Journal Information
Vol. 9. Issue 1.
Pages 629-635 (January - February 2020)
Download PDF
More article options
Vol. 9. Issue 1.
Pages 629-635 (January - February 2020)
Original Article
DOI: 10.1016/j.jmrt.2019.11.003
Open Access
Phase transformation temperatures and Fe enrichment of a 22MnB5 Zn-Fe coated steel under hot stamping conditions
Daniel Alexandre da Costa Ximenesa,b, Luciano Pessanha Moreiraa,
Corresponding author
, José Eduardo Ribeiro de Carvalhob, Duílio Norberto Ferronatto Leitea, Reginaldo Gomes Toledob, Fabio Moreira da Silva Diasb
a Programa de Pós-graduação em Engenharia Metalúrgica, Universidade Federal Fluminense, Av. dos Trabalhadores, 420 Volta Redonda, RJ, CEP 27255-125, Brazil
b Companhia Siderúrgica Nacional, Rodovia BR 393 Lúcio Meira, km 5001, RJ, CEP 27269-900, Brazil
Article information
Full Text
Download PDF
Figures (7)
Show moreShow less
Tables (1)
Table 1. Chemical composition of the 22MnB5 steel (mass %).

In the last years, press-hardening steels have contributed to the automotive industry successfully to meet the increasing regulations for reducing fuel consumption and stringent greenhouse gas emissions while improving passenger safety by manufacturing lightweight car body parts. Zn-Fe coating is an alternative to prevent corrosion or even enhance the corrosion resistance in these steels. However, Zn-Fe coating is prone to liquid melting embrittlement (LME) during the hot forming process. To prevent LME, the coating must be fully transformed into a solid solution before the forming operation, avoiding the contact of the Zinc liquid phase with the steel substrate. This work aimed to determine the phase temperature transformations and critical cooling rate to define the process window for a 22MnB5 sheet with a Zn-Fe coating when submitted to a higher heating rate in comparison to the direct conventional hot forming process. The experimental results indicated that a fully martensitic microstructure is obtained with a cooling rate of 30°C/s. The adopted two-side Zn coating weight of 80g/m2 heated at 53°C/s to 900°C is fully transformed into Fe-α solid solution in 23s, which is an industrial gain compared to longer dwell times required in the conventional heating furnace processes.

Hot forming
Zn-Fe coating
Full Text

Zinc coating has been studied in recent years as a second alternative for the coating of 22MnB5 steel in the hot stamping process, mainly for companies with conventional continuous galvanizing lines, in which the pots for the Al-Si coating are not available. The main advantage of Zn-based coating is the corrosion resistance allowing applications to stamped parts that are exposed to high moisture content [1]. However, Zn-based coating is prone to the occurrence of the LME (Liquid Melting Embrittlement). LME begins during heating and soaking in the furnace, wherein the blank is subjected to a temperature of about 900°C while the pure zinc melt temperature is 419.5°C [2]. During heating and hot stamping, a portion of the liquid zinc contacts the substrate diffusing through the austenite grain boundaries acting as cracking nuclei and, thus, leading to brittle intergranular fracture [2]. The feasible solution to avoid the LME is to use the Zn-Fe alloy since the melt temperature of the galvannealed coating (GA) is higher in comparison to the pure Zn. However, the challenge to the GA coating for the hot stamping of the 22MnB5 steel is to find the Fe content in the Zn layer that reduces the risk of LME but preserving the galvanic corrosion resistance of the Zn-Fe coating.

Extensive research [3–5] has been conducted to evaluate the coating behavior of boron steel sheets submitted to low heating rates, mainly, in the direct hot stamping process. Lee et al. [6] observed using electrical resistance heating that thin continuous layers of Al2O3 in Zn-coated 22MnB5 steel are not damaged by the formation of Fe-Zn intermetallic compounds during a rapid heating (10°C/s) to the soaking temperature. Higher heating rates provide a partial melting of the Zn-coating instead of solidification, which is observed in conventional lower heating rates resulting in the formation of ZnO and Mn3O4 above the damaged Al2O3 layers [6]. Cho et al. [7] investigated the LME cracks in Zn-coated 22MnB5 sheets firstly heated to 900°C at a rate of 30°C/s, then, submitted to uniaxial tensile testing and quenched down to room temperature (60°C/s). From microstructural, quantitative compositional mapping and three-dimensional digital image reconstruction analyses, Cho et al. [7] claimed that the LME crack propagation during the die quenching of Zn-coated 22MnB5 steel is triggered by the presence of thin layers of α-Fe(Zn) at the austenite grain boundaries resulting from the Zn diffusion along these regions.

By means of direct hot stamping tests performed under V-bending deformation mode, Takahashi et al. [8] evaluated the LME formation in GA boron sheets heated to 900°C during 90–300s (heating-rates 10–3°C/s), and observed penetration of the cracks into the substrate for heating times between 120 and 225s and cracks within the coating layer for specimens heated during 90s or longer than 240s. Järvinen et al. [9] evaluated the effect of the steel composition on the resulting phase structures formed in the coating and interface regions of both Zn and Zn-Fe coatings of 22MnB5 and 34MnB5 steels subjected to the direct hot stamping process. After the die-quenching, they have observed the formation of small martensitic constituents α′-Fe(Zn) near to the substrate-coating interface, mainly in both Zn and Zn-Fe coated 34MnB5 steel owing to its higher content of C. The α′-Fe(Zn) layer seems more likely to be formed due to the partitioning from the α-Fe(Zn) to the γ-Fe(Zn) in the heating step. Recently, Peng et al. [10] and Kang et al. [11] investigated by means of hot tensile tests the diffusion phenomena and cracking behavior of Zn and Zn-Fe coatings in 22MnB5 steels, respectively. During the heating step, it was verified that the interdiffusion between the Zn-coating and substrate is controlled by an interfacial Fe2Al5 layer and also that a transition layer is prone to LME cracking, indicating that both γ-Fe(Zn) and α-Fe(Zn) are susceptible to the liquid zinc [10]. The Fe-Zn reaction from different annealing and deformation temperatures (500–900°C) resulted in cracks in the coating that propagated to the substrate along fine α(Zn) grains, mostly at 600 and 700°C [11].

Considering the brief literature survey, few studies investigated the coating evolution of press-hardening steels with the Zn-Fe coatings when submitted to higher heating rates (>30°C/s) which can be achieved by means of electrical resistance heating. Bearing in mind that a full transformation of zinc must occur to avoid the occurrence of LME during the hot stamping, the present work aimed at evaluating the Fe enrichment of a 22MnB5 steel sheet with Zn-Fe (GA) coating submitted to a high heating rate. Firstly, the transformation temperatures of the 22MnB5 steel were determined as a function of the heating rate. Secondly, the continuous cooling diagram (CCT) of the investigated 22MnB5 steel was determined to define the cooling rate that results in a fully martensitic microstructure for its application under industrial conditions. Moreover, the Zn-Fe coatings resulting from different dwell times were evaluated by means of X-ray diffraction measurements and energy-dispersive spectroscopy (EDS) semi-quantitative chemical analysis.

2Experimental procedure

The investigated material is a 22MnB5 press-hardenable steel sheet with the chemical composition listed in Table 1. The cold-rolled sheet has a nominal thickness of 1.5mm with a Zn-Fe two-side coating weight of about 80g/m2, hereafter, referred to ZF80.

Table 1.

Chemical composition of the 22MnB5 steel (mass %).

Mn  Si  Cr  Ti 
0.25  1.14  0.21  0.16  0.0031  0.050 

Scanning electron microscope (SEM) observations and energy-dispersive spectroscopy (EDS) semi-quantitative chemical analysis were performed in a dual-beam scanning electron microscope. The Fe enrichment of the coating was evaluated from ZF80 samples submitted to thermal treatments using a heating rate of 53°C/s to 900°C and then by varying the dwell time at 900°C. After that, the samples were cooled down using a higher cooling rate in comparison to the critical rate of the investigated steel, obtained from the CCT test, that is, the cooling rate under which the resulting microstructure is fully martensitic to attain the targeted tensile strength of 1500MPa. SEM and EDS analyses were performed on the heat-treated samples to identify whether the entire coating layer was transformed in Fe-α and quantify the Fe content. For the non-occurrence of the LME, the coating layer must present a Fe content greater than 54wt.% [12].

The dilatometric and CCT analyses were performed in the Adamel-Lhomargy DT1000 dilatometer using samples taken along the rolling direction with 8mm length and 2mm width. To determine the Ac3 temperature, the samples were heated to 1200°C at different rates from 8 to 95°C/s. The CCT diagram was determined by firstly heating the samples and holding for 5minutes at 950°C and then cooling using different rates, namely, 0.41, 0.5, 1, 2, 3, 6, 10, 17, 30, 50 and 130°C/s. After each cooling condition, all samples were analyzed using light optical and scanning electron microscopy techniques to identify the resulting microstructures. To obtain the entire CCT diagram, the start and finish transformation phase temperatures were calculated from the plot of the dilation versus temperature using the derivative method.

3Results and discussion

Fig. 1 shows the dependence of both transformation temperatures Ac1 and Ac3 with the heating rate. Linear fitting provided somewhat reasonable estimates for both Ac1 and Ac3 temperatures which increased with the heating rate, ranging from 740 to 775°C for Ac1, and from 835 to 890°C for Ac3. The knowledge of these variations is essential, mainly for high heating rates processes in which the blank must be fully austenitic before soaking and hot forming. Also, the intercritical interval temperature varied slightly with minimum and maximum values 95 and 105°C for heating rates of 5 and 95°C/s, respectively. Therefore, heating the blank up to 900°C is enough to achieve a fully austenitic microstructure.

Fig. 1.

Transformation temperatures Ac1 and Ac3 determined for the 22MnB5 steel as a function of the heating rate.


The SEM microstructures determined from three cooling rates are presented in Fig. 2. For the cooling rate of 1°C/s, the microstructure formed is composed of ferrite/pearlite and bainite. At this rate, the start of ferritic/pearlite phase transformation temperature occurred at ∼710°C and finished at 645°C. The bainite was also formed with a start and finish transformation temperatures at 645 and 580°C, respectively. The hardness of the microstructure formed with the cooling rate of 1°C/s is 185 HV. For 10°C/s, bainite, and martensite were formed with a start and finish transformation temperatures of 585 and 405°C, respectively, resulting in a higher hardness value equal to 391 HV. By applying a cooling rate of 30°C/s, the steel presented only the martensite phase with a hardness value of 495 HV. Thus, 30°C/s is the critical cooling rate of the investigated 22MnB5 steel from which the resulting microstructure is fully hardened.

Fig. 2.

22MnB5 microstructures observed by SEM as a function of cooling rates: (a) 1°C/s, (b) 10°C/s and (c) 30°C/s. The letters F, M, B, and P designate ferrite, martensite, bainite and pearlite, respectively.


Fig. 3 resumes the CCT methodology of the dilatometric analysis for the lower cooling rate of 0.41°C/s, from which two diffusional transformations were obtained, namely, austenite to ferrite-pearlite (733°C) and ferrite-pearlite to bainite (655°C). Microhardness measurements were also performed on each tested CCT sample.

Fig. 3.

Dilatometric results of the 22MnB5 steel obtained for the lower cooling rate: (a) normalized length changes during heating and cooling stages as a function of the temperature and (b) normalized length changes and its resulting first derivative with respect to the time as a function of the temperature during the cooling stage.


The CCT diagram determined from dilatometric testing for the 22MnB5 steel is plotted in Fig. 4, in which the start and finish transformation temperatures of ferrite/pearlite, bainite and martensite are indicated with the transformation temperatures Ac1 and Ac3 and the resulting Vickers microhardness values. For the investigated 22MnB5 steel, the start transformation temperature for obtaining a martensitic microstructure is close to 380°C while the start transformation temperature resulting in ferritic-pearlitic-bainitic microstructures is 580°C. Similar start transformation temperatures were determined in the work performed by Maki et al. [13] for a 22MnB5 steel austenitized at 950°C for 5min, namely, about 600°C and 400°C to achieve ferritic-pearlitic-bainitic and martensitic transformation start temperatures, respectively. For comparison purposes, the microhardness values determined from the present CCT tests are equal to 504 HV (130°C/s) and 190 HV (2°C/s) whereas the corresponding values obtained from the work of Maki et al. [13] are 445 HV (130°C/s) and 193 HV (2°C/s).

Fig. 4.

Continuous cooling transformation diagram of the 22MnB5 steel.


In the as-received condition, the coating is mainly formed by δ1-phase with a Fe content of about 10mass%, as indicated in Fig. 5. To determine the minimum dwell time at 900°C required to the 22MnB5 steel with a Zn-Fe (GA) coating to complete the transformation into a solid solution of zinc enriched by iron (α-Fe), the samples were heated with a rate of 53°C/s which is higher than the heating rate of conventional furnace process, usually 20°C/s. After that, the resulting sample coatings were evaluated by SEM and EDS analyses.

Fig. 5.

As-received 22MnB5 Zn-Fe coating: (a) cross-sectional SEM micrograph and (b) EDS elemental distribution for Fe and Zn.


Fig. 6 shows the cross-sectional microstructures and corresponding elemental distributions obtained from the 22MnB5 Zn-Fe coatings for the as-received condition and dwell times equal to 10, 30 and 45s. In the as-received condition the coating layer is mainly composed by zinc showing a clearly defined interface with the steel substrate. After 10s of dwell time, one can observe that the coating been submitted to a level of Fe enrichment which, in turn, was not sufficient to totally transform the coating layer. It is also worth to note that the Fe content, revealed by the dark region of the coating, is qualitatively close to the corresponding Fe content within the steel substrate. However, the bright region of the coating layer (Zn) has not been exposed to a substantial Fe-enrichment, which can be obviously seen from the combined mapping of Fe and Zn. On the other hand, for larger dwell times, see the mapping images obtained for 30s, the coating layer is considerably enriched by Fe which is evident due to the absence of a well-defined enrichment interface that was clearly noticeable for smaller dwell times. This indicates that during 30s the investigated coating is fully transformed in α-Fe, which is shown in Fig. 7, from the semi-quantitative chemical analysis (EDS) and α-Fe thickness measurements results. It is also possible to observe that the 30s dwell time resulted in a homogeneous Fe-enrichment in the ZF80 coating layer. After 45s, oxygen and zinc are observed owing to a longer time exposure at high temperature, and, thus, resulting in the formation of a zinc-oxide layer at the coating surface.

Fig. 6.

ZF80 coating cross-sectional microstructures and corresponding elemental distributions for the as-received condition and from dwell times of 10, 30 and 45s.

Fig. 7.

ZF80 coating cross-sectional microstructures obtained from different dwell times and the corresponding α-Fe layer thickness, α-Fe and Γ-Fe wt.% contents..


The minimum dwell time at 900°C obtained for the Fe-enrichment of ZF80 coating was 23s, as can be seen in the plot of Fig. 7. For dwell times lesser than 23s, two phases were identified in the coatings, namely, one rich in zinc with Zn content greater than 70wt.% and the other one rich in iron. According to the Fe-Zn diagram, the zinc rich phases are the δ-delta, Γ-gamma and Γ1-gamma phases whereas the iron-rich phase is α-Fe. From Fig. 7, it is worth to observe that the Γ-Fe phase appeared only in the GA coatings submitted to dwell times up to 17s.

The α-Fe thickness, determined from SEM observations and also plotted in Fig. 7, were slightly altered with the iron enrichment observed between 17 and 45s times. Conversely, such layer changes usually occur when the 22MnB5 coated steel sheet is exposed to the conventional furnace heating process (20°C/s for 600s at 900°C). Actually, with increasing heating rates the zinc layer is melted in the first few seconds and the diffusion process of the iron takes place between the solid (steel substrate), and liquid (Zn-Fe) phases being faster in comparison to a low heating rate. As a result, the diffusion front occurs between two solid phases and the melting temperature of the formed Zn-Fe layer increases due to the favored iron diffusion and coating enrichment.


In the present work, experimental procedures and microstructural characterization were performed to determine the phase transformation temperatures and the required process window to apply higher heating rates to a 22MnB5 sheet with a Zn-Fe coating. The transformation temperatures Ac1 and Ac3 increased with the heating rate and their differences varied slightly in the range of 8–95°C/s, thus, allowing to identify the temperature of 900°C for completing the austenitization stage. From the CCT diagram, the start and finish phase transformation temperatures of the 22MnB5 steel were defined in the range of low (0.41°C/s) and high (130°C/s) cooling rates. For this steel, a fully targeted martensitic microstructure is obtained with a cooling rate of 30°C/s. The adopted two-side Zn coating weight of 80g/m2 heated at 53°C/s to 900°C is completely transformed into α-Fe solid solution in 23s. This short dwell time can be viewed as an improvement from the industrial cost-effectiveness standpoint in comparison to much longer dwell times required to the production of press-hardening steels by means of conventional direct hot forming process with heating furnaces. Therefore, Zn-Fe alloy can be effectively applied as an interesting alternative to the current blank coatings used in the hot stamping of ultra-high strength steel parts.


The authors express their sincere thanks to CSN R&D (Brazil) which supplied the 22MnB5 steel. D.N.F. Leite acknowledges CAPES for granting the Ph.D. scholarships and L.P. Moreira acknowledges FAPERJ for the research grant E26-211.760/2015.

A. Robert, L. Gerald, K. Siegfried, F. Josef, W.H. Achim.
Surface and coating analysis of press-hardened hot-dip galvanized steel sheet.
Steel Res Int, 83 (2012), pp. 1005-1011
C.W. Lee, D.W. Fan, R. Sohn Il, S.-J. Lee, B.C. De Cooman.
Liquid-metal-induced embrittlement of Zn-coated hot stamping steel.
Metall Mater Trans A, 43 (2012), pp. 5122-5127
J. Kondratiuk, P. Kuhn, E. Labrenz, C. Bischoff.
Zinc coatings for hot sheet metal forming: comparison of phase evolution and microstructure during heat treatment.
Surf Coat Technol, 205 (2011), pp. 4141-4153
A. Sengoku, H. Takebayashi, K. Matsumara.
Microstructural and phase evolution of galvannealed coating during hot stamping heating.
CHS2, organizer. Proceedings of the 5th international conference on hot sheet metal forming of high-performance steel, pp. 363-370
C. Chiriac, R. Sohmshetty.
The effects of the heating rate and the incoming microstructure on the phase transformation temperatures of 22MnB5 Steel.
CHS2, organizer. Proceedings of the 6th international conference on hot sheet metal forming of high-performance steel, pp. 403-413
C.W. Lee, W.S. Choi, Y.R. Cho, B.C. De Cooman.
Surface oxide formation during rapid heating of Zn-coated press hardening steel.
ISIJ Int, 54 (2014), pp. 2364-2368
L. Cho, H. Kang, C. Lee, B.C. De Cooman.
Microstructure of liquid metal embrittlement cracks on Zn-coated 22MnB5 press-hardened steel.
Scr Mater, 90–91 (2014), pp. 25-28
M. Takahashi, M. Nakata, K. Imai, N. Kojima, N. Otsuka.
Liquid metal embrittlement of hot stamped galvannealed boron steel sheet – effect of heating time on crack formation.
ISIJ Int, 57 (2017), pp. 1094-1101
H. Järvinen, M. Honkanen, M. Patnamsetty, S. Järn, E. Heinonen, H. Jiang, et al.
Press hardening of zinc-coated boron steels: Role of steel composition in the development of phase structures within coating and interface regions.
Surf Coat Technol, 352 (2018), pp. 378-391
H. Peng, W. Peng, R. Lu, G. Wu, J. Zhang.
Diffusion and cracking behavior involved in hot press forming of Zn coated 22MnB5.
J Alloys Compd, 806 (2019), pp. 195-205
J.-H. Kang, D. Kim, D.H. Kim, S.-J. Kim.
Fe-Zn reaction and its influence on microcracks during hot tensile deformation of galvanized 22MnB5 steel.
Surf Coat Technol, 357 (2019), pp. 1069-1075
Z.N. Ghanbari.
Fe-Zn phase evolution and cracking behavior in Zn-coated press-hardened steel.
Colorado School of Mines, (2017),
T. Maki, M. Amino, K. Hirano, H. Murai.
Mechanical link servo press for hotforming.
CHS2, organizer. Proceedings of the 5th international conference on hot sheet metal forming of high-performance steel, pp. 179-187
Copyright © 2019. The Authors
Journal of Materials Research and Technology

Subscribe to our newsletter

Article options
Cookies policy
To improve our services and products, we use cookies (own or third parties authorized) to show advertising related to client preferences through the analyses of navigation customer behavior. Continuing navigation will be considered as acceptance of this use. You can change the settings or obtain more information by clicking here.