% Mn, there are virtually no dimples and only locally cleavage facets. Most of the fracture surface shows intergranular failure; therefore, the low impact energy data are not surprising. At 2% Mn, the proportion of this undesirable failure mode is lower, and at 1% Mn, ductile and cleavage fracture are observed – but at a significantly lower hardness –, which shows the huge effect of the Mn content.
Impact fracture surfaces of PM steels with 0.6% admixed carbon, compacted at 700 MPa, sintered for 60 minutes at 1250 °C in N2-10% H2. (a) CrA-0.6C (O content = 140 µg/g, IE = 37.6 J·cm−2, 172 HV30). (b) CrA-3Ni-0.6C (O content = 90 µg/g, IE = 33.3 J·cm−2, 362 HV30). (c) CrA-3Mn-0.6C (O content = 110 µg/g, IE = 5.8 J·cm−2, 407 HV30). (d) CrA-4MA-0.6C (O content = 230 µg/g, IE = 22.6 J·cm−2, 381 HV30).
The tendency of sintered Mn steels to intergranular failure has been described in the literature, for example, by Hryha 32 and Hryha et al. 33 for Fe–Mn–C. They observed it in particular when Mn was added as elemental powder – that is, as in the present case – and attributed it to the interaction of Mn vapour with the oxide layers covering the powder particles. These layers are transformed from iron oxides to significantly more stable Mn oxides that are difficult to reduce during sintering. It is not surprising that this effect occurs still more pronouncedly in the present case, since the oxide layers covering Cr prealloyed steels are thermodynamically more stable than iron oxides and thus are retained up to temperatures at which gas phase transport of Mn is very pronounced. Mn then results in at least partial manganothermic reduction of the Cr-containing oxides and thus in formation of oxide layers in the grain boundaries and sintering contacts, which are accordingly weakened. This theory is corroborated by the fact that in the steels into which Mn has been introduced through a masteralloy, this intergranular failure is not observed, because of the lower chemical activity of Mn in this alloying route. The total oxygen content is not the relevant criterion here; it is even higher in the MA types than in the plain Mn alloyed grades. It is rather the distribution of oxygen in the critical areas that is responsible for the different impact behaviour. From the combined carbon data, it is further evident that the deoxidation is primarily determined by the base powder particles, the carbon loss during sintering being in the range of 0.15% virtually regardless of the admixed alloying additives.
Functional properties
For characterising steels, also sintered steels, functional properties can be helpful (e.g. Cytermann 27 ), to gain insight into homogenisation processes as well as in the heat treatment state, possibly also for quality control. Therefore, electrical conductivity, magnetic saturation and coercive force were measured. For the conductivity, a clear effect of the alloying elements is visible: compared to the reference material, all additives lowered the conductivity, the more the higher the content was. Elemental Mn seemed to have a slightly higher effect than Ni or the MA, but here the effective alloy element content has to be considered (4% MA = 1.6% Mn–0.3% Si) as well as the lower density of the Mn alloyed variant.
The coercive force Hc agrees well with the hardness, at least for Mn and MA. For the Ni alloyed steels, the increase of Hc is lower than would be expected from the gain in hardness. In general, however, the qualitative correlation of mechanical and magnetic hardness stands out clearly. For the magnetic saturation, finally, the effect of Ni is rather low, being attributable primarily to the presence of retained austenite. For Mn and Si, the negative effect on the theoretical saturation as described, for example, by Exner 34 and Hoselitz, 35 which is much higher than that of Ni, can be assumed to be dominant.
Experimental results and discussion – Part 2: Effect of the carbon content
As stated above, PM steel sinter hardening grades require higher alloy element contents than oil hardening grades to compensate for the lower cooling rates. However, also the carbon content plays a significant role. This is known from wrought steels, for which not only the peak hardness but also the hardenability depends on the carbon content. For sintered steels, this relationship has been shown, for example, by Kalss et al., 36 with focus on Cr prealloyed steel grades.
In the present study, the carbon content in hybrid alloyed steels with selected content of admixed alloy components was investigated. In this case also the carbon-free variant was included. One focus of this test run was to check which minimum carbon content is required to result in sinter hardening under the cooling conditions applied here. The second target focused on the low ductility of well-hardening grades. It was to be checked if improved toughness might be attainable by lowering the carbon content without losing the sinter hardening capacity. To reveal the trends, the properties are shown in Figure 4, and the data are also listed in Table S2 in the Supplemental Material.
Properties of steels with different levels of admixed carbon, compacted at 700 MPa, sintered for 60 minutes at 1250 °C in N2-10% H2. (a) Green density; (b) sintered density; (c) dimensional change; (d) apparent hardness; (e) impact energy (unnotched); (f) electrical conductivity; (g) coercive force; (h) dynamic Young's modulus; (i) magnetic saturation; (j) carbon content and (k) oxygen content.
Dimensional properties
Both for green and sintered density, the effect of carbon is hardly visible for the plain Cr steel and the Ni alloyed variants, but slightly more for the Mn and MA alloyed types, respectively. In any case, it is markedly lower than that of the admixed metallic alloying components. In the case of admixed elemental Mn, the lowering of the sintered density compared to the green density is evident once more. This agrees well with the data for the dimensional change: the Mn variant is the only one showing expansion during sintering while all other grades shrink. Higher carbon content slightly increases swelling for the Mn type and reduces shrinkage for the MA variant, while shrinkage is increased both for the plain Cr steel and the Ni alloyed grade. This is particularly evident when comparing it to the carbon-free reference, which shows only very slight shrinkage. The reason could, in principle, be insufficient oxygen removal, but this disagrees with the oxygen content, which is surprisingly low. It can be assumed that the higher reduced sintering temperature (absolute temperature relative to the respective solidus temperature) in the carbon-containing variants plays a role here.
Mechanical properties and microstructures
Young's modulus
For the Young's modulus, once more, a correlation to the density can be observed; the effect of the lower density in the case of Mn alloying is evident. Furthermore, it is also visible that the slightly lower values registered for the Ni alloyed variant, as described in the previous chapter, are found only for 0.6% C admixed, but not for the lower C levels. This corroborates the hypothesis that the drop of E is related to retained austenite, which also agrees with the data for the magnetic saturation.
Hardness
The data for the apparent hardness document that all admixed additives result in sinter hardening effects, as indicated by the pronounced differences between the plain Cr steel CrA–C and the hybrid variants. The hardness levels attained, however, strongly depend on the carbon content. For 0.4% C admixed, the hardness levels range between 250 and 300 HV30, as compared to 150 HV for the Cr alloyed reference steel. For 0.5% C admixed the hybrid types are in the range 300–360 HV30; the hardening effect becomes more evident since the hardness of the Cr steel increases just to 160 HV30. At 0.6% C admixed, finally, the apparent hardness of the hybrid variants is in the range 350–410 HV30 as compared to just 170 HV for the reference.
Impact energy
The impact energy, as the strictest and most sensitive criterion for sintered steels, is highest for the carbon-free steel. This indicates that the reducing component of the atmosphere – hydrogen – has been sufficient to reduce the natural oxide layers covering the powder particles, which agrees well with the low oxygen content of this material. The C-containing plain Cr steel, as well as the Ni alloyed variant, are not much lower, with only a slight decrease in the values with higher carbon content. For the Mn alloyed steel, in contrast, low impact energy data are recorded for all carbon levels studied here. As expected, the impact energy values decrease with higher C content, but even for 0.4% C admixed, in which case the hardness is still moderate, the impact energy is below 20 J·cm−2, which is not a favourable combination of the properties. For the steel alloyed with the Mn-Si masteralloy, there is also a drop in the impact energy with higher C content, but at a significantly higher level, the values ranging between 23 and 33 J·cm−2. This once more confirms that adding Mn through the masteralloy route is much more effective than using elemental powder.
Metallography
Metallographic investigations (Figure 5) showed that the carbon-free reference material is plainly ferritic, as was to be expected, and the Cr steel containing carbon is pearlitic-ferritic, the ferrite fraction depending on the carbon content. As evident from Figure 2(a), at least 0.6% C admixed is required to attain a fully pearlitic microstructure. The hybrid alloyed steels with lower C content show mostly bainitic microstructures, which agrees with the lower hardness levels. For the Mn alloyed steel, some martensite is visible at 0.5% C admixed, which agrees with the fact that this material has the highest hardness at this carbon level. Comparing these microstructures with those shown in Figure 2 indicates that 0.6% nominal carbon content is required to attain a primarily martensitic microstructure, at least at the low cooling rates applied here. For mesh belt furnaces with a separate gas-quench facility, however, also the lower C levels might be sufficient for sinter hardening (though at the expense of less homogeneous microstructure as a consequence of the lower sintering temperatures possible).
Metallographic sections of PM steels with varying carbon content, compacted at 700 MPa, sintered for 60 minutes at 1250 °C in N2-10% H2; nital etched. (a) Astaloy CrA, 71.3 HV30. (b) CrA-0.4% C, 144 HV30. (c) CrA-3Ni-0.4C, 266 HV30. (d) CrA-3Mn-0.4C, 302 HV30. (e) CrA-3Mn-0.5C, 363 HV30. (f) CrA-4MA-0.4C, 277 HV30.
Fractography
Fractographic studies (Figure 6) showed ductile fracture with dimple structures for the carbon-free variant as well as the C-containing Cr steels and the Ni alloyed grades, regardless of the carbon content. For the masteralloy variant, a mix of ductile and cleavage fracture was observed (see also Figure 3(d)), the fraction of cleavage facets increasing with higher C content. For the Mn steel, some dimples as well as cleavage are observed at the lowest C content, but even in this material, intergranular fracture is observed locally, and the fraction of this undesirable failure mode increases with higher C content. This shows that the sensitivity of the steel alloyed with elemental Mn to grain boundary embrittlement, apparently caused by the ‘internal getter’ effect described by Gierl-Mayer et al., 37 is slightly diminished by lowering the carbon content, but the combination of hardness and toughness is still less attractive than with the other hybrid alloyed variants.
Impact fracture surfaces of PM steels with varying carbon content, compacted at 700 MPa, sintered for 60 minutes at 1250 °C in N2-10% H2. (a) Astaloy CrA (O content = 0.029%, IE = 55.0 J·cm−2). (b) CrA-4MA-0.5C (O content = 0.025%, IE = 30.3 J·cm−2). (c) CrA-3Mn-0.4C (O content = 0.017%, IE = 17.6 J·cm−2). (d) CrA-3Mn-0.5C (O content = 0.017%, IE = 13.4 J·cm−2).
Functional properties
When comparing the functional properties, conductivity, saturation and coercive force, it stands out clearly that in all cases the effect of the metallic alloy elements is much more pronounced than that of the carbon content, which indicates the stronger effect of substitutional additives compared to interstitials. This is especially noteworthy for the electrical conductivity, which is virtually constant for all carbon levels, while the differences between the alloying systems are clearly visible. In particular, the difference between the reference CrA–C and the hybrid types is pronounced, while the carbon-free and the carbon-containing CrA show virtually the same level.
Also, for the magnetic saturation, the effect of carbon is small compared to that of the metallic alloy elements, except for some stabilisation of retained austenite. For the coercive force, finally, the effect of carbon on this property is similar to that on the mechanical hardness, once more confirming the relationship between both properties.
Conclusions
Hybrid alloying, that is, combining Cr prealloyed base powders with admixed further alloy elements, has been shown to be a suitable route for producing steel grades with sinter hardening capability even at fairly low cooling rates. Compared to the plain base powder containing different amounts of carbon, the addition of Ni or Mn as elemental powders or of a Mn–Si masteralloy powder resulted in sinter hardening effects, that is, formation of predominantly martensitic microstructures. This occurred even at linearised cooling rates below 1 K/s if a certain additive content was exceeded, which is lower for Mn than for the other variants, as well as a sufficiently high carbon content.
However, it showed that adding elemental Mn, which is most beneficial for high sintered hardness, is problematic regarding the impact energy, because of a pronounced tendency to intergranular failure. Here it was shown that adding Mn via the masteralloy route is preferable to the elemental powder route, offering an attractive combination of hardness and impact toughness.
The functional properties, electrical conductivity and magnetic saturation, were found to be influenced much more by the metallic additives than by the carbon content, while the coercive force is roughly related to the mechanical hardness.
Conclusively, it can be stated that combining Cr prealloyed base powder with an Mn–Si masteralloy offers sinter hardening capability and attractive combinations of hardness and toughness without the cost and health problems associated with using nickel.
Generally, the masteralloy approach can be regarded as a highly flexible alloying route, in particular for introducing oxygen-sensitive alloy elements, avoiding the penalties both of elemental mixing – high reactivity to oxygen – as well as of full prealloying – poor compactibility. This offers benefits for the industry regarding both economic and health aspects. The attractivity of the masteralloy route is further enhanced by the availability of ultra-high pressure water atomisation on an industrially relevant scale. This technique results in a high yield of fine fractions <25 µm, which is well suited for alloying PM steels, and in surprisingly low oxygen contents. The Mn–Si masteralloys in particular result in property profiles well comparable to those of traditional low alloy steels but without the penalties associated with Cu and Ni, especially when combined with suitable base powders. This can be regarded as a potential solution to current problems of the ferrous PM parts manufacturing industry, overcoming which necessitates advances in economy, health and safety as well as performance.
Supplemental Material
sj-docx-1-pmg-10.1177_00325899251398417 - Supplemental material for Effect of admixed alloying components on the properties of sintered steels based on Cr prealloyed powder
Supplemental material, sj-docx-1-pmg-10.1177_00325899251398417 for Effect of admixed alloying components on the properties of sintered steels based on Cr prealloyed powder by Milad Hojati, Christian Gierl-Mayer and Herbert Danninger in Powder Metallurgy
Footnotes
Acknowledgments
The authors wish to thank Atomising Systems Ltd, in particular J.J. Dunkley, for preparing and supplying the masteralloy powder used.
Authors’ contributions
Herbert Danninger and Christian Gierl-Mayer: conceptualisation; Herbert Danninger and Milad Hojati: methodology; Milad Hojati: software; Herbert Danninger and Christian Gierl-Mayer: validation; Milad Hojati: formal analysis; Milad Hojati: experimental investigations; Herbert Danninger and Christian Gierl-Mayer: resources; Herbert Danninger, Christian Gierl-Mayer and Milad Hojati: data curation; Herbert Danninger: writing–original draft preparation; Herbert Danninger, Milad Hojati, and Christian Gierl-Mayer: writing–review and editing; Milad Hojati: visualisation; Herbert Danninger: supervision; Herbert Danninger: project administration; Herbert Danninger: funding acquisition.
Funding
The authors received no financial support for the research and authorship. Open access publication was supported by TU Wien.
Declaration of conflicting interests
The authors declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Prof. Herbert Danninger is Editor of Powder Metallurgy.
Supplemental material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
