Influence of Thermal Processing Factors, Linked to the Destabilisation of Austenite, on the Microstructural Variation of a White Cast Iron Containing 25% Cr and 0.6% Mo

Abstract

Hypoeutectic white cast irons containing 25% Cr are used in ore-processing industries due to their high resistance to erosive wear. Applying a Design of Experiments (DoE), the aim of this study is to analyse the influence of thermal processing factors on the microstructural variation of a white cast iron containing 25% Cr and 0.6% Mo. The carbides present in the as-cast state are of the M₇C₃, M₂C, and M₃C types. M₂C carbides precipitate on the eutectic M₇C₃ carbides favoured by heterogeneous nucleation conditions. Two kinetics compete during the destabilisation of austenite. One dissolves those eutectic carbides precipitated as a result of non-equilibrium solidification (M₇C₃ and M₂C), while the other enables the precipitation of secondary M₇C₃ and M₂₃C₆ carbides. The M₇C₃ carbides begin to precipitate first. Low destabilisation temperatures and short dwell times are insufficient to dissolve the precipitated eutectic carbides from non-equilibrium conditions, thus favouring the presence of M₂C carbides, which are associated with Mo. The factor that has the greatest influence on hardness is the tempering temperature. The optimal tempering temperature is found to be 500 °C. Short tempering times maintain the distortion of the ferrite unit cell. The precipitation of Cr carbides during tempering requires a temperature of 500 °C and the prior dissolution of the carbon precipitated during the initial stages of said tempering. With short tempering times, the Cr atoms still remain dissolved in the ferrite, distorting its unit cell and increasing the hardness of the matrix constituent of the alloy.

1. Introduction

The microstructure of hypoeutectic white cast irons containing between 15% and 30% Cr is mainly composed of eutectic carbides in an austenite matrix, or of any of the phases or constituents resulting from their transformation [1,2]. The austenite will be supersaturated mainly in Cr and C. This austenite can be destabilised by means of a high-temperature heat treatment, causing the precipitation of Cr-rich carbides. This precipitation results in an increase in the Ms temperature, thus reducing the risk of cracking during cooling [3,4]. Wiengmoon et al. [5] conclude that the secondary carbides which precipitated during the destabilisation of austenite at 1000 °C are M₇C₃-type carbides in cast irons containing 20% Cr. However, secondary carbides with M₇C₃ and M₂₃C₆ stoichiometries would be jointly present in cast irons containing 27% Cr. Guitar et al. [6] conclude that M₇C₃ carbides are the first to precipitate during the destabilisation of the austenite heating process. During the dwell time at the destabilisation temperature, they would transform into M₂₃C₆ carbides. Furthermore, Guitar et al. [6] state that additional precipitation of secondary carbides takes place during cooling in the 980–750 °C range. Hardness increases when M₇C₃ secondary carbides predominate over M₂₃C₆ carbides [5]. Bedolla et al. conclude that the volume fraction of secondary carbides increases at low destabilisation temperatures (900 °C) and high dwell times (8 h) [7]. Excessive dwell times at the destabilisation temperature lead to thickening of the secondary carbides and a reduction in hardness [3]. Khanitnantharak et al. conclude that high destabilisation temperatures favour the presence of retained austenite [8]. The presence of Mo in the chemical composition favours the precipitation of eutectic M₂C carbides [9,10,11,12,13], leading to an increase in both hardness and wear resistance [9,12,14]. Mo can also dissolve in mixed carbides of the M₇C₃ type [15], replacing the Fe atoms and producing a distortion of their unit cell without having a significant influence on hardness. A suitable tempering temperature favours the reduction in retained austenite via a second destabilisation of this austenite [16,17,18]. However, an excessive tempering temperature could result in the redissolution of the C atoms in the tempered martensite matrix [12]. This carbon would come from carbides precipitated at the lowest tempering temperatures. The optimal tempering temperature to increase the hardness and wear resistance appears to be around 500 °C [4]. The presence of hard carbides gives these cast irons a high resistance to wear, which is why they are widely used in ore-processing industries [1,19]. Alloys in the 25–30% Cr range have a very high hardness [20]. The tempering temperature is decisive in both the wear resistance and hardness of the constituent matrix of these cast irons. Low tempering temperatures do not completely destabilise the austenite [21]. Solidification and thermal processing parameters can alter the microstructure of the material and its properties and hence influence its in-service performance [22]. The aim of this study is to analyse the influence of thermal processing factors on the microstructural variation of a white cast iron containing 25% Cr and 0.6% Mo via the application of a Design of Experiments (DoE), analysing how the modifications of those factors that affect the destabilisation of austenite influence the percentage of the different carbides that are formed, the percentage of martensite, and the percentage of retained austenite. A secondary goal is to determine how the variation of these processing factors would affect the hardness of the material. Five factors were analysed, performing eight experiments [23]. The results will allow the manufacturers of these alloys containing 25% Cr to correlate the parameters of thermal processing with the microstructure and hardness of the material.

2. Materials and Methods

Applying a DoE, the aim is to modify certain thermal processing parameters in a deliberate and controlled manner so as to analyse the microstructural variations produced in the material. In other words, it is a matter of determining which factors have a significant influence on this microstructure and in what way the variation of these factors exerts an influence.

Table 1. Factors and Levels.

Factors		Levels
Code	Metallurgical parameter	Level −1	Level +1
A	Destabilisation temperature of austenite (°C)	900	1000
B	Dwell time at the destabilisation temperature (h)	4	8
C	Tempering temperature (°C)	500	600
D	Number of tempers	1	2
E	Tempering time (h)	2	4

shows the analysed factors and the levels of analysis. The proposed dwell times for the destabilisation of austenite, Factor B, are higher than those usually employed. These longer dwell times are intended to promote the precipitation of secondary carbides and an increase in the Ms Temperature so as to avoid the presence of retained austenite after quenching as far as possible.

The effect of a factor is measured as the consequence of its variation on a certain microstructural feature of the material. This type of effect is called a main effect. The effect of one factor may possibly depend on the value that another takes. When this occurs, these factors are said to interact. The influence of the main effects is greater than that of the 2-factor interactions. In turn, the influence of the 2-factor interactions is greater than that of the 3-factor interactions, and so on successively. In practical terms, it is sufficient to consider only the main effects and the 2-factor interactions, which allows the number of experiments to be reduced [23]. In this study, the effect of 5 factors performing 8 experiments has been analysed. If the aim were to analyse all the possible interactions with these 5 factors, it would be necessary to perform 32 experiments (2⁵ = 32). However, in this case, only 8 effects (2⁵⁻²) have been estimated, which therefore means a 1/4 (32/4 = 8) fractional factorial design.

Table 2. Array of Experiments.

Experiment	A	B	C	D	E	Confounding Pattern
1	−1	−1	−1	+1	+1	A + BD + CE B + AD C + AE D + AB E + AC BC + DE BE + CD
2	+1	−1	−1	−1	−1
3	−1	+1	−1	−1	+1
4	+1	+1	−1	+1	−1
5	−1	−1	+1	+1	−1
6	+1	−1	+1	−1	+1
7	−1	+1	+1	−1	−1
8	+1	+1	+1	+1	+1

shows the array of experiments. Columns D and E have been respectively constructed from the product of columns A × B and A × C [23]. The “Confounding Pattern” column indicates only the main effects and those 2-factor interactions whose effects are confounded with the main effects.

The effects are linear combinations of the analysed responses or features. Therefore, according to the Central Limit Theorem, they will follow a normal law. The representation of the distribution function of the N(0,σ) law on a normal probabilistic plot scale will be a straight line that passes through the coordinate point (0,50). If any of the effects followed a different normal law, it would not appear aligned along this straight line. Those factors associated with effects that do not follow the distribution function of the N(0,σ) law are considered significant and will hence be found at a certain distance from the straight line on the normal probabilistic plot. If any effect deviates to the right of this straight line, that would mean that this factor would produce an increase in the analysed feature when placed at its +1 level. Likewise, if any effect deviates to the left of this straight line, that would indicate that the factors associated with this effect produces an increase in the measured feature when placed at their −1 level. The features studied in this paper are:

• The Vickers hardness of the material. The applied load was 300 N, while the hardness value was calculated as the average value obtained from 10 indentations.
• The Vickers hardness of the constituent matrix, formed by tempered martensite and secondary carbides. In this case, the applied load was 0.5 N, while the hardness value was calculated as the average value obtained from 20 indentations.
• The following microstructural features:

  ○

  Percentage by weight of austenite

  ○

  Percentage by weight of tempered martensite (ferrite)

  ○

  Volume of the unit cell of the tempered martensite (ferrite).

  ○

  Percentage by weight of the M₇C₃, M₂₃C₆ and M₂C carbides.

The microstructural features were determined by X-ray diffraction on a SEIFERT XRD 3000 T/T diffractometer (Baker Hughes, Celle, Germany). The radiation was emitted via a fine-focus Mo tube at a working power of 40 kV × 40 mA and monochromatised to the K α doublet: λ₁ = 0.709316 Å and λ₂ = 0.713607 Å. The diffracted intensity was determined in a 2θ range from 7° to 57°. The Rietveld structural refinement method was used to determine the percentage of the crystalline phases via fitting of the diffractograms [24]. To this end, following the recording of the diffraction figures, a structural refinement was carried out using the crystallographic information files present in the Inorganic Crystal Structure Database (ICSD), FIZ Karlsruhe, Germany. The program employed for this purpose was FullProf.2k, version 6.20 (2018). The increase in width observed in the peaks of the majority phases were modelled using Stephens’ formulation [25], which is implemented in the aforementioned analysis program.

The optical microscope used was a NIKON Epiphot 200 (Nikon, Tokyo, Japan), and the images were obtained using Beuhler Omnimet Enterprise image analyser software (OmniMet Enterprise, Beuhler). The scanning electron microscope employed was a JEOL JSM-5600 (JEOL, Nieuw-Vennep, The Netherlands), equipped with the characteristic energy-dispersive X-ray (EDX) microanalysis system.

3. Results

Table 3. Chemical Composition (% by weight).

C	Si	Mn	Cr	Mo
2.7	1.1	0.8	25.0	0.6

shows the chemical composition of the alloy under study. Figure 1 shows the microstructure in the as-cast state following solidification in a sand mould. Figure 1a,b shows that the microstructure is mainly composed of proeutectic austenite and a eutectic constituent. The proeutectic austenite presents a dendritic growth model, the presence of the eutectic constituent being observed between the dendritic “arms”. M₇C₃ eutectic carbides can be seen in the latter constituent. Figure 1c,d show the presence of carbides, both in the proeutectic austenite and in the austenite of the eutectic constituent. These carbides have been identified as belonging to two different types. Mixed carbides with a grey colouring and diameter greater than 1 micron are of the M₇C₃ type. However, carbides of the M₃C type are also observed, which have a brighter colouring and a smaller size: less than 1 micron.

Table 4. Results of the microanalysis performed by SEM-EDX (energy-dispersive X-ray), corresponding to Figure 1d. (at. %).

Spectrum	%C	%Cr	%Fe	%Mo
1	42.91	31.13	25.03	0.94
2	41.15	32.46	25.69	0.69
3	22.42	6.58	71.00	-
4	29.47	8.76	61.77	-

shows the results of the microanalysis performed on these two types of carbides (spectra 1 to 4). The hardness of the material in this state was 310 HV.

Figure 2 shows the presence of eutectic carbides of the M₂C type, associated with Mo. Figure 2a was taken with secondary electrons and Figure 2b was taken with backscattered electrons. The difference in contrast in the latter image reveals the presence of M₂C carbides associated with Mo. Their precipitation seems to be favoured by heterogeneous nucleation conditions from the eutectic M₇C₃-type carbides [4].

Table 5. Results of the microanalysis performed by SEM-EDX, corresponding to Figure 2a (at %).

Spectrum	%C	%Cr	%Fe	%Mo
1	24.23	7.54	40.31	27.92

shows the results of the microanalysis performed on this type of carbide (spectra 1)

Figure 3 shows the overall fittings obtained by means of Rietveld structural refinement. The red marks indicate the observed intensities; the black line indicates the intensity calculated according to the Rietveld structural model; the blue line indicates the difference between the two; while the vertical segments indicate the angular positions of the different identified phases.

Table 6. Percentage by weight of the precipitated phases and the volume of the unit cell of the tempered martensite.

Experiment	Rietveld Fitting	Phases	wt %	vol. (Å3)
1	Rwp = 15.7 Rexp = 28.16 Chi2 = 0.31	Martensite (α’)	74	23.638
		Cr7C3 carbide	18
		Cr23C6 carbide	2
		Mo2C carbide	5
2	Rwp = 13.3 Rexp = 28.11 Chi2 = 0.22	Martensite (α’)	77	23.781
		Cr7C3 carbide	19
		Cr23C6 carbide	1
		Mo2C carbide	2
3	Rwp = 13.5 Rexp = 27.71 Chi2 = 0.24	Martensite (α’)	84	23.626
		Cr7C3 carbide	13
		Cr23C6 carbide	1
		Mo2C carbide	2
4	Rwp = 13.1 Rexp = 28.82 Chi2 = 0.21	Martensite (α’)	88	23.801
		Cr7C3 carbide	9
		Cr23C6 carbide	-
		Mo2C carbide	3
5	Rwp = 13.9 Rexp = 28.74 Chi2 = 0.23	Martensite (α’)	80	23.718
		Cr7C3 carbide	16
		Cr23C6 carbide	-
		Mo2C carbide	4
6	Rwp = 16.9 Rexp = 27.63 Chi2 = 0.37	Martensite (α’)	72	23.722
		Cr7C3 carbide	25
		Cr23C6 carbide	1
		Mo2C carbide	1
7	Rwp = 12.8 Rexp = 27.99 Chi2 = 0.21	Martensite (α’)	87	23.682
		Cr7C3 carbide	12
		Cr23C6 carbide	1
		Mo2C carbide	-
8	Rwp = 14.3 Rexp = 28.23 Chi2 = 0.26	Martensite (α’)	87	23.737
		Cr7C3 carbide	13
		Cr23C6 carbide	-
		Mo2C carbide	-

shows the percentages by weight and the network parameters of the main crystalline phases detected by XRD in each of the experiments. The goodness of fit is defined by the agreement factor, Rwp, the index, Rexp, and the relationship between their squares: Chi² = (Rwp/ Rexp)²

Table 7. Calculation of Effects.

(a) Percentage by Weight of Martensite and Volume of Its Unit Cell
Experiment	Martensite (α’)						Calculated Effects
	(wt. %)	Effect	(vol. Å3)		Effect
1	74	81.12	23.638		23.713		Average
2	77	−0.2	23.781		0.0942		A + BD + CE
3	84	10.7	23.626		−0.0032		B + AD
4	88	0.7	23.801		0.0032		C + AE
5	80	2.2	23.718		0.0207		D + AB
6	72	−3.7	23.722		−0.0647		E + AC
7	87	0.2	23.682		−0.0072		BC + DE
8	87	1.7	23.737		0.0047		BE + CD
(b) Percentage by Weight of Carbides
Experiment	Cr7C3 Carbide		Cr23C6 Carbide		Mo2C Carbide		Calculated Effects
	(wt %)	Effect	(wt %)	Effect	(wt %)	Effect
1	18	15.62	2	0.75	5	2.12	Average
2	19	1.7	1	−0.5	2	−1.25	A + BD + CE
3	13	−7.7	1	−0.5	2	−1.75	B + AD
4	9	1.7	0	−0.5	3	−1.75	C + AE
5	16	−3.2	0	−0.5	4	1.75	D + AB
6	25	3.2	1	0.5	1	−0.25	E + AC
7	12	−0.2	1	0.5	0	−0.75	BC + DE
8	13	−0.7	0	−0.5	0	−0.25	BE + CD
(c) Overall Hardness and Microhardness of the Microstructural Matrix
Experiment	Overall Hardness (HV)		Hardness of the Matrix (HV)				Calculated Effects
	Kgf/mm2	Effect	Kgf/mm2		Effect
1	484	517.62	371		398.00		Average
2	650	74.2	488		63.0		A + BD + CE
3	589	24.7	412		2.5		B + AD
4	623	−137.7	472		−75.5		C + AE
5	408	−44.2	333		−26.5		D + AB
6	479	−25.7	395		−25.5		E + AC
7	441	−14.2	350		−10.0		BC + DE
8	467	21.7	363		2.0		BE + CD

shows the results of the effects, corresponding to the confounding pattern given in the array of experiments outlined in Table 2. The row corresponding to the average shows the average value obtained considering all 8 experiments.

Figure 4 shows the representation of these effects on a normal probability plot, highlighting those that have a significant effect on these responses.

Figure 4a shows that the factors that have a significant effect on the percentage by weight of tempered martensite are B (dwell time at the austenisation temperature) and E (tempering time). Thus, if the aim is to increase this percentage, factor B should be placed at its +1 level (8 h) and Factor E should be placed at its −1 (2 h) level. Figure 4b shows that Factor A (austenisation temperature) and Factor E (tempering time) have a significant effect on the volume of the unit cell of the tempered martensite. An increase in the distortion of this phase would be obtained by placing Factor A at its +1 level (1000 °C) and Factor E at its −1 level (2 h). The significant effect of interaction AB is also observed, which means that the effect of the austenitisation temperature is reinforced if factor B (austenitisation time) is placed at its +1 level (8 h). It can de deduced from both these results that 2 complementary kinetics compete during the destabilisation treatment of austenite. One of them is related to the destabilisation of austenite itself, and the other is related to the dissolution of those eutectic carbides that would have precipitated as a result of a possible non-equilibrium solidification [22]. High austenitisation temperatures and long dwell times at this temperature favour the dissolution of these eutectic carbides, precipitated in non-equilibrium conditions. This involves the dissolution not only of C atoms, but also of Cr atoms. The former precipitate in the early stages of tempering, forming ε (cementitic) carbides. However, the latter continue to occupy substitution positions in the unit cell of the tempered martensite (ferrite). Carbon atoms need to be available for the precipitation of the Cr carbides during tempering. This requires the prior dissolution of the cementitic carbides precipitated during the tempering process following the decomposition of their constituent atoms [26]. The precipitation of the Cr carbides during tempering requires longer dwell times that allow the prior dissolution of the carbon precipitated during the initial stages of the tempering process and their subsequent precipitation as M₇C₃ carbides. All this takes time, which justifies the finding that the tempered martensite presents network distortions due to Cr atoms which have not precipitated during this tempering when short tempering times are employed.

Figure 4c shows that Factor B (dwell time at the destabilisation temperature of austenite) has a significant effect on the percentage by weight of the M₇C₃ carbides: placing this factor at its −1 level (4 h) leads to an increase in the content in these carbides. Dwell times at the austenitisation temperature of 4 h may be too short for dissolution of the eutectic carbides, solidified under non-equilibrium conditions, to be observed. Therefore, it turns out that the dwell time of 4 h would be sufficient for the destabilisation of the austenite, but it is too short for the dissolution of the precipitated carbides in non-equilibrium conditions. This evidences that the first carbides formed by the destabilisation of austenite are of the M₇C₃ type. This effect is reinforced by interaction AB: placing Factor A (austenitisation temperature) at its +1 level (1000 °C) would lead to an increase in the rate of precipitation of M₇C₃ carbides. These results are in line with the recent conclusions reached by Guitar et al. [6].

Figure 4d shows that no effect is found to significantly influence the percentage by weight of M₂₃C₆-type carbides. This does not contradict the previous results reported by Guitar et al. [6] or by Wiengmoon et al. Rather, it is possible that the redissolution of eutectic carbides precipitated due to a possible non-equilibrium solidification, observed with long dwell times at the austenitisation temperature, does not enable us to verify the partial transformation of M₇C₃ carbides into M₂₃C₆ carbides, or their precipitation.

Figure 4e shows the significant influence of interaction AB on the precipitation of M₂C eutectic carbides.

Table 8. Analysis of the effects of interaction AB on the percentage by weight of the M₂C carbides.

AB (wt % Mo2C)	−1	+1
−1	4.5	1
+1	1.5	1.5

shows the analysis of this interaction. It is concluded that their presence is favoured when Factor A (austenitisation temperature) and Factor B (dwell time at said temperature) are simultaneously placed at their −1 levels. It is probable that the M₂C eutectic carbides were formed as a result of non-equilibrium solidification and that higher dwell times or higher temperatures enable the dissolution of these carbides.

Figure 4f shows the significant effect of Factor C (tempering temperature) on the overall hardness of the material: placing this factor at its −1 level (500 °C) produces an increase in hardness. This figure also shows the significant effect of Factor A (austenisation temperature) and of interaction CE.

Table 9. Analysis of the effects of Factor A and the interaction CE on the hardness of the material.

A	HV	CE (HV)	−1	+1
−1	481	−1	637	537
+1	555	+1	425	473

simultaneously analyses both possibilities; placing Factor E (tempering time) at its −1 level (2 h) is found to reinforce the effect of Factor C at its −1 level (500 °C). That is, short tempering times reinforce the effect of the tempering temperature of 500 °C. It is precisely at this temperature that the precipitation of M₇C₃ carbides occurs during tempering [26]. With shorter tempering times, the tempered martensite (ferrite) maintains a greater distortion of its network.

Figure 4g shows the significant effect of Factor C (tempering temperature) on the overall hardness of the constituent matrix, which is formed by tempered martensite and secondary carbides. Placing this factor at its −1 level (500 °C) leads to an increase in this hardness. The significant effect of Factor A and of interaction CE can also be seen in this figure. Similar to the previous analysis,

Table 10. Analysis of the effects of Factor A and the interaction CE on the hardness of the matrix constituent.

A	HV	CE (HV)	−1	+1
−1	367	−1	480	391
+1	429	+1	342	379

simultaneously analyses both possibilities; placing Factor E (tempering time) at its −1 level (2 h) is found to reinforce the effect of factor C at its −1 level (500 °C). On the other hand, high temperatures, or excessively long tempering times, can lead to the coalescence of secondary carbides precipitated during tempering, thereby negatively affecting hardness [26].

Figure 5 shows a representative micrograph of each experiment that allows us to appreciate the general microstructure of these cast irons.

4. Conclusions

This paper has analysed the influence of different thermal processing factors on the microstructural variation of a white cast iron containing 25% Cr and 0.6% Mo via the application of a Design of Experiments (DoE). The main conclusions may be summarised as follows:

• The carbides present in the microstructure are of the M₇C₃, M₂₃C₆ (both associated with Cr), and M₂C (associated with Mo) types. The M₇C₃ and M₂C carbides form part of the eutectic constituent. Part of this eutectic constituent was formed as a result of non-equilibrium solidification.
• Two complementary kinetics compete during the high-temperature destabilisation treatment of the austenite. One of them is related to the destabilisation of austenite itself, via the precipitation of secondary carbides, and the other is related to the dissolution of those eutectic carbides that would have precipitated as a result of possible non-equilibrium solidification. The latter kinetics is favoured if the treatment is carried out at 1000 °C, compared to lower temperatures.
• The carbides that form during the destabilisation of austenite at high temperatures are M₇C₃ and M₂₃C₆ carbides. The M₇C₃ carbides begin to precipitate first.
• Mo tends to segregate into the eutectic fluid and to precipitate via heterogeneous nucleation on the M₇C₃ carbides. Low destabilisation temperatures and short dwell times are insufficient for the dissolution to take place, which means that these conditions favour the presence of M₂C carbides.
• The factor that most influences hardness is the tempering temperature. A temperature of 600 °C is found to be excessive. The optimal tempering temperature is 500 °C. The highest hardness would be achieved with short tempering times (2 h).
• Short tempering times maintain the distortion of the tempered martensite unit cell. Precipitation of the Cr carbides during tempering requires longer dwell times that allow the prior dissolution of the carbon precipitated during the initial stages of the tempering process and their subsequent precipitation as M₇C₃ carbides.