Regular Article
Aluminizing of High-carbon Steel by Explosive Welding and Subsequent Heat Treatment
2015 Volume 55 Issue 1 Pages 272-277
Details
2015 Volume 55 Issue 1 Pages 272-277
A carbon steel plate containing 0.87 mass% C was coated with an aluminum foil with a thickness of 100 μm by explosive welding. This aluminum-coated steel was heat-treated in the temperature range of 973–1273 K for up to 7.2 ks in the air to investigate reactions between molten aluminum and high-carbon steel from the viewpoint of aluminide coating. An aluminized layer was basically composed of Fe2Al5, FeAl, Fe3Al containing carbon (Fe3Al(C)), and ferrite stabilized by aluminum diffusion (α-Fe(Al)). FeAl2 was detected with Fe2Al5 at heating temperatures of more than 1223 K, whereas FeAl had two layers due to its chemical composition. In addition, defects like a crack and a void were observed in the vicinity of the FeAl/α-Fe(Al) interface after heating at 1273 K for 3.6 ks. To reduce brittle Fe2Al5 and FeAl2 in the aluminized layer, the steel coated with aluminum with a thickness of 50 μm was prepared and then heat-treated under the same conditions. At 1173 K, the region consisting of FeAl, Fe3Al(C), and α-Fe(Al) accounted for a large part of the aluminized layer. This aluminized steel was subjected to quenching and tempering on the basis of features of the used bare steel. As a result, the hardness near the surface of the aluminized layer was approximately equal to that of the steel substrate.
Aluminized steel, which is commercially produced by immersing steel in molten aluminum, is used as structural members in marine structures and incinerators because of its excellent corrosion and oxidation resistances. The reactive diffusion between liquid aluminum and solid iron have been extensively studied, and it is known that hard and brittle compounds like Fe2Al5 form at the interface.1,2,3,4,5,6,7) On the other hand, an aluminizing method using an aluminum foil has been proposed to facilitate thickness control of an aluminized layer on the steel.8,9) This is composed of the following two steps: (1) the aluminum foil is bonded to the steel at temperatures below the melting point of aluminum, and (2) the aluminum-coated steel is held at temperatures above the melting point of aluminum, and then aluminizing is completed. This process also has an advantage that Fe-rich Fe–Al compounds like FeAl and Fe3Al, which have good mechanical properties and excellent oxidation resistance,10,11,12,13) can be obtained by heat treatment at a relatively low temperature in a short time, because the amount of aluminum contributing to reactions with the steel is purposefully selected. Yamada et al. have applied this process to the production of electric razor blades.8) Specifically, the bonding of aluminum to martensitic stainless steel was carried out by cold rolling, and the joint was heated in the temperature range of 1173–1323 K for 60 s. The obtained blade material was characterized by forming a hardened layer of Fe–Al compounds on the stainless steel surface.
We had an interest in the formation of Fe–Al compounds in the Fe–Al–C ternary system, and attempted to aluminize the surface of high-carbon steel by using bonding and heat treatment. Explosive welding, which is a bonding technique using explosive energy, was applied to the bonding of a steel plate to an aluminum foil. This is capable of forming strong bonding at the collision interface, and the bonding strength of the explosively-welded joint is known to be higher than that of the joint fabricated by any other bonding methods. Furthermore, the region affected by heat generation during explosive welding is very small.14) Such a situation seems to be suitable for investigating interfacial reactions during heat treatment.
In the present study, a high-carbon steel (carbon tool steel) plate was coated with an aluminum foil by explosive welding, and then heat-treated in the air to investigate the reactive diffusion between molten aluminum and high-carbon steel. In addition, quenching and tempering were conducted for the aluminized steel on the basis of features of the used bare steel, and their influence on the microstructure and the hardness was examined.
A commercially-produced high-carbon steel plate with a thickness of 2 mm was used as a base metal. This is designated as carbon tool steel, SK85, according to Japanese Industrial Standards (JIS G 4401), and contained 0.87 mass% C, 0.24 mass% Si, 0.40 mass% Mn, 0.010 mass% P, 0.007 mass% S, 0.01 mass% Cu, 0.02 mass% Ni, and 0.15 mass% Cr. The steel plate was cut into rectangular shapes with dimensions of 90 mm × 40 mm. This bonding surface was finished with #1200 emery paper. As a cladding metal, aluminum foils with thicknesses of 50 and 100 μm were prepared. The purities of these foils were over 99 mass%. These were also cut into rectangular shapes with dimensions of 90 mm × 40 mm. Before explosive welding, these materials were degreased in acetone using an ultrasonic cleaner and dried with hot air.
Explosive welding of the aluminum foil to the high-carbon steel plate was carried out using a special apparatus, as shown in Fig. 1. The apparatus was composed of an explosive, a flyer plate, and a base plate. The aluminum foil was preliminarily bonded to a supporting plate of aluminum with a thickness of 2 mm by using an organic adhesive to play a role as a flyer plate. This flyer plate was fixed parallel to the high-carbon steel as a base plate. The distance between the aluminum foil and the steel plate was 5 mm. In addition, a powdery explosive consisting mainly of ammonium nitrate was placed on top of the flyer plate. The detonation velocity of the explosive was about 2400 m·s–1, and the amount of it used was 43.2 g. The velocity of the flyer plate during explosive welding was calculated to be approximately 540 m·s–1.15) The bonding of the aluminum foil to the high-carbon steel plate was momentarily achieved and, simultaneously, the supporting plate separated from the aluminum foil.
Schematic illustration of experimental assembly for explosive welding.
The aluminum-coated steel was cut into square shapes with dimensions of 5 mm × 5 mm. To investigate reactions between molten aluminum and high-carbon steel, the specimens were inserted into an electric furnace, which was heated in the temperature range from 973 to 1273 K, and held for 0.6 to 7.2 ks in the air. After a holding step, the specimens were allowed to cool in the furnace to room temperature. Several specimens were rapidly cooled with water from the heating temperatures. Subsequently, tempering treatment was also carried out, as we will describe later.
After heat treatment, X-ray diffraction (XRD, JEOL JDX-8030) measurement was conducted on the surfaces of the aluminum side in the specimens. Thereafter, the specimens were mounted in polyester resin and cut in half to reveal the microstructures. The cross sections were ground with emery paper and then mirror-finished using diamond slurry with a particle size of 0.5 μm. They were examined by using optical microscopy, scanning electron microscopy (SEM, JEOL JSM-5600) in conjunction with energy dispersive X-ray spectroscopy (EDX), and electron probe X-ray micro analysis (EPMA, Shimadzu EPMA-1720H). Before optical microscopy and SEM, the steel was etched in methanol containing nitric acid (CH3OH : HNO3 = 97 : 3 v/v). In some cases, a scanning ion microscope (SIM) in a focused ion beam system (FIB, FEI Quanta 3D 200i) was used. A Vickers hardness test was also performed on the cross sections in the specimens using a hardness tester, Shimadzu HMV-2000. A load of 0.147 N was applied at room temperature for 15 s.
Figure 2(a) shows an optical micrograph of the cross section in high-carbon steel coated with aluminum with a thickness of 100 μm by explosive welding. A sound specimen without defects like a void was obtained. The thickness of aluminum was slightly decreased down to 80–90 μm after explosive welding. Since the steel was used in the as-received state, globular cementite was observed as shown in Fig. 2(b). Figure 2(c) shows an SIM image of the collision interface after FIB milling. A reaction region with a thickness of approximately 1 μm, indicated by white arrows, existed at the interface. This is considered to be a trace of melting and subsequently rapid solidification at the contact surfaces of aluminum and high-carbon steel.16) Such a microstructure would contribute to strong bonding at the interface. In fact, aluminum was attached firmly to the steel even after the specimen with dimensions of 45 mm × 10 mm was bent like Fig. 2(d).
(a) and (b) Optical and SEM micrographs of the collision interface between high-carbon steel and aluminum with a thickness of 100 μm. (c) SIM image of the collision interface finished with FIB milling. (d) External view of aluminum-coated steel after bending test.
Figure 3(a) shows an XRD pattern of the surface of the aluminum side in high-carbon steel coated with aluminum with a thickness of 100 μm by explosive welding. By heating this specimen at 973 K for 3.6 ks in the air, an XRD pattern of the surface was changed to Fig. 3(b). All diffraction peaks in Fig. 3(b) were identified as Fe2Al5, and a (002) peak was very intense. Such a result was also obtained from the specimens heat-treated at 1073 and 1173 K for 3.6 ks. This would have a close relation with a distinctive structure (a tongue-like structure) of Fe2Al5 resulting from preferential diffusion in the direction of the c-axis of its crystal structure.4,5)
XRD patterns of the surface of the aluminum side in aluminum-coated steel [aluminum thickness = 100 μm]. (a) After explosive welding. (b) After heat treatment at 973 K for 3.6 ks in the air.
Figures 4(a) to 4(d) show optical micrographs of aluminized layers formed on the high-carbon steel after heating at 973, 1073, 1173, and 1223 K for 3.6 ks in the air, respectively. At 973 K, there was no residual aluminum region, as can be expected from Fig. 3(b), and Fe2Al5 occupied a large portion of the aluminized layer. Needle-shaped products were also observed near the aluminized layer/steel interface. As mentioned later, these are thought to be FeAl. As the heating temperature increased, multiple reaction regions became clear in the steel side. Eventually, five kinds of the reaction regions, indicated as I to V in Fig. 4(d), were confirmed. Figure 5 shows the results of SEM-EDX analysis for the aluminized layer after heating at 1223 K for 3.6 ks in the air. Regions I to V in an SEM image are equal to those in Fig. 4(d). The formation of FeAl2 was suggested by an analytical value at point 1 in region I. In an XRD pattern of this surface, diffraction peaks for not only FeAl2 but also Fe2Al5 were recognized. Therefore, Fe2Al5 and FeAl2 would exist in region I after heat treatment at 1223 K. As mentioned above, only Fe2Al5 was detected by XRD measurement of the surface after heating at 973, 1073, and 1173 K. Such a change is probably due to the limitation of aluminum reacting to high-carbon steel. The thickness of region I decreased with increasing heating temperature. In addition, a few voids, which may be caused by the formation of Fe2Al5 in molten aluminum, were observed in this region.
Optical micrographs of aluminized layers formed on high-carbon steel after heating at (a) 973, (b) 1073, (c) 1173, and (d) 1223 K for 3.6 ks in the air [aluminum thickness = 100 μm].
SEM-EDX analysis results of an aluminized layer on high-carbon steel after heating at 1223 K for 3.6 ks in the air [aluminum thickness = 100 μm].
According to the Fe–Al binary phase diagram,17) it is thought that analytical results of regions II and III in Fig. 5 correspond to FeAl. The interface between these regions was located in the vicinity of equiatomic composition in the Fe–Al system. Therefore, we refer to regions II and III as Al-rich FeAl and Fe-rich FeAl, respectively. In Al-rich FeAl, a large number of voids were observed. It is known that supersaturated vacancies are easy to be introduced in B2-type intermetallic compounds like FeAl.18) The void formation in region II may be related to structural vacancy in Al-rich FeAl.
In region IV in Fig. 5, the composition ratio of iron to aluminum was approximately 3 to 1. The existence of carbon was also confirmed by using EPMA. Thus this is thought to be Fe3Al containing carbon, and is hereafter denoted as Fe3Al(C). In fact, Fe3AlCx forms in the Fe–Al–C ternary system.19) On the other hand, region V was identified as ferrite containing aluminum. Aluminum is known to be a ferrite former, and this region is hereafter denoted as α-Fe(Al).
Accordingly, the aluminized layer formed on the high-carbon steel was basically composed of five reaction regions: I. Fe2Al5 (Fe2Al5 + FeAl2 at 1223 K), II. Al-rich FeAl, III. Fe-rich FeAl, IV. Fe3Al(C), and V. α-Fe(Al). Sasaki et al. have dealt with an aluminized layer formed between aluminum with a thickness of 50 μm and carbon steel containing 0.45 mass% C at a heating temperature of 1248 K.20) They have reported that Fe2Al5, FeAl2, FeAl, and α-Fe(Al) were observed in the aluminized layer. Compared with the results by Sasaki et al., the formation of two FeAl regions and a Fe3Al(C) region is pointed out as a microstructural feature in the present study. In particular, Fe3Al(C) would result from the use of high-carbon steel.
Figure 6 shows SEM micrographs of the aluminized layers formed on the high-carbon steel after heating at 973 K for 0.6 and 7.2 ks. Needle-shaped products in Fig. 4(a) were already produced in a holding time of 0.6 ks, and were identified as FeAl by compositional analysis. It has been reported that a Fe2Al5 layer consists of relatively large columnar grains almost perpendicular to the steel surface, and shows a tongue-like structure.5) Therefore, FeAl is thought to develop along crystal grains of Fe2Al5 indicated as I. In Fig. 6(a), it was difficult to draw a clear line between Al-rich FeAl (region II) and Fe-rich FeAl (region III). However, as the holding time increased, these became distinguishable. It is noteworthy that there were many voids in Al-rich FeAl, as well as Fig. 5. In addition, a reaction region indicated by a double arrow was also observed, and seems to be Fe3Al(C). These observation results indicate that the configuration of the aluminized layer is essentially identical in the heating temperature between 973 and 1223 K.
SEM micrographs of aluminized layers formed on high-carbon steel after heating at 973 K for (a) 0.6 and (b) 7.2 ks in the air [aluminum thickness = 100 μm].
Figure 7(a) shows an optical micrograph of the aluminized layer on the high-carbon steel after heating at 1273 K for 3.6 ks in the air. Reaction products in regions I to V were consistent with those in Fig. 4(d), although the whole thickness of the aluminized layer was remarkably increased at 1273 K. In region I, the formation of Fe2Al5 and FeAl2 was confirmed by XRD measurement. Furthermore, it is notable that island-shaped Fe3Al(C), indicated as IV, was observed in α-Fe(Al). This led to the indistinct interface between Fe-rich FeAl and α-Fe(Al). To examine the microstructures in a high-temperature state, aluminum-coated steels were heated at 1223 and 1273 K for 3.6 ks and then rapidly cooled with water. Optical micrographs of the aluminized layers after quenching are shown in Fig. 8. In each case, the aluminized layer was retained on the high-carbon steel without breaking. Regions I to V was equal to those in Figs. 4(d) and 7(a). At 1223 K, martensitic transformation occurred in the high-carbon steel, whereas the configuration of the aluminized layer did not significantly change. However, the island-shaped Fe3Al(C) was not observed after rapid cooling from 1273 K. This means that Fe3Al(C) indicated as IV in Fig. 8(a) vanishes at temperatures above 1223 K. Therefore, the formation of island-shaped Fe3Al(C) would be caused by the diffusion of carbon from high-carbon steel, when the specimen was cooled in an electric furnace from 1273 K to room temperature.
(a) Optical and (b) SEM micrographs of an aluminized layer on high-carbon steel after heating at 1273 K for 3.6 ks in the air [aluminum thickness = 100 μm].
Optical micrographs of aluminized layers formed on high-carbon steel after heating at (a) 1223 and (b) 1273 K for 3.6 ks in the air [aluminum thickness = 100 μm]. These were rapidly cooled with water from the heating temperatures.
On the other hand, defects like a crack and a void were observed in the position indicated by white arrows in Fig. 7. The indistinct Fe-rich FeAl/α-Fe(Al) interface should be located in the vicinity of this position, and its influence on the defect formation cannot be ignored. Interestingly, such defects were confirmed even after quenching, as shown in Fig. 8(b). It is thought that the defects were already produced in heating and/or holding steps in heat treatment, and the details are still under investigation.21)
3.4. Use of Aluminum with a Thickness of 50 μmTo reduce brittle compounds like Fe2Al5 and FeAl2 in the aluminized layer, the high-carbon steel coated with aluminum with a thickness of 50 μm was fabricated by explosive welding, and then heat-treated at temperatures above the melting point of aluminum. Figures 9(a) to 9(c) show optical micrographs of the aluminized layers after heating at 1173, 1223, and 1273 K for 3.6 ks in the air, respectively. From the results of XRD measurement and compositional analysis, reaction products in regions I to V were identified as I. mainly FeAl2, II. Al-rich FeAl, III. Fe-rich FeAl, IV. Fe3Al(C), and V. α-Fe(Al). In Fig. 9(a), although black products, indicated by a double white arrow, were observed along the α-Fe(Al)/steel interface, these were remarkably etched regions and were not reaction products. At 1173 K, the region consisting of Al-rich FeAl, Fe-rich FeAl, Fe3Al(C) and α-Fe(Al) occupied a large part of the aluminized layer. Furthermore, the specimens heat-treated at 1223 and 1273 K had Fe-rich FeAl and α-Fe(Al) at the outermost surface, respectively.
Optical micrographs of aluminized layers formed on high-carbon steel after heating at (a) 1173, (b) 1223, and (c) 1273 K for 3.6 ks in the air [aluminum thickness = 50 μm].
High-carbon steel used in the present study is designated as carbon tool steel. In general, this is practically used via quenching and tempering treatments. As shown in Fig. 9(a), the reduction of Fe2Al5 and FeAl2 in the aluminized layer was reasonably achieved at a heating temperature of 1173 K, and we applied the quenching and tempering treatments to aluminized steel. Three types of heat treatment processes are shown in Fig. 10. In process A, aluminum-coated steel was heated at 1173 K for 3.6 ks in the air and then cooled in an electric furnace to room temperature. This corresponded to Fig. 9(a). Process B was a two-step heat treatment: (1) the aluminum-coated steel was heated at 1173 K for 3.6 ks in the air, and then cooled in the furnace to 1073 K, and (2) it was kept at 1073 K for 0.9 ks and rapidly cooled with water. In process C, the specimen obtained by process B was tempered at 473 K for 3.6 ks. Figure 10 also shows a cross-sectional image of the aluminized layer, which was subjected to process C. There was no microstructural change in the aluminized layer. Cracks were slightly introduced in the aluminized layer by water quenching, as indicated by a white arrow. However, they were blocked in the vicinity of α-Fe(Al) and were not propagated into the steel substrate.
Optical micrograph of an aluminized layer on high-carbon steel obtained through process C in the diagram of heat treatment process [aluminum thickness = 50 μm].
Figure 11 shows the relationship between the hardness distribution of the aluminized layer and the heat treatment process in Fig. 10. The measurements were conducted seven times at the specific distance from the surface, and their average values were plotted in the figure. Regions I to V are also equivalent to reaction regions shown by the same number in Fig. 9(a). The hardness of the aluminized layer was slightly increased by quenching and tempering treatments. In process C, the hardness near the surface of the aluminized layer was approximately equal to that of the steel substrate. Although α-Fe(Al) indicated as V had the minimum hardness in the aluminized layer, its value was close to HV = 400 and was relatively high. Therefore, aluminizing of carbon tool steel is beneficial from the practical viewpoint, and suitable quenching temperature and cooling agent are necessary to obtain an aluminized layer without quenching cracks.
Hardness distributions in the vicinity of the aluminized layer/high-carbon steel interfaces prepared by processes A, B and C in Fig. 10 [aluminum thickness = 50 μm].
High-carbon steel plates were coated with aluminum foils by explosive welding, and then heat-treated in the temperature range of 973–1273 K for up to 7.2 ks in the air from the viewpoint of aluminide coating. The main results were as follows.
(1) In the use of aluminum with a thickness of 100 μm, an aluminized layer was basically composed of Fe2Al5 (Fe2Al5 + FeAl2 at more than 1223 K), Al-rich FeAl, Fe-rich FeAl, Fe3Al(C) and α-Fe(Al). At 1273 K, island-shaped Fe3Al(C) formed in α-Fe(Al), and the interface between Fe-rich FeAl and α-Fe(Al) became indistinct. Defects like a crack and a void were also observed in this vicinity.
(2) The high-carbon steel coated with aluminum with a thickness of 50 μm was subjected to quenching and tempering treatments after aluminizing. Although cracks were slightly introduced in the aluminized layer by water quenching, they were blocked in the vicinity of α-Fe(Al). In addition, the hardness near the surface of the aluminized layer was approximately equal to that of high-carbon steel, which was hardened by quenching and tempering.
The authors would like to express their appreciation to Dr. T. Yamamuro and Dr. S. Tanaka of Kumamoto University for their kind assistance in the experiments. We greatly appreciate support from the JSPS Grant-in-Aid for Scientific Research (C) (No. 21560749) and the ISIJ Research Promotion Grant.
