2016 Volume 56 Issue 2 Pages 319-325
This study investigates the γ→α transformation kinetics of Fe–Mn–C dual-phase steels at a temperature above the melting point of an Al–8.2Mg–4.8Si (wt.%) alloy coating formed on their surface by hot-dipping process. Using an experimentally determined time-temperature-transformation (TTT) diagram for a model steel of Fe–1.5Mn–0.1C (wt.%), the volume fraction of martensite is controlled through the intercritical heat treatment incorporated into the hot-dipping process. The microstructural observations confirm that this combined process route makes it possible to fabricate hot-dipped Al alloy-coated dual-phase steels with a controlled microstructure.
The potential for Zn prices to increase due to limited resources of the metal has created a need to develop alternatives to galvanizing steel such as a hot-dipped Al–Mg–Si alloy coating. Tsuru et al.1,2,3,4,5) have found that an Al–8.2Mg–4.8Si (wt.%) alloy coating with a fine eutectic microstructure of α-Al and Mg2Si produces a sufficient sacrificial anodic effect, with hydrogen entry and embrittlement being a much lower risk than with galvanized steel.4,5) This hot-dipping Al–Mg–Si alloy coating process is therefore expected to increase automotive applications of high strength steels with a dual-phase microstructure of martensite and ferrite by effectively replacing galvanized high strength steel sheet.6,7) However, such application requires an ability to control the steel microstructure through the hot-dipping coating process, as well as control the formation and growth of Fe–Al intermetallic layers at the interface between the steel substrate and Al alloy coating.8,9,10)
Conventional continuous hot-dipping galvanizing process6) involves heating steel sheet/strip to around 800°C in a N2/H2 reducing atmosphere prior to immersion in a Zn bath.11,12,13,14,15,16) However, for this same process to be applied to the fabrication of an Al alloy coating dual-phase steels, it is necessary to control the steel microstructure through a sequence of heating and cooling during the hot-dipping process. A representative heat profile for the hot-dipping of an Al–Mg–Si alloy coating and the time-temperature-transformation (TTT) diagram for a plain low-carbon steel are shown in Fig. 1(a), together with the corresponding microstructures illustrated in Figs. 1(b)–1(e). In this process, the ferrite (α) in the steel sheet transforms to austenite (γ) by heating to the reduction temperature corresponding to a single γ phase (or γ + α two phase) region (Fig. 1(b)). After holding at the reduction temperature, the steel sheet is then cooled to a dipping temperature above the melting point of the Al alloy coating. During this cooling period, the α phase nucleates on grain boundaries in the γ matrix, resulting in a two-phase microstructure of γ and α phases at the dipping temperature (Fig. 1(c)). Following hot-dipping of the steel sheet in a molten Al alloy bath, Fe–Al intermetallic layers are formed at the interface between the steel and molten Al alloy (Fig. 1(d)). The steel sheet is then rapidly cooled by gas wiping to control the coating thickness11,14) and an additional forced air cooling,6) which causes any remaining γ phase to transform to martensite. This produces a final dual-phase microstructure of ferrite and martensite (α + α’) (Fig. 1(e)). In the conventional process, γ + α two phase microstructure often changes during the cooling and hot-dipping processes after the reduction treatment, whereby it would be difficult to control both the steel microstructure (γ + α two phase microstructure) and the interface microstructure between the steel sheet and liquid Al alloy (Fe–Al intermetallic layers) all together. One of the potential routes to control both microstructures independently is the controlled cooling process or the additional heating (and holding) process above the dipping temperature (to control γ + α two phase microstructure) prior to the hot-dipping process (to control the Fe–Al intermetallic layers). Thus, making control over the kinetics of the γ→α transformation is one of most important issues to achieve these processes for the fabrication of Al alloy-coated dual-phase steels.
(a) Thermal profile of a steel sheet during hot-dipping process, together with its time-temperature-transformation (TTT) diagram and (b–e) corresponding microstructures during hot-dipping.
This study investigates the kinetics of the γ→α transformation that occurs in a model dual-phase steel with Fe–Mn–C ternary compositions. Based on the experimentally determined time-temperature-transformation (TTT) diagram, an attempt has been made to fabricate the Al–Mg–Si alloy-coated dual-phase steels. These results are herein discussed in terms of developing an appropriate sequence of heat treatment to control the microstructure of Al–Mg–Si alloy-coated dual-phase steels during the hot-dipping process.
The chemical compositions of the steels used in this study were Fe–(1.5, 3.5)Mn and Fe–(1.5, 3.5)Mn–0.1C (all compositions are given in weight percent unless specified otherwise). These compositions were plotted on a Fe–Mn binary system and a 0.1% C section of the Fe–Mn–C ternary system was prepared using the thermodynamic database of PanIron17) (Fig. 2). For each steel composition, 50 kg ingots were prepared by vacuum melting and casting, which were homogenized at 1200°C for 7.2 ks and then hot-rolled to plates with a thickness of approximately 30 mm. These plates were held at 1230°C for 7.2 ks, hot-rolled to sheets with a thickness of approximately 4 mm, and then cold-rolled to 1 mm in thickness. Pieces cut from these cold-rolled sheets were mechanically polished with #800 emery paper, and then hot-dipped in molten Al–8.2Mg–4.8Si (wt.%) alloy saturated with Fe. The method used to prepare this Fe-saturated Al–Mg–Si alloy has been described elsewhere.9,18)
Chemical compositions of the steels studied plotted on (a) Fe–Mn binary phase diagram and (b) 0.1 at.% C vertical section of the Fe–Mn–C ternary phase diagram.
Hot-dipping experiments were carried out under a reducing atmosphere using the customized glove box system, which as shown in Fig. 3. This system contains both a reducing furnace and Al alloy bath furnace (Figs. 3(b), 3(c)) so as to allow pre-heating (austenitizing) and hot-dipping to be performed under an Ar-3vol.%H2 gas atmosphere at a constant gas flow rate of 3.3 × 10−5 m3/s (2.0 l/min) with a controlled dew point of approximately −20°C. Note that, before the heating or hot-dipping experiments, the ambient air in the glove box was fully replaced with the reducing Ar–H2 gas using the equipped gas inlet and purge systems (Fig. 3(b)). In order to cut off the ambient air supply during these experiments, the inside of the glove box was kept in a constant positive pressure of approximately 2 × 104 Pa by controlling the gas purge flow rate. This reducing atmosphere served to prevent oxidation of the sample surface makes it possible to achieve the defect-free Al alloy coatings, as shown in Fig. 3(d). The container for these two furnaces was cooled by flowing water (Fig. 3(a)). To determine the γ→α transformation temperature under continuous cooling, each of the steel types was heated (austenitized) at 900°C for 600 s, and then cooled under the reducing atmosphere inside the glove box. To prepare a time-temperature-transformation (TTT) diagram, the samples were austenitized at 900°C for 600 s, and then quenched in a bath of molten Al alloy at temperatures ranging from 650 to 780°C for 2–1800 s, followed by a water quench. The heat profiles were measured by K-type thermo-couples welded on the sample surface. Examples of the heat profiles obtained by hot-dipping in the molten Al alloy (within the glove box) for 10 s at temperatures of 650, 700, and 750°C are provided in Fig. 4. Upon immersion, the sample temperature rapidly decreased from 900°C (i.e. within 5 s) to the temperature of the Al alloy melt.
Customized glove box system used in this study: (a) schematic, (b) appearance, (c) two furnaces in the storage box cooled by flowing water, (d) appearance of samples prepared using the system.
Representative thermal profiles of steel samples hot-dipped in Al alloy melt at different temperatures to produce a TTT diagram.
Trial fabrication of Al alloy-coated dual-phase steel sheets was carried out using an Rhesca hot-dip process simulator.11,19) The steel sheets with a dimension of 150 mm × 50 mm × 1 mm were heated to 900°C for 600 s, and then cooled down to 750°C under an Ar-5 vol.% H2 atmosphere with a dew point set to approximately −30°C, followed by holding at 750°C for 600 s. After that, these sheets were hot-dipped in molten Al alloy at 670°C for different lengths of time, followed by rapid cooling by gas wiping. These heat profiles of the sheets were measured using K-type thermo-couples welded on the sheet surface. The precise heat profile experimentally measured will be described in further detail later.
The microstructure of the prepared samples was observed by optical microscopy. The observed sample surfaces were mechanically polished and etched with a 3% Nital solution. The sample surface ion-polished by a cross section polisher at 5 kV for over 36 ks was observed by a scanning electron microscope (SEM) operating at 15 kV. In order to evaluate the volume fraction of α phase in microstructures of the samples, the average area fractions of α phase were measured using five optical micrographs for one sample. A series of optical micrographs for the present measurement were taken at a fixed magnification of ×500.
Figure 5 shows the temperature profiles of the Fe–Mn–C steel samples during the air-cooling after austenitizing pre-treatment at 900°C for 600 s. These profiles show that the temperatures decreases at a rate of approximately 15°C/s above 800°C for all samples. There is a noticeable difference in the cooling profiles below 800°C. In the case of the Fe–1.5 Mn sample, the rate of cooling slows at 770°C and the temperature remains stable for a few seconds, followed by a period of cooling at a constant rate of approximately 10°C/s to below 500°C. This reduction in the rate of cooling is caused by the exothermal reaction of γ→α transformation, and can also be seen with the profiles of the Fe–3.5Mn and Fe–1.5Mn–0.1C samples at 640°C and 740°C, respectively. However, no change in slope was observed with the Fe–3.5Mn–0.1C sample.
Cooling profiles of Fe–Mn–C steel samples after austenitizing at 900°C.
The temperatures at which γ→α transformation was observed to start in Fig. 5 were plotted together with the A3 temperatures from the phase diagrams (Fig. 2) to prepare the continuous-cooling-transformation (CCT) diagram of Fe–1.5Mn, Fe–3.5Mn and Fe–1.5Mn–0.1C steels. This result is shown in Fig. 6. This demonstrates that the addition of 0.1%C to a Fe–1.5 Mn alloy makes the γ→α transformation sluggish, with the addition of 2% Mn also retarding the transformation kinetics by an order of magnitude. This importantly shows that the present steels would allow the two-phase microstructure of γ and α phases to be controlled in a cooling rate ranging from 10 to 100°C/s, which corresponds to the initial air-cooling period of the continuous hot-dipping coating process.6,7)
Continuous-cooling-transformation (CCT) diagram of γ→α transformation in Fe–Mn–C steels.
As was discussed in the previous section 3.1, the γ→α transformation kinetics of Fe–1.5Mn–0.1C steel is favorable for controlling the two-phase microstructure of γ and α phases during air-cooling, so that this composition was selected as a model case for the dual-phase steel. In order to experimentally determine the time-temperature-transformation (TTT) diagram for this steel, the microstructure of sheet samples was observed after being hot-dipped in Al alloy at various temperatures. Figure 7 presents optical micrographs showing the microstructure produced by hot-dipping at 750°C for various times (followed by water quenching). The fully martensitic structure (α’) was observed in the sample hot-dipped for 2 s (Fig. 7(a)) indicating that a single γ-phase microstructure was maintained at 750°C. Increasing the dipping time to 10 s produced a number of elongated ferrite (α) grains with a mean width of approximately 5 μm along the prior austenite grain boundaries in the martensite structure (Fig. 7(b)), which indicates that the α grains grow by consuming the γ phase at 750°C. The volume fraction of this α phase increases with dipping time to approximately 70% of the sample after 600 s (Figs. 7(c), 7(d)).
Optical micrographs of Fe–1.5Mn–0.1C steel dipped in molten Al alloy at 750°C for (a) 2, (b) 10, (c) 60 and (d) 600 s. All samples were austenitized at 900°C for 600 s.
Figure 8 shows optical micrographs of steel samples hot-dipped at 700 and 650°C for different times (followed by water quenching). The samples hot-dipped at 700°C contain the granular α grains on prior austenite grain boundaries even after 2 s (Fig. 8(a)). These α grains grow to approximately 10 μm in size after 600 s (Fig. 8(b)). In the sample hot-dipped at 650°C, a number of α grains are also formed along the prior austenite grain boundaries after 2 s (Fig. 8(c)), but these grow much faster at 650°C to occupy more than 90% in volume fraction after 600 s (Fig. 8(d)). A few grains of lamellar pearlite were locally observed after 600 s.
Optical micrographs of Fe–1.5Mn–0.1C steel dipped in molten Al alloy at (a, b) 700 and (c, d) 650°C for (a, c) 2 and (b, d) 600 s. All samples were austenitized at 900°C for 600 s.
Figure 9 shows the change in the volume fraction of ferrite (α phase) with dipping time at various temperatures. The volume fraction increases with increasing time. It reaches to 73% after 600 s at 750°C, which corresponds to the equilibrium volume fraction of α phase calculated using the thermodynamic database of PanFe.17) In the samples held at 700°C, the average volume fraction of the α phase is 13% after 2 s, and then increases with time to a value of 80% at 600 s that is comparable to the equilibrium volume fraction of 83%. In the samples held at 650°C, the volume fraction of the α phase rapidly increases to become almost saturated at 93% after 600 s, which is also equivalent to its equilibrium volume fraction (95%). These results demonstrate the slower γ→α transformation kinetics at higher temperature within a temperature range from 750 to 650°C.
Change in the volume fraction of ferrite in Fe–1.5Mn–0.1C steel with dipping time in the molten Al alloy at different temperatures.
Figure 10 shows the time-temperature-transformation (TTT) diagram for Fe–1.5Mn–0.1C steel determined in this study. In this figure, the numbers beneath the various symbols represent the volume fraction of the α phase. The diagram shows how the γ→α transformation starts in less than 20 s at temperatures between 650 to 750°C (Fs). The point at which γ→α transformation finishes (Ff) can be deduced from the change in the volume fraction of α phase with time (Fig. 9), creating a line located above 300 s, which shows the transformation time is reduced at lower temperature at least within a temperature range from 750 to 650°C. Note that the localized pearlite structure in the sample hot-dipped at 650°C for 600 s (Fig. 8(d)) also gives a pearlite transformation start line (Ps).
Time-temperature-transformation (TTT) diagram of Fe–1.5Mn–0.1C steel. The numbers beneath the symbols represent the volume fraction of ferrite.
In order to fabricate hot-dipped Al alloy-coated dual-phase steel with a volume fraction of martensite of ~30%, Fe–1.5Mn–0.1C steel sheets were austenitized at 900°C for 600 s, and then cooled down to 750°C (at a cooling rate of approximately 7°C/s), followed by holding for 600 s to equilibrate in the two-phase region of α and γ phases. Subsequently, the sheets were hot-dipped in Al alloy at 670°C, and then rapidly cooled by gas wiping. The measured cooling rate of the steel surface till 400°C (the martensite start temperature of this steel is approximately 450°C20,21)) was approximately 25°C/s, which suggests that the γ phase would transform to martensite during the rapid cooling. A representative heat profile for this is plotted on the TTT diagram in Fig. 11, which shows how well the temperature was controlled by the cooling/heating system of the hot-dipping process simulator used. The photographs in Fig. 12 also show that the steel sheets were uniformly coated with Al alloy without any discernible defects, regardless of the hot-dipping time. The dark areas on the coating surface are an oxide film believed to be from the surface of the molten Al alloy.
Thermal profile of the heat treatment process with the aim of fabricating the Al alloy coated dual-phase steel with 30% volume fraction of martensite (austenitizing at 900°C for 60 s → cooling → holding at 750°C for 600 s → hot-dipping at 670°C → rapid cooling).
Appearance of Al alloy-coated steels prepared using a hot-dip process simulator.
Figure 13 presents the optical micrographs showing the steel microstructure and the interface between the steel sheet and the Al alloy coating in the prepared sample. These optical micrographs illustrates the dual-phase microstructure of fine martensite (α’) grains distributed in an equiaxed ferrite (α) microstructure with a mean grain size of approximately 15 μm (Figs. 13(a), 13(b)). Most of the martensite grains are equiaxed, whereas some interconnect to form an irregular shape that is seen in the sample hot-dipped after 2 and 100 s. The minute observation revealed the fine substructures in these martensite grains, corresponding to the packet or block structures. The volume fraction of martensite is approximately 26% in all samples. These results indicate that there is scarcely any change in the α and γ two-phase microstructure during hot-dipping at 670°C. At the interface between the Al alloy coating and steel substrate, a Fe–Al alloy layer with a mean thickness of 6 μm is formed after 2 s of hot-dipping (Fig. 13(c)), which grows to a thickness of 20 μm after 100 s (Fig. 13(d)). A back-scattered electron image of the Fe–Al alloy layer in the sample hot-dipped for 2 s is shown in Fig. 14. This indicates a dual layer structure of θ-FeAl3 phase on the Al alloy coating and a thick η-Fe2Al5 phase layer on the steel substrate. The structure corresponds well to that formed on pure iron sheet hot-dipped in Al–Mg–Si alloy at 750°C (followed by water quenching),9) whereas the layer thickness is somewhat smaller than that in the sample hot-dipped at 750°C, which is likely due to lower hot-dipping temperature in the present study. Note that the η phase layer has a relatively uniform interface with the dual-phase steel substrate, which suggests the steel microstructure has a slight effect on the growth of the η phase into the steel substrate.
Optical micrographs showing the microstructure of (a, b) dual-phase steel and (c, d) the interface between this sheet and an Al alloy coating produced by hot dipping at 670°C for (a, c) 2 s and (b, d) 100 s.
Backscattered electron image showing the Fe–Al alloy layer of an Al alloy-coated dual-phase steel (dipped at 670°C for 2 s).
Based on the results presented thus far, a proposed process concept to control the microstructure of dual-phase steels during hot-dipping with an Al alloy coating is presented in Fig. 15. This is a two-step heat treatment/hot-dipping approach, whereby steel sheets are first austenitized under a reducing atmosphere, and then cooled to an intercritical temperature between A1 and A3 in order to equilibrate in a two-phase region of α and γ phases (the heat profile needs to pass the γ→α transformation finish line (Ff) in the TTT diagram). This allows the volume fraction of the α phase to be controlled by changing the holding temperature, after which the steel sheets are immediately hot-dipped in Al alloy and rapidly cooled to transform the γ phase to martensite and create a dual-phase microstructure. The thickness of the Fe–Al alloy layer formed at the interface between the Al coating and steel substrate can, if needed, be controlled by changing the dipping time (or dipping temperature). It should also be noted here that reducing the difference between the intercritical holding temperature and dipping temperature can prevent any change in the steel microstructure during hot-dipping, as this temperature difference provides the undercooling to provide the driving force for the γ→α transformation. In order to minimize the amount of undercooling at the dipping temperature, a two-phase region of α and γ phases can also be controlled through the addition of alloying elements in terms of ferrite or austenite stabilizers. The addition of alloying elements can also change the γ→α transformation kinetics to provide greater control over the two-phase microstructure of α and γ phases through the intercritical heat treatments. In addition, considering the industrial continuous hot-dipping coating line,6) it could be required to achieve the austenitizing and intercritical treatments at lower temperature for shorter time. In order to solve these issues, optimizing the alloying elements is needed to control two kinetics of α→γ and γ→α transformations as well. Thus, further alloy modification is desirable, especially when it comes to controlling the multi-phase microstructures of advanced high strength steels such as TRIP and TWIP22) through a combination of heat treatment and hot-dipping.
Proposed heat treatment process for the microstructural control of hot-dipped Al alloy-coated dual-phase steels.
From an engineering point of view, the adhesiveness of the Al alloy coating layer on the dual-phase steel sheet is an important consideration with advanced high strength steels. The present process can control the layer thickness of the Fe–Al intermetallic phase on the dual-phase steels with various volume fraction of martensite. This awaits future work in order to understand that the martensite in the α matrix plays a role in the delamination behavior of the Al alloy coating.
This study has identified the γ→α transformation kinetics of Fe–Mn–C steels at a temperature above the melting point of an Al–Mg–Si alloy used for hot-dip coating. From experimentally determined CCT and TTT diagrams, hot-dipped Al–Mg–Si alloy-coated dual-phase Fe–1.5Mn–0.1C steel sheets with a controlled volume fraction of martensite have been successfully fabricated by combining a suitable heat treatment with the hot-dipping process. This entails first equilibrating the steel at an intercritical temperature in the α+γ two-phase region to control the volume fraction of α-Fe, with rapid cooling after the hot-dipping being used to transform the γ phase to martensite and create the required dual-phase microstructure. Microstructural observations have confirmed that this processing route is effective for coating dual-phase steels with Al alloy by hot-dipping process.
The support of the Ministry of Education, Sports, Culture, Science, and Technology of Japan through the Element Science Technology Project “Development of Hot-Dipped Aluminum Alloy Coated Steels” is gratefully acknowledged.
