Effect of Ni on microstructure and wear behaviour of 13Cr-W-Mo-2C white cast iron

3.1. Microstructure and hardness

SEM micrograph of hypoeutectic as-cast 13Cr-W-Mo-Mn-2C white cast iron is shown in Fig. 1. Microstructure comprised a typical eutectic structure (matrix) and large hexagonal-shaped M₇C₃ carbides. The as-cast state had M₇C₃ carbides in the austenitic matrix. After a destabilization heat treatment, the matrix microstructure became heavily formation of secondary carbides within the austenite dendrite (Fig. 2). The typical microstructures of samples having different Ni concentration of 1, 2, 4, 6, 7 wt.% and as-heat treated circumstances (6 h, 1000 ^oC, air cooling) are shown in Fig. 2a-e. After the heat treatment, the dendritic structure disrupted, the continuity of dendritic carbides was cut and the carbides started to shrink. A significant fraction of retained austenite still remained after heat treating between M₇C₃ carbides which is thought to be due to the alloys higher Ni content (˃4). The presence of high Ni content around the austenite dendrite will tend to stabilize the remaining austenite where Ni is completely rejected by the carbide phase. The heat treatment time duration was effective especially on the derogation of the M₇C₃ carbides. The rejection of Ni to austenite phase softened the transformation of austenite to martensite, and suppressed the martensite start temperature. The most fundamental phases in white cast irons were austenite, martensite and pearlite. The hardness, toughness and abrasive characteristics of the structure depend on the ratio of these phases (Biner, 1985Biner, S.B. (1985). The role of eutectic carbide morphology on the fracture behaviour of high-chromium cast irons-I. Austenitic alloys. Canadian J. Metal. Mater. Sci. 24 (2), 155-162. https://doi.org/10.1179/cmq.1985.24.2.155.). The samples studied here were hypoeutectic white cast irons, and the first matrix phase was the austenite. This phase forming as the first nucleolus during solidification evolve to the form of dendrite arms with different shapes based on the thermal gradient. Austenite crystals had a certain carbon equivalent. This carbon equivalent is created by elements such as C, Mn, Ni and Mo. As the C concentration in the matrix structure increases, the lattice parameter in the body-centered cubic atomic crystal also increases. The carbon concentration in the austenite may change between 2.11% and 0.30% based on the concentration of Cr in the austenite. At room temperature, the solubility of the C equivalent in the austenite was approximately zero. C had the most effective role in the stability of the austenite phase in room temperature. C was the most important factor in reducing the martensite initiation temperature to lower values. If the transition to the pearlite phase is wanted to be prevented from the austenite phase with high-C content, then the structure is cooled down to room temperature, and the appearing phase remainder becomes austenite (Biner, 1985Biner, S.B. (1985). The role of eutectic carbide morphology on the fracture behaviour of high-chromium cast irons-I. Austenitic alloys. Canadian J. Metal. Mater. Sci. 24 (2), 155-162. https://doi.org/10.1179/cmq.1985.24.2.155.; Lin et al., 2010Lin, C.M., Chang, C.M., Chen, J.H., Wu, W. (2010). The effects of additive elements on the microstructure characteristics and mechanical properties of Cr-Fe-C hard-facing alloys. J. Alloys Compds. 498 (1), 30-36. https://doi.org/10.1016/j.jallcom.2010.03.127.; Chandan et al., 2022Chandan, A.K., Kishore, K., Hung, P.T., Ghosh, M., Chowdhury, S.G., Kawasaki, M., Gubicza, J. (2022). Effect of nickel addition on enhancing nano-structuring and suppressing TRIP effect in Fe40Mn40Co10Cr10 high-entropy alloy during high-pressure torsion. Int. J. Plast. 150, 103193. https://doi.org/10.1016/j.ijplas.2021.103193.).

SEM micrographs of the microstructure showing the changes created by Ni addition are shown in Fig. 2. The presence of Ni had a substantial impact on the arm spacing of the carbides. There were microstructural differences in the arm length of carbides and carbide vol.%. The additions of Ni were increased arm spacing of carbides, and the carbide volume fraction from 0.20 to 0.30 for 0.5-7.0 wt.%. The increase of the Ni concentration caused destabilisation of the M₃C phase, replaced by M₇C₃ eutectic carbide phase, and broke the dendritic morphology (Fig.2e) (Laird et al., 2000Laird, G., Gundlach, R., Röhrig, K. (2000). Abrasion resistant cast iron handbook. American Foundry Society, USA, pp. 1-222.). In addition, the presence of 2-7 wt.% Ni in the structure allows the virgin austenite to approach the Ms temperature unimpeded by the creation of pearlite. The microstructure of as-cast structure having 2-7 wt.% Ni included austenite-martensite phase mixture. Together with Ni, the concentration of Cr was important in the formation of carbide type of the structure during solidification (Laird et al., 2000Laird, G., Gundlach, R., Röhrig, K. (2000). Abrasion resistant cast iron handbook. American Foundry Society, USA, pp. 1-222.). The presence of Cr over 5 wt.% in this alloy having high Ni concentration formed the refined M₇C₃ carbides. Ni concentration above 6 wt.% produced a homogeneously dispersed microstructure with M₇C₃ carbides lower than 2 µm (Fig. 2e).

The heat-treated samples shown in Fig. 2 had a microstructure that differed significantly from the cast structure (Fig. 1). The biggest difference lies in transformation of austenite to martensite and the increase in M₃C secondary carbide concentration. The hardness results of experimental samples are given in Fig. 3. The increase of Ni had a great influence on hardness according to the as-cast condition. However, there were a little rise on the hardness for the additions of Ni over 2 wt.% in the heat-treated condition. Heat treatment eventuated in a significant hardness change. The matrix hardness was 400 HV in the as-cast sample, it was 781 HV in the 5 wt.% Ni sample. This rise was caused probably by solid solution treatment leading to dissolution of Si in the austenitic structure. Microhardness increased from 576 to 731 HV in the sample with 7 wt.% Ni. Fig. 3 also indicates that the hardness decreased adjacent to the surface due to fine, rod and blade type M₇C₃ carbides and small proeutectic austenite dendrites. Further away from the surface, the austenite dendrites became larger and carbide morphology became more blades like. The fine dendritic structure transformed to coarser structure as function of solidification rate. The increase in hardness presented an interesting relationship between the microstructures of the samples during heat treatment. This relation was not only depending on the concentration of martensite, the presence of tiny secondary carbides was primarily effective on the degree of hardness and toughness.

The XRD pattern of the steel with 7 wt.% Ni content (S5) is displayed in Fig. 4. The samples contained hard eutectic carbides embedded in an austenite matrix. The matrix consisted mainly of austenite dendrites. The austenite was surrounded by M₇C₃ carbides. This carbide was embedded in a martensitic matrix which contains retained austenite as an outcome of heat treatment in the microstructure. Austenite and martensite phases were supersaturated with carbon and other alloying elements. This led to a decrease in Ms (initial temperature of the martensitic transition) and the preservation of a high amount of austenite at room temperature for the samples having Ni over 6 wt.%.

Chemical composition clarified by EDS of the steel with 7 wt.% Ni content (S5) is shown in Fig. 5 and results verify the phases as M₇C₃. Figure 6 demonstrates the SEM mapped image analysis of the sample S5. Mn, Si, W, Mo, Ti, Ni, C, Cr and Fe elements are presented in the sample. Note that there was a hypoeutectic composition for 0.5 wt.% Ni within the austenite phase interval in the alloy. The 2 wt.% Ni alloy (S2) was near to the eutectic phase, while the alloys with 4-5 wt.% Ni content had a hypereutectic composition where solidification began with the formation of M₇C₃. The addition of up to 6 wt.% Ni (S4) increased the amount of austenite retained in the cast structure from about ∼65 vol.% to ∼70 vol.%. For samples containing 7 wt.% Ni (S5), the composition was stable at ∼25% under heat-treated conditions. The presence of more than 6 wt.% Ni in the structure stabilized the austenite.

The DTA plot is given in Fig. 7. DTA analysis values obtained from four samples are given in Table 3. The solidification limit was observed based on the liquid and solidification temperatures in samples having increased Ni concentration from 0.5 to 6 wt.% (Fig. 2 and 5). All eutectic carbides and phases were identified at room temperature. Under unstable conditions, the phase transition interval was shorter than under equilibrium conditions. The liquidus temperature of the steel with 7 wt.% Ni (S5) content was higher than other samples. The sample having higher Ni concentration showed higher formation temperature of M₇C₃ and austenite. Additionally, in this alloy, M₃C precipitated earlier than the corresponding eutectic transformation in low-grade Ni samples (Biner, 1985Biner, S.B. (1985). The role of eutectic carbide morphology on the fracture behaviour of high-chromium cast irons-I. Austenitic alloys. Canadian J. Metal. Mater. Sci. 24 (2), 155-162. https://doi.org/10.1179/cmq.1985.24.2.155.; Chandan et al., 2022Chandan, A.K., Kishore, K., Hung, P.T., Ghosh, M., Chowdhury, S.G., Kawasaki, M., Gubicza, J. (2022). Effect of nickel addition on enhancing nano-structuring and suppressing TRIP effect in Fe40Mn40Co10Cr10 high-entropy alloy during high-pressure torsion. Int. J. Plast. 150, 103193. https://doi.org/10.1016/j.ijplas.2021.103193.). The presence of primary M₇C₃ carbides in our hypoeutectic samples were not expected. However, during the eutectic solidification, the carbides will completely reject Si and Ni into the remaining liquid. The remaining liquid then becomes supersaturated with Si and Ni, rising the activity coefficient of carbon, and the carbon content of eutectic boundary increases as the carbon concentration of hypereutectic liquid. Consequently, primary M₇C₃ carbides formed during solidification reactions (Table 3). If the Si level is not kept above 1.5%, the eutectic reaction (Rx3) drift down forming Rx4 which is not preferred.

3.2. Wear test

The change of friction coefficient as a function of load is shown in Fig. 8. The friction coefficient (µ) was determined to be more than 0.4 for various sample loads. An increase in the friction coefficient from the first value to the stable value was observed following ~1000 m sliding. For base samples, a major decrease in the friction was observed with load from ~1.0 for 40 N to ~ 0.62 for 140 N.

The worn surfaces for loads of 40, 90 and 140 N are shown in Fig. 9 a-c. A mild increase in the friction was observed at all loads with the increase in the austenite phase concentration. The austenite concentration nearly obtained the identical wear values in cast and heat-treated samples for 40 and 90 N loads. With the increase in the load to 140 N, certain differences were acquired in the as-cast and heat-treated conditions. At the above-mentioned loads, samples with 7 wt.% Ni (S5) exhibited the best wear resistance. Sample of as-cast form showed the weakest wear resistance. No differences in the composition surface structure were observed. In Fig. 9a, it could be observed that the worn surface of the reference sample without Ni content at 40 N was presented local grooves. The remaining regions were determined to have a uniform structure and smooth oxide layer. A carbide component was determined within the oxide film of the surface at the load of 40 N.

Wear rates versus sliding distance are given in Fig. 10 a-d. It was determined to be Fe₂O₃ in Fig. 10b. Phases α-Fe and carbide M₇C₃ were also present. Increasing the load to 90 N resulted in fewer abrasive grooves (Fig. 10b) and a smoother surface. Mostly Fe₂O₃, few traces of carbide M₇C₃ and wear residues occurred. However, α-Fe was not detected at 40 N load. Oxide layers and fine particulate oxide deposits formed the surface morphology. Almost no change in the structure of the eroded surface was observed at higher loads. However, surface roughening was detected, which is a sign of local adhesion. Together with Fe₂O₃, the Fe₃O₄ phase was present in the debris. Moreover, more intense plates seen on the wear surface caused an increase in the density of the oxide layer. The oxides were effortlessly solved at the load of 90 N. However, the surface did not change at 40 N load. At 140 N load, the oxide layer with M₇C₃ fragments was thicker (Fig. 10c). The expansion of carbide fragmentation was determined to be ∼25 µm at the above-mentioned load. There were also incipient stages of carbide cracking by fragmentation of carbide rods adjacent to the surface. Nevertheless, a more diffuse interface was acquired. Numerous M₇C₃ particles obtained from an extremely deformed substrate layer were present in the oxide. The effect of M₇C₃ size and morphology as a result of Ni addition was clearly observed. The smallest carbide is equals to 7 wt.% Ni, and it caused a thinner and steadier deformation zone.

However, bigger sizes of cracks were observed in the surface deformation layer in the 6 wt.% Ni sample (S4) with bigger carbide particles. The development of the microstructure on the surface apart from the cracking grade was not affected in any way by the material composition. Deformation measurement were performed on the worn surface. Addition of up to 5 wt.% Ni resulted in reduced the carbide size, changed connectivity and an increase in volume fraction. An alteration in the solidification range resulted in the improvement of the structure because of Ni addition. The high Cr iron, hypoeutectic, but Ni changed the chemical content of the alloy adjacent to the single-phase structure and ultimately, a fully austenitic zone for sample with >6 wt.% Ni. There is a shorter solidification range in the single-phase structure, and the following tendency is observed: a finer structure of the alloy is acquired in case of a shorter solidification range. Additionally, the austenite phase is stabilized by Ni, subject to the impact on solidification range, and an increase in the rate of Ni and a decline in the austenite content demonstrate this. This alteration was graded to 2 wt.% Ni (S2), yet heavy for 7 wt.% Ni (S5). A significant difference was observed in the microstructure of the 6 wt.% Ni (S4) addition and materials with eutectoid transformation properties created pearlitic structures. An alteration in the ratio of secondary carbides represents a complementary impact of adding Ni. The accessible the carbon consequent for deposition on tempering was increased by Ni, subject to carbide fraction (Lin et al., 2010Lin, C.M., Chang, C.M., Chen, J.H., Wu, W. (2010). The effects of additive elements on the microstructure characteristics and mechanical properties of Cr-Fe-C hard-facing alloys. J. Alloys Compds. 498 (1), 30-36. https://doi.org/10.1016/j.jallcom.2010.03.127.; Chandan et al., 2022Chandan, A.K., Kishore, K., Hung, P.T., Ghosh, M., Chowdhury, S.G., Kawasaki, M., Gubicza, J. (2022). Effect of nickel addition on enhancing nano-structuring and suppressing TRIP effect in Fe40Mn40Co10Cr10 high-entropy alloy during high-pressure torsion. Int. J. Plast. 150, 103193. https://doi.org/10.1016/j.ijplas.2021.103193.). The mechanism of wear was predominantly oxidational at every load. However, at lower loads, there was an indication of abrasive grooving. Load generally increased the oxide layer thickness. At low loads, metallic abrasion was demonstrated by the wear debris composition, however, just oxidational wear was observed at the load of 140 N. There was an alteration in the composition of the oxide debris by the load. XRD patterns of worn particulates obtained from the surface of sample S5 during wear tests are displayed in Fig. 11. At low loads, it was totally Fe₂O₃, and at high loads, it was the Fe₂O₃ and Fe₃O₄ combination.

With an increase in the load, the M₇C₃ content in the oxide increased. A reduction in the secondary dendrite arm spacing constituted the single important microstructural impact (Biner, 1985Biner, S.B. (1985). The role of eutectic carbide morphology on the fracture behaviour of high-chromium cast irons-I. Austenitic alloys. Canadian J. Metal. Mater. Sci. 24 (2), 155-162. https://doi.org/10.1179/cmq.1985.24.2.155.; Laird et al., 2000Laird, G., Gundlach, R., Röhrig, K. (2000). Abrasion resistant cast iron handbook. American Foundry Society, USA, pp. 1-222.). Nevertheless, there was no major change in the carbide volume fraction. No significant impact of Ni on wear behavior was observed particularly at 140 N. A load 90-140 N slightly reduced the wear coefficient for 1-2 wt.%.