Predicting the intergranular corrosion of austenitic stainless steels

Predicting the intergranular corrosion of austenitic stainless steels

Corrosion Science, 1968, Vol. 8, pp. 9 to 18. Pergamon Press. Printed in Great Britain PREDICTING THE I N T E R G R A N U L A R CORROSION OF AUSTENIT...

436KB Sizes 0 Downloads 11 Views

Corrosion Science, 1968, Vol. 8, pp. 9 to 18. Pergamon Press. Printed in Great Britain

PREDICTING THE I N T E R G R A N U L A R CORROSION OF AUSTENITIC STAINLESS STEELS* W. D. FRANCE, JR. t a n d N. D. Gmmr,m Corrosion Kesearch Laboratory, Materials Research Centre, R.ensselaer Polytechnic Institute, Troy, New York AbstraetIThe precise environmentalconditions necessary for the intergranularcorrosion of austenitic stainless steels have been determined by potentiostatie methods. Intergranular corrosion of sensitized 18 Cr-8 Ni stainless steel only occurs in limited potential regions. These results have been used to develop a new method for rapidly predicting the intergranular corrosion tendencies of the steels in various sulphuric acid environments. It is also shown that sensitized stainless steels may be used in many media without the occurrence of intergranular attack. R&um&--On a pr~cis~ potentiostatiquement les circonstances de corrosion intercristalline d'aciers inoxydables anst~nitiques. Dans le cas de l'acier 18 Cr-8 Ni, elle n'apparalt que dans des intervalles d6fiuis de tension. On a appliqu6 ces r~sultats au d6veloppement d'une nouvelle m~thode de pr~d6termination rapide de la propension tt la corrison intercristalline des aciers dans divers milieux sulfuriques. On a ~galement montr6 que les aciers inoxydables sensibilis6s peuvent ~tre utilis~s darts de nombreux milieux sans souffrir de corrosion intercristalline. Zusammenfassung--Die genauen Angriffsbedingungen, die zur interkristallinen Korrosion von anstenitisehen, rosffreien Sffihlen fiihren, wurden mit potentiostatischen Methoden ermittelt. Es zeigte sich, dab interkristalline Korrosion von sensibilisierten 18/8 Mn--Cr-Ni-St~len nur in einem engen Potentialbereich auftritt. Diese Ergebnisse werden benutzt, um ein neues Verfahren anzugeben, das die schnelle Bestimmung der Komzerfallsanffilligkeit derartiger Stiihle in Schwefels~iureuntersehiedlicher Konzentration erm6glieht. Es kann ferner gezeigt werden, dab sensibilisierte, rosffreie St~hle unter vielen Augriffsbedingungen benutzt werden k6nnen, ohne dab interkristallineKorrosion auftritt. Pe~epaT - - ToqH~ae napaMeTphl oHpyrRatomell cpeau, Heo6xo~HMHe ~ a

BOSHHHHOBeHHH

Me~HHpHcTaJIJItITHOt~ Hoppoa~H ayCTeHllTHIxIX HepmaBemm~tx cTa~elt, onpe~eJ~eH~ HOTeH-

I~HOCTaTHqecHHMHMeTO~a~. MemKpHcTaJIJr~THa~Koppoan~ aHTHBHpOBaHHO~/HepmaBemm~e~ CTaJIH 4_8 Cr-~Ni IIpoIICXORHTTOJIbHOB oPpaHHqeHHO~ 05JIaCTH IIOTeHI~HaJIOB. ~TH pesyJIblaThI HCIIOJII~8OBaHKI~IIJIHpa3BHTHH HOBOF0 MeTO~a IIpe~BapHTeJIl~HOl~IoI~eHHH CHJIOHHOCTH cTaJIei~ H MeH~HpHCTaJIJIHTHOI~Hoppo3HH B pa3JIHqHhIX cpe~ax~ eoRepmanlHx cepaym HHCJIOTy. [IoHa3aHO, Taah~e, qT0 aHTHBHpOBaHHbIe HepmaBem~IHe CTanH MOryT HCIIOJIbSOBaTI~CH B0 MHOFHX cpe~Iax, He no~BepraHcb MeH~HpHETaJIJIHTHOMypaapy~ueHH~O. T i m C h r o m i u m Depletion Theory 1 has been successfully employed to predict n u m e r ous measures for preventing the i n t e r g r a n u l a r corrosion o f austenitic stainless steels. Examples include solution quenching, stabilization with t i t a n i u m a n d n i o b i u m a d d i t i o n s a n d the use of low c a r b o n alloys. 2 A l t h o u g h very useful, this theory has a serious flaw: it specifies the necessary b u t n o t sufficient c o n d i t i o n s for i n t e r g r a n u l a r attack. A c o n t i n u o u s grain b o u n d a r y precipitate o f c h r o m i u m - r i c h carbide is con*Manuscript received 7 March 1967. fPresent address: Research Laboratories, General Motors Corporation, Warren, Michigan.

10

W.D. FRANCE,JR. and N. D. GREENE

sidered necessary for intergranular corrosion.* However, m a n y environments do not selectively attack the grain boundaries o f sensitized stainless steels, 4 and the use o f costly preventative measures is unnecessary. Thus, knowledge o f the environmental conditions which cause intergranular attack would permit more efficient and economical application o f austenitic stainless steels. Recent electrochemical studies 5-9 have demonstrated that the intergranular corrosion of austenitic stainless alloys occurs only in limited potential regions. I f these regions were precisely and completely characterized, it would be possible to predict, apriori, whether a given environment will cause intergranular attack on the basis o f a corrosion potential measurement. The purpose o f this investigation was to precisely determine the potential dependence of intergranular attack, and to test the possibility o f predicting the occurrence o f this type o f attack in various environments. Sulphuric acid solutions were chosen for this study since stainless steels are frequently employed in such media. EXPEKIMENTAL A cast, austenitic stainless steel o f the CF-8 type was used for most experiments since the large grain size facilitated visual observation o f intergranular attack. Nickel content was slightly above normal specifications to ensure absence of ferrite phase. W r o u g h t Type 304 stainless steel was also included for comparison purposes. The compositions o f these two alloys are given in Table 1. Rectangular specimens approximately 7 cm ~ were water quenched after 1 h at 1100°C (2000°F) and sensitized by heating at 675°C (1250°F) for 25 h. Specimens were ground to 00 emery finish, and their dimensions were measured within -t-0"001 cm. Prior to testing, samples were degreased in boiling benzene, rinsed with doubly distilled water, dried and weighed to the nearest 0.1 rag. All solutions were prepared f r o m reagent grade chemicals and doubly distilled water. TABLE I.

STAINLESSSTEEL COMPOSITIONS (Wt. ~)

Cast alloy (CF-8 Type)

Type 304

C . . . . . 0'08 Mn . . . . 0-74 Si . . . . . 1"22 Cr . . . . 18-08 Ni . . . . 12'20 N . . . . . 0.05 Fe .... Balance

C . . . . . 0.072 Mn . . . . 0"60 P . . . . . . 0.024 S . . . . . . 0'014 Si . . . . . 0"37 Cr . . . . 18"75 Ni . . . . . 8'65 Mo . . . . 0-46 Cu . . . . 0'30 Co . . . . 0'22 N . . . . . 0"032 Fe . . . . Balance

*This is not completely accurate, since intergranular corrosion of austenitic stainless steels have been observed in the absence of grain boundary carbide precipitates. 3 These observations are limited to very specific environments, and thus, the Chromium Depletion Theory can be considered generally correct.

Predicting the intergranularcorrosion of austeniticstainlesssteels

11

Two types of corrosion tests were employed: controlled potential and conventional weight loss measurements; the former is similar to the conventional test but the potential of the specimen is maintained constant with a potentiostat,l°,n All tests were conducted in a polarization celln-13 which facilitated measurement and control of specimen potential (measured relative a S.C.E. with a Luggin-Haber probe). Temperature was thermostatically controlled within q-l°C. Corrosion test duration depended on the severity of attack with a maximum period of 100 h. Extended tests up to 1000 h yielded results nearly identical to those obtained during 100 h. After testing, specimens were dried, weighed and examined microscopically. To detect fine grain boundary attack, specimens were plastically deformed before visual examination. Controlled potential corrosion tests were conducted in pure 1, 5 and 10N HzSO4 at 25 °, 40° and 90°C at various potentials maintained within 4-10 mV. Conventional corrosion testing was performed in these same media with and without oxidizer additions (Ce +4, Fe +a and Cu+2). In some corrosion tests, specimens were galvanically coupled to platinum or copper. RESULTS Controlled potential tests

Intergranular corrosion only occurs in specific potential regions as shown in Figs 1-3. For convenience, the observed attack is divided into three categories: general corrosion, fine intergranular corrosion, and coarse intergranular corrosion, and examples of these three categories are illustrated in Figs 7-9, respectively. Interpolation between points has been used to map zones of intergranular attack, which enlarge with acid concentration and temperature. Tests with sensitized Type 304 stainless steel (Table 2) yielded nearly identical results. Overall corrosion rates calculated from weight loss are plotted in Figs 4-6. These data indicate average overall rates rather than true penetration, since intergranular attack is present in some cases. These curves are very similar to those obtained by potentiostatic anodic polarization.14,15 This is expected since the applied anodic current approximates closely to the actual corrosion rate. 18,17 Comparisons of Figs 1-3 with Figs 4-6 shows that susceptibility to intergranular corrosion begins at approximately the maxima of Figs 4--6. This demonstrates that selective grain boundary attack is restricted to the passive and transpassive regions and does not occur in the active state in these media. Conventional corrosion tests

Results of these tests, conducted in sulphuric acid with and without oxidizer additions, are .tabulated in Table 3; corrosion potentials were measured during testing and are also given in the table. The presence or absence of intergranular attack was predicted on the basis of these potentials and the data in Figs 1-3. There is complete agreement between observed and predicted results. That is, controlled potential and conventional corrosion tests are directly comparable at constant potential, acid concentration, and temperature. The correspondence between controlled potential and conventional corrosion tests is demonstrated more vividly by the detailed comparisons presented in Figs 7-9.

12

W . D . FRANCE, JR. and N. D. GREENE I'60

1.40

I

I

I

®

I

I

I

I

°

I

CF-8 25 ° C

o

1"20 0

1.00

laJ

--

o

0

0"80

> .

0.60

--

0

0.40

--

0

o

0"20

0 ¢

0"00

-0'20

-O'40

I 0

Normality

PIG. I.

I

I

I

5

I0

H.zSO 4

Intergranular corrosion of sensitized cast stainless steel (CF-8) in sulphuric acid at 25°C. o : General corrosion. : Fine intergranular corrosion. • : Coarse intergranular corrosion. TABLE 2. CONTROLLEDPOTENTIAL CORROSION TESTS (Sensitized Type 304 Stainless Steel in 1N H,SO4)

Temp. °C

Potential (V)

Immersion time (h)

Corrosion rate (mpy)

25

0"300

100

0'9

25 25

--0.240 --0"280

100 98

8.3 33

90 90

0-940 --0.150

I00 43

230 3700

90

--0.200

18

7500

Observations general corrosion fine I G A fine IGA, grain dislodgement. coarse I G A coarse IGA, grain dislodgement. coarse IGA, grain dislodgement

Predicting the intcrgranular corrosion of austenitic stainless steels

13

otto 1"40

0

CF-8

4O" C

ILl > o

Q.

0

5

I0

Normality H2SO4

Fro.

2. Intergranular corrosion of sensitized cast stainless steel in sulphuric acid at 40°C. o: General corrosion. : Fine intergranular corrosion. • : Coarse intergranular corrosion.

Here, conventional corrosion tests are compared with controlled potential studies performed at similar potentials. These examples have been selected to show uniform (Fig. 7), fine intergranular (Fig. 8), and coarse intergranular (Fig. 9) attack. In each case, both the type of attack and corrosion rate are almost identical. The correspondence of visual appearances is particularly striking. The slight discrepancies in corrosion rates are due to variations in potentials during the two types of experiments. DISCUSSION The above experiments demonstrate that both the rate and type of intergranular corrosion in a given environment are unique functions of electrode potential. Consistent with other studies, za,t9 small amounts of non-complexing oxidizers such as eerie, ferric and cupric ions, only influence electrode potential and do not alter dissolution behaviour. Oxidizers which form complexes with iron, chromium or nickel

W . D . FRANCE,JR. and N. D. GREENE

14

1'20

I'00

0"80

0.80

"~ bJ

O'40

"~ Q.

0'20

0.00

-0"20

--0"40

0

5 Normollfy

FIG. 3.

I0 HzSO 4

Intcrgranular corrosion of sensitized cast stainless steel in sulphuric acid at

90°C. • : General corrosion. : Fine intergranular corrosion. o : Coarse intergranular corrosion. 160

f

I I IIIll

I

I

i ltiIH

I

I

I IIl~ll

I

I

~ I ~

|

I f IIIfl

i

i i ill

o J~ H~SO,

140

[] 5 N H=SO 4 120

A 10~ H2S0 ,

I 00

o)

.~ 0 8 0 >

.-" 0 6 0

,2

CF-8 25 ° C

.~. 0 4 0

g. 0 2O 000 - 0 2 0 --

-0401 0.I

i

i I iiii1~ 1.0

I

z Is

till I0

I

I Jllgall

J

I tJlIHI

I0 z

I,l 10 "I

I rltfjl i04

Corrosion rote (mpy)

FIG. 4.

Potential vs. corrosion rate of sensitized cast stainless steel in sulphuric acid at

25°C.

105

Predicting the intergranular corrosion of austenitie stainless steels 1"60

--

I

i ' I111' i

1.40

0

IN

I

I I illll

I

I

I i,,'ri,

I

i

i llll,~

I

i

I ~ Lilll{

HaSO 4

I

I

I

I

IIII

. ~

a s ~ u, so, ~*zo

15

~

- -

,', tON HzSO 4

1.00 Ij

CF-8

0'80

> 0"60 ,?-

0.40 S 0.

0-20

0"00

-0.20 -0'40

0.1

~

)

~]lilll

I'0

I

i I ILLIti

IO Corrosion

FIG. 5.

I

I I IIITI[

102

I

I I tlllll

103

rote

I11

104

i05

(mpy)

Potential vs. corrosion rate of sensitized cast stainless steel in sulphuric acid at 4O°C.

"20

~

l -W T T ; I T ~

0-80

N 0.60 ~ ~

0.40

-

o.~o

1 -

p

~ ,L J J ~ q

.... r~-TTrnq

L "T~-,~rTrrl~--]~

Q

i'O0 u.J (j

-T-'-F"I'TTFTT

~-TTrrrT

90 ° C

o I N HaSO4 0 5 N HaSO4 ~, tO ~ H~ SO4

o.oo -0,20

-0.40

i 0.1

i IIIl,1J

I

1.0

I llij11]

i

lilirlll

iO Corrosion

FIG. 6.

I

l]llllll

102 rote

I

103

lil~t"~l

i IIII

104

(mpy)

Potential vs. corrosion rate of sensitized cast stainless steel in sulphuric acid at 90°C.

~C~

W. D. FgANCE,JR. and N. D. G ~ E

16

TABLE3. PREDICTINGINTERGRANULARCORROSION (Sensitized CF-8 Type Stainless Steel (I00 h maximum test duration))

Test

Environment

1. 11'74HaSO4 25°C 2. llq I'~SO4 and Pt couple, 25°C 3. IIKIH~SO4 + 0"2 g/l CuSO(. 5I-I~Oand Pt couple, 25°C 4. Iiq I-I~SO4 + 0-3 g/1 CuSO4. 5H~O and Cu couple, 25°C 5"l'. Iiq I-I~SO4 + 2'0 g/1 CuSO,. 5I-I~O, 25°C 6. llq I-~SO4 + 2'0 g/l Fe2 (SO4)s. 61-I20,25°C 7. IN H~SO4 q- 1.0 g/l Ce (SO~)a. 4I-~O, 25°C 8. 51q H~SO4, 25°C 9. 10N H~SO4, 25°C 10. 10Iq H~SO4 --b 1.0 g/l CuSO4.5HzO, 25°C l l t . 10N H~SO4 + 2.0 g/l Fe2 (SOt)3. 6H20, 25°C 12. 1N H~SO~ + 1-0 g/l CuSO4.5H20, 40°C 13. IN H2SO4 + 0.5 g/l Ce (SO4)z. 4HzO, 40°C 14. 5/',7I-~SO4 -k 8'0 g/l Ce (SO~)2. 4H20, 40°C 15. 10N I~SO4 + 5-0 g/l Ce (SO4)~. 4I--I~O,40°C 16. 1N H2SO4, 90°C 17. lit4H2SO, and Pt couple, 90°C 18. IN H~SO4 + 0'3 g/l CuSO,. 5H~O, 90°C 19. IN I-~SO4 + 4.0 g/l CuSO4.5H~O, 90°C 20. IN H~SO4 + 10-0 g/l Ce (SO4)s. 4H~O, 90°C 21. 10N H2SO, q- 3"0 g/l CuSO4.5H~O, 90°C 22. 10b-/H~SO4+ 5.0 g/l Fe~ (SO4)s. 6H~O, 90°C 23. 10N H3SO4 + 5"0 g/I Ce (SO~)2. 4HzO, 90°C

Corrosion potential (V)

Type of corrosion* Predicted Observed

--0-44 to --0"41

G

G

--0"31 to --0'29

G

G

--0"24 to --0.26

IGA

IGA

0"00 to 0-01

G

G

0"22 to 0"26

G

G

0.30 to 0-54

G

G

0'87 to 0"94 --0.41 to --0.36 --0.38 to --0"36

G G G

G G G

0"24 to 0-25

G

G

0.59 to 0.61

G

G

0-14 to 0-18

G

G

0"92 to 0-93

G

G

1.00 to !.03

IGA

IGA

1"04 to 1.06 --0.43 to --0'42

IGA G

IGA G

--0.37 to --0-36

G

G

--0"06 to --0-34

IGA

IGA

0-23 to

0"25

IGA

IGA

0-92 to

0.95

IGA

IGA

G

G

--0"34 to --0.32 0"57 to

0.64

IGA

IGA

0-99 to

1.03

IGA

IGA

•l-1000 h tests G: General corrosion; IGA: Intergranular attack. (e.g. chloride or chlorine-containing c o m p o u n d s ) would preclude such predictions, since they strongly influence dissolution kinetics. The above results a n d techniques would be useful in engineering applications to predict the i n t e r g r a n u l a r susceptibility of a given e n v i r o n m e n t . There are, however, several precautions which m u s t be observed. First, the data presented in Figs 1-3

Predicting the intergranulax corrosion of austenitic stainless steels

17

contains extrapolated points and may be influenced by time of sensitization. These data may be considered as representative of an extreme condition of sensitization which would be rarely achieved in normal practice. Although there is little difference between wrought and cast alloys of the 18Cr-8Ni type, these results nhay not apply to other austenitic stainless steel alloys. Exposure time may also influence results and, of course, this should be determined before applying these methods to long-term industrial exposure. However, 100 and 1000 h corrosion tests yielded identical results and intergranular attack, when present, was usually visually apparent after 10-20 h. Finally, it is necessary that corrosion potentials remain in a region corresponding to general corrosion if intergranular attack is to be avoided. This may be difficult to achieve or control in some industrial processes where environmental conditions fluctuate widely. Aside from these practical problems, this investigation clearly demonstrates the accurate prediction of intergranular corrosion susceptibility of sulphuric acid solutions. Other media, such as phosphoric and acetic acids, could also be characterized by controlled potential corrosion testing. Each environment probably possesses unique characteristics, since Streicher 2° has shown that there is no correlation between intergranular susceptibility in different acid solutions at nearly constant potential. Intergranular corrosion can be prevented by altering the environment as shown in Table 3. Removing copper sulphate from 1N H2SO4 at 90°C (Test 18) prevents intergranular corrosion (Test 16). Also, anodic or cathodic protection could be employed to shift the potential to a region of general corrosion. For example, stainless steel is immune to intergranular attack and corrodes at less than 2 mpy in 5N H~SO4 at 90°C in the potential range 0.40-0.60 V vs. S.C.E. (Figs 3 and 6). This potential could be maintained by appropriate oxidizer concentration or by externally applied current. SUMMARY (1) The intergranular corrosion of sensitized 18Cr-8Ni stainless steels in sulphuric acid media occurs in limited potential regions whose limits vary with acid concentration and temperature. (2) It is possible to rapidly predict the intergranular susceptibility of various sulphuric acidoxidizer mixtures on the basis of corrosion potential measurements and controlled potential corrosion tests. Acknowledgements--We thank the Office of Naval Research for the support of this programme; Ohio State University and the Alloy Casting Institute for supplying cast stainless steel samples. REFERENCES I. E. C. BAINand R. H. ABORN,Trans. Am. Soc. Steel Treat. 18, 837 (1930). 2. E. C. BA/N,R. H. ABORNand J. J. Rtrr~.RFORD, Trans. Am. Soe. Steel Treat. 21,481 (1933). 3. C. EDELEANU,Chem. Ind. 42, 1360 (1958). M. A. STREICHER,J. electrochem. Soc. 1-6, 161 (1959). 4. Symposium on Evaluation Tests for Stainless Steels, Tech. Pub. No. 93, ASTM, Philadelphia, 1949. Intergranular Corrosion o f Chromium-Nickel Stainless Steel--Progress Report No. 1, Bulletin No. 93, Subcommittee on Field Corrosion Tests, Welding Research Council, January, 1964. 5. E. BRAUNSand G. PIER, Stahl Eisen. 75, 579 (1955). 6. V. Cm,~Land M. PRAZAK,ttutn. Listy 11,225 (1956). 7. L. CLERROIS,F. CLERSOISand J. MASSAR%Electrochim. Acta 1, 70 (1959). 8. H. J. SC~rULLER,P. SCHWAABand W. SCrtWENK,Arch. EisenhiittWes. 33, 853 (1962). 9. W. SC~rW'~NK,H. J. SCHULLERand P. SCHVCAAB,Werkstoffe Korros. 15, 621 (1964). 10. N. D. GRF.F.NF.,C. R. BISHOPand M. STERN,d'. electrochem. Soc. 108, 836 (1961). 11. N. D. GRF.ENE, Experimental Electrode Kinetics, p. 37. Rensselaer Polytechnic Institute, Troy, N.Y. (1965).

18 12. 13. 14. 15. 16. 17. 18. 19.

W . D . F~NCE, JR. and N. D. GREENE

M. S'r~N, J. electrochem. Soc. 102, 609 (1955). M. STERNand A. C. MAra~DES, J. electrochem. Soc. 107, 782 (1960). C. EDELEAN~r,J'. Iron Steel Inst. 188, 122 (1958). N. D. GREENE, Corrosion 15, 369t (1959). M. S'rEaN and A. L. G ~ t Y , J'. electrochem. Soc. 104, 56 (1957). N. D. Gl~a~F.~n~,Corrosion 18, 136t (1962). N. D. GR~ENB, d. electrochem. Soe. 107, 457 (1960). J. M. KOt.OTY~dN, N. J. BUNEand G. M. FLOmANOWCH,Symposium Europ~en sur les Inhibiteurs de Corrosion, p. 493, Universit/t degli Studi di Ferrara, Italy (1961). 20. M. A. S ' r n ~ c ~ , Corrosion 20, 57t (1964).