Effect of Heat Treatment and Silicon Concentration on Microstructure ...

EFFECT OF HEAT TREATMENT AND SILICON CONCENTRATION ON MICROSTRUCTURE AND FORMATION OF INTERMETALLIC PHASES ON HOT DIP ALUMINIZED COATING ON INDIAN RAFMS A. SARADA SREE* and E. RAJENDRA KUMAR Institute for Plasma Research, Bhat, Gandhinagar-382428, India

Received April 5, 2013 Accepted for Publication July 19, 2013 http://dx.doi.org/10.13182/FST13-673

Hot dip aluminizing was tried on Indian reduced activation ferritic martensitic steel. This experiment was performed with aluminum (Al) melt, with three different silicon (Si) concentrations (3%, 5%, and 7%). Samples were dipped into the Al-Si melt, at 750uC for 30 s, which produced a hard and brittle Fe2Al5 intermetallic layer on the samples. These samples were subjected to two types of heat treatments: (I) 760uC for 30 h and (II) 980uC for 0.5 h, followed by 760uC for 1.5 h to convert the intermetallic layer into more ductile phases. The width of the Fe2Al5 layer was v10 mm for all the samples with different Si concentrations, and for the pure Al melt, it was *35 mm. For both the heat treatments, FeAl and aFe(Al) layers were observed. Out of the two heat treatments, heat treatment I gives thinner FeAl and

a-Fe(Al) layers compared to heat treatment II. X-ray diffraction measurements confirmed the formation of an a-Al2O3 layer on the surface, for 3% and 5% Si concentrations for heat treatment I and for all Si concentrations for heat treatment II. The hardnesses of the Fe2Al5, FeAl, and a-Fe(Al) layers were found to be 972 to 1089 HV (hardness value)/0.01, 324 to 384 HV/ 0.01, and *200 HV/0.01, respectively.

I. INTRODUCTION

a corrosion rate of 90 mm/year was measured for Eurofer steel with lead-lithium eutectic at 480uC, flowing at a velocity of 22 cm/s. The corrosion rate increases drastically with temperature, and a corrosion rate of 400 mm/year was measured at 550uC (Refs. 1 and 2). From the Institute of Physics of University of Latvia (IPUL) corrosion experiments under joint collaboration of India and IPUL, a corrosion rate of *200 mm/year was measured for P91 samples, with lead-lithium eutectic, at 550uC, flowing at a velocity of 15 cm/s. In the presence of magnetic field (1.7 T), the corrosion rate was *320 mm/year for the same operating conditions.3 These high corrosion rates would lead to enormous amounts of corrosion products, which preferentially

Indian reduced activation ferritic martensitic steel (IN-RAFMS) is a possible candidate structural material for the Indian test blanket module (TBM) for ITER. Leadlithium eutectic is used as the breeder as well as the coolant in the Indian lead-lithium ceramic breeder (LLCB) TBM. The major concerns for using Pb-Li as the coolant are corrosion of the structural material, tritium permeation through the structural material, and magnetohydrodynamic pressure drop. From PICOLO loop experiments, *E-mail: [email protected] 282

KEYWORDS: hot dip aluminization, intermetallic layer, heat treatment Note: Some figures in this paper may be in color only in the electronic version.

FUSION SCIENCE AND TECHNOLOGY

VOL. 65

MAR./APR. 2014

Sree and Kumar

precipitate in the low-temperature regions of the leadlithium loop, thus plugging the pipes. This could be a serious problem for the safe and reliable operation of the TBM. Therefore, it is necessary to have an anticorrosion coating over the structural material. In the LLCB TBM, aluminum-based coatings have been chosen as the reference solution for attacking corrosion and tritium permeation issues. Different kinds of coatings are being developed to address the above issues, such as hot dip aluminizing (HDA), chemical vapor deposition, physical vapor deposition, magnetron sputtering, sol-gel technologies, and plasma spray techniques.4–8 Nowadays, electrochemical coating technologies are being developed for the production of improved Al-based corrosion barriers.9 The HDA process10–15 is one of the potential coating methods, tried for fusion blanket application, to produce anti-corrosion and anti-permeation coating on ferritic martensitic steels (Eurofer, F82H-mod, and MANET II) in the lead-lithium environment. For the Indian LLCB blanket, made of IN-RAFMS, HDA is one of the possible techniques being considered for aluminide coating. The LLCB blanket has complex internal structures that are in contact with flowing lead-lithium, which will require anticorrosion and anti-permeation coating. Application of other coating techniques may not ensure complete coating on all the internal parts of the TBM. In the HDA coating technique, aluminum melt can be flowed through all the complex internal structures to form the intermetallic layer, and by giving heat treatment, this intermetallic layer can be converted into softer phases. Therefore, present research and development in HDA has been taken up on IN-RAFMS samples; the results are discussed in this paper. This method involves two steps: (a) dipping and (b) heat treatment. IN-RAFMS samples were dipped into the Al-Si melt, which produces the hard and brittle intermetallic phase Fe2Al5. To transform this into smoother ductile phases and optimize the microstructure, heat treatment is necessary. Therefore, two types of heat treatments were tried: (I) 760uC for 30 h and (II) 980uC for 0.5 h, followed by 760uC for 1.5 h. Heat treatment I was carried out at the tempering temperature (760uC) for INRAFMS. This temperature is far below the Ac1 temperature (831uC) of IN-RAFMS (Ref. 16). Heat treatment II was chosen for the following reasons. First, if 760uC is not sufficient to convert the hard and brittle intermetallic phase

HOT DIP ALUMINIZED COATING ON INDIAN RAFMS

into softer ductile phases, then a higher temperature would be required, which should be compatible with the admissible heat treatment for the structural material, to guarantee the original mechanical properties of the material. Second, the tempering temperature (760uC) may not be sufficient to form a-Al2O3 at the surface, which is essential for the coating to act as a corrosionresistant layer as well as a tritium permeation barrier. II. SAMPLE PREPARATION

Samples were prepared from normalized and tempered IN-RAFMS plate. Normalization of the steel was carried out at 980uC for 0.5 h, and tempering was carried out at 760uC for 1.5 h. The sample size was 35|9.5|2.3 mm. The composition of the IN-RAFMS material is given in Table I. IN-RAFMS fresh samples were initially cleaned in acetone and then with ethanol in an ultrasonic cleaner. To improve the wetability of the steel surface with the Al melt, samples were dipped into the flux solution consisting of KCl, NaCl, and Na3AlF6 in the ratio 5:4:1 in water and dried.

III. EXPERIMENTAL PROCEDURE

Samples were dipped into pure aluminum (Al) melt and Al melt with three different silicon (Si) concentrations (3%, 5%, and 7%) for 30 s, in each case. The temperature of the aluminum melt was maintained at 750uC. To carry out the heat treatments, hot dip aluminized samples were placed in an alumina crucible and kept in a muffle furnace. Time and temperature regimes for both the heat treatments are given in Table II. After the heat treatment, the samples were cooled naturally. Cross sections of the hot dip aluminized samples as well as heat-treated samples were prepared. The cross-section surface was polished with different emery papers and finally with alumina paste. Analytical investigation was carried out using scanning electron microscopy (SEM) and energy dispersive X-ray (EDX) analysis. Heat-treated samples were also characterized using X-ray diffraction (XRD) to get the information on the compounds present on the surface.

TABLE I Chemical Composition of IN-RAFMS* Cr

C

Mn

V

W

Si

P

S

Ta

Nb

Mo

Ni

Fe

9.15

0.08

0.53

0.24

1.37

0.026

v0.002

0.002

0.08

v0.001

v0.002

0.004

Balance

*In weight percent. FUSION SCIENCE AND TECHNOLOGY

VOL. 65

MAR./APR. 2014

283

Sree and Kumar


Al melt, the thickness of the Fe2Al5 layer reduced in comparison to the case of pure Al melt. The microhardness of the Fe2Al5 layer was *1000 HV (hardness value)/0.01.

IV. RESULTS AND ANALYSIS IV.A. Hot Dipped Samples

The samples were fully covered with Al melt without any gaps. To study the composition and morphology of the different layers formed, cross sections of the samples were characterized using SEM and EDX analysis. The SEM micrograph of the hot dipped sample for the pure Al melt is shown in Fig. 1. SEM micrographs of the hot dipped samples for different Si concentrations (3%, 5%, and 7%) are shown in Fig. 2. In all cases, microstructures revealed three distinct regions: an outer Al layer, an inner Fe2Al5 intermetallic layer, and the substrate. In the micrographs of Figs. 1 and 2, the Al layer is denoted by ‘‘A’’ and the Fe2Al5 layer is denoted by ‘‘B.’’ For the samples dipped into pure Al melt, the thickness of the Al layer was *12 mm and that of the Fe2Al5 layer was *35 mm. For the hot dipped samples, with different silicon concentrations, the aluminum layer thickness varied from 5 to 35 mm. It was also observed that irrespective of Si concentration, the thickness of the Fe2Al5 layer was v10 mm. With the addition of Si to the

TABLE II Time and Temperature Regimes for Hot Dip Aluminization and Heat Treatments HDA Condition

Heat Treatment I

750uC/30 s

760uC/30 h/air cooling

Heat Treatment II 980uC/0.5 h/air cooling 760uC/1.5 h/air cooling

Fig. 1. SEM micrograph of the sample dipped in pure Al melt. ‘‘A’’ - Al layer, ‘‘B’’ - Fe2Al5 layer. 284

Fig. 2. SEM cross section of hot dipped samples, with different Si concentrations: (a) 3% Si, (b) 5% Si, and (c) 7% Si. ‘‘A’’ - Al layer, ‘‘B’’ - Fe2Al5 layer.


VOL. 65

MAR./APR. 2014

Sree and Kumar

IV.B. Results of Heat Treatment I

SEM micrographs, for heat treatment I, for different Si concentrations (3%, 5%, and 7%) are shown in Figs. 3a, 3b, and 3c, respectively. On the micrographs, ‘‘C’’,


and ‘‘D’’ represent FeAl and a-Fe(Al) layers, respectively. For all the samples, for depth measurement, zero was set at the end of the porous zone near the surface. SEM EDX measurements were taken at a 2-mm interval. Thicknesses of the FeAl and a-Fe(Al) phases for both the heat treatments (I and II) are tabulated in Table III. For 3% Si concentration, three layers were observed. In the first layer, the Fe2Al5 phase still existed, which indicates incomplete transformation of the Fe2Al5 intermetallic layer. This layer extended up to *70 mm. Beneath this layer, FeAl and a-Fe(Al) layers were observed. The total thickness of the coating was *110 mm. For 5% Si concentration, the Fe2Al5 layer was completely transformed into FeAl and a-Fe(Al) phases. The total thickness of the coating was *50 mm. For 7% Si concentration, the porous zone extended up to *20 mm. No specific phase could be observed close to the surface, and the FeAl layer started at a depth of 16 mm. Beneath this layer, an a-Fe(Al) layer was observed. In all the three cases (3%, 5%, and 7% Si concentrations), a smooth interface was observed between the substrate and the a-Fe(Al) layer. In this heat treatment, irrespective of Si concentration, finer Kirkendall pores were observed compared to heat treatment II. IV.C. Results of Heat Treatment II

SEM micrographs, for heat treatment II, for different Si concentrations (3%, 5%, and 7%) are shown in Figs. 4a, 4b, and 4c, respectively. FeAl and a-Fe(Al) layers are denoted by ‘‘C’’ and ‘‘D,’’ respectively, on the micrograph. For all the samples, for depth measurement, zero was set at the end of the porous zone near the surface. For 3% and 5% silicon concentrations, FeAl layer was observed from a depth of 6 to 10 mm from the surface and extended up to *70 mm. Beneath this surface, a-Fe(Al) phase was observed. In both the cases, complete transformation of Fe2Al5 into FeAl and a-Fe(Al) phases was observed. In the above two cases, the total thickness of the coating was *120 mm. TABLE III Thickness of Different Phases for Both Heat Treatments

Heat Treatment Parameters 760uC/30 h

Fig. 3. SEM cross section of heat-treated samples for heat treatment I, with different Si concentrations: (a) 3% Si, (b) 5% Si, and (c) 7% Si. ‘‘C’’ - FeAl, ‘‘D’’ - a-Fe(Al). FUSION SCIENCE AND TECHNOLOGY

VOL. 65

980uC/0.5 h/air cooling 760uC/1.5 h/air cooling

MAR./APR. 2014

Si (%)

FeAl Layer Thickness (mm)

a-Fe(Al) Layer Thickness (mm)

3 5 7

22 24 6

8 8 8

3 5 7

70 62 34

16 16 22 285

Sree and Kumar


For 7% silicon concentration, the total thickness of the coating layer was *230 mm. The first layer extended up to *140 mm. Composition of elements in this layer could not be attributed to any specific layer. In this

case, the Fe2Al5 layer, which was formed after hot dipping, was not completely transformed into FeAl and a-Fe(Al) phases, even after heat treatment. Beneath this layer, FeAl and a-Fe(Al) phases were observed. For comparison of thicknesses of different layers, all the micrographs, Figs. 4a, 4b, and 4c, are presented with 2000|magnification. As it is not possible to present the full micrograph with the total coating width (230 mm), at 2000|magnification, only the part of the micrograph showing the FeAl and the a-Fe(Al) layers is presented in this particular case, Fig. 4c. For all the three cases (3%, 5%, and 7% Si), smooth interface was observed between the substrate and the a-Fe(Al) layer. Coarse Kirkendall pores were observed in the FeAl layer, and fine Kirkendall pores were observed in both the FeAl and the a-Fe(Al) layers. Kirkendall pore formation closer to the surface indicates aluminum was diffusing faster than iron. Variation in concentration of Fe, Cr, Al, and Si with depth, for the case of 5% Si concentration, for heat treatments I and II, are shown in Figs. 5a and 5b,

Fig. 4. SEM cross section of heat-treated samples for heat treatment II, with different Si concentrations: (a) 3% Si, (b) 5% Si, and (c) 7% Si. ‘‘C’’ - FeAl, ‘‘D’’ - a-Fe(Al).

Fig. 5. Variation in concentration of Al, Fe, Si, and Cr with depth: (a) with heat treatment I and (b) with heat treatment II.

286


VOL. 65

MAR./APR. 2014

Sree and Kumar

respectively. The Al content was slowly decreasing and the iron (Fe) content was slowly increasing from the surface to the substrate. The total coating layer thicknesses for heat treatments I and II were found to be *46 and *110 mm, respectively. To form thinner intermetallic layers, heat treatment I is better than heat treatment II. Thinner coatings are preferred as aluminum is an activating element and is not desired in the matrix of low-activation steels.


V. XRD MEASUREMENTS

XRD analysis was carried out on all the heat-treated samples, and FeAl peaks were observed on all of them. a-Al2O3 peaks were observed for 3% and 5% Si concentrations in heat treatment I and for all cases (i.e., 3%, 5%, and 7%) in heat treatment II. Prominent a-Al2O3 peaks were observed for 5% Si concentration, irrespective of heat treatment. Figures 6 and 7 show

Fig. 6. XRD diffractogram of heat-treated sample for heat treatment I, with 3% Si concentration. Lines are added to the spectra at a-Al2O3 peaks. FUSION SCIENCE AND TECHNOLOGY

VOL. 65

MAR./APR. 2014

287

Sree and Kumar


Fig. 7. XRD diffractogram of heat-treated sample, for heat treatment I, with 5% Si concentration. Lines are added to the spectra at a-Al2O3 peaks.

the diffractograms for 3% and 5% Si concentrations, respectively, for heat treatment I. Figures 8 and 9 show the diffractograms for 3% and 5% Si concentrations, respectively, for heat treatment II. Lines are added to the diffractograms for easy identification of a-Al2O3 peaks.

due to its smaller thickness. The hardness of the Fe2Al5 layer was *972 to 1089 HV/0.01, that of the FeAl layer was 324 to 384 HV/0.01, and that of the a-Fe(Al) layer was *200 HV/0.01.

VII. DISCUSSION AND CONCLUSIONS VI. MICROHARDNESS MEASUREMENTS

Hardness measurements of the layers were carried out with a Mitutoyo HM 211, Micro Vickers hardness testing machine. Hardness of the Fe2Al5 and FeAl layers was measured with both 50-g and 10-g loads. The hardness of the a-Fe(Al) layer was measured with only a 10-g load 288

After hot dipping, the thickness of the Fe2Al5 intermetallic layer was *35 mm for the case of pure Al melt and v10 mm for all the samples with different silicon concentrations. After the heat treatments, FeAl and a-Fe(Al) phases were observed in all the cases. A band of pores was observed between the FeAl and a-Fe(Al)


VOL. 65

MAR./APR. 2014

Sree and Kumar


Fig. 8. XRD diffractogram of heat-treated sample for heat treatment II, with 3% Si concentration. Lines are added to the spectra at a-Al2O3 peaks.

layers. In heat treatment I, finer pores were observed compared to heat treatment II. In heat treatment I, thinner FeAl and a-Fe(Al) layers were formed compared to heat treatment II. For example, the FeAl layer thickness was *22 mm for heat treatment I and *70 mm for heat treatment II. For heat treatment I, the thickness of the a-Fe(Al) layer was *8 mm, whereas for heat treatment II, the thickness was *16 mm. In heat treatment I, XRD diffractograms show the formation of a-Al2O3, in the case of 3% and 5% Si FUSION SCIENCE AND TECHNOLOGY

VOL. 65

concentrations, whereas no such peaks were observed for 7% Si concentration. In heat treatment II, a-Al2O3 was observed for all cases (3%, 5%, and 7% Si concentrations). As per our observations, heat treatment I is better compared to heat treatment II, as thinner intermetallic layers can be achieved with heat treatment I and the heat treatment temperature (760uC) is sufficient to form an aAl2O3 layer at the surface, which is essential for the coating to act as corrosion-resistant layer and tritium permeation barrier. But, further experimentation is

MAR./APR. 2014

289

Sree and Kumar


Fig. 9. XRD diffractogram of heat-treated sample for heat treatment II, with 5% Si concentration. Lines are added to the spectra at a-Al2O3 peaks.

required to check the performance of this coating as a corrosion-resistant layer/tritium permeation barrier. Though heat treatment II can also produce the required FeAl and a-Fe(Al) layers and the a-Al2O3 layer at the surface, heat treatment I is still better compared to heat treatment II, as at the corresponding temperature (980uC) of heat treatment II, complete austenitization of steel occurs, and therefore, tempering also needs to be carried out. During these heating and cooling cycles, there is a chance of deformation of the end product. 290

ACKNOWLEDGMENTS The authors would like to thank T. Kamble for support for the experiment, H. Agravat for technical help during the experiment, V. Rajulapati Koteswar Rao and his team [School of Engineering Sciences and Technology (SEST), Hyderabad] for supporting XRD measurements, V. S. S. Shrikant (SEST, Hyderabad) and his team for supporting microrange SEM/EDX measurements, N. L. Chauhan (Facilitation Centre for Industrial Plasma Technologies) for supporting SEM/EDX measurements, and J. Chauhan for measuring microhardness.


VOL. 65

MAR./APR. 2014

Sree and Kumar


REFERENCES

Environment,’’ Fusion Eng. Des., 87, 1483 (2012); http:// dx.doi.org/10.1016/j.fusengdes.2012.03.042.

1. J. KONYS et al., ‘‘Compatibility Behavior of EUROFER Steel in Flowing Pb–17Li,’’ J. Nucl. Mater., 386–388, 678 (2009); http://dx.doi.org/10.1016/j.jnucmat.2008.12.271.

10. E. SERRA et al., ‘‘Hot-Dip Aluminium Deposit as a Permeation Barrier for MANET Steel,’’ Fusion Eng. Des., 41, 149 (1998); http://dx.doi.org/10.1016/S0920-3796(98)00224-5.

2. J. KONYS et al., ‘‘Flow Rate Dependent Corrosion Behavior of Eurofer Steel in Pb 15.7Li,’’ J. Nucl. Mater., 417, 1191 (2011); http://dx.doi.org/10.1016/j.jnucmat.2010.12.277.

11. H. GLASBRENNER et al., ‘‘Alloying of Aluminum and Its Influence on the Properties of Aluminide Coatings: Oxidation Behavior and the Chemical Stability in Pb-17 Li,’’ J. Nucl. Mater., 233–237, 1378 (1996); http://dx.doi.org/10.1016/ S0022-3115(96)00183-3.

3. E. PLATACIS et al., ‘‘Corrosion Phenomena of FMS (P-91) Steel in PbLi Flow in Magnetic Field,’’ Magnetohydrodynamics, 48, 2 (2012). 4. J. KONYS et al., ‘‘Development of Advanced Al Coating Processes for Future Application as Anti-Corrosion and T-Permeation Barriers,’’ Fusion Eng. Des., 85, 2141 (2010); http://dx.doi.org/10.1016/j.fusengdes.2010.08.018. 5. G. BENAMATI et al., ‘‘Development of Tritium Permeation Barriers on Al Base in Europe,’’ J. Nucl. Mater., 271–272, 391 (1999); http://dx.doi.org/10.1016/S0022-3115(98)00792-2. 6. D. LEVCHUK et al., ‘‘Erbium Oxide as a New Promising Tritium Permeation Barrier,’’ J. Nucl. Mater., 367–370, 1033 (2007); http://dx.doi.org/10.1016/j.jnucmat.2007.03.183. 7. D. L. SMITH et al., ‘‘Development of Coatings for Fusion Power Applications,’’ J. Nucl. Mater., 307–311, 1314 (2002); http://dx.doi.org/10.1016/S0022-3115(02)00985-6. 8. X. LI et al., ‘‘Al Based Coating on Martensitic Steel,’’ J. Nucl. Mater., 329–333, 1407 (2004); http://dx.doi.org/ 10.1016/j.jnucmat.2004.04.207. 9. J. KONYS et al., ‘‘Impact of Heat Treatment on Surface Chemistry of Al-Coated Eurofer for Application as AntiCorrosion and T-Permeation Barriers in a Flowing Pb–15.7Li


VOL. 65

12. H. GLASBRENNER et al., ‘‘Comparison of Microstructure and Formation of Intermetallic Phases on F82H-Mod. and MANET II,’’ J. Nucl. Mater., 258–263, 1173 (1998); http:// dx.doi.org/10.1016/S0022-3115(98)00181-0. 13. J. KONYS et al., ‘‘Comparison of Corrosion Behavior of Bare and Hot-Dip Coated EUROFER Steel in Flowing Pb– 17Li,’’ J. Nucl. Mater., 367–370, 1144 (2007); http://dx.doi.org/ 10.1016/j.jnucmat.2007.03.205. 14. S. HAN et al., ‘‘Influence of Silicon on Hot-Dip Aluminizing Process and Subsequent Oxidation for Preparing Hydrogen/Tritium Permeation Barrier,’’ Int. J. Hydrogen Energy, 35, 2689 (2010); http://dx.doi.org/10.1016/j.ijhydene.2009.04.033. 15. H. GLASBRENNER et al., ‘‘Comparison of Hot Dip Aluminised F82H-Mod. Steel After Different Subsequent Heat Treatments,’’ J. Nucl. Mater., 257, 274 (1998); http://dx.doi.org/ 10.1016/S0022-3115(98)00456-5. 16. B. RAJ et al., ‘‘Progress in the Development of Reduced Activation Ferritic-Martensitic Steels and Fabrication Technologies in India,’’ Fusion Eng. Des., 85, 1460 (2010); http://dx.doi.org/10.1016/j.fusengdes.2010.04.008.

MAR./APR. 2014

291