Corrosion of CrMn based austenitic stainless steels by the lithiumlead eutectic Pb17Li

Corrosion of CrMn based austenitic stainless steels by the lithiumlead eutectic Pb17Li

Fusion Engineering and Design 14 (1991) 309-319 North-Holland 309 Corrosion of Cr-Mn based austenitic stainless steels by the lithium-lead eutectic ...

6MB Sizes 0 Downloads 4 Views

Fusion Engineering and Design 14 (1991) 309-319 North-Holland

309

Corrosion of Cr-Mn based austenitic stainless steels by the lithium-lead eutectic Pb-17Li V. C o e n a, H. K o l b e a, L. Orecchia a, M. Della Rossa b Commission of the European Communities, Joint Research Centre, Ispra Establishment, a Institute for Advanced Materials, b Safety Technology Institute, 21020 Ispra (Va), Italy

Submitted May 1990; accepted 1 August 1990 Handling Editor: G. Casini

The compatibility between the lithium-lead eutectic and different Cr-Mn steels has been studied in the temperature range 723 K in capsules in a rotating furnace for times up to 9500 h and in a thermal convection loop for 4200 h with a z~Tof 50 K. The corroded specimens have been examined by metallographic and SEM analysis. It is shown that the corrosion mechanism consists essentially of the dissolution of Mn in Pb-17Li with the formation of a porous ferrite layer in which penetration of Pb and Li has been observed. No grain boundary attack has been observed. The behaviour of the different steels is reported together with semiquantitative analysis of corrosion layers.

1. I n t r o d u c t i o n

Thermonuclear reactions do not intrisically lead to the generation of radioactive products but only to induced radioactivity in the structural and blanket materials. In most conceptual design studies for fusion reactors it is envisaged to use either austenitic stainless steel of the type AISI 316, ferritic or martensitic steels as structural and blanket containing material. It is important for safety requirements to minimize the problems associated with the induced radioactivity of the above mentioned materials and to develop "Low Activation" materials. The term low activation is used [1] to denote any material exhibiting a residual activity below that of conventional stainless steels. In order to obtain this result three route can be followed - Elemental substitution in existing and well-tested materials such as steels, - D e v e l o p m e n t of novel alloys and ceramic based materials, - Use of isotopically-separated elements. The most immediate solution is the development of

alternative steels with modification to eliminate the elements that lead to high levels of long lived activity after irradiation with 14 MeV neutrons. One of the requisites [1] is that post service contact-dose rate should not exceed 25 mSv. h -1 after a decay period of 100 years. This means that Ni should be replaced by Mn, Mo by W and Nb by Ta. To decrease the production of 14C arising from N it is also necessary to reduce the N content of the steels. JRC-Ispra is widely involved in the characterisation of C r - M n austenitic stainless steels for possible use as low activation materials in fusion application [2-6]. In this study we report on the compatibility of this family of steels with the lithium lead eutectic Pb-17Li which is envisaged as a liquid tritium breeder in the European Fusion Technology programme.

2. M a t e r i a l s

2.1. C r - M n steels

The different C r - M n steels investigated were developed under an Ispra-Creusot-Loire collaboration.

0 9 2 0 - 3 7 9 6 / 9 1 / $ 0 3 . 5 0 © 1991 - Elsevier Science Publishers B.V. ( N o r t h - H o l l a n d )

310

1I. Coen et al. / Corrosion of Cr-Mn based austenitic stainless steels

Table 1 Main constituents of the different alloys studied Type

Cr

Mn

Ni

Mo

C

N

Si

V

W

AMCR 0033 AMCR 0035 IF-A IF-B IF-C IF-D IF-E

10.12 14.09 13.59 12.39 12.98 10.20 17.95

17.50 19.88 11.04 10.43 17.42 17.51 10.5

0.10 0.265 1.950 0.228 2.081 0.132 2.084

0.060 0.030 0.022 0.041 0.025 0.045

0.105 0.029 0.110 0.31 0.11 0.305 0.085

0.19 0.048 0.045 0.033 0.04 0.07 0.052

0.555 0.63 0.165 0.2 0.22 0.445 0.28

0.75 0.622 0.022 0.03 0.75

1.4 1.465 2.2 2.1 2.09

The alloy AMCR-0033 comes from a batch of about 1600 kg, A M C R 0035 of about 500 kg and the others from batches of about 300 kg. The composition of the various alloys is given in Table 1. The alloys of the 00 series were delivered in the form of hot-rolled, flat bars, mill annealed at 1373 K for 1 h. and water quenched. A M C R 0035 contains 15 vol% ferrite as a second phase dispersed in an austenite matrix [4]. The alloys of the series IF have optimized compositions in order to minimize long lived radioactivity. Detailed information on their fabrication and metallurgical characteristics can be found in ref. [5]. Briefly: alloys I F - A and I F - C contain Ni up to 2% to stabilize the austenite while IF-B and I F - D have a high C content (0.3%). The composition of I F - E was balanced to obtain intentionally, for purpose of comparison, a duplex austenite + ~ ferrite microstructure. Discs (15 mm diameter) were cut from all the alloy samples; they were then metallographically polished, degreased with acetone and heat treated for 2 h at 873 K in a U H V furnace.

3. Experimental apparatus and procedure 3.1. Capsule tests

The specimens were introduced in AISI 316 L stainless steel containers, filled with Pb-17Li. The capsules were then sealed by electron beam welding. The operi

-L~ 1 -

0

2

1

/~-3

2.2. Pb-17Li

For the Thermal Convection Loop experiment the lithium-lead eutectic used came from a batch of semiindustrial production (1000 kg) prepared by the C E A for distribution to C E A / F o n t e n a y - a u x - R o s e s , J R C / Ispra and C E N / M o l . Both Pb and Li used were of industrial quality. The batch nr. D-2536 had a Li content in wt% of 0.625 _+ 0.02. For the capsule experiments the P b - 1 7 L i was prepared in our laboratory starting from very pure materials, the method of preparation is reported in [7]. The Li content in wt% was 0.70 +_ 0.02.

o

i

'

SO mm

'

Fig. 1. Thermal convection loop (1 = Pb-17Li, 2 = sample, 3 = thermocouple).

311

V. Coen et al. / Corrosion of Cr-Mn based austenitic stainless steels

welding device. T h e c o n t a i n e r s were t h e n h e a t e d in a resistance furnace, u n d e r v a c u u m ; they were frequently rotated.

ations were carried out in a stainless steel glove box which can b e o p e r a t e d u n d e r either high v a c u u m or pure argon a n d is e q u i p p e d with a n electron b e a m

Fig. 2a. SEM micrograph of cross-section of AMCR 0033 heat treated at 673 K in Pb-17Li for 9245 h in 316L capsule (long face). wt%

Mn

Cr

Pb

Ni

Fe

Si

MATRIX Chem. Analysis

17.50

10.12

-

0.10

71.43

0.555

1.5

11.5

8.4

3.4

74.5

0.8

POINT A Semi-quantit. EDX-Mieroanal.

312

V. Coen et al. / Corrosion of Cr-Mn based austenitic stainless steels

Fig. 2b, SEM micrograph of cross-section of AMCR 0033 heat treated at 673 K in Pb-17Li for 9245 h in 316L capsule (short face).

3.2. Thermal convection loop

The TCL is made of AISI 316 L, fig. 1; eleven thermocouples are welded to various positions in order to have accurate temperature measurement. The dimensions are such that the device can be introduced into the glove box through the air lock. The samples are fixed in the isothermal segment of the hot leg, the TCL is filled with Pb-17Li and closed through a sealed cover. The loop is then removed from the glove box. The heating of the TCL is achieved through four independent resistance elements controlled via electronic thermo regulators. After the tests both the capsules and the TCL were opened in the glove box, the Pb-17Li melted and the samples extracted. The specimens were then sectioned, metallographically mounted, polished and examined by Scanning Electron Microscopy, wavelength dispersive X-ray elemental distribution imaging and semi-quantitative energy dispersive X-ray microanalysis.

4. Results and discussion

The capsule tests were carried out at 673 K for 9245 h. The steels studied were AMCR 0033 and 0035. The Thermal Convection Loop was operated for

4200 h with hot and cold leg temperature of respectively 723 and 673 K. The liquid metal velocity was calculated to be 0.1 cm s-1. The steels investigated were AMCR 0033 and 0035 and the new alloys IF-A, IF-B, IF-C, IF-D and IF-E. The following micrographs (Figs. 2-11) show the result of the tests, the composition of the corroded layer has also been determined in most cases. The thickness of the corrosion layer appears to depend on the composition of the different alloys. It clearly appears that the corrosion of the C r - M n steels by the lithium-lead eutectic is essentially governed by a dissolution process. Mn and, to a lesser extent, Cr (this can be seen in the TCL results) are dissolved. Pb and Li penetrate in the corroded layer, the presence of Pb being evident from the wavelength dispersive X-ray elemental distribution imaging - that of Li has not been experimentally determined in the case of the C r - M n steels but there is no reason why these alloys should differ in behaviour from AISI 316 L where Li has been identified by ion microanalysis [7] in the regions where Pb has penetrated. On corrosion the surface layer looses its austenitic structure, being transformed to ferrite essentially on account of the disappearance of the Mn which is the main austenite stabilizer. The analogy with the behaviour of AISI 316 in Pb-17Li [7,8] is striking. The corrosion mechanism is identical with the difference that while in the Ni rich steel this element dissolves in the lithium lead in the case of the C r - M n steels it is the Mn that practically disappears from the corroded layer. In both cases the Cr content in the corroded zone is also lowered but the decrease is not of the same order of magnitude as in the case of Mn or Ni. This is evident from the semiquantitative analysis of the composition of the TCL corroded layers. The solubility of Mn in Pb-17Li, though lower than that of Ni is of the order of 500 wppm at 723 K (Ni = - 3000 wppm). The solubility of Cr at the same temperature is much lower and does not seem to exceed the value of 10 wppm [9]. If we compare the behaviour of the C r - M n steels in Pb-17Li with that in Li, that has been extensively studied, [3,6] we see that the dissolution process is similar. The main relevant difference is that while in the case of Li, nitrogen enhances grain boundary attack, this has not been observed with Pb-17Li (Fig. 4). In both C r - M n and AISI 316 type steels nitrogen whether present as an impurity in the Li or as a steel constituent promotes grain boundary penetration with the formation of Li9frN 5. It has been shown that in the case of AISI 316 [10] Pb-17Li could corrode that steel at 879 K to form

313

V. Coen et al. / Corrosion of Cr-Mn based austenitic stainless steels

Li9CrN 5 only at a nitrogen partial pressure of > 1012 Pa. The free energy of the corrosion reaction, in the case of the C r - M n steels, should be in the same order of

magnitude than that of A I S I 316. The same should thus apply to the nitrogen pressure above Pb-17Li. If we examine the corroded zones we observe that

Fig. 3. SEM micrograph of cross-section of AMCR 0035 heat treated at 673 K in Pb-17Li for 9245 h in 316L capsule (long face). wt~

Mn

Cr

Pb

Ni

Fe

Si

MATRIX Chem. Analysis

19.88

14.09

-

0.265

65.0

0.63

2.5

12.5

5.6

2.9

75.7

0.8

POINT A Semi-quantit. EDX-Microanal.

314

V. Coen et al. / Corrosion of C r - M n based austenitic stainless steels

Fig. 4. SEM micrograph of cross-section of AMCR 0033 heat treated at 773 K for 1500 h in Pb-17Li with addition of 300 wppm of Li3N. the attack is generally not uniform, we often find localized penetration and the corrosion zones are often thicker on the short face of the sample. The localized penetration or rugged corrosion front

has been explained [11] in the case of type 316 stainless steel by the use of a surface destabilization model based on the work of Harrison and Wagner [12]. A surface that undergoes preferential dissolution (Mn, Cr in our case) develops a very non uniform rugged corrosion front. Surface destabilization is also invoked to account for the difference in behaviour of the long and short faces of the sample [11]. On the short face of our specimens the superficial cold worked layer due to machining has not been removed, during the metallographical preparation of the discs prior to the corrosion experiments. The perturbation of the starting planar surface "triggers" the destabilization and subsequent growth of an irregnlar interface. In the wavelength dispersive X-ray elemental distribution imaging a clear increase in the Ni content is observed in the porous corrosion layer. This is due to the nickel-coming from AISI 316 L capsules or loop and dissolved in the lithium lead eutectic. If we consider the ferrite thickness layer in the case of AMCR 0033, which is the most representative of the alloys studied, we can see that the value of 25 ~m after 4200 h at 723 K with a AT of 50 K is slightly less than the value of - 33/~m for AISI 316 L corroded in Tulip 1 TCL loop for 3000 h at the same temperature with a AT of 60 K [8]. The Pb-17Li velocity in Tulip 1 is 0.12 m s-1 in our case it is estimated to be of 0.1 cm s-1.

Fig. 5. SEM micrograph of cross-section of AMCR 0033 heat treated in Pb-17Li at 723 K for 4200 h in AISI 316L TC loop. AT= 50 K (long face).

315

V. Coen et al. / Corrosion of Cr-Mn based austenitic stainless steels

dissolution of Mn and to a M ~ extent Cr, formation of a ferrite layer and penetration of Pb and possibly Li in this layer. Cold work enhances corrosion. - The behaviour of this family of steels in Pb-17Li is at least similar if not better than that of AISI 316. - Some of the alloys of the series IF with optimized composition seem to be less corroded. Further systematic analysis will be carried out once one or two of these steels are selected for further study on the basis of considerations other than corrosion behaviour. - It clearly appears that in the dual austenite, delta ferrite structure the phase richer in Cr is less attacked by Pb17Li.

Alloy 0035 which contains - 15~ of delta ferrite is less corroded as are the other alloys of the IF series. It is interesting to note that for alloy IF-E which has a composition balanced to obtain a duplex austenite + delta ferrite microstructure, it clearly appears that the zones richer in Cr are less corroded than the zones containing less chromium. 5.

CondMom

The compatibility tests on C r - M n austenitic stainless steels in presence of Pb-17Li in the temperature range of interest for fusion applications have shown that: - The corrosion mechanism is essentially based on the

Fig. 6. SEM micrograph of cross-section of AMCR 0035 heat treated in Pb-17Li at 723 K for 4200 h in AISI 316L TC loop. A T K ( a = long face, b - s h o r t f a c e ) . wt~

MATRIX Semi-quantitative EDX-microanalysis AVERAGE VALUE CORRODED ZONE Semi-quantitative EDX-microanalysis AVERAGE VALUE

Mn

Cr

Ni

$i

19.80

14.23

0.00

0.77

0.30

7.90

2.80

0.95

=

50

316

14. Coen et al. / Corrosion of C r - M n based austenitic stainless steels

Fig. 7. SEM micrograph of cross-section of IF-A heat treated in P b - 1 7 L i at 723 K for 4200 h in AISI 316L TC loop. A T = 50 K (a = long face, b = short face). wt% MATRIX Semi-quantit. EDX-microanal. AVERAGE VALUE CORRODED ZONE Semi-quantit. EDX-microanal. AVERAGE VALUE

Mn

Cr

Ni

Si

V

W

11.37

13.57

2.07

0.77

0.63

2.13

0.10

8.30

1.13

1.03

0.67

2.80

Fig. 8. SEM micrograph of cross-section of IF-B heat treated in P b - 1 7 L i at 723 K for 4200 h in AISI 316L TC loop. A T = 50 K (a = long face, b = short face).

V. Coen et al. / Corrosion of C r - M n based austenitic stainless steels

317

Fig. 9. SEM micrograph of cross section of IF-C heat treated in P b - 1 7 L i at 723 K for 4200 h in AISI 316L T C loop. /iT = 50 K (a = long face, b = short face).

Fig. 10. SEM micrograph of cross-section of IF-D heat treated in P b - 1 7 L i at 723 K for 4200 h in AISI 316L T C loop. A T = 50 K (a = long face, b = short face).

318

V. Coen et al. / Corrosion of C r - M n based austenitic stainless steels

Fig. 11. SEM micrograph of cross-section of IF-E heat treated in Pb-17Li at 723 K for 4200 h in 316L TC loop. AT = 50 K (long face). wt%

Mn

MATRIX Dark zone Semi-quantit. EDX-microanal. AVERAGE VALUE MATRIX Light zone Semi-quantit. EDX-microanal. AVERAGE VALUE

Cr

Ni

Si

V

W

9.25

19.60

1.55

1.30

0.95

3.55

11.85

15.70

2.50

0.75

0.60

2.05

References [1] W. Gulden, C. Ponti, Ph. Guetat, D. Ochem, K. Broden, G. Olsson and G.J. Butterworth, Fusion Waste Management - Safety and Environment Studies 1985-86, EUR-

FU/XII-80/89-92 (Jan. 1989). [2] P. Fenid, D. Boerman; V. Coen, E. Lang, C. Punti and W. Schule, Properties of C r - M n austenitic stainless steels for fusion reactor applications, Nucl. Eng. and Design/Fusion 1, No. 2 (April 1984). [3] V. Coen and P. Fenid, Compatibility of structural materials with liquid breeders - a review of recent work carried out at JRC, Ispra, NucL Eng. and Design/Fusion 1, No. 2 (April 1984). [4] G. Piatti and P. Schiller, Thermal and mechanical properties of the C r - M n (Ni free) austenitic steels for fusion reactor applications, J. Nucl. Mater. 141-143 (1986). [5] G. Piatti, D. Boerman and J. Heritier, Development of

low activation C r - M n austenitie steels for fusion reactor applications, Proceedings of the 15th Symposium on Fusion Technology Utrecht, The Netherlands, 19-23 Sept. 1988 (Elsevier, Amsterdam, 1989). [6] V. Coen, Corrosion problems in nuclear fusion reactors, European Federation of Corrosion Publications No. 1, A Working Patty Report on Corrosion in the Nuclear Industry, The Institute of Metals Book Number 481 (1989). [7] V. Coen, P. Fenici, H. Kolbe, L. Orecchia and T. Sasaki, Compatibility of AISI 316 stainless steel with the Li17Pbs3 eutectic, J. Nucl. Mater. 110 (1982). [8] M. Broc, P. Fanvet, T. Flament, A. Terlain and J. Sannier, Compatibility of 316 L stainless steel with the liquid alloy Pb-17Li, Proceedings of the 14th Int. Conf. on Liq. Met. Engineering and Technology, 17-21 Oct. 1988, Avignon (F). [9] M.G. Barker and T. Sample, The solubilities of some metals in Pb-17Li and lead, Fusion Engrg. Des. 14 (1991) 219-226, in this issue.

V. Coen et al. / Corrosion of C r - M n based austenitic stainless steels

[10] W.R. Watson and R.J. Pulham, The chemical reactions of Li-Pb alloys with nitrogen, lithium nitride and 316 steel, Liq. Met. Eng. and Technology, Vol. 3, Proceedings of the 3rd Int. Conf. held in Oxford on 9-13 April 1984. [11] P.F. Tortorelli, Corrosion and mass transfer of ferrous

319

alloys in Pb-17Li at% Li, Vol. 3, Proceedings of the 4th Int. Conf. on Liq. Met. Eng. and Technology, 17-21 October 1988, Avignon. [12] J.D. Harrison and C. Wagner, The attack of solid alloys by liquid metals and salt melts, Acta Metall. 7 (1959).