A Study of Carbide Precipitation in a H21 Tool Steel

Transcription

1 , pp A Study of Carbide Precipitation in a H21 Tool Steel Meilinda NURBANASARI, 1,2) * Panos TSAKIROPOULOS 1) and Eric J. PALMIERE 1) 1) Department of Materials Science and Engineering, The University of Sheffield, Sheffield S1 3JD, United Kingdom. 2) Department of Mechanical Engineering, Institut Teknologi Nasional, Bandung, Indonesia. (Received on November 3, 2013; accepted on March 7, 2014) Carbide precipitation in a H21 tool steel during conventional heat treatment was studied. The aim of this work was to study the exact microstructure and for better understanding of carbide formation during double tempering process of the H21 tool steel. The steel was austenised either at C or C for 1 hour, and water quenched. Double tempering was performed at 650, 750 and 800 C for 1 hour with air cooling in the first and second temper for each austenising temperature. The results showed that the double tempered microstructure consisted of tempered martensite, lower bainite and carbides. The current study confirmed previous findings and contributed to existing knowledge that depending on the tempering temperature, the types of carbide formed during double tempering were M 2C, Fe 3C, M 6C and M 23C 6 carbides. The present study findings add substantially to our understanding of the carbide formation sequence in the H21 tool steel during double tempering. No secondary peak hardening was observed, and the highest hardness (505 HV) was obtained after austenising at C and double tempering at 650 C, which implies that the double tempering of the H21 tool steel should be carried out below 650 C. KEY WORDS: carbides; tempered martensite; heat treatment; electron microscopy. 1. Introduction * Corresponding author: meilinda@itenas.ac.id DOI: Tungsten hot work tool steels are widely used for diecasting dies and hot forging dies due to having high hot hardness, and strength and resistance to softening at high working temperatures. The ability of these tool steels to meet the property requirements is strongly affected by the presence of carbides. The type of carbides, as well as their formation sequence in the tool steels depends on the chemical composition of the latter together with the heat treatment conditions. A typical heat treatment for tool steels involves austenisation to dissolve the carbide networks that formed during solidification and tempering to improve the toughness. In tempering, non-equilibrium carbides are formed that are subsequently replaced by more stable carbides with increasing tempering time and temperature. 1 3) In particular, the secondary carbides that precipitate during tempering play an important role for the final mechanical properties of tool steels. In the group of tungsten hot work tool steels, the H21 tool steel is the most popular one. This tool steel is not resistant to water quench. 4) Despite the fact that previous researchers have reported the carbides formed during tempering of the H21 tool steel are the W 2C, M 6C, and M 23C 6 (M stands for W, Fe, or Cr) carbides 5,6) and the hardness of steels after tempering was strongly correlated to the formation of carbides, 7) however the exact microstructure analysis with different heat treatment process still needs further investigation for better understanding of the carbide precipitation to control final properties of the H21 tool steel. Hence, the objective of the work presented in this paper was to examine the effect of different austenising and double tempering temperatures on carbide precipitation and hardness in a H21 tool steel. Three different tempering temperatures were selected for double tempering process. The selected tempering temperature (650 C) corresponds to maximum temperature expected in use and the two higher tempering temperatures (750 and 800 C) were selected to improve the toughness of the tool steel. 2. Experimental Method The chemical composition of the investigated H21 tool steel was 0.3C-0.3Si-0.3Mn-3.1Cr-7.5W-0.4V-0.3Ni-0.03P S and Fe-balance (wt-%). This tool steel was melted using a vacuum induction laboratory furnace with air cooling to produce an ingot with a size of cm 3. The samples for heat treatment were taken from the middle top of the ingot and their size was 1 cm 3. The schedule for conventional heat treatment can be seen in Fig. 1. The austenisation of samples was carried out to homogenise the as cast microstructure and to dissolve the primary carbides. The samples were either austenised at T γ = 1100 C or 1250 C for one hour at each temperature under an argon atmosphere and were water quenched. Afterwards, the as quenched samples from the respective austenising temperature were double tempered isothermally at 650, 750 and 800 C for 1 hour holding time with air cooling in between the first and second temper treatments. Optical microscopy was carried out using a MET Polyvar microscope. The etchant was 5% picral plus 2% HCl and some drops of tee ISIJ

2 pol. The SEM studies were conducted on a JEOL 6400 equipped with EDS and INCA software and operated at 20 kv. The samples for TEM studies were thinned by twin-jet electro polishing using a solution of 5% perchloric acid, 35% butoxyethanol and 65% methanol. The TEM work was performed on a Philips 420 microscope operated at 120 kv and a JEOL 2010F operated at 200 kv. The hardness of all samples was measured using the CV Instrument Vickers hardness tester with a 10 kg load and a 15 s dwell time. Six measurements per sample were made to calculate mean values and standard deviations. The average diameter and the volume fraction of the carbides (V c) were measured in every condition using ImageJ Fig. 1. Schedule of conventional heat treatment. software without carbide differentiation. Average equivalent diameters of the carbides were calculated by the length of the major and minor axes. Only carbides larger than 0.1 μ m were taken into account and more than 500 carbides for every condition were quantified to ensure reliability of the measurements. XRD was also used to confirm the phases using a Siemens D5000 diffractometer and Co radiation. The samples were scanned in the 2θ range 30 to 130 with step size 0.02 and counting time 1 per minute. The peaks were identified using the STOE WinXPow software program and ICDD (International Centre for Diffraction Data) files. 3. Results and Discussion 3.1. As Quenched Microstructures Figures 2 and 3 show SEM images and EDS spectra of matrix and carbides of the as quenched H21 tool steel. After quenching from austenising at 1100 C (Fig. 2), the martensitic matrix was not clear and the microstructure was dominated by a high volume fraction of carbides dispersed within the matrix. At the higher austenising temperature (Fig. 3), the martensite was very clear and the carbides had coarsened. The austenising temperature (T γ ) did affect the dissolution of the carbides, as at the higher austenising temperature, the diffusion rate of the carbide forming elements increased and as a result more carbides dissolved Fig. 2. Secondary electron image of the H21 tool steel after water quenching from T γ = C and EDS spectra of matrix and M 6C carbides. Fig. 3. Secondary electron image of the H21 tool steel after water quenching from T γ = C and EDS spectra of matrix and M 6C carbides ISIJ 1668

3 Table 1. SEM-EDS data of the as quenched H21 tool steel (wt%). Element T γ ( C) Phase Comment Fe Cr V W Matrix 87.1 ± ± ± ± 1.3 Martensite Carbide 60.2 ± ± ± ± 7.0 M 6C * Matrix 88.0 ± ± ± ± 0.9 Martensite 1250 Carbide 41.9 ± ± ± ± 5.8 M 6C * * The quantitative data should be considered with caution given the size of the carbides. Fig. 4. Isopleth phase diagram of the H21 tool steel taking into account six alloying elements 0.3C 0.25Mn 3.2Cr 0.3Ni 7.6W 0.4V (wt%). The dashed line indicates the investigated tool steel. enriching the matrix with alloying elements. 8) Figure 4 shows calculation of phase diagram of the H21 tool steel using ThermoCalc software (TCEF6 database) to support the microstructural analyses. The calculated phase diagram for the investigated tool steel (Fig. 4) shows that at the lower austenising temperature (1100 C), the stable phases are γ and M 6C carbide. At the higher austenising temperature (1 250 C), the only stable phase is γ, which indicated that austenising temperature at C was above the carbide dissolution temperature. It can be said that increasing austenising temperature will make the steel temperature close to the carbide dissolution temperature. However, in this study, a high volume fraction of primary carbides still existed, even at the higher austenising temperature, which indicated that the carbides did not dissolve completely during austenisation. Comparison between prediction and experimental results shows that the calculated phase diagram was not in agreement with the solidification microstructure of the H21 tool steel. It is well known that the phase diagrams do not provide details of phase transformation mechanisms under actual kinetic conditions. Figures 2 and 3 also suggested that there was only one type of coarse carbide formed in the as quenched H21 tool steel. This was confirmed by back scatter electron imaging, which identified them as M 6C carbides with their typical EDS spectra showing the high W and Fe peaks. The morphology of the M 6C carbide was rod-like, spherical and irregular, and the higher the austenising temperature, the lower was their volume fraction. At the austenising temperature of 1100 C, a high volume fraction of fine spherical M 6C carbides existed in the matrix, Fig. 2, which was attributed to the formation of secondary M 6C carbides. The presence of secondary M 6C carbides in the as quenched steel from C was not observed with SEM-EDS. Therefore, two categories of the M 6C carbides existed in the as quenched H21 tool steel for the C austenising temperature, one was the undissolved primary carbide with larger size and the other one was secondary carbide with finer size that formed during austenisation. The formation and growth of carbides is controlled by diffusion. At the lower austenising temperature nucleation of secondary carbides was possible, thus there were many more carbides but their growth was slower and as a result there was a high volume fraction of secondary fine carbides but with small size. These fine secondary carbides were far from the equilibrium condition and less stable due to having higher surface to volume ratio than the larger carbides and hence they dissolved easily at the higher austenising temperature. The primary M 6C carbides grew and became coarser at the higher austenising temperature and did not dissolve completely during austenising at C, presumably because of the short austenising period. SEM-EDS quantitative data of the matrix and carbides is given in Table 1. The alloying content in carbides was affected by the austenising temperature, as can be seen in Table 1. The M 6C carbide was rich in W, and its W content increased with increasing austenising temperature. With respect to the M 6C carbide, which was around 0.5 μm in average, interference between the interaction volume (~1 μm spatial resolution at 20 kv) and the matrix occurred when the carbides were analysed. Therefore, the analysis for the M 6C carbides could have shifted to that of the matrix depending on the size of the carbides. The EDS result for the matrix was more accurate. The composition of the matrix was in good agreement with that calculated using ThermoCalc TCFE 6 database (90.7Fe, 3.1Cr, 0.6V, 5.6W (wt%) at T γ = C and 88.5Fe, 3.1Cr, 0.7V, 7.7W (wt%) at T γ = C). The microstructure of the H21 tool steel was also studied by TEM. Figure 5(a) provides strong evidence that the microstructure of the H21 tool steel after austenising and water quenching consisted of lath martensite and carbides. The martensite was heavily dislocated. The width of the martensite laths varied between 0.2 to 0.6 μ m, and they were slightly parallel to each other. The morphology of martensite did not change significantly as the austenising temperature increased from 1100 to C, which would suggest that high austenising temperature cannot change the type of martensite. The TEM studies also confirmed that lower bainite was present after austenising at C and water quenching (Fig. 5(b)). The phase transformation during cooling from the austenising temperature led to the formation of lower bainite in which fine carbides formed in one direction across the laths and at an angle of 60 to the lath axis. The internal stresses that built up during the transformation from austenite to martensite accelerated the formation of bainite. 9,10) It should be noted that Fig. 5 was taken using TEM JEOL ISIJ

4 Fig. 5. Bright field TEM images of the as quenched H21 tool steel showing (a) the presence of lath martensite, (b) lower bainite (indicated by arrow) and the MC carbide (indicated by circles), and (c) diffraction pattern of the MC carbide. 2010F. The clear diffraction patterns (Fig. 5(c)) arising from the MC carbides and actually there were extra faint spots from the matrix. Those diffraction patterns were related to the smallest area for SAD, which is limited by the size of the aperture and the spherical aberration present in the lenses of the microscope (~500 nm). The MC carbide exhibited a spherical morphology. According to Karagoz and Fischmeister, 11) the rate of carbide dissolution during austenisation increases in the sequence MC, M 6 C, M 7 C 3, M 23 C 6. The MC carbide is thermodynamically stable and does not completely dissolve even at high austenising temperature. 8) Therefore, it is suggested that the MC carbide in the as quenched microstructure after austenisation was that of an undissolved primary carbide Double Tempered Microstructures It is strongly recommended to temper the quenched tool steels at least twice to improve their toughness and the thermal stability of their microstructure. 4) Figure 6 shows the microstructures of the H21 tool steel after double temper at different temperatures. Optical microscopy did not reveal any significant changes in the microstructure after double tempering, even at 800 C, Fig. 6. The prior austenite grain boundaries (PAGBs) were not revealed as they were covered by carbides. In order to study the effect of double tempered condition on the matrix and the precipitation of carbides, a TEM study was carried out. The bright field TEM images (Figs. 7, 8 and 9) showed that the carbides precipitated along the lath martensite boundaries (interlath) and within laths (intralath). The nucleation of most carbides was heterogeneous and took place preferentially at lath boundaries and PAGBs, as these are energetically favorable sites. The carbides exhibited rodlike, spherical and irregular morphology. It is apparent from Fig. 7 that the double tempered microstructure of the H21 tool steel at 650 C from both austenising temperatures consisted of lath tempered martensite, and carbides. The lath martensite morphology was still present with a low dislocation density in some areas, and coarse needle or rod bar cementite carbides were located in the interlath and intralath boundary. A tempered lower bainitic microstructure was also observed after austenising at 1100 C and double temper at 650 C, as indicated by the major axis of the cementite forming parallel arrays at about 60 to the axis of the bainitic lath. 12) The thickness of the cementite carbides was finer at the lower austenising temperature. The double tempered microstructure at 650 C showed that the dislocations formed networks through recovery, and this restoration seemed to be more effective at the higher austenising temperature. The microstructure in Fig. 7 indicated that the recovery process occurred heterogeneously as the dislocation density was still high in some areas, especially after austenising at C and double tempering at 650 C. During tempering, the recovery process in the matrix was associated with the migration of boundaries containing relatively low dislocation density (i.e., lath boundaries) and dislocation cells toward higher dislocation density boundaries. 13) The partial recovery of lath martensite during double temper at 650 C was attributed to incomplete lath boundary migration owing to insufficient time for completing the recovery process. Furthermore, it seemed that the presence of secondary carbides (M 2 C carbides) within the matrix delayed the recovery process during tempering by impeding the movement of dislocations through a pinning effect (see also Fig. 11 for the presence of the M 2 C carbide). Figure 7 also indicated that the recovery area at the lower austenising temperature was less than at the higher austenising temperature because the volume fraction of fine lath boundary carbides qualitatively was greater at the lower austenising temperature than at the higher austenising temperature and as a result the dislocation movement became more difficult at the lower austenising temperature. Figure 8 shows TEM images after double tempering at 750 C. It can be seen that the morphology of interlath boundary carbides changed as the continuous rod-like carbides tended to split in spherical or oval shape ones and the fine lath boundary carbides dissapeared. The lath martensite coarsened and tended to dissapear, especially at the higher austenising temperature until the very late stages of tempering, when equiaxed grains gradually developed. The martensite laths also exhibited an extensive reduction in dislocation density, which was attributed to the dissolution of fine lath boundary M 2 C carbides; and, as a result, the dislocation movement accros the lath boundaries becomes easier. It is suggested that the carbides made a significant contribution to the recovery of the lath martensite matrix. The double tempered microstructure at 800 C, Fig. 9, shows that the lath morphology of the matrix had dissapeared and equi ISIJ 1670

5 axed grains were formed with clear boundaries. At the lower austenising temperature of 1100 C, the carbides coalesced on the grain boundaries. Larger spherical carbides were observed at the prior austenite grain boundaries and inside the grains at both austenising temperatures. The carbides that were located along the grain boundaries were larger than those within the grain interiors, which would suggest that the stable carbides nucleated first in the grain boundaries and grew rapidly. The TEM study also indicated that the growth rate of the carbides was faster at the grain boundary. 14) The formation of alloy carbides to replace cementite carbides occurred by nucleation at PAGBs that are energetically favorable nucleation sites. 15) It can be said that the size and morphology of lath boundary carbides and the mor- Fig. 6. Optical images of the H21 tool steel with different austenising (T γ) and double tempering temperatures (T DT). Fig. 7. Bright field TEM images of the H21 tool steel after austenising at different temperatures and double tempering at 650 C. Fig. 8. Bright field TEM images of the H21 tool steel after austenising at different temperatures and double tempering at 750 C. Fig. 9. Bright field TEM images of the H21 tool steel after austenising at different temperatures and double tempering at 800 C ISIJ

6 Fig. 10. Secondary electron images of the double tempered H21 tool steel showing morphology of the M 6C carbides with different magnification. Table 2. SEM-EDS analyses* of the M 6C carbides in the H21 tool steel with different austenising and double temper temperatures (wt%). Chemical composition of the M 6C carbide T γ ( C) T DT ( C) Fe Cr V W ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± 3.4 * The quantitative data should be considered with caution given the size of the carbides. phology of the matrix were a function of the tempering temperature Identification of Carbide in the Double Tempered Microstructure The type of carbide in the double tempered H21 tool steel was identified by SEM EDS, TEM-EDS and selected area electron diffraction. a. M 6C Carbide The M 6C carbide was observed in all double tempered microstructures of the H21 tool steel as can be seen in Fig. 10. Back scatter electron imaging showed that the carbides exhibited only one contrast and were identified by EDS as M 6C carbides. This M 6C carbide had spherical, irregular and rodlike morphologies ranging in size between 20 to nm. The fishbone morphology of the M 6C carbide was also found, Fig. 10(c). Table 2 gives the SEM-EDS quantitative data for M 6C carbides in the double tempered condition. It can be seen from the data in Table 2 that the lowest W content in the M 6C carbide was observed after austenising at 1100 C and double temper at 650 C. At this condition the M 6C carbides were fine. The W content of the M 6C carbides Fig. 11. Bright field TEM image showing the presence of M 2C with the typical thin needle morphology and rod-like or bar shape Fe 3C carbides. The SADPs were taken at the interface between Fe 3C and matrix with [010] Fe3 C//[011] α and between M 2C and matrix with [ 2110] //[001] α. MC 2 increased with increasing austenising and tempering temperatures as the carbides grew and became more stable, see Fig. 9. Thus, the growth of the alloy carbides toward equilibrium involved considerable solute diffusion and was time and temperature dependent. Concerning the size of carbides, there was interference of interaction volume with matrix in all double tempered conditions and the strongest interference occurred when the carbides were analysed for the condition T γ = 1100 C and T DT = 650 C where the carbides size was smaller than that at the higher austenising and double tempering temperature. Therefore, the quantitative data should be considered with caution given the size of the carbides. b. M 3C and M 2C Carbides The M 3C carbide is rich in Fe, has an orthorhombic crystal structure, and as a consequence it is often referred to as Fe 3C or cementite. The Fe 3C carbides were identified based 2014 ISIJ 1672

7 on their specific coarse needle or rod-shape morphology, as reported by Fujita and Bhadeshia, 16) and by electron diffraction. The cementite carbide was first formed in the early stages of tempering owing to the rapid diffusion of C in Fe and the low diffusivity of the substitutional alloying elements. This carbide can take into solution very considerable amounts of other alloying elements, in particular Cr can substitute for Fe gradually as the tempering temperature increases. 4) The TEM image in Fig. 11 shows different locations of cementite precipitation, which has been also reported by Furuhara et al., 17) and the presence of M 2C carbides. Figure 11 indicates that the cementite and M 2C carbides precipitated from martensite during tempering with the Bagaryatsky orientation relationship with the matrix. The cementite carbides were observed to precipitate independently in the PAGB, the interlath and intralath boundaries. The presence of coarsened cementite within martensite laths as well as at lath boundaries indicated that the cementite carbides grew rapidly at the double tempering temperature (T DT) of 650 C. The cementite carbide, which has a rod-like morphology, was slightly coarser at the interlath compared with that at the intralath boundaries. The thickness of cementite carbides in the interlath and within boundaries was in the range 10 to 15 nm, which is slightly finer than that reported by Furuhara et al. 17) This is probably due to the W content restricting the growth of cementite during tempering. From the aforementioned locations, the appearance of bar-shaped cementite carbides indicated that cementite precipitation occurred in the early stages of tempering, and the precipitation of cementite was controlled by diffusion to the interlath boundaries (fine interlath cementite) and dislocations (finer intralath cementite), and thus the cementite grew with increasing tempering temperature. 13,18) The M 2C carbides were identified by TEM (Fig. 11) using selected area electron diffraction and also based on their morphology, which was fine and needle-like. 2,16) The M 2C carbides had a high aspect ratio, typically higher than 10. 1) Figure 11 also shows that the M 2C carbides nucleated at the ferrite matrix. The M 2C carbides were rich in W and thus their formation was attributed to the high W content of the H21 tool steel. This type of carbide can also dissolve a considerable amount of Cr 19) and therefore their formation in the H21 tool steel was enhanced by the steel s W and Cr content. In this study, the M 2C carbides were only observed after a double temper at 650 C. According to Cai et al., 20) the M 2C carbide in a steel containing 5 wt% Cr was not stable and rapidly transformed to stable M 23C 6 carbide. M 2C carbides were also observed by Asadabad et al. 1) in a 4.5Cr-2W-0.25V-0. 1C steel (wt-%) in the early stages of tempering at 600 C for 0.5 hour. c. MC and M 23C 6 Carbides The fine MC carbides were identified by TEM-EDS, see Fig. 12. The presence of the MC carbide was identified by the highest peak of V in the TEM-EDS spectrum. The MC carbides had a nearly spherical morphology and their size was around 35 nm to 45 nm. There was no significant difference in size compared with the as quenched condition. In alloy steels, the transformation of carbides can take place by in situ nucleation at pre-existing carbides or by separate nucleation within the matrix or at grain boundaries and sub grain boundaries. 2) Figure 12 shows in-situ transformation of carbides in the H21 tool steel after austenising at C and double tempering at 750 C. Note that the carbide types in Fig. 12 were identified by TEM-EDS qualitative analyses using a JEOL 2010F. The letter A in Fig. 12 indicates that in situ nucleation had occurred and the letter B indicates coarsened M 23C 6 carbides. Details of both processes can be seen in Figs. 13 and 14, showing Fe 3C carbides. There was a change of cementite morphology with increasing tempering temperature. At a double temper at 650 C (Fig. 11), most cementite carbides maintained a rod-like morphology with an aspect ratio of around ~6.5, and when the tempering temperature increased to 750 C (Fig. 13), the cementite began to change toward a spherical morphology. The presence of the M 23C 6 carbides was confirmed by qualitative analyses using TEM-EDS, see Fig. 13. The M 23C 6 carbides were secondary carbides and formed during double temper. The spectrum of M 23C 6 carbides had similar height peaks to that of Fe. During tempering, the formation of the M 23C 6 carbides, which had a high Cr and W content, occurred as the Cr replaced Fe gradually in M 3C and as the Cr content was getting higher, in order to become more stable. The M 3C carbide transformed to Cr based carbides (either M 7C 3 or M 23C 6 carbide). The diffusivity of W in ferrite is lower than that of other alloying elements. and all elements (except C) have very low solubilities in M 3C. Kuo 21) reported that the transformation of M 3C carbides to Cr rich carbides started by the increase of Cr concentration in the former until a saturation value was reached, and then the M 3C transformed to Cr rich carbide (M 7C 3 or M 23C 6) and this transformation was accompanied by an abrupt increase in the concentration of Fig. 12. Bright field TEM images (JEOL 2010F) taken from thin foil after austenising at C and double temper at 750 C showing MC carbide and its spectrum and in-situ transformation (for meaning of letters see text) ISIJ

8 Fig. 13. A Bright field TEM images showing in-situ transformation of M 3C to M 23C 6 carbides indicated by A in Fig. 12 and their typical EDS spectra (for the size of carbides see Fig. 12). Fig. 14. Bright field TEM images showing M 23C 6 carbide after in situ nucleation indicated by B in Fig. 12 (for the size of carbides see Fig. 12). were identified using the ICDD cards given in Table 3. The diffractograms for the same double tempering temperature and different austenising temperatures were different regarding the number of peaks and intensities. The peak intensity of each carbide was higher at the higher austenising temperatures. The XRD suggested that the main carbide formed in all double tempered conditions was the M 6C carbide. The number and intensity of the M 6C carbide peaks increased with increasing double tempering temperature. This was attributed to the formation of secondary M 6C carbides. Peaks of the Cr 23C 6 carbides were detected after double tempering at 750 C at both austenising temperatures. Increasing the tempering temperature to 800 C caused the intensity of the Cr 23C 6 carbide peak to increase further. Thus, the Cr 23C 6 carbides nucleated above 650 C and grew with increasing tempering temperature. The MC, M 2C and M 3C carbides were not clearly identified by XRD, owing to their volume fraction being very low. Cr. Secondary carbides usually do not form at tempering temperatures below 500 C because the alloying elements cannot diffuse rapidly enough to allow alloy carbides to nucleate. 8) The presence of M 23 C 6 carbide was also observed after a short tempering (1 2 h) of a steel containing 10 wt% W. 22) Figure 13 also shows that the M 23 C 6 carbides nucleated at the Fe 3 C/ferrite interfaces. In situ nucleation of alloy carbides at pre-existing cementite carbides is a common occurrence due to these cementite carbides being widely spaced at temperatures above 500 C. 2) The high W content in the H21 tool steel encouraged the formation of the M 23 C 6 carbide by enhancing its nucleation. 2) The Fe 3 C carbides were completely replaced by Cr carbides when tempering was done at 800 C. Figure 14 indicates that the M 23 C 6 carbides grew toward the M 3 C carbides to consume them and increase their size and volume fraction. This mechanism has been also reported by Inoue and Matsumoto, 3) who observed the in situ transformation of M 3 C to Cr rich carbide X-ray Diffraction of the Double Tempered H21 Tool Steel Figure 15 shows the XRD diffractograms after austenising and double tempering of the H21 tool steel. The peaks 3.3. Carbide Evolution In alloy steels, the transformation of carbides can take place by in situ nucleation at pre-existing carbides or by separate nucleation within the matrix or at grain boundaries and sub grain boundaries. 2) The evolution of carbides during heat treatment is strongly related to the chemical composition of tool steels, the diffusivities of alloying elements and the selected heat treatment. The carbides found after double tempering at 650 C were the M 6C, M 2C and Fe 3C carbides. The carbides found after double tempering at 750 C were the Fe 3C, M 6C and M 23C 6 carbides. TEM investigation of the H21 tool steel after double tempering at 800 C confirmed the presence of M 23C 6 and M 6C carbides but not that of cementite. The MC carbides were present in the as quenched and double tempered samples. Significant coalescence of the MC carbides was not observed. This would suggest that the MC carbide had high thermal stability and was an equilibrium phase once formed. Yongtao et al. 23) reported an increase of average size and dissolution of MC carbides in a steel containing 0.25%-wt V after tempering at 710 C for 10 hours. In this study, the V content of the H21 tool steel was higher than that of Yongtao et al., which would suggest that the stability of MC was due to the presence of V in the tool steel. In this study, the M 2C carbide 2014 ISIJ 1674

9 Fig. 15. XRD diffractograms of the double tempered H21 tool steel. Table 3. The ICDD cards used for the identification of phases in the H21 tool steel after double temper. Phase ICDD card α Fe Cr M 6C (Fe 3W 3C) M 23C 6 (Cr 23C 6) was not the first carbide to precipitate, but formed after the Fe 3 C carbides. This conclusion was based on the size of the Fe 3 C carbides that were coarser than the M 2 C carbides, and indicated that the Fe 3 C has formed earlier than the M 2 C and grew with increasing tempering time. Bała and Pacyna 24) studied the tempering behaviour of a steel containing 4.14 wt-% Cr and 6.55 wt-% W and observed that the Fe 3 C carbides started to precipitate at ~280 C, the M 2 C carbides formed at ~600 C and the former transformed to more stable Cr rich carbides when the Cr content of the Fe 3 C increased with time. The M 2 C carbides, which were observed at tempering at 650 C, were not stable. With increasing tempering temperature they went back into solution and the cubic M 6 C carbides formed predominantly at grains boundaries as massive carbides and grew quickly. Thus, the M 2 C carbides were not observed after double tempering at either 750 or 800 C. Even though the M 6 C carbides can form at the expense of M 23 C 6 carbides, in this study the double tempering time (one hour) was insufficient for the latter carbides to transform to the M 6 C carbide, which occurred after prolonged tempering. Thus, the sequence of carbide evolution in the H21 tool steel was as follows: separate transformation : Matrix Fe 3 C M 2 C M 6 C and in situ transformation : Matrix Fe 3 C M 23 C 6 It must be emphasized that the sequence of carbide evolution was time and temperature dependent and both carbide formation sequences occurred simultaneously. Fig. 16. The effect of austenising and double tempering temperatures on the hardness of the H21 tool steel The Effect of Double Tempering on Hardness The tempering hardness is linked with precipitation strengthening of carbides, and the decomposition of martensite and austenite. Tempering curves of hardness as a function of tempering temperature for austenising temperatures of 1100 and C are shown in Fig. 16, 25) which shows that the hardness after double tempering increased with increasing austenising temperature and decreased with increasing double tempering temperature. The highest double tempered hardness was 505 HV and was obtained after austenising at C and double tempering at 650 C. This is because at the higher austenising temperature, the more carbides were dissolved in the matrix, giving a higher supersaturation and driving force for precipitation, which led to more extensive carbide precipitation during the double tempering. The maximum hardness in this study is lower than reported by Min-Xian 7) due to the difference in heat treatment process. The tempering in the range 650 C to 800 C did not show a secondary hardening peak or in other words the highest double tempered hardness (505 HV) obtained in this study can not be concluded as the peak hardness. The secondary hardening has been attributed to the precipitation of fine M 2C carbides. 4) Although TEM investigations of the H21 tool steel after double tempering at 650 C, Fig. 11, indicated that some M 2C carbides were still present, but their volume fraction of M 2C carbide were very low, which were not identified by XRD (Fig. 15). As mentioned before, the M 2C carbide was not stable and rapidly transformed to stable car ISIJ

10 Table 4. Tool T γ steel ( C) bide and double tempering at 650 C for 1 hour promoted the dissolution of the M 2 C carbides and favoured the precipitation and coarsening of the M 6 C carbides. It is suggested that the formation of a high volume fraction of the M 2 C carbide occurred below 650 C. The precipitation and growth of cementite and M 6 C carbide was accompanied by a decrease in hardness significantly from double tempering temperature of 650 C to 750 C. Furthermore, the decrease of hardness with increasing tempering temperatures was due to the recovery of lath structure and ferrite grain growth. This implies that the double tempering the H21 tool steel should be carried out below 650 C. Quantitative data of carbide size and volume fraction for the three double tempering temperatures is given in Table 4. It can be seen from Table 4 that the volume fraction of carbides slightly increased with increasing tempering temperature. As mentioned before, this is was attributed to the precipitation of secondary carbides that was higher at the higher tempering temperature owing to the higher diffusivity of alloying elements. There was no correlation between the average size of carbides with increasing tempering temperature, probably due to coalesced of carbides forming networks carbides that made the measurement of carbide size less accurate. 4. Conclusions Volume fraction (V c) and average size (μ m) of carbides in the H21 tool steel at different austenising and double tempering temperatures. V c (%) Double tempering temperature ( C) Mean size (μm) V c (%) Mean size (μm) V c (%) Mean size (μm) ± ± ± ± ± ± 0.2 H ± ± ± ± ± ± 0.1 The following conclusions can be drawn from this study. a) The current study confirmed previous findings and contributed to existing knowledge that two types of carbide formed in the as quenched microstructure of the H21 tool steel, namely the M 6C carbides and the MC carbides, whereas the M 23C 6 carbide found in the H21 tool steel after double temper. b) Depending on the tempering temperature, the carbide formation sequence in the H21 tool steel during double tempering, e.g.: a. In situ transformation : Matrix Fe 3C M 23C 6 and b. Separation transformation : Matrix Fe 3C M 2C M 6C c) Higher austenising temperatures accelerated the approach of the carbide precipitates to their stability during tempering. d) A lath morphology of martensite still existed in the microstructure of the steel double tempered at 650 C. The M 2C carbides contributed retaining the lath morphology of the martensite matrix. e) The highest double tempered hardness of the H21 tool steel was 505 HV, and was obtained after austenising at C followed by double tempering at 650 C. f) In this study, secondary hardening as a consequence of secondary precipitation of fine carbide particles did not occur, which implies that the double tempering of the H21 tool steel should be carried out below 650 C. Acknowledgements One of the authors (M. Nurbanasari) gratefully acknowledge to the Directorate General of Higher Education, Ministry of National Education, Indonesian Government and the Institut Teknologi Nasional, Bandung, Indonesia for their financial support. REFERENCES 1) M. A. Asadabad, S. Kheirandish and A. J. Novinrooz: J. Iron Steel Res. Int., 17 (2010), 57. 2) H. K. D. H. Bhadeshia and R. W. K. Honeycombe: Steels Microstructure and Properties, 3rd ed., Elsevier Ltd, Oxford, (2006). 3) A. Inoue and T. Matsumoto: Metall. Trans. A, 11A (1980), ) G. A. Roberts, G. Krauss and R. L. Kennedy: Tool Steels, 5th ed., ASM International, Materials Park, OH, (1998). 5) T. Sato, T. Nishizawa and K. Murai: Tetsu-to-Hagané, 44 (1958), ) K. Kuo: J. Iron Steel Inst., 174 (1953), ) W. Min-xian, W. Shu-qi, W. Lan and C. Kang-min: J. Iron Steel Res. Int., 18 (2011), 47. 8) G. E. Totten: Steel Heat Treatment: Metallurgy and Technologies, 2nd ed., Taylor and Francis, New York, (2006). 9) R. T. Howard and M. Cohen: Trans. Metall. Soc. AIME, 176 (1948), ) S. V. Radcliffe and E. C. Rollason: J. Iron Steel Inst., 191 (1959), ) S. Karagoz and H. Fischmeister: Steel Res., 58 (1987), ) H. K. D. H. Bhadeshia: Bainite in Steels: Transformation, Microstructures and Properties, 2nd ed., The Institute of Materials Communication Ltd, London, (2001). 13) R. N. Caron and G. Krauss: Metall. Trans., 3 (1972), ) M. V. Speight: Acta Metall., 16 (1968), ) M. Durrand-Charre: Microstructure of Steels and Cast Irons, Springer- Verlag Berlin, (2004). 16) N. Fujita and H. K. D. H. Bhadeshia: ISIJ Int., 42 (2002), ) T. Furuhara, K. Kobayashi and T. Maki: ISIJ Int., 44 (2004), ) G. R. Speich: Trans. Metall. Soc. AIME, 245 (1969), ) J. R. Davis: Tool Materials, Illustrated ed., ASM International, Materials Park, OH, (1995). 20) G. J. Cai, H. O. Andren and L. E. Svensson: Mater. Sci. Eng. A, 242 (1998), ) K. Kuo: J. Iron Steel Inst., 173 (1953), ) R. J. Tunney and N. Ridley: Met. Sci., 13 (1979), ) Z. Yongtao, M. Lede, W. Xiaojun, Z. Hanqian and L. Jinfu: Mater. Trans., 50 (2009), ) P. Bała and J. Pacyna: J. Achiev. Mater. Manuf. Eng., 28 (2008), ) M. Nurbanasari and P. Tsakiropoulos: Mater. Sci. Eng. A, 570 (2013), ISIJ 1676