Materials Processing
Erosion Resistance of Heat-Treated Aluminum Cast Iron to Aluminum Alloy Melt
2024 Volume 65 Issue 5 Pages 534-540
Details
2024 Volume 65 Issue 5 Pages 534-540
Improvement in the erosion resistance of permanent molds to aluminum alloy melt is required. Since chemically and mechanically stable layer containing aluminum oxides can be formed on the cast irons by aluminum addition and heat treatment, the layer would improve the erosion resistance of the cast iron in the running melt. In this study, cast irons with different aluminum contents were fabricated, then they were heat-treated to form the oxide layer on the surface. The optimum heat treatment conditions to form the stable layer and the erosion behavior of the cast irons were clarified. Heat treatment in air was found to result in the formation of a layer that consists of the oxides, such as Fe2O3, Fe3O4 and Al2O3 on the cast irons. Based on the castability of the cast irons and morphology of the layer, it was concluded that the addition of 3% aluminum in the cast iron and heat treatment at 1173 K for 10 hours was the most suitable treatment. The heat treatment drastically improved the erosion resistance of the cast iron. The addition of magnesium to the melt temporarily decreased the time to erosion, but increased the time to erosion again when more than 0.75% magnesium.
This Paper was Originally Published in Japanese in J. JFS 94 (2022) 11–17. A typographical error is modified in Sec. 2.
Aluminum (Al) alloy castings are mainly produced by the permanent mold casting techniques such as the die casting and low-pressure casting. Since the long-term use of the permanent molds often undergoes erosion due to the aluminum alloy melt flow, the improvement in the erosion resistance of the molds is required. Although coating techniques have been used for improving the erosion resistance,1–3) the drawbacks, such as high cost and complex repair after the damage of the coating film, still remain. Although the addition of alloying elements, such as chromium, molybdenum and vanadium, has also been performed,4–6) these elements are expensive and have geopolitical concerns. Based on these characteristics, the inexpensive materials and processes for improving the erosion resistance are required. The addition of Al to cast iron, which is the popular alloy as the mold material, followed by heat treatment in air to form alumina (Al2O3) can be expected to improve the erosion resistance of the cast iron with a low cost. Takamori et al.7–9) reported that the addition of Al and heat treatment improved the oxidation resistance at high temperature and wear resistance of the cast iron. However, the erosion resistance of the cast irons in the Al alloy melt have not been reported by them. The erosion resistance of the cast irons with chromium and molybdenum was investigated,10) and it was reported that the erosion resistance was improved due to the formation of an alumina layer. In this case, the decarburized layer was formed due to the heat treatment in air followed by the reaction between the residual oxygen in the layer and Al alloy melt to form the alumina layer. However, this technique cannot be used for the vacuum die cast, in which oxygen to form the alumina was negligible. Although the chemical composition of the Al melt would affect the erosion resistance of the cast iron, it was not reported by them.
In the present study, Al cast irons were heat-treated in air, and their erosion resistance was investigated. The effects of the Al content in the cast iron, heat-treatment condition and the chemical composition of the Al alloy melt on the erosion resistance of the cast iron were discussed.
The raw materials, which are the pig iron ingot, steel scrap, Fe-Si alloy and pure (99.99%) Al ingot, were melted at 1723 K in a graphite crucible using a high-frequency induction furnace, and cast in permanent molds for emission spectro-photometry and sand molds. The composition of the cast iron specimens was 3.2 mass%C, 2.0 mass%Si (hereinafter, mass% is abbreviated as %), and the Al content varied from 1 to 7 mass%. The chemical composition of the cast iron specimens is shown in Table 1.
;%0A%09%09%09newWindow.document.open();%0A%09%09%09newWindow.document.write('<img src=%22./Graphics/F-M2024804tab01.jpg%22>');%0A%09%09%09newWindow.document.close();%0A%09%09)
The sand-cast specimen (30-mm diameter and 120-mm length) was machined to obtain bar-shaped test pieces (8.0-mm diameter and 50-mm length). The test piece was heat-treated in air using a muffle furnace. The holding temperatures were 973, 1173, and 1373 K, and the holding times were 1, 4, and 10 h. After the holding period, the specimens were furnace-cooled. The appearance of a heat-treated test piece is shown in Fig. 9 and compared to that after immersion in the Al alloy melt.
A schematic illustration of the erosion test is shown in Fig. 1. The test piece was immersed in the Al-9Si-xMg (x = 0∼1.5) alloy melt in the depth of 25 mm at 973 K in the graphite crucible and stirred by rotation. The test piece was removed from the melt after every hour of immersion, then the melt that adhered on the surface of the test piece was removed using a metal spatula to observe the erosion feature of the test piece, and the diameter of the immersed part was measured. The immersion, stirring and measuring the diameter were repeated until the erosion occurred. It was determined that the erosion began when more than a 10% decrease in the diameter was confirmed compared to that of the test piece before the immersion. The rotating velocity of the test piece was 377 mm/s (180 rpm), assuming the melt flow velocity through the practical mold parts. The Mg content in the melt, which is reported as that would severely affect the behavior of the reaction with the oxides,11–14) was varied from 0 to 1.5%. The Mg content was based on that in the commercially-available Al alloy castings. Since the long-time stirring promotes the oxidation of the melt and increases the viscosity of the melt, the melt in the crucible was replaced every several hours. Optical microscopy, X-ray diffractometry (XRD) (MiniFlexII, RIGAKU) and Electron Probe Microanalysis (EPMA) (JXA-8530F, JEOL) were carried out to analyze the microstructure of the test pieces.
Schematic illustration of erosion test.
Figure 2 shows the optical micrographs of the cast iron specimens in the as-cast condition. The Al-free specimen (FC) consisted of the A-type flake graphite and the pearlite matrix. The pearlite decreased and the ferrite increased as the Al content increased, and the matrix in specimen 3AL was almost ferrite. For specimens 5AL and 7AL, the thin flake and bulky graphite and a gray phase, which was different from the pearlite in the matrix, were clearly observed. Based on the qualitative analysis using EPMA and XRD and the previous reports of the Al cast irons,7,15) this phase observed in specimens 5AL and 7AL was determined to be the κ phase (Fe3AlCX). The elemental mapping using EPMA revealed that the Al was detected not only from the κ phase, but also from the matrix, indicating that some amount of the Al had dissolved in the matrix.
Optical micrographs of cast iron specimens.
Figure 3 shows the cross-sectional micrographs in the vicinity of the surface of the cast iron test pieces heat-treated at 973 K. The difference in the thickness of the oxide layer or the decarburized layer between the test pieces with different Al contents was not clear after a 1-h heating. The thickness of these layers increased after a 10-h heating for the 1%Al cast iron test piece (1AL). The layers in the test pieces with the higher Al content were thinner than that of 1AL.
Cross-sectional microstructure near cast iron specimen surface (heat-treated at 973 K).
In order to increase the thickness of the oxides layer, the heat treatment temperature was increased to 1173 and 1373 K. Figure 4 shows the SEM and X-ray images of the cross-section near the surface of the cast iron specimens heat-treated at 1173 and 1373 K. The dark parts in the right-hand side of the SEM images are embedded resin for polishing the test pieces. When the 1AL test piece was heat-treated at 1173 K for 1 h (Fig. 4(a)), oxygen (O), iron (Fe), Al and silicon (Si) were detected from the surface, indicating that the surface was covered with oxides consisting of these elements. Although the oxide layer can be more clearly observed when the heat treatment time was increased to 10 h (Fig. 4(b)), the luminance of Al was not high near the surface due to the low Al content. The gap between the oxides layer and cast iron substrate was observed when the heat treatment temperature was increased to 1373 K (Fig. 4(c)). Although the thickness of the oxides layer increased by increasing the heat treatment time, the gap was more remarkable (Fig. 4(d)). The increase in the Al content to 3% made the detection of Al near the surface stronger (Fig. 4(e)). Increase in the heat treatment time to 10 h increased the luminance and distribution area of O near the surface, indicating the promotion of the oxidation (Fig. 4(f)). Under this condition, the interface between the oxides layer and substrate was indistinct, showing a gradual compositional distribution with a decrease in the oxide toward the inside and suggesting a strong interfacial bond. When the heat treatment was carried out at 1373 K, voids were clearly observed near the interface between the oxide layer and the substrate (Fig. 4(g)(h)), suggesting that the bonding strength between them was insufficient. Based on the distribution of carbon (C) and O, the oxide layer and decarburized layer in the 3AL test piece were thinner than those in the 1AL test piece.
SEM and X-ray images of cross-section near surface of heat-treated cast iron specimens.
Enlarged SEM and X-ray images of the cross-section near the surface of the heat-treated cast iron specimens (3AL, 1173 K), in which the formation of oxides near surface was clear and the interfacial bond was sound without any voids or cracks, are shown in Fig. 5. While the existence of the residual graphite (C in the figure), which was surrounded by the oxides containing Al, indicated in the specimen heat-treated for 1 h, most of the graphite was replaced with Al and O after the heat treatment for 10 h.
Enlarged SEM and X-ray images of cross-section near surface of heat-treated cast iron specimens (3AL, 1173 K).
The diameter of the test piece after the heat treatment was greater than that of the as-received one; the diameter of the 3AL test piece was 8.8 and 10.5 mm after the heat treatment at 1173 K for 10 h and 1373 K for 10 h, respectively.
Figure 6 shows an XRD pattern of the oxides chips obtained by turning near the surface of the 5AL test piece heat-treated at 1373 K for 4 h. α-Fe2O3, Fe3O4, and α-Al2O3 were detected from the chips, indicating that these oxides were contained near the surface of the heat-treated test piece.
XRD pattern of oxides machined from heat-treated cast iron test piece (5AL, 1373 K-4 h).
By combining these results with the previous reports16,17) on the oxidation behavior at high temperature of the flake graphite cast iron, it can be concluded that the reactions expressed by the following equations occurred during the heat treatment and a layer consisting of the multiple oxides was formed in this study.
| \begin{equation} \text{4Fe} + \text{3O$_{2}$} \rightarrow \text{2Fe$_{2}$O$_{3}$} \end{equation} | (1) |
| \begin{equation} \text{3Fe} + \text{2O$_{2}$} \rightarrow \text{Fe$_{3}$O$_{4}$} \end{equation} | (2) |
| \begin{equation} \text{4Al} + \text{3O$_{2}$} \rightarrow \text{2Al$_{2}$O$_{3}$} \end{equation} | (3) |
As shown in Figs. 3 and 4, the increase in Al content decreased the thickness of the oxides and the decarburized layers would be attributed to the Al2O3, which was formed on the surface and suppressed the diffusion of oxygen into the decarburized layer. Although Takamori et al.8,9) reported similar tendency of the effects of the heat treatment on the oxidation behavior of the Al cast irons, the chemical composition and heat treatment condition are different from those in this study. Considering the castability and bonding of the oxide layer, the optimal condition in this study was the 3%Al content in the cast iron heat-treated at 1173 K for 10 h.
3.3 Erosion resistance of specimensFigure 7 shows the relationship between the heat treatment time at 973 K and the time to erosion of the cast iron test pieces in the Al-9Si-0.5Mg alloy (JIS-AC4A) melt. The time to erosion increased as the heat treatment time increased, and the increase in time to erosion with the Al content in the specimen was more remarkable as the heat treatment temperature increased.
Relation between heat treatment time (973 K) and time to erosion of cast iron specimens in Al-9Si-0.5Mg alloy melt.
The effect of the Mg content in the Al alloy melt on the erosion resistance of the cast iron specimens is shown in Fig. 8. In this case, the specimens with 3%Al and heat-treated at 1173 K for 10 hours, which was the optimal condition from the viewpoint of the formation feature of the oxide layer, was used. The time to erosion was 49 h when the Mg-free alloy melt was used. It decreased as the Mg increased, was the shortest when the Mg was around 0.75%Mg, and became longer again when the Mg exceeded that level. The time to erosion in the JIS-AC4A (0.5%Mg) alloy melt of the Al-free cast iron (FC) and 3AL test pieces without heat treatment is also shown in this figure, indicating that the effect of the Al addition and heat treatment improved the erosion resistance.
Effect of Mg content in Al alloy melt on erosion resistance of cast iron specimens (HT: heat-treated at 1173 K for 10 hours).
Figure 9 shows the appearance of the heat-treated (1173 K-10 h) 3AL test pieces lifted from Al-9Si-0.5Mg alloy melt after a 9-h immersion then the residual Al attached on the specimen surface was removed using an aqueous solution of sodium hydroxide. As shown in Fig. 8, the erosion of the test piece had not yet occurred at this time. In order to clarify the erosion feature, the test piece before the immersion is also shown in Fig. 9(a). Although discoloration of the immersed part was observed, its shape was not changed before and after the immersion (Fig. 9(b)).
Appearance of heat-treated (1173 K-10 h) 3AL test pieces ((a) before immersion and (b) extracted from Al-9Si-0.5Mg alloy melt after 9 h immersion).
Figure 10 shows the SEM and X-ray images of the cross-section near the surface of the heat-treated cast iron specimen (3AL, 1173 K-10 h), which was immersed in the Al alloy melts with various Mg contents and removed from the melts before erosion. These images indicated that the MgAl2O4 for the 0.75%Mg alloy and the mixture of MgAl2O4 and MgO for the 0.9%Mg alloy were mainly formed in the vicinity of the oxides and melt during the immersion. Since these newly-formed oxides were not detected from the heat-treated specimens before the immersion, these would be formed due to the chemical reaction between the oxides layer and the melt during the immersion and stirring. The thickness of these products, MgAl2O4 and MgO, was irregular, indicating that some parts of the products had delaminated.
SEM and X-ray images of cross-section near surface of heat-treated cast iron specimen (3AL, 1173 K-10 h) picked up from Al alloy melts with various Mg contents before the erosion.
The Al alloy melt generally reacts with the steels as shown by the following equations to form a seizure.18)
| \begin{equation} \text{2Fe} + \text{5Al} \rightarrow \text{Fe$_{2}$Al$_{5}$} \end{equation} | (4) |
| \begin{equation} \text{Fe} + \text{Al} \rightarrow \text{FeAl} \end{equation} | (5) |
Also in this study, the formation of Fe2Al5 and FeAl were confirmed near the interface between the Al alloy melt and the dissolved cast iron test piece without heat treatment by EPMA.
In contrast, as previously shown (for example, Fig. 10), the chemical reactions as shown by the following equations (6) to (14) would occur when the heat-treated Al cast iron was immersed and maintained in the melt. The reactions shown by eqs. (6) and (7), the reactions by eqs. (8) to (11), and reactions shown by eqs. (12) to (14) would occur in the melt without Mg, with a moderate amount (up to 0.75%) of Mg, and with an additional amount of Mg, respectively.
| \begin{equation} \text{Fe$_{2}$O$_{3}$} + \text{2Al} \rightarrow \text{Al$_{2}$O$_{3}$} + \text{2Fe} \end{equation} | (6) |
| \begin{equation} \text{3Fe$_{3}$O$_{4}$} + \text{8Al} \rightarrow \text{4Al$_{2}$O$_{3}$} + \text{9Fe} \end{equation} | (7) |
| \begin{equation} \text{4Fe$_{2}$O$_{3}$} + \text{6Al} + \text{3Mg} \rightarrow \text{3MgAl$_{2}$O$_{4}$} + \text{8Fe} \end{equation} | (8) |
| \begin{equation} \text{Fe$_{3}$O$_{4}$} + \text{2Al} + \text{Mg} \rightarrow \text{MgAl$_{2}$O$_{4}$} + \text{3Fe} \end{equation} | (9) |
| \begin{equation} \text{4Al$_{2}$O$_{3}$} + \text{3Mg} \rightarrow \text{3MgAl$_{2}$O$_{4}$} + \text{2Al} \end{equation} | (10) |
| \begin{equation} \text{Al$_{2}$O$_{3}$} + \text{MgO} \rightarrow \text{MgAl$_{2}$O$_{4}$} \end{equation} | (11) |
| \begin{equation} \text{Fe$_{2}$O$_{3}$} + \text{3Mg} \rightarrow \text{3MgO} + \text{2Fe} \end{equation} | (12) |
| \begin{equation} \text{Fe$_{3}$O$_{4}$} + \text{4Mg} \rightarrow \text{4MgO} + \text{3Fe} \end{equation} | (13) |
| \begin{equation} \text{Al$_{2}$O$_{3}$} + \text{3Mg} \rightarrow \text{3MgO} + \text{2Al} \end{equation} | (14) |
Tsujikawa et al.14) reported that the addition of SiO2 particles into the Al-12.6Si-xMg alloy melt caused a reaction between the melt and particles, and MgAl2O4 was formed in the melt with less than 10%Mg and MgO was formed in the melt with 20%Mg. They reported that the MgO layer formed at the particle-melt interface was thinner than the MgAl2O4 layer, indicating that the reaction between the particle and melt got slower as the Mg content in the melt increased. They also reported that the volume of the oxide decreased due to the reaction, leading to the fact that the volume ratio of the oxides was 0.876 for (MgAl2O4/SiO2) and 0.993 for (MgO/SiO2). This indicated that the MgO is denser than MgAl2O4, and they concluded that the increase in the Mg content in the melt promoted the formation of the dense MgO, then the diffusion control was replaced from the supply of the Al melt into the SiO2 particle through the gaps in the products, leading to the decrease in the reaction velocity. Based on these findings and chemical reactions shown by eqs. (6) to (14), the volume ratio of oxides formed by the reaction with the Al alloy melt to oxides before the reaction in this study was calculated and shown in Table 2. These results showed that the volume ratio when MgO was formed is greater than that when MgAl2O4 was formed for every oxide before the reaction, indicating that the MgO is denser than MgAl2O4 in this study.
;%0A%09%09%09newWindow.document.open();%0A%09%09%09newWindow.document.write('<img src=%22./Graphics/F-M2024804tab02.jpg%22>');%0A%09%09%09newWindow.document.close();%0A%09%09)
Based on these results, the mechanism of improvement for the erosion resistance by heat treatment and the effects of the chemical composition of the Al alloy melt on the erosion resistance of Al cast iron are schematically illustrated in Fig. 11. Heat treatment of the Al cast iron forms multiple oxide layer due to the reaction as shown by eqs. (1) to (3). This layer prevents contact between the melt and the cast iron substrate and prevents the reactions as shown by eqs. (4) and (5), leading to the improvement of the erosion resistance. As shown in Fig. 4, although the interface between the oxide layer and the cast iron substrate becomes indistinct under some heat treatment conditions, the reaction with the melt would be suppressed due to the existence of oxides, especially alumina (Al2O3) in the decarburized area, even if the oxide layer is dissolved or fractured during the immersion or stirring. When the Al alloy melt does not contain Mg, the reaction between the iron oxides, such as Fe2O3 and/or Fe3O4, in the layer and the melt would occur to form alumina, as shown by eqs. (6) and (7). As shown in Table 2, the volume ratio of these oxides would be decreased due to these reactions; 84.3% from Fe2O3 to Al2O3 and 76.4% from Fe3O4 to Al2O3. Therefore, structural gaps would be observed in the Al2O3 formed by these reactions. In contrast, the Al2O3 in the oxide layer originally formed by the heat treatment would not react with the melt without Mg. Consequently, the melt would react only with the iron oxides in the layer and the reaction with the melt would be totally suppressed (Fig. 11(a)). It is reported that the Mg in the melt is condensed near the interface with the oxide layer to lower the interfacial energy of the system and the condensed Mg would be involved in the interfacial reaction14) as shown by eqs. (8) to (11), leading to form MgAl2O4 (Fig. 11(b)). These reactions would decrease the volume of oxides (Table 2) and gaps in the structure of the oxide layer would be formed. The layer with many gaps would be easily fractured or delaminated due to the shear stress of the melt flow. These phenomena would promote the penetration of the melt into the substrate. As the Mg in the melt further increased, MgO, which is denser than MgAl2O4, was formed, as shown by eqs. (12) to (14). This would prevent the melt from penetrating into the substrate during the immersion and prevent cracking and delamination of the oxide layer, thereby preventing erosion.
Schematic illustration of erosion behavior of heat-treated cast iron in Al alloy melts with various Mg contents.
For the cast iron specimens obtained in this study, the oxide layer would be reproducible by the heat treatment if the erosion is slight. In the Al alloy die casting, it is reported that the addition of Fe to the alloy melt suppresses the seizure of the die.19) Since the heat treatment of Al cast irons obtained in this study can be expected to suppress the seizure, the improvement in toughness of the Al alloy castings due to the decrease in the Fe content can be expected by using these cast irons as a die. Meanwhile, it was confirmed that the increase in the Al content in the cast iron decreases the fluidity of the cast iron melt and lowers the tensile strength. Future topics of investigation would be to clarify the effect of Fe in the Al alloy melt on the erosion resistance of the Al cast irons and to obtain a cast iron having a superior erosion resistance with a lower Al content taking into account the melt fluidity.
This study was supported by JSPS KAKENHI Grant Number JP18K04802 and The Die and Mold Technology Promotion Foundation.
The Kindai Joint Research Center provided the opportunity for operating the SEM, EPMA and XRD instruments.
