Market forces are challenging traditional compositions typically used in powder metallurgical alloys. Molybdenum (Mo), nickel (Ni) and copper (Cu) are the predominant alloying elements used in ferrous PM due to their low affinity for oxygen.
In addition, molybdenum has little effect on compressibility and copper rapidly alloys by way of a liquid phase at sintering temperatures. All three elements also increase the hardenability of steels, allowing for sinterhardening of parts produced with a combination of the elements. Price pressures are causing a re-evaluation of powder chemistries using these elements.
The challenge for alloy development is to advance alloying systems that optimise the balance between mechanical properties and overall production cost. Sinter-hardenable compositions play a key role in this regard. Research led by Bruce Lindlsey of US powder giant Hoeganaes Corporation looked at alternative alloys to those primarily used in the European market.
Traditional sinter-hardening PM steel compositions use comparatively high levels of molybdenum, nickel, copper and carbon to achieve martensitic microstructures in the as-sintered condition. In the past high cooling rates could not be achieved in the sintering furnace, and this led to only the most highly alloyed materials being used for sinter-hardening. With the development of accelerated cooling zones and the adoption of the technologies associated with them, lower alloy contents can be used.
When alloy prices are low, it is easier to use heavily alloyed materials, ensuring a martensitic microstructure forms regardless of section size and cooling rate in the part.
As price pressures force a re-assessment of alloy selection, it is more cost effective to invest in additional processing to reduce the content of high priced alloying elements. The added processing may include longer sintering times to fully use admixed ingredients, accelerated cooling, and an analysis of the actual cooling rate in each part to determine the lowest alloy content necessary for sinter-hardening.
There are several approaches that can be taken to address raw material costs. The first is to introduce lower-cost alloying elements, such as chromium and/or manganese. These oxygen-sensitive elements provide excellent hardenability, but may lead to higher processing costs associated with powder production and sintering.
One benefit of these more effective alloying elements is that lower carbon levels can be used while maintaining a martensitic microstructure. These lower carbon martensitic alloys provide lower dimensional variation and enhanced mechanical properties [1]. The chromium-containing Ancorsteel® 4300 and Ancorsteel 4300L (0.3% molybdenum) are examples of such alloys, where as-sintered martensitic microstructures are common with sintered carbon contents less than or equal to 0.6 wt% [2].
Another approach to lower cost sinter-hardening is the development of alloys that have intermediate levels of alloying elements. Earlier alloys used high levels of molybdenum and nickel as powder costs were relatively low compared to secondary heat treating steps.
Ancorsteel 737SH (MPIF FL-4800) has the combination of good compressibility and excellent hardenability, but at current price levels, the 1.25% molybdenum and 1.4% nickel prealloyed in the powder make it cost prohibitive. However, when processing larger parts or where accelerated cooling is not an option, slow cooling rates within the part require these high levels of alloying for sinterhardening. The diffusion alloyed materials containing 4% nickel and either 0.5% or 1.5%
molybdenum also have high cost, and given that the Ni is not prealloyed, do not take full advantage of the alloying elements present. Those parts producers that have the ability to cool components at higher rates than conventional cooling needn’t pay for extra alloying when a leaner alloy would suffice. With that in mind, a developmental alloy is being explored for lower cost sinter hardening.
This new prealloyed composition was water atomised and annealed to obtain a particle size distribution typical of that present in the industry. The developmental alloy was compared to alloys FL-4600 and FL-4800 in premixes containing 0.7 to 0.9% graphite, 0% to 2% admixed copper, and 0.75% EBS wax. TRS and dogbone tensile specimens were compacted at either 690 MPa or to a green density of 7.0 g/cm3.
The laboratory specimens were sintered in a belt furnace for 15 minutes at 1120°C in an atmosphere of 90N2-10H2 (vol%). Temperature was measured using a thermocouple embedded in a test bar. Time at temperature was measured when the sample was within 5°C of the set temperature. Two average cooling rates, measured between 650°C and 315°C, were obtained during the study: 0.7°C/sec, and 1.6°C/sec. A tempering temperature of 205°C for one hour was used for all samples. Dimensional change was measured from die size.
FL-4400 (0.85% Mo) has also been included for comparison. Mixes of base alloy with 0.7% graphite and 0.75% EBS wax were used. It can be seen that the new alloy has improved compressibility compared with FL-4800, while both alloys exceed the compressibility of alloy FL-4600. The compressibility of the FL-4400, which contains no nickel, is the highest. Given that the new alloy is essentially a lean version of FL-4800, it follows that its compressibility lies between that of FL-4400 and FL-4800.
Mixes of FL-4800 and the developmental alloy with 0.7% graphite and either nil or 1% Cu were tested at 690 MPa under conventional and accelerated cooling. It is clear from the data that alloy FL-4800 is a superior sinter-hardening alloy to the developmental alloy. Hardness values are higher for the FL-4800 at the slower cooling rate and full hardenability can be achieved without the addition of copper at the faster cooling rate. Nevertheless, the developmental alloy is fully hardened with the addition of 1% copper and accelerated cooling. If a cooling rate of 1.6°C/sec can be achieved in a part, Alloy 1 + 1% copper could be used in lieu of the more costly FL-4800.
In fact, it appears that the hardenability of the developmental alloy + 1% copper is similar to that of the FL-4800 with no admixed copper. At a given hardness, the mechanical properties of the two alloys are quite similar. The dimensional change of the alloys is also comparable.
The sinterhardening alloy MPIF FLC-4608 commonly used in North America contains the FL-4600 base material and 2% admixed copper and 0.9% graphite. The developmental alloy compares favourably to the FL-4600 in mixes with 2% copper. The apparent hardness is quite similar for the two alloys at the different carbon contents and cooling rates.
The growth is significantly higher for the FL-4600 alloys, in part due to the higher compaction pressure used to achieve the target green density of 7.0 g/cm3. The tensile properties were higher for Alloy 1 compared with the FL-4600, indicating that it may be possible to reduce the amount of admixed copper and/or graphite with Alloy 1 to achieve the same strength properties as the FLC-4608. Such a modified composition could also be used to produce similar dimensional change values as the FLC-4608 as well. As the two base alloys appear to provide similar mechanical properties, they could likely be substituted for each other in applications depending upon cost of the alloy system. Alloy 1 contains 0.4% more molybdenum and 1.3% less nickel than FL-4600, so the alloys will have similar costs when the molybdenum is roughly three times the price of nickel. The high price and volatility of alloying elements have hurt the PM industry, as many of the alloy surcharges cannot be passed onto the end user. In high-volume applications it would therefore be beneficial to approve multiple alloy compositions so that the part producer could choose the alloy and process that results in the lowest-cost production. This would help to mitigate the unpredictable alloy costs.
The microstructure of the FL-4800 alloy with 0.7% graphite (FL-4805) is fully martensitic when the sample was cooled at 1.6°C/sec, which agrees with the hardness value of 70 HRA (39 HRC). The developmental alloy is fully bainitic under the same conditions, while with the addition of 1% admixed copper, the microstructure is predominately martensitic with a small amount of bainite, estimated to be 5% of the microstructure. The bainitic microstructure results in good tensile properties at an intermediate hardness level. Although all samples were tempered in this study, the bainitic structure with the absence of any martensite may allow this alloy and cooling rate to be used in the untempered condition.
The microstructure of the FLC-4608 and Alloy 1 with 2% copper and 0.9% graphite are quite similar at both cooling rates. Both alloys are shown in the conventionally cooled (0.7°C/s) condition. The resulting microstructure is a mixture of high carbon martensite with retained austenite and pearlite / bainite. The amount of martensite is similar in both alloys, suggesting the hardenability is comparable. A more in depth study is required to quantify the hardenability of the developmental alloy.