Powder metallurgy is a continually evolving technology utilizing advanced compaction techniques and new alloy systems to optimize mechanical performance of the resulting powder-metallurgy (PM) part.
For many applications, the use of hybrid-alloyed FLN2-4405 powder is specified to meet the heat-treated part performance requirements. Development of a leaner, heat-treatable alloy powder enhances PM’s competitiveness by delivering a cost-effective PM component with reduced material cost. Several alternative PM steels were evaluated in both test specimens and automotive components, and multiple candidates were found to meet the required performance.
Powder-metallurgy steels are used in several components throughout the automotive industry, accounting for more than 40 pounds per average North American vehicle. These components include but are not limited to sprockets, connecting rods, bearing caps, VVT, oil pumps and mechanical diodes.
The steel PM process consists of water-atomizing iron alloys, annealing the powder to remove oxide, mixing the iron with graphite and other alloying powders, compacting components in a mechanical or hydraulic press, sintering components at nominally 1120°C (2050°F) to join the powder particles and diffuse alloying elements, and any secondary operations required. These secondary operations include heat treatment and limited machining because the parts are near-net shape after compaction/sintering.
Alloys are typically limited to Mo, Ni and Cu because powder oxygen control is difficult with more traditional steel alloying elements of Mn, Cr, Si and V during water atomization and annealing. Certain elements are preferentially added in the steel melt prior to atomization (pre-alloyed), whereas other elements are added as separate powders (admixed) during mixing. Iron powders must maintain good compressibility (low flow strength) so that the powders can be compacted to target densities. This further limits the composition of atomized steels.
PM steels pre-alloyed with Mo have been widely used in the industry for the last two decades. Molybdenum is a desirable alloying element for PM steel applications for a number of reasons, including its relatively minor effect on compressibility, high contribution to hardenability and an easily reducible oxide.
Figure 1 shows a plot of hardenability factor as a function of percentage alloying content. Mo has by far the greatest contribution to hardenability of the elements shown, with nearly 1.5x more benefit than Mn and 2x more than Cr at an addition level of 1.0 wt.%. A synergistic effect with Ni at levels above 0.75 wt.% (designated by Mo*) contributes an increase of approximately 25% more hardenability than at lower Ni levels.
This unique synergy between the two alloying elements has been successfully exploited in PM, and several grades were adopted by the industry, including alloy FLN2-4405. This particular alloy (0.85% Mo, 2% Ni, 0.6% C) offers a good combination of compressibility and physical properties, leading to its acceptance as one of the most popular hybrid-alloy PM steels. It is well suited for induction-hardening applications because the pre-alloyed Mo base alloy provides the necessary hardenability while the Ni addition provides good toughness.
Ongoing cost pressures in the automotive industry force a constant re-assessment of alloy selection and the need for leaner alloys. The approach used in this work is to reduce the current alloying elements to the lowest level that will still meet required properties. Powder grades with lower molybdenum content (0.3 and 0.5% Mo) have been introduced to provide increased flexibility in alloy selection. The chief concern with this approach is the reduction in hardenability. Fortunately, the rapid heating and cooling rates employed with induction hardening mitigate the need for high hardenability, and these lower Mo alloys have been found to harden quite well under oil-quench conditions.
The objective of the current work is to illustrate the performance capabilities of an alloy with a reduced Mo content under production conditions. A comparative analysis of a lean 0.3 wt.% Mo steel with 0.85 wt.% Mo and Mo-free steels was conducted. Mechanical-property data and metallography developed from the processing of a crankshaft sprocket are presented in the as-sintered, oil-quenched and tempered, and induction-hardened conditions at densities from 6.8-7.2 g/cm3.
Pilot-scale premixes of four alloys containing varying levels of Mo and Ni are shown in Table I. Steel powders containing 0.85, 0.3 and 0.0 wt.% pre-alloyed molybdenum were combined with Ni and graphite powders to produce the alloy premixes shown. All alloys contained 0.75 wt.% Lonza Acrawax C lubricant and Asbury 3203H graphite. Inco 123 was used as the Ni source.
Test specimens were compacted to densities of 6.8, 7.0 and 7.2 g/cm3 at room temperature. The samples were sintered in a production belt furnace at Cloyes Gear & Products for 25 minutes at 1120°C (2050°F) in 95N2-5H2 (vol.%). Crankshaft sprockets were also compacted from the four alloys at a nominal density of 7.0 g/cm3. Test specimens at 7.0 g/cm3 and sprockets were oil quenched (designated hereafter as Q&T) by austenitizing at 840°C (1550°F) for one hour with a 0.6% C potential (0.8% C potential was used for mix 4) and quenching in oil at 93°C (200°F). The second set of crankshaft sprockets were induction hardened by heating for 6 seconds at 28.4 kW on a 60-kW, 450-kHz induction machine and quenching in oil at 60°C (140°F). Heat-treated samples were tempered at 205°C (400°F) for one hour.
Percent dimensional change, sintered density and apparent hardness were measured from the transverse rupture samples using standard MPIF (Metal Powder Industries Federation) procedures. Tensile testing was performed using a crosshead speed of 0.065 cm/min (0.025 inch/minute). The machine is equipped with a 25 mm (1 inch) extensometer, which was left on until failure.
Impact testing was conducted at room temperature on un-notched Charpy samples. Rotating bending-fatigue data was collected on as-sintered and Q&T samples of mixes 1 and 2 using a speed of 8,000 rpm at a nominal density of 7.0 g/cm3. Specimens were machined from blanks processed under the conditions described above. The median fatigue endurance limit (50% FEL) was calculated by using the “staircase” method until failures and runouts occurred at a minimum of two stress levels.
Results and Discussion
Plots of tensile properties as a function of sintered density are shown in Figure 2 for the four alloys tested in the as-sintered condition. A summary of the remaining mechanical properties that were tested from 7.0 g/cm3 green bars is shown in Table 2. As expected, the higher pre-alloyed Mo level in mix 1 provided a significant increase in strength and hardness. Likewise, this alloy showed lower elongation and impact performance than the leaner mix 2.
Of interest to note is the dramatic effect of pre-alloyed Mo on performance compared to the presence of 2.0 wt.% admixed Ni. Mix 2 provides an increase in yield strength of 5-10% compared to a similar alloy with no admixed Ni (i.e., mix 3). An increase in the pre-alloyed Mo level from 0.3-0.8 wt.% results in a yield strength increase of 25-30%.
Because yield strength is directly related to microstructure, this property is attributable to the hardenability of the alloy. Referring to Figure 1, an increase in Mo from 0.3-0.8 wt.% is expected to increase the hardenability factor much more versus an increase in Ni from 0.0-2.0 wt.%. This expected increase correlates well with the data observed in Figure 2 and Table 2. A reduction in pre-alloyed Mo level from 0.8-0.3 wt.% increased ductility approximately 50%. As in the yield-strength data, however, 2.0 wt.% admixed Ni had only a minor influence on elongation.
The higher C level in mix 4 provided high-strength performance despite having no Mo present. Although this material had relatively high as-sintered yield and tensile strength, it also provided a fair amount of ductility and impact toughness.
Although the majority of PM parts that are fabricated from alloys such as these are typically heat-treated, this data was acquired to serve as a baseline to identify the relative hardenability levels of each of the compositions. Performance capabilities of the alloys under heat-treating conditions are shown in the following sections.
The four alloys evaluated in the as-sintered condition were also investigated under oil quench and tempering. Table 3 (online) summarizes the Q&T properties for each of the materials. Fatigue data for mixes 1 and 2 are shown in Table 4 (online). Figure 3 shows plots of yield strength, tensile strength and apparent hardness in as-sintered and Q&T conditions.
Under Q&T processing, mix 2 provides very similar static and dynamic properties compared to mix 1 despite the lower Mo content. Mix 4, however, which contains no pre-alloyed Mo, had yield strengths that were approximately 15-30% lower than that of the other three mixes. It is evident that even a small addition of Mo provides significantly improved strength for Q&T processing.
Important to note, however, is that this data was developed on test specimens with relatively small cross-sectional areas. For components with small distances between the case and core, alloys containing 0.3 wt.% Mo will perform similarly under Q&T processing to alloys containing 0.8 wt.% Mo. Larger components, however, will require a significantly higher hardenability in order to form martensite throughout the cross section. In these instances, higher Mo contents may be required.
Crankshaft Sprocket Analysis and Metallography
A crankshaft sprocket frequently used to evaluate new materials was characterized in the as-sintered, Q&T, and induction hardened and tempered conditions at a nominal density of 7.0 g/cm3. The component, used in the GM small-block V-8 engine, first debuted in 1967 and was used by the OEM until 1998. It was produced in a variety of displacements, as well as the 4.3L V-6 for a shorter duration.
The inverted-tooth chain that meshes with the tooth form on this type of sprocket requires high wear resistance. The high flank hardness for this sprocket is obtained by heat treating, typically via induction hardening. Figure 4 (online) shows a schematic of the sprocket and the location of the tooth tested for strength. A diagram of the sprocket in the test rig is shown in Figure 5.
Tooth strength and apparent hardness measured on the component teeth for all four mixes are shown in Figure 6. In general, Q&T processing provided the highest strength and hardness, followed by induction hardening and as-sintered. Apparent hardness levels were nearly identical for all mixes processed with Q&T. Tooth strength was found to vary with alloy chemistry.
Similar to the tensile properties in Figure 3, mixes 1 and 2 had comparable tooth strengths, while mixes 3 and 4 had slightly lower tooth strength under the heat-treating conditions. In contrast to the as-sintered tensile properties in Figure 2, however, tooth strengths for mixes 1 and 2 were equal despite the much higher hardness for mix 1. Table 5 demonstrates that all mixes with 0.6 wt.% graphite had the same microindentation hardness value (650-660 HV100) regardless of alloy content. Higher microindentation hardness values were achieved in mix 4, which contained 0.8 wt.% graphite.
Microstructures recorded near the component teeth surfaces under the various processing conditions are shown in Figure 7 (online). As would be expected, the microstructures in the induction-hardened and Q&T conditions are comprised of 100% martensite. This hard microstructure resulted in the dramatic increase in apparent hardness that was observed for all alloys (Fig. 6) compared to the as-sintered data.
In mixes 1, 2 and 4, bright Ni-rich regions exist. Mix 3, with no admixed Ni, is comprised of a homogeneous microstructure. Also of interest is the refined microstructure in the as-sintered condition for both mix 1 (due to higher Mo content) and mix 4 (due to higher C). This refined microstructure, relative to the coarser 0.3% Mo as-sintered mixes 2 and 3, resulted in the highest yield strengths (Fig. 2).
Apparent hardness depends on microstructure, carbon content and density. At a density of 7.0 g/cm3 for all alloys, a 100% martensite microstructure at the teeth surfaces results in the same hardness level under either the Q&T or induction-hardened process.
Tooth strength is dependent upon the microstructure in the junction area between the component teeth and hub, as can be seen in the fracture surface shown in Figure 8. The microstructure in this area is of critical importance because it is subjected to bending stress from loading of the tooth.
Figure 9 (online) shows microstructures recorded for mixes 1-4 in the Q&T condition in such areas. These images clearly indicate the importance of alloy hardenability on tooth strength. Mixes 1 and 2 are fully martensitic in Q&T processing and thus have identical tooth strengths (Fig. 6). Because mixes 3 and 4 both contain large amounts of bainite, these two mixes have significantly lower tooth strengths.
Similar observations can be made for the induction-hardened condition. The mixture of divorced pearlite and martensite in mixes 1 and 2 provide higher tooth strengths than the ferrite-containing mixes 3 and 4, although the strength levels were lower relative to the Q&T condition.
These results indicate that the more economically favorable mix 2 can serve as a suitable replacement for mix 1 in both induction-hardening and Q&T applications. Furthermore, the addition of Ni is critical to enhance hardenability and toughness (mix 2 compared to mix 3). In situations in which wear resistance (i.e., apparent hardness) is the main criteria, mixes 3 and 4 are more inexpensive alternatives that provide similar apparent hardness compared to the alloys containing both pre-alloyed Mo and admixed Ni. The microindentation hardness of mix 3 could be increased by adding 0.2 wt.% graphite. The economic benefit of mix 3 versus 4 is dependent upon fluctuations of raw-material prices (Mo and Ni).
A comparative analysis of 0.3 wt.% pre-alloyed Mo steel and more conventional 0.85 wt.% Mo and Mo-free steels was performed. A crankshaft sprocket was used as a prototype component. The study demonstrated that components with small cross sections need only 0.3 wt.% Mo to provide desired strength and wear-resistance properties in the heat-treated condition.
In heat-treat applications in which wear (i.e., apparent hardness) and tooth strength are both critical, the more economically favorable Fe-0.3Mo-2Ni-0.6gr (mix 2) was found to have performance levels comparable to Fe-0.8Mo-2Ni-0.6gr (mix 1). In situations in which wear resistance is the main criteria, Fe-0.3Mo-0.6gr (mix 3) and Fe-2Ni-0.8gr (mix 4) are more cost-effective alternatives that provide similar apparent hardness compared with alloys containing pre-alloyed Mo and admixed Ni.
The comparative economic benefit of these two alloys is dependent upon fluctuations in Ni and Mo raw-material prices. Listed below are several other findings:
- Q&T provided higher apparent hardness over the induction-hardening process for all materials.
- The crankshaft sprocket tooth strength of a Fe-0.3Mo-2Ni-0.6gr steel (mix 2) is comparable to Fe-0.8Mo-2Ni-0.6gr (mix 1), regardless of processing (as-sintered, Q&T or induction hardening).
- Hardness levels were the same for all alloys when using either Q&T or induction hardening.
- The tooth and tensile strengths of a Fe-2Ni-0.8gr steel are significantly lower than that achieved by compositions containing pre-alloyed Mo for all processing conditions.
- Addition of Ni to a 0.3 wt.% Mo pre-alloyed steel is critical to enhance hardenability and toughness. Fe-0.3Mo-2Ni-0.6gr (mix 2) had significantly higher tooth and tensile strengths compared to Fe-0.3Mo-0.6gr (mix 3) in all three processing conditions.
The authors thank these individuals: Steve Vogt, Gilbert Schluterman Jr., Ed Elsken and Dewayne Askins of Cloyes Gear; Ken Gick, William Bentcliff, Steve Kolwicz and Gerald Golin of Hoeganaes.
For more information: Contact Bruce Lindsley, director, Advanced Engineering Materials, Hoeganaes Corporation, 1001 Taylors Lane, Cinnaminson, NJ 08077; tel: 856-829-2220; e-mail: email@example.com; web: www.hoeganaes.com
- A. F. deRetana and D. V. Doane, “Predicting the Hardenability of Carburizing Steels,” Metals Progress, p. 65, 1971
- B. Lindsley and H. Rutz, “Effect of Molybdenum Content in PM Steels,” Advances in Powder Metallurgy & Particulate Materials, compiled by R. Lawcock, A. Lawley, and P. McGeehan, MPIF, Princeton, NJ, 2008, part 7, p. 26