Materials Processing
Influence of Nb and V Addition on Abrasive and Impact Wear Properties of 16%Cr–3%Mo White Cast Iron
2020 年 61 巻 12 号 p. 2363-2370
詳細
2020 年 61 巻 12 号 p. 2363-2370
Investigations were conducted on the microstructure and behaviors of abrasive wear and impact wear of 16%Cr–3%Mo white cast iron without and with 2.7%Nb–1%V addition heat-treated at 1253 K followed by forced-air cooling to clarify the difference between two kinds of wear mechanisms. Microstructures were analyzed by means of scanning electron microscope equipped with energy dispersive X-ray spectrometer, X-ray diffractometer and ferrite meter. Results reveal that since MC spheroidal carbides are formed by addition of Nb and V, the other primary carbides are refined and martensite in the matrix becomes more stable. Such microstructure improves abrasive wear resistance, but deteriorates impact wear resistance. Under the severe test condition, self-tempering softening of martensite in the matrix caused by the impact energy of abrasive media and it increases the impact wear rate.
High chromium white cast irons are known as effective structural materials for applications in the cement manufacturing, slurry pumping, mining, mineral ore processing and paper manufacturing that require excellent wear resistance. Numerous studies point out that the wear performance of high chromium white cast irons depends on the type, size, morphology, volume fraction of primary carbides and matrix phase.1–3) Also, the life cycle of cast iron products is remarkably influenced by not only the relative movement of contact surface, but also kind and size of abrasive media. Strong carbide-forming elements such as niobium, vanadium, tungsten and so on are reported to be effective for improving wear resistance.4–6) Recently, corresponding to the use environment, some of the cast irons developed by individual or combined addition of these elements have been mass-produced industrially.
With regard to wear behavior, many researches have been done about the effect of alloying elements on abrasive wear resistance.7–9) However, the playing roles of phase in matrix and carbides to the impact wear behaviors are not completely clarified. Moreover, published data on the comparison between abrasive wear and impact wear attributed by microstructure is limited. From a metallurgical point of view, understanding the difference between both wear behaviors is of considerable practical significance. Therefore, in the present work, 16%Cr–3%Mo white cast iron was chosen to investigate how Nb and V give influence on impact wear and abrasive wear performances. Thereby, the difference between both wear mechanisms were clarified through microstructural parameters.
Experimental cast irons were produced using a 100 kg capacity high-frequency induction furnace. After melting raw materials, the molten metals were superheated up to 1773 K. Each molten metal was poured into an alumina sand mold of JIS G0307 a test coupon (Y block) after holding at the temperature for 0.6 ks. The chemical compositions are listed in Table 1.
As shown in Fig. 1, disk shape samples with 37 mm in diameter and 17 mm in thickness for abrasive wear test, and rectangular samples with 35 mm in width, 40 mm in length and 15 mm in thickness for impact wear test were machined from a bottom of Y block, and they were heat-treated in an electric furnace at 1253 K for 3.6 ks followed by forced-air cooling.
Dimensions of rectangular sample for impact wear test and disc shape sample for abrasive wear test (unit: mm).
Abrasive wear test was carried out on a metallographic grinding/polishing machine. Three samples were set in a 108 mm diameter sample holder to be automatically ground on 80 grit SiC paper in tap water under a load of 100 N at a constant rotational speed of 40 rpm for 3.6 ks in each cycles. The total time for 10 cycles abrasive wear test was 36.0 ks. On the other hand, impact wear test was performed by an air blast machine, as described in detail in previous work.10) Iron grit (Fe–1.0C) which have 850 HV of the hardness and 1.2 mm of average diameter was chosen as abrasive media. 15 kg of iron grit were accelerated at the pressure of 0.5 MPa from a 6 mm diameter nozzle tip to collide the surface sample for 0.85 ks in each cycle. Impingement angle was 90° and blast distance from the nozzle to the sample surface was 100 mm. The total time for 20 cycles impact wear test was 17.0 ks. The spark that occurred was observed visually when abrasion media collided with the sample surface during the test, as shown in Fig. 2. Weight loss and decrease in thickness of samples were measured using electronic balance and depth profilometer, respectively.
(a) Inside the chamber of abrasive blasting machine and (b) Observation of spark generation during test.
Initial microstructure and worn surface and cross-section of the tested samples which were etched with Villela’s reagent (5 ml HCl, 1 g picric acid in 100 ml ethanol) for 7∼10 s, were examined using scanning electron microscope (SEM) equipped with energy dispersive X-ray spectrometer (EDS). Also, an X-ray diffractometer (XRD) was used to identify the structural phases. Measurement of martensite volume was performed using ferrite meter.
The start transformation temperature (Ms) and finish temperature (Mf) of martensite were calculated based on the dilatometric curves measured by thermomechanical analyzer (TMA). The cylindrical test pieces with 5 mm in diameter and 15 mm in height were cut from samples heat-treated at 1253 K for 3.6 ks by wire cutter. Under argon atmosphere, the test pieces were heated up to 1273 K and then cooled to room temperature. The heating rate and cooling rate were 10 K/min. The samples were not kept at 1273 K.
Measurement of bulk hardness was performed using a Vickers hardness tester with a load of 294 N (30 kgf) in time of 30 s. The hardness profile from the most severe worn surface to the depth direction in each sample after the wear test was measured by a micro Vickers hardness tester with a load of 2.94 N (0.3 kgf). Additionally, hardness of only austenitic/martensitic matrix with secondary carbide was also measured with a load of 0.98 N (0.1 kgf).
Figure 3 shows the microstructures and hardness of experimental samples heat-treated at 1253 K for 3.6 ks followed by forced-air cooling. It can be clear that there is a difference in microstructure between both samples. Precipitates are consisted of two types of primary carbides in the matrix of the base cast iron: fish bone-like carbide (white) and massive carbide (dark gray). These carbides become finer and a large amount of fine spheroidal carbides disperse in the matrix by adding Nb and V. Based on the EDS compositional maps shown in Fig. 4 and XRD patterns shown in Fig. 5, fish bone-like carbide and massive carbide are M2C and M7C3 type carbide, respectively. On the other hand, crystallized spheroidal carbides in the Nb–V added cast iron are confirmed as MC-type carbide. Unlike the enrichment of alloying elements detected in each carbide, Mo does not distribute homogenously in M7C3. Enrichment of elements in three types of carbides is summarized in Table 2. Moreover, remarkable diffraction peaks attributed to retained austenite and martensite are observed in the base cast iron. Addition of Nb and V results in lower peaks intensity of austenite and stronger peaks intensity of martensite. Measurement of martensite amount in both samples were performed by using ferrite meter. As a result, the measured value of the base cast iron is 29.4 vol%, which is smaller than that of the Nb–V added cast iron. As shown in Fig. 3, the macro hardness of base cast iron was 783.9 HV, but its Nb–V cast iron was 725.5 HV, quite lower.
Microstructures of the samples heat-treated at 1253 K for 3.6 ks.
Backscattered electron images and EDS compositional maps showing the distribution of alloying elements in the primary carbides.
X-ray diffraction patterns of the samples heat-treated at 1253 K for 3.6 ks.
Precipitation of secondary carbides in the matrix was observed by SEM in a high magnification to clarify the difference of the hardness. As shown in Fig. 6, fine granular carbides mainly composed of Cr homogenously precipitated in the matrix of the base cast iron. Size of these carbides is approximately 0.5∼1.0 µm. No Mo-rich secondary carbide was detected. According to H. Gasan et al., the secondary carbide in 16%Cr–3%Mo cast iron heat treated between 1223 K and 1323 K is M7C3 type.8)
Secondary electron images and EDS compositional maps showing secondary carbides in matrix of samples.
On the other hand, in Nb and V added cast iron, it was easily seen that secondary carbides increased not only in volume but also in types. The coexistence of secondary carbides of three different types was detected in a matrix. Shape of Cr-rich secondary carbide was like as rod shape (white arrows), different from that of M7C3 in the base cast iron. V plus Cr exist in several µm of fine spheroidal carbides which are presumed to be MC type, indicated by black arrows. Besides, a small amount of Nb–Mo-rich secondary carbides (marked by circles) was also confirmed. It should be noted here that surface analysis of the region where consisted of matrix with secondary carbide in Nb–V added cast iron showed that V content was about 0.8 mass% but Nb content was less than 0.1 mass%, which is a much smaller result. This is an evidence that lots of V-rich secondary carbides precipitated in matrix.
Hardness of only matrix with secondary carbides in both samples was also measured using micro Vickers hardness tester. The hardness of matrix in base cast iron was 835 HV, whereas that of Nb–V added cast iron was 742 HV. This result exhibits a good correlation with the macro hardness measurement, and it is considered that the decrease in hardness due to the addition of Nb–V is dominated by the microstructure of the matrix, especially secondary carbides.
Sasaguri et al. investigated the effect of Mo, Nb and V individual addition on the hardness of 16%Cr–(2.8∼3.6)%C cast iron after heat treated at 1273 K followed cooling with and without sub-zero treatment.11) According to their results, hardness of the heat-treated cast iron did not increase despite the variation of Mo amount in the range of 1.0∼4.0%. The reason is that Mo is one of austenite stabilizing elements and secondary carbides are not formed, so it has little effect on the hardness of heat-treated samples. In case of V addition, besides primary MC type carbides were formed in as-cast condition, a part of V amount dissolved in austenite, and when heat treatment was performed, precipitation of secondary carbide contributed to a remarkable increase in hardness. However, in Nb added 16%Cr cast irons, most of Nb, which has a low solubility limit in austenite, and a large amount of C were consumed for crystallization of primary MC (NbC), so the hardness of martensite was decreased by reduction of carbon in matrix. Increment of the hardness due to crystallization of primary MC carbides was presumed to be smaller than decrement of the hardness of martensite, resulting in the decrease in hardness of cast iron.
In this study, the hardness of Nb–V added cast iron is smaller than that of base cast iron, which could be explained by the same idea as Sasaguri et al. It was supposed that the increase in hardness due to crystallization of primary MC (M = Nb, V, Mo) and M2C (M = Mo, Nb, V, Cr) as well as precipitation of secondary carbides was lower than the decrease in hardness of martensite hardness because solute carbon in the matrix was lower, when Nb and V was added to 16%Cr–3%Mo cast iron.
Dilatometric curves of both samples measured by TMA are shown in Fig. 7. The start temperature Ms of martensitic transformation of base cast iron was 434 K, but the finish temperature Mf could not be measured well because it might be near or below room temperature. Addition of Nb and V shifted Ms and Mf to higher temperature. Therefore, metastable austenite was considered to transform to martensite easily under rapid cooling, as compared with base iron. In other words, it was supposed that amount of retained austenite might be comparatively low while martensite might exist stably in Nb–V added cast iron if quenched from the heat-treatment temperature of this study.
Dilatometric curves measured by TMA.
From the above results, it can be concluded that the addition of 2.7%Nb–1.0%V to 16%Cr–3%Mo cast iron results in the formation of spheroidal MC primary carbides and increase in the amount of martensite and secondary carbides, but the hardness was reduced.
3.2 Wear behaviorMeasurement results of abrasive wear test of experimental samples are shown in Fig. 8. Each value is the average of three samples measured weight loss. It is seen that weight loss of both cast iron samples linearly increases with polishing time. At 36.0 ks, weight losses in the Nb–V added cast iron and the base cast iron are 0.35% and 1.0%, respectively. Wear rate of the Nb–V added cast iron is about one third of the base cast iron, and it is comparably small.
Change in weight loss of samples during abrasive wear test.
Weight loss and reduction of thickness during impact wear test are shown in Fig. 9. There is little difference in weight loss and reduction of thickness between both samples at the beginning stage. It was obvious that weight loss and reduction of thickness of the Nb–V added cast iron increased linearly, but those of the base cast iron was slight when impact time exceeded 12.0 ks. Therefore, the gap of weight loss and reduction of thickness between both tested samples became larger as impact time elapsed. Figure 10(a) illustrates the visual appearance of both samples after impact time of 18 ks (test cycle of 20). The cross-sections of samples cut along the broken line (abc line) by wire cutter are shown in Fig. 10(b). It can be visually confirmed that a large hole (impact crater) at center of the Nb–V added cast iron was made by the blasting of iron grit, whereas the surface of the base cast iron sample was worn gradually. Moreover, reduction of thickness (Δh) of the Nb–V added cast iron was more remarkable than that of base cast iron. From the results, it should be noted that abrasive wear and impact wear behaviors are completely different. Simultaneous addition of Nb and V to 16%Cr–3%Mo white cast iron improves abrasive wear resistance but gives a bad influence on impact wear resistance. This reason will be explained in detail below.
Change in (a) weight loss and (b) reduction of thickness of samples during impact wear test.
(a) Visual appearance of surface and (b) cross-section of samples after impact time of 18 ks.
Microstructures of the surface and cross-section of samples after abrasive wear test are illustrated in Fig. 11. It is seen that numerous cracks exist in M7C3 compared with other carbides. Primary carbides, especially M7C3 tend to be broken by the friction with SiC grit on emery paper before detaching from the matrix surface. And then, the matrix is gradually scratched. As mentioned in section 3.1, the addition of 2.7%Nb–1.0%V to 16%Cr–3%Mo cast iron stabilize martensite in matrix with a large amount of various types of fine secondary carbides. It is considered that such matrix which does not easily undergo plastic deformation due to polishing, suppresses the detachment of carbides from surface of Nb–V added sample. In addition, the distribution of not only spheroidal primary MC carbide but also fine M2C carbides is considered to be effective for reducing abrasive wear rate. Therefore, their roles cannot be ignored. As a result, Nb–V added cast iron exhibits an excellent abrasive wear resistance rather than base cast iron.
Microstructures of worn surface and cross-section after abrasive wear test.
Hardness profiles from surface to the depth direction of the region where impact wear was most severe, were measured by micro Vickers hardness tester under a load of 2.94 N (0.3 kgf). Each measured value shown in Fig. 12 reflects the hardness of the mixture of matrix and primary carbides. It is found that there was the surface hardening phenomenon caused by the blasting of abrasive media in the vicinity of surface of both samples from the figure. Before impact wear test, the results of macro and micro hardness measurements revealed that Nb–V added cast iron was quite lower than base cast iron, as described above. After the test, it was also found that the maximum hardness of the base cast iron was larger than that of Nb–V added cast iron. Measurement of matrix hardness was performed at regions where was a depth approximately 50 µm below the impacted surface under micro load of 0.98 N (0.1 kgf). For comparison, the results of matrix hardness of both samples before and after impact wear test were summarized in Table 3. The increase in hardness of matrix due to work hardening, (ΔHV = HVafter test − HVbefore test) was calculated. As a result, ΔHV of the Nb–V added cast iron is remarkably smaller than that of base cast iron.
Hardness profiles of the samples after impact wear test.
The difference in ΔHV of both cast iron can be explained as follows. As shown in Fig. 13, it was clear that there was increase in martensite amount of both samples after abrasive wear and impact wear tests. These values can be considered as the amount of stress induced martensite transformed from retained austenite. Noted that the increment of martensite amount after impact wear test was larger, as compared with abrasive wear test. This reason can be explained that work hardening generated by the impact of iron grit was significant. Furthermore, it is found that the increment of martensite in the Nb–V added cast iron was comparably small after both tests, which was supposed by the volume fraction of martensite and austenite in the matrix. When the amount of retained austenite was larger, the work-induced martensite transformation was likely to occur, so the occurrence of surface hardening became more remarkable. On the other hand, since iron grit collided with the sample surface under condition of impact wear test, spark generation continuously occurred, as observed in Fig. 2(b). It is considered that the generated amount of impact energy (Joule heat) was large enough to promote the self-tempering softening of the local martensite in the matrix. Therefore, it easily accelerated plastic deformation of the matrix. Martensite in the matrix became stable by simultaneous addition to Nb and V, so it is supposed that self-tempering softening rather than work hardening dominantly occurred by impact of abrasive media.
Change in volume fraction of martensite in samples after abrasive wear and impact wear tests.
Figure 14 shows the region where impact wear was most severe. A lot of cracks exist inside not fish-bone like M2C, but massive M7C3 carbide in the base cast iron under test condition. Therefore, it can be said that the morphology of M2C is stable to not only abrasive wear, but also impact wear. As mentioned above, retained austenite presented stably in the base cast iron, so the amount of work hardening was relatively larger. As a result, the matrix surrounding primary carbides provides a sufficient mechanical support to suppress the primary carbides from cracking or spalling. Addition of Nb and V stabilizes martensite in the matrix. However, even though a large volume of secondary carbides strengthen the matrix, it was softened by self-tempering when iron grit collided with the surface. It can be concluded that the cause of such better impact wear behavior is not martensite but metastable austenite, and there was little effect of secondary carbides. Moreover, separation of fine carbide tended to be relatively easier than that of large carbide.
Cross-sectional microstructures of the most severe worn region after impact wear test.
Obtained results are as follows.
The authors express their appreciation to Emeritus Prof. M. Nakayama for helpful discussions.
