The Cenos simulation of induction heating showed that there were significant differences in the hardening depth, but the influence on the microstructure and wear properties could not be simulated. Figure 3 shows that there were substantial changes in the temperature-depth along the diameter of the individual roll depending on the feeding rate. The feed rate seems to be a decisive factor in determining the hardening depth related to the temperature at any given position of the individual roll. When we increased the feeding rate from 24 to 42 mm/min, the temperature-depth decreased as the thermal energy transferred, which is heavily related to the electrical current frequency used in this experiment.
Figure 4 shows that in each disc from 1 to 4, the hardening depth ranges from 45 to 60 mm for the lowest feeding rate of 24 mm/min. On the contrary, the surface hardness varies only within the magnitude of a few HRc (i.e., from around 58 HRc for the feed rate of 36 mm/min to 62 HRc for the feed rate of 24 mm/min, which is the lowest). Additionally, the decrease in the hardening depth and hardness itself, as depicted by the curves in different colors, shows that in the case of the lowest feed rate, the decrease was less steep and more continuous than that in the case of the highest feed rate.
Table 3 shows the hardness depth measurements from a stationary Rockwell tester. Disc 1, which had a feed rate of 24 mm/min and an austenitizing temperature of 7.5 minutes, had a hardness depth of 50 mm according to the standard DIN 10328, or 25 mm according to the usual customer requirements in the roll industry. For disc 2, which had a 30 mm/min feed rate, the hardness depth was measured at 46 mm according to DIN 10328 and 22 mm according to standard customer requirements. Disc 3, which had a feed rate of 36 mm/min, had a hardening depth of 41 mm according to DIN 10328 and 16 mm according to standard customer requirements.
Table 3
Hardening depth measurements
|
Feed rate [mm/min]
|
Time at austenitization temperature [min]
|
Hardening depth acc. to DIN 10328 [mm]
|
Hardening depth acc. to customer requirements (min. 58 HRc) [mm]
|
Disc 1
|
24
|
7.5
|
50
|
25
|
Disc 2
|
30
|
6
|
46
|
22
|
Disc 3
|
36
|
5
|
41
|
16
|
Disc 4
|
42
|
4.3
|
38
|
16
|
The same hardness depth according to standard customer requirements was also measured on disc 4 with a feed rate of 42 mm/min, but there is a difference in hardness depth according to DIN 10328, with disc 4 having a 3 mm lower hardness depth compared to disc 3 with feed rate 36 mm/min. With the results of the hardness measurements in mind, we turned to SEM to determine and evaluate the condition of the tested rolls on the microscale before and after induction hardening. SEM was performed on specimens from all four discs. During the microscopic examination, we discovered significant differences in the samples’ microstructures, which were mainly composed of the martensitic matrix and primary and secondary carbides. Images of the microstructure are shown in Fig. 5. However, despite the significant differences in microstructure associated with carbide streaks, we did not find substantial variations in hardness between all four specimens. See Fig. 6.
Table 4 shows the chemical compositions of various strains. As can be seen from the table, the matrix (i.e., Spectra 9 and 10) has almost the same chemical composition as that in Table 1, while the carbides have two different chemical compositions: the first group contains 27.20–34.62% Cr, 10.79–23.77% V, 3.10–5.76% Mo, and 1.17–3.09% W, while the second group or grade contains 6.79–7.96% Cr, 41.48–55.18% V, 5.09–6.31% Mo, and 4.19–9.99% W.
Table 4
Chemical composition of spots on disc 1 with feed rate 24 mm/min
Spectrum Label
|
Spectrum 1
|
Spectrum 2
|
Spectrum 3
|
Spectrum 4
|
Spectrum 5
|
Spectrum 6
|
Spectrum 7
|
Spectrum 8
|
Spectrum 9
|
Spectrum 10
|
C
|
23.21
|
23.95
|
18.46
|
19.86
|
23.16
|
19.07
|
19.59
|
23.86
|
|
|
Si
|
|
|
|
|
|
|
|
|
1.03
|
1.02
|
Ti
|
0.36
|
0.95
|
|
0.60
|
|
|
|
1.27
|
|
|
V
|
55.18
|
53.99
|
23.77
|
59.05
|
41.48
|
11.95
|
10.79
|
50.31
|
0.76
|
1.03
|
Cr
|
7.53
|
6.97
|
27.20
|
7.39
|
7.96
|
34.66
|
34.62
|
6.98
|
6.79
|
7.13
|
Mn
|
|
|
|
|
|
|
|
|
0.44
|
0.48
|
Fe
|
2.20
|
4.26
|
21.72
|
1.83
|
2.30
|
29.73
|
30.70
|
8.30
|
88.92
|
88.23
|
Mo
|
6.31
|
5.32
|
5.76
|
6.05
|
15.11
|
3.41
|
3.10
|
5.09
|
1.05
|
1.40
|
W
|
5.20
|
4.55
|
3.09
|
5.21
|
9.99
|
1.17
|
1.20
|
4.19
|
1.01
|
0.70
|
Total
|
100.00
|
100.00
|
100.00
|
100.00
|
100.00
|
100.00
|
100.00
|
100.00
|
100.00
|
100.00
|
The compositions of different streaks of carbides revealed that there were at least two groups of complex carbides present in the samples. One group was based on chromium, and the other was based on vanadium. Both groups contain at least three to four other chemical elements, adding to the hardness of the examined rolls. The carbon content has not been accounted for in the EDS analysis presented in Table 1. Nevertheless, based on our previous knowledge of the formation of carbides in complex steels, we can classify the first group of carbides as the (Cr, V, Mo, W)xCy type and the second group as the (V, Cr, Mo, W)xCy complex type of carbides, actually expecting MC and M7C3 type of carbides as it will be shown later. The first group of carbides is based on chromium while the other is based on vanadium. Both elements, along with some others, are known to be very potent elements for the formation of carbides. Of course, the carbides are not the only microstructural constituents that influence the properties of the steel. The constitution of the matrix of the investigated steel can have significant impacts on the properties of the steel. Therefore, we turned our attention to the EDS analyses and the distribution of the elements in the carbides and the matrixes of the investigated steel. Figure 7 shows the overall EDS plot on the sample from disc 1 at a feed rate of 24 mm/min, confirming the presence of two main types of carbides. The distribution of the individual alloying elements is shown in Fig. 8. Using light microscopy, we analyzed 200 individual spots on samples from discs 1–4 and measured the average size of the carbides (see Fig. 9).
Our analysis results for carbide sizes, which are shown in Fig. 10 and Table 5, revealed that the largest average carbide size was present in the samples from disc 1, which had a feed rate of 24 mm/min, and the smallest average carbide size was present in the samples from disc 4, which had a feed rate of 42 mm/min.
Table 5
Average size of carbides – image size 548.91x410.33 µm
|
Disc 1 with feed rate 24 mm/min
|
Disc 2 with feed rate 30 mm/min
|
Disc 3 with feed rate 36 mm/min
|
Disc 4 with feed rate 42 mm/min
|
Average size of carbide [µm2]
|
2.265
|
2.163
|
2.018
|
1.843
|
When we analyzed the size of carbides in the range of 0–1 µm2, we found that the samples of disc 1 with a feed rate of 24 mm/min had the smallest average carbide size.
The analyses of the sizes of carbides in the 1–5-µm2 range showed that the carbides in the samples from disc 1 with a feed rate of 24 mm/min had the largest size, confirming the presence of the Ostwald ripening process. This led to the supersaturation of the matrix, which together with the feed rate, affects the shape of carbides. The specimens from disc 1 were hardened at a lower feed rate compared to other three discs, giving the carbides additional time to develop and become spherical (the carbides become more rounded), and the smaller ones dissolved. When measuring the macro hardness, we found that the microhardness was the highest in the sample from disc 1 with a feed rate of 24 mm/min and the lowest in the sample of disc 4 with a feed rate of 42 mm/min, which was hardened at the highest feed rate. Because we identified two types of carbide, we wanted to determine which types of carbide they were. For this purpose, we used the well-established EBSD technique, and the results are presented in Figs. 11 and 12, with the EBSD IPF Z map and the phase for the confirmation of the proposed phases shown. The analysis results confirmed the presence of two main types of carbides in the sample, namely MC and M7C3. The IPF Z map shows the typical microstructure of martensitic steel with needle-like structures and carbides in the matrix. In contrast, the phase map shows the martensitic bcc phase in yellow, M7C3 in green and MC in red.
Figure 12 contains an EDS map of the same spot as EBSD, where different amounts of elements (e.g., Cr, Fe, V, W, Mo, and C) in the carbides and the matrix so it can corroborate the different carbides have an additional amount of alloying element.
X-ray diffraction was performed on samples after induction heating and hardening and first and second tempering to confirm the theory that they had lower residual austenite content than in the production of rolls from standard grades. Figure 13 shows the X-ray diffraction pattern of the sample from disc 1 with a feed rate of 24 mm/min after second tempering at 480°C for 24 hours. The sample contained 99.6% martensite and 0.4% austenite. The XRD results are presented in Table 6. The content of retained austenite (RA-Retained Austenite, in vol.%) was calculated using Rietveld methods and use of PowderCell v.2.4 software.
Table 6
Measurements of RA on a sample from disc 1 with a feed rate of 24 mm/min
Process
|
Measured content of RA [%]
|
Induction heating and hardening
|
12.70
|
Tempering 1–515°C/24 h
|
1.00
|
Tempering 2–480°C/24 h
|
0.40
|
Tribological tests that involved using the pin-on-the-disc method gave us clear information about wear volume and specific wear rate of all discs with different feed rates. We applied other loads with 1,800 cycles per specimen. The wear volume and specific wear rate were lowest for disc 1 with a 24 mm/min feed rate. However, the differences in wear were minimal and are more correlated to material homogeneity and its local characteristics, as presented in Figs. 14 and 15.
The wear volume and specific wear rate of a ball made of 100Cr6 steel were highest on the disc with a 24 mm/min feed rate. The specific wear rate was found to be highest at a feed rate of 24 mm/min for all three loads. Furthermore, the specific wear rate decreases as the feed rate increases. As we increased the load for the feed rate of 24 mm/min, the counter body penetrated deeper into the surface of the sample, leaving a narrower track with a lower specific wear rate. On the contrary, the specific wear rate decreased when an increasing feed rate was used and appeared to be similar at the 36 and 42 mm/min feed speeds. We hypothesize that this is due to the lower amount of annealing of the matrix that occurred at the lower feed rates, which generated the heat required to anneal the iron-based matrix. Again, the observed differences from other discs were minor and do not represent significant improvements regarding wear resistance. Images of wear are presented in Fig. 16.