Tempering is a heat-treatment technique used to improve the strength, ductility and toughness of hardened carbon steels. Tempering occurs by heating the steel to below its critical temperature in order to transform the metastable body-centered tetragonal martensite structure that is formed during quenching into a more stable structure of fine carbide particles. Choosing the correct parameters for tempering is critical to achieving the desired balance of properties. It is important to understand the role tempering has on strength and toughness properties, as well as the microstructural changes experienced over a range of tempering temperatures. This study examined the mechanical performance of two different alloy systems commonly used to manufacture powdered metal (PM) gears. A pre-alloyed and a diffusion-bonded material were investigated after being heat treated and tempered at various tempering parameters.
The production of PM gears often requires a secondary heat-treat process in order to meet the requirements for high performance applications. Heat treating can optimize performance properties such as hardness, strength, fatigue, and wear resistance and extend the overall life of a gear. Depending on the specific requirements of the gear application, different types of heat treatments may be used. Some heat treatments focus on specific areas, for example the teeth, while others require heat treatment of the whole gear. While there are various types of heat-treatment processes used for PM gears, case hardening and through hardening are two common methods used throughout the industry.
Case hardening, also known as case carburizing, is a process in which low carbon steels are heated to their austenitizing temperature, typically between 850°C and 950°C (1,560°F – 1,740°F), in a carbon rich atmosphere environment. Due to the high solubility of carbon in austenite, the carbon is absorbed at the surface of the component. This high carbon layer is then quenched to form a martensitic case, typically in an agitated oil medium . The purpose of this method is to develop a case with good surface hardness while preserving the relatively soft but tough inner core. Case hardening is used for components that require high surface wear capabilities, good fatigue life, and shock load resistance.
Through hardening is a heat-treatment method typically reserved for components that will be exposed to axial stresses and that do not require particularly high surface hardness . Through hardening is typically conducted using a quench and temper process where the component is heated to its austenitizing temperature in a carbon neutral atmosphere, then rapidly quenched in a medium such as agitated oil. Unlike case hardening, through hardening uses the carbon within the material for hardening, with the goal of martensite formation throughout the whole component and not just at the surface.
The component size and geometry play a significant role in the overall amount of martensite developed in through-hardened components. The cooling rate for any component is directed by thermal conduction. For large components with thick cross sections, the surface of the component will be subject to a different cooling rate compared to the inner core where the cooling rate is limited . This limitation of cooling in the core may hinder the development of a fully martensitic component. In comparison, thin cross-sectional areas will see more consistent and faster cooling throughout the section, leading to easier transformation to fully martensitic components.
The purpose of any type of heat treatment is to improve the mechanical properties by changing the microstructure to martensite, which contributes to high strength and high hardness properties. During rapid cooling from the austenitizing temperature, there is insufficient time for carbon atoms to precipitate out of the crystal lattice. This causes the face-centered cubic (FCC) crystal structure of austenite to transform into a supersaturated solid solution of carbon trapped in a body-centered tetragonal structure . This structure is known as martensite. The rapid transformation introduces a large amount of dislocations within the crystal structure, which causes high levels of internal stress. This stress results in a very hard but extremely brittle material. In order to relieve the stresses, tempering is used as a technique to restore some ductility and toughness back into the material while giving up some hardness.
Tempering occurs when a material is reheated to a temperature below its eutectoid point for a specified amount of time, allowing for the rearrangement of atoms and precipitation of carbon to relieve the internal stresses and modify the martensitic structure. During the tempering phase, the rearrangement of atoms and the precipitation of carbon results in the arrangement of spherical carbides being dispersed within ferrite. This arrangement is known as a tempered martensite structure. Tempering for PM steels is typically carried out at temperatures between 150-595°C (302-1,103°F) . Precise control of the time and temperature during the tempering phase is important to achieve the desired final mechanical performance.
The hardenability of a material also plays an important role in achieving the desired properties. Hardenability is defined as the ability of a material to achieve a certain hardness level at a given depth after heat treatment . It is a measure of how easily a material will form martensite, and the depth in which martensite will develop when quenched. High hardenability materials will form martensite not only at the surface, but also throughout the core of a component. The depth of hardening is an important factor in a part’s toughness and fatigue performance and is largely influenced by the materials carbon level and alloying elements. The most common types of alloying elements used in the PM industry are molybdenum, nickel, manganese, chromium, and copper. The effect of various alloy elements and additions on the hardenability of a material are shown in Figure 1.
The diagram depicts a multiplication factor which describes the depth of hardening when adding a certain quantity of the alloying element. As seen from the diagram, molybdenum, chromium, and manganese have a strong influence on a materials hardenability, while the influence of nickel is much less. The choice of an alloying element is directed by the alloys ability to raise the hardenability of the material, the quantity needed, and also its ability to diffuse consistently throughout the material .
Regardless of the type of heat treatment a component receives, the material alloying method plays a significant role in the final properties. In pre-alloyed systems, addition of alloying elements during the melting process creates a chemically homogenous alloyed particle. Due to the homogeneity of the alloying content, a pre-alloyed material system typically generates a homogenous microstructure in the as sintered phase. An alternative method to introduce alloying elements is through a diffusion bonding process. In this process, the alloying elements are thermally bonded to the surface of the iron particle. This method provides the benefit of having the alloy additions without compromising the soft, easily compressible iron particle core. However, because the diffusion bonded materials are not one homogenous chemistry as a pre-alloyed material is, the as-sintered microstructural formation is heterogeneous, with various island-like phases which are dependent on the specific alloying element present in the area. In the heat-treated form, pre-alloyed and diffusion bonded materials are both capable of achieving a martensitic microstructure. Figure 2 depicts the microstructures of a pre-alloyed (a) and diffusion alloyed (b) base iron in the as-sintered condition.
The ever-increasing performance requirements of PM gears require a combination of high hardenability and good compressibility in order to achieve the densities needed to obtain high strength. The choice of alloy, alloying method, and component size all strongly effect the final gear properties.
In this study, two alloy systems with similar chemical identities, one pre-alloyed and one diffusion bonded material, were through hardened and tempered at various tempering parameters after conventional sintering. The purpose of this study is to investigate the response of alloying method and component size to heat treat and tempering conditions and to determine how the tempering temperature affects their mechanical properties.
One commercially available pre-alloyed base iron and one diffusion alloyed base iron commonly used to manufacture heat-treated PM gears were chosen for this study. The Astaloy® 85 Mo is pre-alloyed while the D.AB is diffusion bonded base iron. The alloy compositions are listed in Table 1.
Mixes were manufactured from the two alloys according to MPIF material designations. In the FLN2- 4405 the nickel is admixed in. In the FD-0205 mix, no additional alloying elements were added. Both mixes contain the same type and amount of graphite and lubricant. The chemical compositions are shown in Table 2.
Each mix was compacted into 10 mm (0.39 inch) x 10 mm (0.39 inch) x 75 mm (2.95 inch) specimens to a green density of 7.25 g/cm3. They were conventionally sintered at Vision Quality Components. All of the specimens were heat-treated at Bluewater Thermal Solutions, which is based in St. Marys, Pennsylvania. The heat treatment parameters are shown in Table 3. All sintered and heat treated specimens were prepared for mechanical testing in accordance with MPIF standards .
The heat-treated specimens were then tempered for one hour in air in various temperatures ranging from 160 to 275°C (325-525°F). The heat-treated and tempered specimens were evaluated for ultimate tensile strength, yield strength, impact energy, apparent hardness, microhardness and microstructure.
A secondary evaluation for mass effect was conducted on two different size puck specimens manufactured from the FLN2-4405 material. The mix was compacted into 100 mm (4 inch) diameter x 25 mm (1 inch) height pucks, and 40 mm (1.5 inch) diameter x 25 mm (1 inch) height pucks respectively. The puck specimens were all compacted to a green density of 7.25 g/cm3. The specimens were sintered in a 6-inch laboratory belt furnace at 1,120°C (2,050°F) for 30 minutes in a 90/10 N2/H2 atmosphere. The sintered pucks were then heat treated at Bluewater Thermal Solutions at the same parameters as shown in Table 3. The heat-treated pucks were tempered at the same temperatures used for the tensile and impact specimens. One heat-treated puck without tempering was also evaluated. The pucks were evaluated for apparent hardness, microhardness profile and microstructure. Phase mapping was completed on the 100 mm puck specimen tempered at 200°C (400°F) to determine the percentage of martensite at incremental distances from the part surface.
The tensile strength at each tempering temperature is shown in Figure 3. The pre-alloyed FLN2-4405 material resulted in overall higher tensile strengths at each tempering temperature compared to the diffusion bonded FD-0205 material. The tensile strength for each material decreases as tempering temperature increases. Both material systems follow a similar trend, where a sharp decline in tensile strength is observed as the tempering temperature reaches beyond 220°C (425°F).
The yield strength at each tempering temperature is shown in Figure 4. The pre-alloyed FLN2-4405 material resulted in overall higher yield strengths at each tempering temperature compared to the diffusion bonded FD-0205 material. The yield strength for both materials increases as tempering temperature increases, until the temperature reaches 250°C (475°F). When the temperature exceeded 250°C (475°F) the yield strength began decreasing.
The impact energy at each tempering temperature is shown in Figure 5. The impact energy of both material systems is similar within the tempering range from 160°C (325°F) to 200°C (400°F). As the tempering temperature increased above 200°C (400°F), the impact energy begins to drop significantly with each tempering temperature increment. ]
The apparent hardness is shown in Figure 6. Higher apparent hardness levels at all tempering temperatures were obtained from the pre-alloyed FLN2-4405 material when compared to the diffusion bonded FD-0205 material. The apparent hardness decreases as tempering temperature increases for both materials.
The microhardness is shown in Figure 7. The microhardness between the FLN2-4405 and FD-0205 material systems was similar at each tempering temperature. The microhardness levels decreased as tempering temperature increased.
In the mass effect study, a microhardness profile on the martensite was measured on the FLN2-4405 specimens. The profile was developed at 1 mm increments into the core of the components, shown in Figure 8. The microhardness decreased at each 1 mm increment below the surface. The microhardness also decreased as the tempering temperature increased. The untempered pucks resulted in the highest microhardness due to stresses developed on the matrix by the heat treatment.
The mechanical property evaluation shows that the pre-alloyed material system has higher ultimate tensile strength (∼3 percent), yield strength (∼5 percent), and apparent hardness (∼6 percent) at all tempering temperatures compared to the diffusion bonded material system. The diffusion bonded material has slightly higher impact energy compared to the pre-alloyed material. These results are in line with the MPIF standard 35 published data.
The tensile strength of both materials stays relatively constant at tempering temperatures between 160-220°C (300-425°F). A sharp decline in tensile strength occurs as tempering temperatures reach beyond 220°C (425°F), from where both materials measured a 7 percent drop in tensile strength from 220°C (425°F) to 275°C (525°F). In contrast, the yield strength steadily increases as tempering temperatures approach 230°C (450°F), before declining again as temperatures go beyond 250°C (475°F).
The impact energy shows a steady decrease at tempering temperatures above 200°C (400°F) for both material systems. Both material systems show a linear decrease in apparent hardness and microhardness over the full range of the tempering temperatures. Based on this study, the optimum tempering temperatures range between 220°C (425°F) and 250°C (475°F), where both materials achieve high tensile strengths while maintaining good yield strength, impact energy, apparent hardness and microhardness. A summary of the properties for each material system is shown in Table 4.
The microstructures of the tensile specimens tempered at 200°C (400°F) are shown in Figure 9. The FLN2-4405 microstructures at each tempering temperature consist of martensite with nickel rich austenite where the nickel was present. The FD-0205 tensile specimens also consist of martensite with nickel-rich austenite where the nickel was diffusion bonded.
In order to understand how the mass of a component affects the microstructure at each tempering temperature, a study was conducted using two different size puck specimens manufactured from the FLN2-4405 material. For this investigation, the pucks were heat treated and tempered using the same parameters as the specimens submitted to mechanical property testing. The microstructure of the surface and core was evaluated on the pucks and compared to the tensile specimens. The measurements of each specimen are shown in Table 5.
The quenching process during heat treatment relies on the transfer of heat to the quenching medium in order to rapidly cool the components. The mass effect of a component will determine how quickly a component is cooled, and various cooling rates may be seen throughout different sections of a component. A component with small cross sections will cool more rapidly compared to large cross sections. In the same respect, the outer surface layer of larger components will cool faster than the core, resulting in variations in hardness and microstructure the closer to the center of the part.
The microstructures of the core of the 40 mm and 100 mm pucks at different tempering temperatures after heat treat are shown in Figure 10. This study shows that the 40 mm pucks were able to achieve a fully martensitic microstructure, with no bainite observed in the core. The 100 mm pucks have a martensitic surface, however bainite was observed within the core after heat treatment (∼20 percent). The tempering has no effect on the microstructures. Comparatively, the microstructures of the tensile specimens with the smaller cross section also developed a fully martensitic microstructure for both material systems. All specimens showed a similar drop in microhardness from the surface to the core due to stress relief as a result of tempering. This study shows that there is no mass effect on components with a cross sectional area less than 1000 mm2, but a mass effect is observed on components with a cross sectional area of 2,500 mm2. The mass effect is due primarily to heat treatment, and not tempering temperature.
Figure 11 shows the phase mapping percentage of martensite at incremental distances from the part surface into the core on a 100 mm puck specimen tempered at 200°C (400 °F). The component is 100 percent martensite at the part surface and 1 mm below the surface, with no bainite observed. At 2 mm below the surface, a small amount of bainite is observed and the martensite percentage slightly decreases. At 3 mm below the surface, a large amount of bainite is present and the martensite percentage drops significantly. The core of the component measures approximately ∼80 percent martensite.
Figure 12 shows images taken by a scanning electron microscope to compare the differences between untempered martensite and the tempered martensite. The untempered martensite consists of thick, plate- like needles throughout its structure. This structure is a result of carbides being trapped in the crystal lattice during heat treatment (quenching), resulting in high internal stresses throughout the structure. As the tempering temperature increases, alterations occur within the microstructure as the carbon is precipitated out of the crystal lattice. The atoms rearrange and form dispersed, spherical carbides in the martensite. This alteration forms a new structure called “tempered martensite” and has lower internal stresses compared to the untempered martensite . The tempered martensite structure supports the mechanical properties obtained that as the tempering temperatures increase, the internal stresses decrease and also the strength, apparent hardness, impact energy and microhardness decrease.
- The pre-alloyed material system has higher ultimate tensile strength (∼3 percent), yield strength (∼5 percent), and apparent hardness (∼6 percent) at all tempering temperatures compared to the diffusion bonded material system.
- The diffusion bonded material has slightly higher impact energy compared to the pre-alloyed material.
- A sharp decline in tensile strength occurs as the tempering temperature goes beyond 220°C. Both material systems saw a 7 percent drop in tensile strength above 220° C tempering temperature.
- The yield strength increases as the tempering temperatures approach 250°C (475°F). It decreases as temperatures reach above 250°C (475°F).
- The impact energy of both material systems is similar within the tempering range from 160°C (325°F) to 200°C (400°F). As the tempering temperature increased above 200°C (400°F), the impact energy of FLN2-4405 decreased 24 percent, while the impact energy of FD-0205 material decreased 29 percent.
- The apparent hardness levels decrease linearly as the tempering temperature increases. The FLN2-4405 material exhibited higher apparent hardness levels at all tempering temperatures compared to the diffusion bonded FD-0205 material
- The microhardness of the tensile specimens decreases linearly as the tempering temperature increases. The microhardness levels of both FLN2-4405 and FD-0205 are similar.
- Based on this study, optimum tempering temperatures to obtain the highest ultimate tensile strengths while retaining good yield strength, impact energy and hardness levels are between 220°C and 250°C (475°F).
- The mass effect study of the FLN2-4405 specimens shows the microhardness exhibits a linear decrease as the tempering temperature increases. The microhardness also decreases at each 1 mm increment below the surface.
- The 40 mm puck specimens were able to achieve fully martensitic microstructures while the 100 mm puck specimens have a mixture of bainite and martensite within the core.
- For tempering, no mass effect can be observed. The mass effects are primarily seen during heat treatment (quenching).
- The surface of the 100 mm puck specimens tempered at 200°C is 100 percent martensite. At 3 mm below the surface, a significant drop in martensite content is observed. The core of the component is approximately ∼80 percent martensite.
- The microstructure of the untempered puck specimens consists of plate-like martensite needles. At the highest tempering temperature, the martensite contains dispersed spherical carbides typical of tempered martensite.
- The microstructure arrangements are consistent with the mechanical properties obtained. As the tempering temperature increases, the internal stresses decrease, and the strength, apparent hardness, impact energy and microhardness also decrease.
- F. Fillari, T. Murphy, I. Gabrielov “Effect of Case Carburizing on Mechanical Properties and Fatigue Endurance Limits of P/M Steels” Hoeganaes Corporation, USA, Borg Warner Automotive, USA.
- S. Saritas, R.Causton, B. James, A. Lawley “Effect of Microstructural InHomogeneities on the Fatigue Crack Growth Response of a Pre-alloyed and Two Hybrid P/M Steels” Proceedings for PM2002, Hoeganaes Corporation, USA, Gazi University, Turkey, Drexel University, USA 2002
- “Heat Treatment of Plain Carbon and Low-Alloy Steels: Effects on Macroscopic Mechanical Properties” Massachusetts Institute of Technology Department of Mechanical Engineering, Cambridge, MA 2004
- T. Digges, S. Rosenberg “Heat Treatment and Properties of Iron and Steel” U.S. Department of Commerce National Bureau of Standards Monograph 18 1960 Pgs 10-18
- S. Ropar, R.Warzel III, B. Hu “Martensitic PM Materials” North American Höganäs Proceedings for PM2016
- Dr. H. K. Khaira “Hardenability” Manit, Bhopal www.slideshare.net/RakeshSingh125/f46b-hardenability Nov.2013
- “Höganäs AB Handbook for Metallography” No. 6, Höganäs AB 2015
- Lindskog, P. “Controlling the Hardenability of Sintered Steels” Höganäs AB, Höganäs, Sweden
- MPIF Standard 35 Material Standards for PM Structural Parts. (n.d), MPIF
Reprinted from Advances in Powder Metallurgy & Particulate Materials—2018, ISBN 978-1-943694-18-1, © 2018 Metal Powder Industries Federation, 105 College Road East, Princeton, New Jersey, USA.