Effect of tempering temperature and subzero treatment on microstructures, retained austenite, and hardness of AISI D2 tool steel

The presence of retained austenite in the hardening process of tool steel often causes the lower hardness compared to the hardness requirements and poor dimensional stability in the tool steel. The purpose of the present research is to determine the relationships between the tempering process with and without cryogenic treatment to the hardness and retained austenite amount of as-hardened D2 tool steel. The austenitizing temperature was 1020 °C, the tempering temperatures have variations of 180 °C, 280 °C, 380 °C, 480 °C, and 580 °C, and the subzero treatment has a temperature of −172 °C, followed by tempering at 180 °C, 380 °C, and 580 °C. This study aims to determine the appropriate treatment to obtain a minimum retained austenite percentage to prevent and mitigate the failure of AISI D2 tool steel in the industrial application process. An optical microscope with image processing software (Image-J analysis), as well as Brinell and Vickers hardness testing, is the characterization method used in this work. In general, plate martensite, bainite, retained austenite, and primary and secondary carbides are the phases contained in the microstructure. Tempering temperatures have the effect of increasing the secondary carbide precipitation and decreasing the retained austenite content (γr 3,671%–2,769%). However, the cryogenic treatment can provide a more efficient martensitic phase transformation process and minimal retained austenite content (γr 2,257%–1,199%). The increase in tempering temperature causes a decrease in hardness at a temperature of 180 °C–380 °C. On the other hand, the secondary hardening and phase transformation phenomena cause an increase in the hardness of the as-tempered sample at a temperature of 480 °C, before the sample reexperiences a significant decrease in hardness at a temperature of 580 °C due to diffusion that decreases the carbon content.


Introduction
The advancement of science and technology in various fields of life has developed in an unprecedented manner. Nevertheless, the need for higher precision and quality of tools to achieve the higher demand for manufacturing dies and molds has brought the need for research in the steel and manufacturing industries. The steel tool that is commonly used as a manufacturing die and mold has long been known as the high-quality manufacturing tool steel.
Using tool steel in manufacturing steel, containing high percentage of alloying elements and carbon to be exact, consequently results in the presence of retained austenite. The presence of retained austenite is due to the incomplete transformation of the austenite phase to the martensitic phase during a fast cooling rate (quenching) in the hardening process [1]. The presence of alloying elements such as molybdenum, chromium, vanadium, and tungsten in tool steel lowers the temperature of the martensite start (Ms) and martensite finish (Mf), which are the temperature needed to be reached for the austenite phase to completely transform into the martensitic phase. From the other side, during quenching process, materials should be fast cooled from austenite to room temperature, to avoid cracking due to shock temperature. That is why materials transformation of austenite to martensite does not go to completion when quenching to room temperature after austenitization because the low temperature of Ms . As a consequence, some amount of austenite is retained in the as-quenched microstructure, which otherwise consists of martensite (and sometimes to Bainite) and undissolved primary carbides [2]. To remove the retained austenite, further cooling from room temperature is carried out to reach the MS temperature, which is generally carried out below 0°C, which is called subzero treatment. Subzero treatment was carried out with liquid nitrogen quench media, with temperatures reaching −170°C [3].
Retained austenite is a soft and unstable phase which can undergo a transformation to a brittle martensitic phase, causing volume expansion and component distortion during the operating process in the industries [2,4]. Dimensional changes that occur in the tool steel due to the presence of retained austenite have become a crucial problem as they threaten the safety of steel materials used in forming parts as molds and dies. Furthermore, the presence of a significant amount of retained austenite can cause a major decrease in hardness and mechanical properties, simultaneously bringing unwanted and unmeasurable effects. During the application, the unstable retained austenite will transform to other phases such as martensite, pearlite, or bainite and cause a 4% volume change of material for transformation of retained austenite to martensite, which should be avoided [5]. Retained austenite transformation has been investigated by many researchers for a long time, but characterization and understanding are still a challenge [6]. Some teory explain that the retained austenite transforming into ferrite and cementite [6], martensite [7], or Bainite [8]. Ferrite and cementite form during isothermal treatment at the tempering temperature, while martensite transformations occur during cooling from the tempering temperature. It is some theory explaining that at the isothermal transformation, transformation of retain austenite to ferrite and carbide is a two-step process. First, precipitation of cementite occurs from retained austenite, which is gradually followed by transformation into ferrite and cementite [6].
Although the mechanisms are known, little information is available in the literature about the relation of austenization, tempering and subzero treatment to the percentage and the transformation of retained austenite, and its relation to the hardness of cold work tool steels, include the D2 tool steel. The present research was designed to improve the understanding of retained austenite transformation during tempering and cooling and its relation to the hardness of High Alloy steel.
Several methods can be used to measure the presence of retained austenite, including x-ray diffraction method (XRD) [9], metallographic method with optical microscopy (OM), or scanning electron microscopy (SEM) [10], as well as Mossbauer spectroscopy, dilatometry, and measurement of magnetic saturation (VSM) [11]. The transformation behavior during heat treatments was studied with Metallography optical microscopy combined with Image analysis by Image J software. These methods used to overcome the disadvantage of the XRD and other methods, some of which require very thin sample sizes, which require complex sample preparation, thereby eliminating the effect of the heat treatment performed [12].
In this method, the sample micrograph is compared to a series of reference micrographs. With the color etching applied, the color difference between the microstructures will be an indicator of the presence of a certain phase, retained austenite in this case. With the application of Image analyzer that can distinguish phase colors, equipped with Image J software, the percentage of retained austenite can be calculated accurately.
This research uses an AISI D2 high-carbon high-chromium cold work tool steel, which is a commonly used material as mold and die for forming applications. AISI D2 is most commonly used for press forming dies due to the combination of valuable properties and low cost [13].
This research also aims to study the effect of variations in tempering temperature on the microstructure, the percentage of retained austenite, and the hardness of AISI D2 tool steel. It also aims to compare the samples that received subzero treatment and those that do not receive subzero treatment on the microstructure, the percentage of retained austenite, and the hardness of AISI D2 tool steel. The scope of this research is limited to the material used as a specimen, i.e., the AISI D2 tool steel, as well as the treatments carried out on the specimen, including preheating, austenitizing, quenching, variations in tempering temperature, and subzero treatment.

Method
This work was conducted with the sample preparation that required a cutting tool and the sample marking tool to cut the sample into a size of 2 cm × 2 cm × 1 cm. The materials of the specimen used the AISI D2 tool steel, which corresponds to the JIS standard of DC11 tool steel. Further OES examination was performed to obtain the accurate chemical composition of the specimen, as listed in table 1.
The AISI D2 steel used in this research contains high alloying elements, simultaneously increasing its hardenability. The heat treatment process follows after the sample preparation. A furnace, a container, and a thermocouple are used in this process. The thermocouple is need to accurately measure the real temperature of the furnace, clamp, and agitation machine. The oil agitated with the agitation machine is the quench medium used in this work. After the heat treatment processes, metallographic testing was conducted to obtain the microstructure of the heat-treated sample. Further surface preparation was needed to reveal the microstructure. The tools needed for the surface preparation include containers, grinding and polishing machines, sandpaper with grit size of 80-1500, velvet polishing cloth, and image analyzer software (Image-J). Following the metallographic testing, an etching agent of Beraha's sulfamic acid no 4 (100-mL water, 3 g-K 2 S 2 O 5 , and 2-g NH 2 SO 3 H) was used to reveal the microstructure and further measure its phase content in percentage value. On the other hand, an etching agent of Nital 5% (5-mL nitric acid, 95-mL methanol) was used to give a clear identification of its microstructure morphology. The Brinell and Vickers hardness testing machine was used to conduct the hardness testing. The Brinell hardness testing was conducted according to the ASTM E10-18 standard, with a force load and dwell time of 3000 gf and 10 s, respectively. The Vickers hardness testing was conducted with a force load and dwell time of 100 gf and 10 s, respectively.
The overall heat treatment process involves preheating, austenitizing, quenching, and tempering. First, preheating was conducted with a temperature of 650°C and a holding time of 15 min to achieve a uniform temperature between the surface and the core to avoid thermal distortion. Furthermore, austenitization requires a temperature of 1020°C and a holding time of 1 h before rapid quenching to room temperature to obtain the as-quenched specimen. The as-tempered variables require the specimen to continue with the heat treatment process after quenching with varying temperatures of 180°C, 280°C, 380°C, 480°C, and 580°C, with a holding time of 1 h for each tempering variable. On the other hand, specimens with the requirement of subzero treatment or known as the subzero tempered variables were cooled with liquid nitrogen directly after the quenching process, with a holding time of 5 min, and continuing with the varying tempering temperatures of 180°C, 380°C, and 580°C, with a holding time of 1 h for each tempering variable. Figure 1 shows the heat treatment sequence graph to clearly describe the overall heat treatment process.

Microstructure observation
The tool steel in group D of D2 tool steel, which is used in this research, is supplied in the condition of as annealed [14]. Aside from that, the D2 tool steel has a tendency to form carbide, which was also caused by the  high amounts of carbide-forming elements. The microstructure of the steel specimen examined in this research without treatment revealed that using the 5% Nital as an etching agent revealed the presence of ferrite, pearlite, and chromium primary carbide phase, as shown in figure 2. In the as-received samples, Ferrite and Pearlite phases are present in D2 alloy steel, as a result of full annealing after the tool steel manufacturing process which is a standard process after the steel production process. Alloy carbide, in this case Chrome Carbide is a primary carbide formed since the beginning of the D2 steel production process. Figure 2 shows the Ferrite phase as a bright area between the dark cementite phases. The two phases are arranged together into one phase as dark pearlite phases. The white phase with irregular spheroid shape is alloy carbide phase, which in D2 tool steel is primary chrome carbide. Meanwhile in the as quenched sample, after the austenization and quenching processes, the microstructure of D2 tool steel changes to the Martensite (grey fine needles), retained austenite (light phases) and undissolved Primary Chrome Carbide (white irregular spheroid shape), as shown in figure 3. This is due to the Ferrite and Pearlite phases in the as received sample, transformed into austenite during austenization and then into martensite and the retained austenite during the quenching process. Some of Primary Chromium Carbide also still revealed in the as quenched sample.   Figure 3 shows the microstructure of the AISI D2 tool steel with 5% Nital for the as-quenched sample variable with 1000× magnification.
As shown in figure 3, the phase contents obtained with the hardening treatment of the as-quenched sample variable include the retained austenite, martensite, and Primary chromium carbide phases. However, the martensite obtained in this treatment variable showed a rather coarse structure compared to the as-tempered and subzero tempered samples as shown in figure 4. Furthermore, the presence of the chromium carbide phase itself is a primary carbide. The as-tempered and subzero tempered variables were also analyzed with 5% Nital etching agent to reveal its microstructure, as shown in figure 4, which shows a comparison of the sample without the subzero treatment and that which have undergone the subzero treatment.
The subzero treatment aims to lower the temperature of the specimen after going through the quenching process and to further reach the temperature of the martensite finish to have a full martensitic phase transformation process. Figures 4(a) and (b) shows the as-tempered and subzero tempered variables at a low temperature of the tempering process, i.e., 180°C. Both showed a significant amount of primary carbide and a less amount of secondary carbide, which usually forms in the higher tempering temperature [2]. Moreover, the microstructure morphology showed a similarity to the as-quenched treatment variable. The microstructure of the as-tempered and subzero tempered conditions shows the retained austenite among the martensitic phases as the matrix. As shown in figures 5(a) and (b), tempering at a temperature of 380°C reveals that there is mostly no retained austenite. It shows an initial forming of fine secondary carbide, especially with the lesser amount of primary carbide compared to the lower tempering temperature [6]. The same phenomena also found on research on Cold work tool steel with 5 wt% Cr, which on tempering on 500°C, was obtained phases with smaller carbides (light phase), almost no retained austenite and undissolved large primary carbides [2]. Figures 6(a) and (b) shows the microstructures obtained at a higher tempering of 580°C. Both showed a significant amount of secondary fine carbide, Bainite, as well as a finer martensitic phase structure compared to those shown in figures 4(a) and (b) of lower tempering temperature. This is in accordance with the theory that the transformation of retained austenite during the tempering process of cold work steel occurs in two stages, where the initial stage formation secondary carbide occurs from retained austenite, for the next stage it is gradually transformed into ferrite and cementite, which could be as the Bainite microstructure. Also, during tempering, the as quenched coarse martensite becomes fine tempered martensite [6]. Figure 6(b) shows the microstructure that has undergone the subzero treatment process. It showed a better distribution of secondary fine carbide. The presence of secondary fine carbide significantly increases the steel's mechanical strength owing to its small and hard characteristics. Aside from that, the subzero or cryogenic treatment also gives the effect of (i) homogen redistribution of fine alloy carbide, (ii) precipitation of carbide alloy elements with a very small size (<1 μm), and (iii) homogen redistribution of residual stress [15].

Macrohardness calculation
The macrohardness of the specimen was measured with five random points for each specimen variable to obtain a representative value for each of the variables. Figure 7 shows a graphical plot of the results obtained from the Brinell macrohardness test. The graph shows that the as-quenched sample is the hardest. The hardness of the as-quenched sample tends to be with the microstructures are the coarse martensite, which results from the rapid cooling process with the high level of supersaturated carbon compared to that after the tempering treatment. Tempering causes diffusion to decrease the carbon content in the steel. Martensite that is formed by rapid cooling has a coarse structure and tends to be brittle and hard. However, after going through the tempering heat treatment process, martensite will undergo a transformation into tempered martensite with a finer and neater martensite structure and better ductility. The increasing ductility concurrently decreases hardness during the tempering process, caused by the decreased level of supersaturated carbon [16]. In general, the graph of decreasing hardness in the as-tempered variable heat treatment is shown in figure 7, which it corresponds to the same trend shown by Chandler [17].
The result shown in figure 7 conforms with literature [17], indicating that the hardness will decrease with increasing the tempering temperature. On the other hand, the two graphs show the occurrence of the same phenomenon, i.e., an increase in hardness at tempering temperatures between 400°C and 500°C before proceeding with a significant decrease in hardness. This phenomenon is known as the secondary hardening phenomenon, which can occur in high-temperature tempering heat treatment of alloy steels containing elements such as Cr and Mo [18]. Figure 7 shows that the subzero treatment shows higher hardness for the same tempering temperature. This is consistent with the microstructural analysis that has been carried out, which shows that from the subzero treatment, mostly no retained austenite will be obtained, so that the amount of martensite and bainite is greater, and the distribution of secondary carbides is more homogeneous compared to the hardening process without subzero treatment, for the same tempering temperature. Therefore, the hardness of subzero D2 steel becomes higher.

Microhardness calculation
The micro-Vickers testing was conducted to confirm and identify the presence of a specific phase by comparing the hardness value obtained through the testing with the value stated in the literature. The testing was conducted by identifying the difference between the light-colored phase and dark-colored phase. The hardness of the martensitic phase ranged from 525 to 616 HV [19]. On the other hand, the hardness of the retained austenite phase ranged from 246 to 343 HV [20]. The study conducted by Wang et al found the hardness of the bainitic phase ranged from 550 to 750 HV [21]; however, Park et al found that the hardness of bainitic phase 490 HV and 482 HV [22].
The micro-Vickers test was performed at five random locations for each specimen variable to obtain a representative value of the microhardness. Figure 8 shows the result of the microhardness test performed on two variables of the specimens, i.e., asquenched and as-tempered. The result showed that the microhardness of the dark-colored phase is in line with the range of the martensite and bainitic phases. Therefore, it can be inferred that the dark-colored phase contains a mixture of martensitic and bainitic phases.
On the other hand, the light-colored phase is found to be in line with the range of the retained austenite phase. However, the light-colored phase exhibits a slightly higher hardness, due to the presence of the martensitic and bainitic phases that have not been successfully transformed. Therefore, the light-colored phase can be assumed as the mixture of the retained austenite phase with the combination of martensitic and bainitic phase which altogether cause a slight increase in hardness.

Phase percentage identification
The metallographic examination was performed to reveal the microstructure and to obtain the percentage of each specific phase. Beraha's sulfamic acid no. 4 color etch (100-mL water, 3-g K 2 S 2 O 5 , and 2-g NH 2 SO 3 H) was used to etch the specific phase and acquire color differences. It is especially useful for color metallography for the category of alloy steel [23].
Beraha's sulfamic acid no. 4 can accurately detect the presence of the retained austenite phase selectively. The color difference obtained from the color etching was then used to be analyzed with the color picker software (Image-J).
In general, Beraha's sulfamic acid no. 4 etches the martensitic phase with the color of blue and green and retained austenite with yellow. Chromium carbide is not etched, so it is colored white, as stated in the research performed by Vander Voort [24].
The etching results of Beraha's sulfamic acid no. 4 to color the retained austenite phase show a light brown color. Some microstructural literatures show the difference in the color of the phase produced from the color etching of Beraha's sulfamic acid no. 4 between martensite, residual austenite, and carbide phases. The analytical method used can first determine the content of the carbide phase by analysis with an Image-J software. After obtaining the percentage of the carbide phase, the identification of the percentage of the martensitic phase can be continued. After obtaining the percentage of martensite and carbide phases, the determination of the remaining austenite phase can be calculated by subtracting the percentage value of the entire microstructural area, i.e., 100% minus the percentage of carbide phase and percentage of martensitic phase, so that the percentage of the remaining austenite phase can be obtained.
In this research, the applied method is making the different color of retained austenite among the other phases, by threshold setting from the Image-J software. It can then be calculated as the area fraction of the retained austenite converted as a phase percentage.
To more clearly explain the phase identification, the variable of the as-quenched sample can be taken as an example etched with Beraha's sulfamic acid no. 4 with a 1000× magnification, as shown in figure 9.
As shown in figure 9, the microstructure of the as-quenched sample was etched with Beraha's sulfamic acid no. 4 reveals the presence of martensitic phase (colored green and dark blue) [25], retained austenite (colored yellow), and chromium carbides (white particles). This microstructure is similar with the previous literature [24]. Thereafter, obtaining the color difference through color etching, the specific phase percentage can be obtained using the image analyzer software (Image-J). Figure 10 shows the retained austenite percentage obtained.
The methods and results of phase analysis were conducted using OM and Image-J software with a magnification of 500× in a 50-micron scale and magnification of 1000× in a 25-micron scale.
The analysis method of the retained austenite phase was carried out by giving different colors to different microstructure photos, starting with giving different colors and analysis to the carbide phase, then to the martensite phase, and finally to the retained austenite phase. The other method involves making different colors to each phase and then analyzing with Image-J software, based on area analysis of the color of each phase.
On each variable that was analyzed with an optical microscope, three images were taken to obtain a representative value from one variable. From these three images, a quantitative analysis of the phases was then carried out, and the percentage of the phases was obtained using the method previously described to obtain the phase content of each variable. After obtaining the retained austenite phase content in the three images of each variable, it is possible to average the retained austenite content obtained and carry out comparisons and analysis with the literature.
Calculations based on color differences were carried out with an Image-J software. The data analysis method is shown in the table 2, where the retained austenite is indicated with the red color phase.   Table 2 presents the results of the retained austenite calculation and analysis of the overall phase percentage for the results of the OM analysis of all samples with the Image-J software based on the color difference presented in table 3. Figure 10 shows the average percentage of retained austenite obtained in various heat treatment conditions. As shown in the plotted graph of figure 10, the overall average percentage of the retained austenite phase obtained is lower than 5%, indicating a good result of this research, which aims to obtain the most suitable heat treatment process to minimize the presence of the retained austenite. The as-quenched sample showed the highest percentage of retained austenite, i.e., 4.015%. The as-tempered sample, with a lower tempering temperature of 180°C-380°C, showed a rather stable amount of retained austenite phase without a significant decrease until it reaches the tempering temperature of 480°C. Higher tempering temperature causes the retained austenite phase to undergo a transformation to other phases, such as martensite or bainite, therefore causing a decrease in the retained austenite percentage with the increase in tempering temperature. On the other hand, the subzero treatment gives a significantly lower retained austenite percentage compared to the other. This is majorly caused by the drastic cooling of the specimens altogether lowering its temperature and particularly causing a more efficient martensitic phase transformation.

Conclusion
Along with this research, a few things can be concluded, some of which are as stated below: 1. The microstructure phases obtained in the as-quenched, as-tempered, and subzero tempered sample variables are martensite, bainite, retained austenite, and carbide phase. However, variables involved in the tempering process showed a finer and structured martensitic phase, which is also known as the martensitic tempered phase. Furthermore, finer and more content of fine carbide was found with the increase in tempering temperature.
2. The retained austenite percentage of the as-quenched sample showed a value of 4.127%. As the tempering temperature increases, the amount of retained austenite phase decreases and hence has an inversely proportional relationship to the increase in the tempering temperature. The subzero tempered sample specimen exhibited the lowest retained austenite percentage of 2.769%, with a tempering temperature of 580°C.
3. The as-quenched variable has an overall highest value of hardness, which is caused by the martensite obtained from the rapid quenching, producing a hard and brittle martensite structure. The as-tempered specimen variable experienced a decrease in material hardness, which is caused by the diffusion of carbon elements from the supersaturated carbon. Nonetheless, there is an increase in hardness at the tempering temperature of 480°C, which is caused by the secondary hardening phenomenon and the transformation of the retained austenite phase.
4. The subzero treatment performed for the subzero tempered variable has found a minimum amount of retained austenite phase with a tempering temperature of 580°C with the value of 1.20%. The subzero treatment effectively reduces the residual stress and increases the efficiency of the martensitic phase transformation, therefore reducing the presence of the retained austenite phase. The optimum conditioned has found, that in variable of tempered 480°C without Subzero treatment, with high Hardness of 621 HRB And low percentage of retained austenite 3.39% HRB. With a minimum amount of retained austenite in Quench Tempered 480°C, and subzero tempered 580°C conditions, the risk of transformation of the retained austenite into other phases in the application will decrease, so that the dimensional stability of the material will be better, and the possibility of distortion and cracking during the hardening process will be reduced.