ML20045E877

Forwards Info Supporting Use of ASME Section IX Code Cases 2142 & 2143,per 930607 Conversation.Info Supports Use of New Weld Matls as Preferred Choice for Welding Applications in Fabrication & Installation of Replacement SGs

ML20045E877
Person / Time
Site:	Mcguire, Catawba, McGuire
Issue date:	06/28/1993
From:	Tucker H DUKE POWER CO.
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
NUDOCS 9307060173
Download: ML20045E877 (100)

Text

{{#Wiki_filter:w ll l llake Ibuvr Company lluilDruR' ' l'O Box 1007 &nior Vice President Char!vtte. SC2M'01.NOi kuclear Generation (704}373 1531 DUKEPOWER June 28,1993 ~ i U. S. Nuclear Regulatory Conmdssion Document Control Desk Whshington, D.C. 20555

Subject:

McGuire Nuclear Station Units 1& 2 Docket Nos. 50-369,370 Catawba Nuclear Station Unit 1 Docket No. 50-413 ASME Section IX Code Cases 2142 and 2143 Gentlemen: D. L. Rehn's letter dated April 29,1993 requested approval in accordance with 10 CFR 50.55a(a)(3) to use the alternative rules of ASME Section IX Code Cases 2142 and 2143. Per conversation with your staff on June 7,1993, Duke Power is submitting information supporting the use of these Code Cases. This information supports the use of these new l weld materials as the preferred choice for welding applications in the fabrication and f installation of the replacement steam generators for McGuire (Units 1&2) and Catawba (Unitl). Very truly yaurs, h; h'.,e$ A/A 7 s Hal B. Tucker MlIH Attachments 02006 i r 9307060173 930628 fDR ADOCK 05000369 ' h[ PDR am m mc.. i ) \\

U.S. Nuclear Regulatory Conunission June 28,1993 Page 2 xc: S. D. Ebneter Regional Administrator R. E. Martin, ONRR V. Nerses, ONRR W. T. Orders Senior Resident Inspector (CNS) P. K. Van Doorn Senior Resident Inspector (MNS) .}

i CASE 2142 r CASES OF ASME BOILER AND PRESSURE VESSEL CODE i Approval Date: November 25.1992 See NumericalIndex for expirstion and any reaffirmation dates. Case 2142 F-Number Grouping for Ni-Cr.Fe, Classification UNS N06052 Filler Metal l Section IX Inquiry: What alternate rules may be applied to grouping UNS N06052 Ni-Cr-Fe welding filler metal meeting the chemical requirements of Table 1 but otherwise conforming to AWS 5.14 to reduce the number of welding procedure and perfonnance qual-ifications? Rep &: It is the opinion of the Committee that UNS N06052 Ni-Cr-Fe welding filler metal meeting the chemical requirements of Table 1 but otherwise con-forming to AWS A5.14 may be considered as F-No. 43 for both procedure and performance qualification purposes. Further, this material shall be identified as UNS N06052 in the Welding Procedure Specifica. tion, Procedure Qualification Record and Perform-ante Qualification Records. This Case number shall be shown on the Manu-fac:urer's Data Report. TABLE 1 CHEMICAL REQUIREMENTS (UNS N06052) Element Composition, % Carton, max. 0.04 Manganese, max. 1.00 Phosphoras, max. 0.020 Sulphur, max. 0.015 Silicon, max. 0.50 Chromium 28.D-31.5 Molybdenum, max. 0.50 Nickel Bal. Columeium, max. 0.10 Aluminum, max. 1.10 Aluminum & Titanium, max. 1.50 Copper,, max. 0.30 Iron 7.o-11.0 Titanium, max. 1.0 Other Elements, max. 0.50 371 SUPP. 3 - BPV

5 CASE 2143 CASES OF ASME BOILER AND PRESSURE VESSEL CODE Approval Date: November 25,1992 See NumericalIndex for expiration and any rea!Crmation dates. Case 2143 TABLEl F-Number Grouping for Ni-Cr-Fe, Classification CHEMICAL REQUIREMENTS (UNS W86152) USS W86152 Welding Electrode Section IX C *P 5* "' % Carton. max. 0.05 Inquiry: What alternate rules may be applied to Mancanese, max. s.c0 grouping UNS W86152 Ni-Cr-Fe welding electrodes P h 5* h ' *** - 0 030 meeting the chemical and mechanical properties of [13 Tables 1 and 2 but otherwise conforming to AWS 3 chromium ca.0-31.5 A5.11 to reduce the number of welding procedure Marydeenum, max. o.so and performance quali5 cations? Nicket Bat Columbiam, max. 1.0-2.5 Aluminum, max. 0.50 Cc;rer,, max. 0.50 1ron 7.0-12.0 Titanium, max. 0.50 c:.ner Elemen:s, r.ax. 0.50 Reply: It is the opinion of the Committee that LHS WS6152 Ni-Cr-Fe welding electrodes meeting the chemical and mechanical properties of Tables 1 and 2 but otherwise conforming to AWS A5.11 may be considered as F-No. 43 for both procedure and per-formance qualification purposes. Further, this ma-TABLE 2 terial shall be identified as UNS W86152 in the MECHANICAL PROPERTY REQUIREMENTS Welding Procedure Specification, Procedure Quali. UNS 8 fication Record and Performance Quali5 cation Records. Tensu.e strength, rn n., ksi go Th. Case number shall be shown on the Manu-is ggn;3 io, ;n 2 in,, min,, y, facturer's Data Report. 3D SUPP. 3 - BPV

3 s a ) m Technical Nots - Intergranular Corrosion of High Chromium Nickel Base Alloys

• l C. L BRIANT and E. L HALL
• i f

f ( i i k introduction usun - aner c==p==ee a sa wm Intergranular stress corrosion cracking (IGSCC) has been as cr as p. m. m e. c j reported to pecur in parts of boiling water reactors rnade from a, the nickel base alloys designated as 600 and 182.u As a a ns as im am - im em ur set oss tcsult, research has been directed at trying to understand the causes of this cracking. Studies" have now demonstrated 88" o es ris ees ss u em tse esa oms that in neutral, oxygenated water, and acidic environments this cracking is caused by chromium depletion at the grain i boundaries, which has resulted from the precipitation and growth of intergranular chromium-rich carbides. i One possible way to avoid this cracking might be to use 28.3 wt% chromium, respectively. The chromium content of alloys with chromium contents higher than the 14 to 16 wt% heat A is similar to that found in Alloy 82, whereas the usually found in Alloys 182 and 600. Although chromium deple- , chromium content of heat B is similar 1o that found in Alloy 'j tion could occur in a high chromium alloy, the actual 690. Heat C has a chromium content similar to heat A, but con-chromium concentration a+ the grain boundary might never tains a higher iron concentration. This heat was included become low enough tot make the alloy susceptible to inter-because previous investigationsW8 have suggested that an [ granular corrosion (IGC) and stress corrosion cracking (SCCL increase in the iron concentration will raise the activity of car-t Some research *3 has begun to address this idea. In a bon in solution and therefore enhance chromium depletion. E previous publication,e it was shown that Alloy 82, which con. Heat D is similar to heat A but is specifically doped with phos. tains 19 wt% chromLm, is much more resistant to IGC in boil-phorus. This heat was included because phosphorus can ing 25% nitric acid than either Alloy 600 or182, and Page" has enhance iGC in acidic environments.3 '[ shown that Alloy 82 is more resistant to SCC than the other All samples were annealed at 1200 C for 3 h f.nd water i two alloys. Page and McMinn" have shown that Alloy 690' quenched. Individual samples were then aged at 600,650, or which contains -28 wt% chromium, is immune to IGC in boil-700 C for times of 1,3. 5,18,24, or 100 h and water quenched. ing 25% nitric acid and SCC in oxygenated pure water; this All heat-treatments were conducted with the samples in flow-I alloy showed a small amount of IGSCC when tested in slightly ing argoa 2 13 actoic water at 288 C. Floreen and Nelson reported that the The corrosion resistance of the samples was measured by I SCC resistance of weids made with nickel-base filler metals irb placing them in a boiling solution of 25% nitric acid fo; 2 days. l creased significantly if the chromium content was above 24%. Weight loss was used as a measure of corrosive attack. This ( Their tests were run in acidified water at 316 C-test was chosen because previous researchM ' has demon-j in this r.ote. the authors examine the IGC susceptibility of strated that it is very sensitive to chromium depletion, and it j four high chromium nicket-base alloys given heat treatments has become a standard screening test. for the corrosion that produced significant IGC in Alloys 600 and 182. These susceptibility of NiCr alloys. Also, research has suggested that results were also combined with those reported in Drevious samples that corrode extensively in this test will also undergo - j worge to suggest what combination of chromium and carbon IGSCC when tested in oxygenated water.3 { concentrations is required to produce a material that cannot The microstructures of the samples were examined by op-be made susceptible to IGC through heat-treatment. tical and transmission electron microscopy. The details of the j method of sample preparation for transmission electron t microscopy (TEM) are given in Reference 15. j Experimental Four heats of matenal were tested in this program Table 1 lists their compositions. Heats A and B contain 20.8 and Results The corrosion tests showed that all four alloys were ex. l

• Submitted for publication November 1986.

tremely resistant to IGC. No weight loss could be detected on

• Generat Electric Corporate Research and Development, P.O.

any of the samples in any of the heat-treated conditions. The Box 8, Schenectady, New York 12301. samples were examined by scanning electron microscopy i 0010-9312/87/000239/53.00/0 Vol. 43, No. 7, July 1987 @ 1987, National Association of Corrosion Engmeers 437-

g j ,a g, I y,' r g s g . ; y. e.a u -; ~y m. Tk $'L.x

Q'

? I. . A),. y# y y, j. ,f ~ .. f. s h(g. M $ )".1.',., y* \\ 3* . Q. t 2 y ~4sp: 7 y& [1 "l M .o f. k.+ y?'h. - .. Y T9 . j '- hg ( <. Qv .. f T; [ y- ? \\' .h' _ f. = :. j6 n f W:. ~ IW . :, r (- l q

p..

.h ' W .i. g. .7 FIGURE 1 - An SEM micrograph showing the maximum 4 }.g' * + - .s ~. i extent of corrosion observed on any of the samples used 3 in this study after they had been subjected to the 48 h y *g$ p{ '.,( l boiling 25% nitric acid corrosion test. This micrograph j j is of a sample of Heat A aged at 700 C for 24 h. I*, A i 1 - ' ss 9 l q g (SEM) to determine if any corrosive attack occurred. The only [' ' 8* i attack that was observed was the slight intergranular etching a - shown in Figure 1. 5 h.' l Samples of heats A, B, and C that had been aged at 650 C for 100 h were examined by TEM. The grain boundaries of all + three samples contained very heavy carbide precipitation, as shown in Figures 2(a) and (b). Carbides also appeared on the l dislocations, which were piled up near the grain boundaries. l All of the carbides were chromiurrwich MnC,. It was also l' l found that the carbides along any given grain boundary dis-played a cube-cube orientation relationship, with one of the i grains composing the boundary. The carbides along the dis. j location lines were oriented with respect to the austenite grain j in which they lay. This point is demonstrated in Figure 2(c), [ which is a dark-field image of the carbides at a grain boundary, l "m C diffraction spot. Note that all of the b farmed using an Mne l carbides along the boundary were visible. l Optical metallography was used to follow the carbide

• 9 i

I precipitation in these samples. It was found that the solution l anneal dissolved most or all of the carbices. However, as the Samples were aged, the carbides rapidly precipitated along [ i j j the grain boundaries. 1 g 4 *e=W E.43 ee.

• 19 E 7. ag. '.' g j

Discussion The results presented abbve clearly show that these high ' ~. ' ~ l ing in the 25% boiling nitric acid test. Neither the precipitation ~

i. ' [ M,*hkN cntomium, nickel-base alloys are essentially immune to crack.

t J ~~ ,.i / C' of carbides nor the presence of phosphorus in the alloy i.D l ': w i E.d.8y caused degradation of their corrosive resistance, in a sample g), 7 ~ such as that shown in Figure 2, the chromium level at the L ~ - carbide-matrix interface will be lower than that in the bulk ( ~*J i l l matrix, but it is clearly high enough to prevent attack in the m corrosion test. These results are considered along with ~ E previously publishec" results on Alloys 600, 82, and 182 to N ~ I~ cetermine which compositions will be susceptible to attack in C i this test and which ones will not. Research has demonstrated that immunity to cracking FIGURE 2 - A TEM micrograph of thin foils of samples l will occur if the grain boundary chromium level never falls aged at 650 C for 100 ft (a) shows a bright field image of l below -12 wt%M Ideally, the actual chromium concentra-a sample of Heat A;(b) shows a bright field image of a j tion that should be in equilibrium with the carbide could be sample of Heat B;(c) shows a dark field image of a sam. j calculated from thermocynamic parameters. If this value is pie of Heat A. The. image in (c) was formed using an Mn, detraction spot and shows the carbides along the j below 12 wt%, corrosion would be predicted to occur. How-C I ever, to make predictions for a vareety of alloys, very accurate grain boundaries and within the grains. j values are needed fo/ the activity coefficients of carbon and i chromium in solution, but these are difficult to obtain. Using a j much more empirical approach, which has proven very suc-l 1 i 438 CORROSION-NACE q

I M 1 .p k.'. M. E -; ' h, wg N ( " yv. r a.. ,f ] g I ~ .q,

• f

~W e h ~ > ;# _j$ ~ l* 's

4. 9 y

ew 4 4 L y V ' y'fMG.D. % *. - ) m 1 C g. -.c 5....' n N 1 h. a

• * ~

.f. kw s e ~ p (V j% h ~"TO i m i' _ w . [..p j j. .~ ' ( $., m . 'i Q,p FIGURE 1 - An SEM micrograph showing the maximum

4. }.f'

~ l extent of corrosion observed on any of the samples used 3 l In this study after they had been subjected to the 48-h y *g g,.$ p{ boiling 25% nitric acid corrosion test. This micrograph ',S

• l Is of a sample of Heat A aged at 700 C for 24 h.

7 i *,. pF 4- ,A j N ~~ l 'i W (SEM) to determine if any corrosive attack occurred. The only [ 8O f i attack that was observed was the slight intergranular etching .a i shown in Figure 1. $E l Samples of heats A, B, and C that had been aged at 650 C for 100 h were examined by TEM. The grain boundaries of all i three samples contained very heavy carbide precipitation, as l shown in Figures 2(a) and (b). Carbides also appeared on the l dislocations, which were piled up near the grain boundaries. All of the carbides were chromium-rich M C,. It was also,' n found that the carbides along any given grain boundary dis-I i played a cube-cube orientation relationship, with one of the 3 grains composing the boundary. The carbides along the dis-l location lines were oriented with respect to the austenite grain in wnich they lay. This point is demonstrated in Figure 2(c), which is a dark-field image of the carbides at a grain boundary, formed using an M C, diffraction spot. Note that all of the b 23 carbides along the boundary were visible. Optical metallography was used to follow the carbide precipitation in these samples. It was found that the solution ] anneal dissolved most or all of the carbides. However, as the s samples were aged, the carbides rapidly precipitated along .. 7 - l ' }. c the grain boundaries. p 4 *e W e.43 ee. ' M D ".'. a. e, :, 'A Discussion f The results presented abbve clearly show that thece high .j. ..,g chromium, nickel-base alloys are essentially immune to crack-t s' }j gI ing in the 25% botting nitne acid test. Neither the precipitation

• C',

of carbides nor the presence of phosphorus in the alloy C-A g . '.Mi M caused degradation of their corrosive resistance. In a sample such as that shown in Figure 2, the chromium level at the L. ~ ' carbide-matrix interf ace will be lower than that in the bulk (

• l

~ ~ '.. matrix, but it is clearly high enough to prevent attack in the g= corrosion test. These results are considered along with previously published s results on Alloys 600, E2, and 182 to N ~ i determine which compositions will be susceptible to attack in g this test and which ones will not. Research has demonstrated that immunity to cracking FIGURE 2 - A TEM micrograph of thin foils of samples will occur if the grain boundary chromium level never falls aged at 650 C for 100 h:(a) shows a bright field image of below - 12 wt%.M Ideally, the actual chromium concentra-a sample of Heat A;(b) shows a bright field image of a tion that should be in equilibrium with the carbide Could be sample of Heat B;(c) shows a dark field image of a sam-calculated from thermodynamic parameters. If this vane is pie of Heat A. The image in (c) was formed using an below 12 wt%, corrosion wouM he predicted to occur. How. M C, def raction spot a nd shows the carbidos along the 23 ever. to make predictions for * - ' of alloys, very accurate grain boundarios and within the grains. values are needed for the ace efficients of carbon and chromium in solution, but thesi ..N.nfficult to obtain. Using a much more empirical approach, which has proven very suc-438 CORROSION-NACE

- u ai. n _ e v m e, a T n m. a.mo """' O.6 4Al seeee - weassii Lene "!a" 0.5 e e-wt% wt% ' essumesw% .== N 400 290s tu SOS U Desentagrated 3 Es 0.4 ~ e-i en 2 is2 r tu DJue as ass s m 0.3 -a m + 500 stM 141 DD4 7.1 Doesntegraise 3 0 l 6 .= no ev n .,a e Dovrean.s t ; 0.2 a a. m. n e s o sonstrataan y et tempaa 5 0.1 t t Dmsneense 400 L2 10 0.025 11.1 022 3 ^ u iu om m o a 0 4 8 12 16 20 24 a un e. 2. mi Doctroon ([III i A 2nA 28 tu 0 he mm C 21 3 0 019 17.7 0 he study FIGURE 3 - Cr=" plotted as a function of weight loss o 21 4 ms su o m' "* measured af ter samples had been subjected to the 48 h a 2u ams an a

w. mus, boiling nitric acid test. Information about the data used to construct the plot is given in Table 2.

j cessful in predicting the corrosion resistance of stainless resistance of these alloys and, consequently, the value of Cr*" f steel (SS),ms is proposed, above which corrosion will not be observed. Clearly, the two most important parameters in determin-ing if significant depletion will occur will be the carbon and Conclusions chromium concentrations in the alloy. Theretore, if one follows

1. AHoys containing 20% or more chromium and carbon the procedure often used for SS, then concentrations, at least below about 0.02 wt%, are immune to Cr*" = Cru - AC.

corrosion in the boiling 25% nitric acid test.This conclualon is . not changed when phosphorus is present in the alloy, where Cr*" is a parameter representing the amount of deple.

2. Very dense precipitation of chromium-rich M C. car.

n am and Cwm are the bulk concentra. bides occurs along the grain boundaries ' of these high tion that would occur, Cr tions of chromium and carbon, respectively, and A is a multi, chromium alloys when they are aged in the range of 600 to 700 en C. i plicative factor. If A is allowed to have a value of 200, Cr

3. A parameter designated as Cr*", which considers the values very similar to actual chromium concentrations.

carbon and chromium concentrations in NICr alloys, appears measured at the grain boundaries of various nickel-base alloys af ter aging at 650 C are obtained.s.s to predict whether or not corrosion will occur in the boiling 25% nitric acid test. Table 2 compares the calculated results with the corro-sion behavior, and Figure 3 presents the results graphically. Acknowledgments The corrosion data that are used represent the maximum cor. The authors would like to thank Bruce Knudsen for his resion observed for samples af ter they had been aged between technical assistance. 600 and 700 C. It can be seen that when Cr*" exceeds a value of 12 to 15, corrosion ceases to be observed. References Although the formula for Cr*" appears to work very well. 1, E. Serra. Special Report No. NP-2114-SR, Electric Power several Qualifications should be made. First, it is certainly the Research Institute, Palo Alto, California,1981, 1 simplest formula to use. Other elements present in these

2. E. F. Kearney, T. L Chapman,' D. E. Delwiche. Trans. 8th -

{ alloys could affect the amount of chromium depletion through int. Conf. on Structural Matenals in Reactor Technology, their effect on carbon and chromium activities, and with a Vol. D1/2, Brussels. Belgium. August 1985. l larger test matrix, their effects could be incorporated into the

3. C. L Bnant. C. S. O'Toole, E. L Hall, Corrosion. Vol. 42, j

formula. Second, research has clearly shown that if chromium No.1, p.15,1986. depletson occurs, the presence of phosphorus on the grain

4. G. P. Airey, Metallography, Vol.13, p. 21,1980.

bounoanes can accelerate corrosion in this test.3 in contrast,

5. E. L Hall, C. L Bnant, Met. Trans. A Vol.16A p.1225,-

phosonorus segregation alone has little etfeet. This factor has 1985. not been included in this f ormula. Finally, it has been previous-

6. G. S. Was, H. H. Tischner, R. M. Latanision, Met. Trans. A, j

ly shown that alloys that would normally undergo severe Vol.12A, p.1337,1981. chromium depletion can be heat treated near 1050 C to pro-

7. G. S. Was, J. R Martin, Met. Trans. A. Vol 16A, p. 349, duce a high density of intragranular carbides.15 Aging at 600 1985.

} to 700 C does not produce extensive grain boundary precipita-

8. C. L Bnant, E. L Hall, Corrosion, in press.

tion. If the material has this type of structure, and the corro-

9. P. L Andresen, CORROSIONIB4, Paper No.177, National

{ sion rate will be low. Therefore, the complete heat treatment Association of Corrosion Engineers Houston, Texas, cycle can play an important role in determining the corrosion Apnl1984. i Vol. 43, No. 7, July 1987 439-j 7

10. G. J. Theus, R. H. Emanuelson, Electric Power Research

14. T. Wada, H. Wada, J. F. Elliot, J. Chipman, Met. Trans.,

Institute Report NP-3061 Project 5192-2, Electric Power Vol. 2, p. 2199,1971. Research Institute, Palo Alto, California, May 1983.

15. E. L Hall, C. L Briant, Met. Trans. A, Vol.15A, p. 793, i

11. R. A. Page, Corrosion, Vol. 39, No.10, p. 409,1983.

1984.

12. R. A. Page, A. McMinn, Trans. A, Vol.17A, p. 877,1986.

16. C. L Briant, R. A. Mulford, E. L Hall, Gorrosion, Vol. 38,

13. S. Floreen, J. L Nelson, Final Report on EPRI Contract No. 9, p. 468,1982.

RP-1566-2, Electric Power Research Institute, Palo Alto,

17. V. Cihal, intergranular Corrosion of Steels and Alloys, Calif omia, March 19S4.

Elsevier, Amsterdam, Holland, p. 130,1934.

18. S. M. Bruemmer, Corrosion, Vol. 42, No.1, p. 27,19S6.

Sodium Sulfate-induced Corrosion of Pure Nickel and Superalloy Udimet 700 in a High Velocity Burner Rig at 900 C* A. K. MISRA

• Abstract Sodium sulfate induced corrosion of pure nickel and a commercial nickel-base superalloy, Udimetm 700 (U-700), were studied at 900 C in a Mach 0.3 burner rig with different Na levels in the combustor.The corrosion rate of Ni was ir: dependent of the Na leve!in the combustor and considerably lower than that measured in laboratory salt spray tests. The lower rates are associated with the deposition of only a small amount of Na2SO, on the surface of the NiO scale. Corrosion of U-700 was observed to occur in two stages. During the first stage, the cor-rosion proceeds by reaction of Cr O scale with the Na2SO, and evaporation of the Na:CrO, 2 3 reaction product from the surface of the corroding sample. Cr depletion in the alloy occurs

{ and small sulfide particles are formed in the Cr depletion zone. Extensive sulfidation occurs during the second stage of corrosion, and a thick scale forms.The relationship between the corrosion rate of U-700 and the Na level in the combustor gives a good correlation in the range of 0.3 to 1.5 parts per million by weight (wppm) Na. Very low levels of Na in the combustor cause accelerated oxidation of U-700 without producing the typical hot corrosion morphology. introduction conditions; therefore. they are frequently used to rank various Deposrtion of Na2SO on turbine blades and vanes is alloys for hot corrosion resistance. However, there have been 4 known to cause accelerated corrosion of these components; no mechanistic studies in a high velocity burner rig, this is conventionally known as hot corrosion. From several presumably because of a lack of adequate control over various studies on the hot corrosion mechanisms of metals and operating parameters. Moreover, there is no unique salt level i alloys," the various mechanisms can be broadly classified at which the bumer rig tests are conducted in different labora. under two categones:(1) the fluxing theory in which Na SO, tories, and wide variations in the salt levels are used. The Na 2 dissolves the protective oxides, and (2) the action of Na SO, as level in the combustion gases influences the deposition rate of 2 a source of sulfur, which introduces sulfides into the alloy. Na:SO, on the corroding sample and could presumably also Two different test methods are used in most of the influence the corrosion rate. The effect of the contaminant mechanistic studies-the laboratory salt spray test and the flux rate (CFR) has been emphastred in werk by Hancock.5 crucible test in the former, a fixed quantity of Na:SO, is 7 Work in this laboratory has shown that the depth of internal scrayed onto the samples before oxidation at elevated temper-penetration increases with an increase in the amount of salt in ature, in crucible tests, the alloy is immersed in a crucible con-the combustor, reaches a peak, and then decreases with fur-tanning the molten salt. These procedures are useful for ther increase in the amount of salt. Recently, Lowell and Dead. mechanistic studies because certain parameters, e g., temper. 8 more have studied the kinetics of hot corrosion of Udimet 700 ature, gas atmosphere, etc., can be precisely controlled. (U-700)in a high velocity burner rig, using a nondestructive test However, these test procedures suffer from two major draw. procedure called the inductance techruque. They statistically backs. First, the sait spray test involves one-time application correlated the corrosion rate to the relative rate of salt deposi. of Na SO., where% a conumous or near continuous deposi-tion. To relate the burner rig tests with actual engine condi-2 tion of NaQ occurs under actual service conditions. Sec-tions, a correlation between the corrosion rate and deposition ond, in nrvice conditions, the materials are exposed to high rate must be established. For example, it is important to know velocity combustion gas fiows, which are absent in the labora-il a burner rig test with 5 ppm salt adequately represents an tory furnace tests, Burner rig tests can simulate these two engine with 0.1 ppm salt in the atmosphere. Also, deposition. corrosion relationships would be usefulin the development of

• Submitted f or publication February 1986; revised September any corrosion life prediction model.

1936. Most of the burner rig corrosion tests are performed on a

• Department of Metallurgy and Materials Science, Case carousel of samples, and in many cases, only a fraction of the Western Reserve University, Cleveland, Ohio 44106, and total length of tne sample is exposed to the hot combustion Resident Scientist, National Aeronautics and Space Ad-gases. This results in a temperature gradient in the sample ministration, Lewis Research Center, Cleveland. Chio 44135.

and nonuniforrn corrosion; therefore, it becomes difficult to t" Registered trade name. obtain meaningful corrosion rates. Extensive stucies of the 0010-9312lU:000241!S3.0010 m , a ...:--,....w..~ a n----

l l f Stress Corrosion of High Chromium Nickel-Base Weld Metals and AISI 316 Nuclear Grade Stainless Steel in Simulated i Boiling Water Reactor Environments

• I

/ t l \\ A. McMINN* l l Abstract The stress corrosion cracking ($CC) susceptibility of AISI 316 NG (nuclear grade) stainless W steel (SS), which had been weldeo with three h#gh-chromium nickel-base weld metals (I 72, R 127, and R-135), was investigated by means of a senes of slow stra a strain rate of 2 x 10 7 environments at 2BB C, which simulated either the normal BWR primary coolant chemistry or li the impure (SO.2-) chemistry that results from the intrusion of resin from the deminera zer system. The results indicated that all of the materials were immune to SCC in the norma i enwonment. In the resin intrusion environment, both the 1-72 and R-135 weld metals were m-i mune :c SCC, but the R 127 material exhibited intergranular stress corrosion crack ng king (IGSCC). The AISI 316 NG SS was susceptible to transgranu l i l' immune to SCC. (CF) ions '.s However, the transgranular crack growth rates Introduction for AISI 316 NG SS are lower than the intergranular crack The intergranular stress corrosion cracking (IGSCC) of weld-growth rates for regular AISI 304 and 316 by at least a f acto sensitized AISI 304 stainless steet (SS) piping in bosting water three.s reactors (BWRs) has been a subject of concern since the mid Concurrent with the pipe cracking problem, on a more 1960s. when the problem was first detected. The generic limited scale, BWRs have also experienced IGSCC of their nature of the problem was recognized in the mid 1970s, when recirculation piping safe-ends, which join the piping to the an outbreak of crackmg occurred in the smaller diameter by-pressure vessel The recirculation saf e-ends are either an aus-304,316, or 316 NG) or a nickel-base alloy, in-pass and core spray piping. The problem has continued overtenitic SS (AISI the last decade, but over the same period, much effort has G Alloy 600. Most of the earfy f ailures were in the austen-conel been expended in developing and implementing remedies to itic SS materials that had been furnace sensitized whe One of the remedies has been to develop alter-pressure vessel had been post we!d heat treated. The first the prob!cm '2 native stress corrosion cracking (SCCFresistant materials, off ailure in an Alloy 600 safe end forging occurred in 1978 in the which the nuclear grade (NG) SSs, ALSI 304 NG and 316 NG, are recirculation intet nozzle of the Duany Arnold BWR. Results of by f ar the most important. These steeis have a low carbon con-the Duane Arnold f ailure analysis indicated that the IGSCC in-s tent to avoid sensitization, but contain nitrogen to maintain itiated in a we!d heat aff ected zone and propagated through a 2 Boiler and Pressure weid made with inconel Welding Electrode 182. Since then, a the strength required by the ASME Vessel Code. The NG materials have less than 0.2 wt% carbon number of other cases have occurred in which IGSCC h and between 0.06 and 0.1 wt% nitrogen, compared to up to detected either in the furnacosensitized austenitic SS 0.08 wt% C and less than 0.1 wt% N for the regular AISI 304 end material or in the 1-182 weld metal. As for the pipe cracking problem, one of the reme-and 316 SSs. The NG matenals have been found to be much more IGSCC dies has been to develop more SCC-resistant materials. A pro-resistant than the regutar AISI 304 and 316 SSs in laboratory posed replacement material for Alloy 600 is the high-chromiu tests,' althougn AISI 316 NG SS has been found to be sus-Al!oy 690. Alloy 690 exhibits thermai and mechanical proper-ceptible to transgranular stress corrosion cracking (TGSCO) in ties similar to those of Alloy 600, and it appears to be signifi-impure environments containing suif ate (S0,2-) and chloride cantly more resistant to IGSCC.7 High-chromsum weld alloys (R.127 and R-135) compatible with Alloy 690 have also been j

• Submitted for publication January 1986; revised March 1986. developed that can be used to replace 1-182 and/or incone!

Institute. 6220 Culebra Road, San

• Southwest Research Filler Metal 82.

Antonio, Texas 78284. ". ults from taboratory slow strain rate tests (SSRTs) WRegistered trade name. CAmerican Society of Mechanical Engineers, New York, GRegistered trade name. New York. md 0010-9312/86/000203t53.0010 E CORROSION--NAC

• M (%meinn Encineers

TABLE 1 - Ch3mical Composition of tha Test Altoys This paper describes it'e results of a study to evalutte the SCC susceptibility of AtSI 316 NG SS when welded with either e., f, f 3-R-127, R 135, or 172 weld rnetals. A series of SSRTs was per. sI'~ Element Alst 316 NG R-127tR-135 l72 formed on these weldments at 288 C in both a simulated nor-wt% (Heat 33461) (Heat NX4123H) (Heat CKL1139) mal operating BWR environment and a simulated resin intru. sion BWR environment. The objective of the work was to C 0.016 0.02 0.08 define conditions under which these alloys may be susceptible Mn 1.03 0.40 0.01 to SCC. Fe Dalance 8.63 0.03 s 0.0c5 0 003 0.001 Experimental Procedures Si 0.53 0 44 0.01 Comme,cially melted AISI 316 NG SS in the form of a Ni 10.17 60.72 56.26 47.mm-thick plate was used as the base material for 25-mm-Cr 16.90 29.26 42.80 thick weldments. The plates to be loined were machined at 19 i Al 0.001 0.23 0.20 and 45 degrees to yield an included angle of 64 degrees and a l Ti <0.01 0.30 0.61 19 mm root gap was used. The unusual geometry of this weld Cu 027 0.01 preparation was necessary to enable long, cylindrical tensile-M 2 type SSRT specimens to be machined from the weldments at y 5 specific locations. Welds were prepared using three high. } P 0.026 chromium nickel-base weld metals: 1-72, R 127 and R 135. N 0.079 Table 1 gives the chemical compositions of these weld metals Sn 0.011 Co 0.26 and the AtSt 316 NG SS. The covered welding electrode (R-135) and filter metalIR-127) were both prepared from the same heat of material. Gas tungsten are welding (GTAW) was used for the l-72 and R 127 filler metals. and shielded metal arc welding conducted in both a simulated BWR resin intrusion environ. (SM AW) was used for the R-135 material. Eacn weld was radio-trent and a simulated normal BWR water environment have graphed to ensure that only sound welds were used to prepare I shown that Alloys 600. 690. and 60 * *38 weldments made the SSRT specimens, which were tested in the as-welded con-with inconel Filler Metal R-127 and trL et Welding Electrode dition. R 135 are much more resistant to SCL en those made from SSRT specimens were cut from the weldments at two lo-l182.7 U-bend tests in a similar resm t. trusion test environ-cations. The first specimens (Type A) were cut so that the Qace ment have confirmed these results and, in addition, have length sampled the AISI 316 NG SS, fusion zone, and weld shown that inconel Filler Metal 72 is equally SCC resistant.e o metal. Thus. dependent on their susceptibilities. cracks could This work also demonstrated that cracking susceptibility de-propagate completely through either of these materials or the creased as the chromium content of the weld inetalincreased, fusion zone. The second specimens (Type B) were cut so that Thus. potential remedies to both pipe cracking and safe-the gage lengin consisted almost entirely of the weld metal. end cracking involve the use of alternative materials:(1) AISI although one end of the gage length did partially sample the 316 NG SS for replacement piping and (2) high-chromium AISI 316 NG SS and fusion zone. Thus, any cracks initiated in nickel base weld metals (R 127, R-135. and/or 1-72) for welding the AISI 316 NG SS would intersect the fusion zone and weld the piping to Alloy 600 or 690 saf e-ends. However, the immuni-rnetal before completely propagating through the gage sec-ty of AISI 316 NG SS welded with these weld metals has not tion. Figure 1 shows the geometry of weldments prepared and been demonstrated. the locations of the Type A and B specimens.The specimens I 316 NG 316 NG Weld \\'b :, ' :.y y Type A SSRT l Specimen g ( [ Y (a) l 64' I h + 316 NG 316 hG i "l VL l .%.,. M h s w Type B SSRT l Specimen 25 U f i 19 =*- (b) FtGURE 1 - Schematic diagram showing the weld preparation and location of the (a) Type A SSRT specimen and (b) Type B SSRT specimen. All dimensions are in mm. 5 Vol. 42, No.11, November 1986 683

TABLE 2 - Test Results in a Simulated Norrnal BWR Environmentm Time to Elongation RA Failure UTS Weld Specimen (%) (h) (MPa) Type of Failure Metal Type (%) Ductile in AISI 316 NG 22.2 71.9 312 492 A l-72 B 26.9 47.7 367 592 Ductile in fusion line I-72 A R 127 /s 30.3 47.9 383 502 Ductile in weld metal R-127 B 31.5 57.3 367 486 Ductile in weld metal R-135 A 27.6 67.0 335 483 Ductile in AISI 316 NG R-135 8 33.7 42.2 407 570 Ductile in wetd metal f ] 4 s-'. ) O'2BB C. 7.93 MPa, pH 7. 0.1 ppm 0, and 2 x 102 Failure in the fusion line propagat 2 Tables 2 and 3, which show the data for the norrnal primary were cylindocal one-piece tensile specirnens with 6.35.mm coolant environment and resin intrusion environment. respec-diameters and 25.4-mm gage lengths. tively. All four of the alloys tested (1-72, R-127, R 135. and AtSt The SSRTs were performed in a six specimen autoclave 316 NG) demonstrated immunity to SCC in the normal primary system. enabling two specimens from each weldment to be coolant environment (Table 2). The fracture surf aces were en-tested simultaneously. The specimens were tested in paradel tirely ductile in appearance, and with the exception of the so that the test was not interrupted when one of the speci. R 127 specimens, which exhibited some wefd fissures that had i mens broke. Each of the specimens was galvanically isolated opened upon straining, no surf ace cracking was observed.The from the autoctave. loading frame, load train, and other speci-type of ductile f ailure ditfered, however, between the AISI 316 I mens by use of Zirconia and Teflon

• inserts. Tests were run at NG SS and wefd rnetals. The AIS1316 NG SS exnibited a cla I

a constant crosshead displacement rate equivalent to a strain cai necked cup-and-cone f racture with a high reduction in area s-'. The autoclave was incorporated into a (RA) (67 to 72%), whereas the weld rnetals all exhibited a 4 rate of 2 x 10 hign-pressure /-temperature. recirculating water loop that was 45 degree stant or shear-type ductile fracture with a much operated at a pressure of 7.34 MPa, temperature of 288 C, and lower RA(42 to 57%). It is interesting to note that in the Type A flow rate of 11.4 Uh. Further details on this environmental test specimens, in which failure could occur either entirely in the f acility and the six-specimen SSRT system have been pre-weld rnetal or the AISI 316 NG SS, ductile failure always oc-sented elsewhere.W curred in the SS except for the R-127 weld rnetal, in which The SSRTs were conducted in two environments: (1) af ailure occurred in the weld. Faiture in the SS would be ex-high-purity water environment to simulate normal pnrnary pected because of its lower strength compared to the weld coolant chemistry and (2) a sulfuric acid-dosed high-purity rnetals; therefore, failure in R 127 probably reflected the presence of weld fissures in this specimen, which reduced its water environment to simulate a resin intrusion water chemistry. The undosed high-purity water environment con. tensile strength. 1 sisted of double deionized water with a conductivity of ( Table 3 presents the results of the tests run in the simu-Sicm and an oxygen level of 0.1 ppm. The oxygen level was lated resin environment. In general, very good reproducibility slightly below that under normal BWR operating conditions of the test results was observed for the cupticate tests, even (0.2 ppm). The simulated resin intrusion environment con

• though in some cases, the location of the failure moved from sisted of high-punty double-deionized water containing 1 ppm the weld metal to the fusion line and vice versa. The I 72 and H,SO., which resulted in a room temperature conductivity and R-135 weld metals were found to be immune to SCC in this en-pH value of ~ 8 gS/cm and 4.6, respectively. This environment vironment, with ductile failures occurring either in the weld was oxygenated (6 ppm of dissolved oxygen) to accelerate rnetal or in the fusion line (Type B specimens). The R 127 weld cracking. Duplicate tests were performed in this environment metal on the other hand, exhibited SCC susceptibility in this to determme the repeatability of the test results.

environment. For the Type A specimens, f ailure occurred pref-in a separate test, the electrode potentials of the four erentiaffy in the A!SI 31ti NG SS, with the exception of one in anoys were measured in the simulated resin intrusion environ-the R 127 weldment. Interestingly, the AtSt 316 NG SS f ailed by ment at 288 C and 8 ppm O.The potentials were measured us-SCC in specimens containing R 135 and R-127 weldments, but 2 ing an internal Ag AgCi reference electrode of a design similar it exhibited totally ductile failure when welded with the I 72 to that described by Magar and Morris." Once stable potential weld metat. readings were obtained at 8 ppm 0, the water in the makeup Fractographic analysis of the R-127 weld metal fracture 2 tank was deaerated with argon, and the change in electrode surf aces indicated that f ailures occurred by the combination potential was monitored as the oxygen level decreased to 0.05 of a single stress corrosion crack and ductile failure (Figure 2). The depths of the stress corrosion cracks were between 1.6 ppm. After the SSRTs were completed,the specimens were ex-and 1.9 mm. The SCC was found to have an intergranular mor, amined, using conventional light microscopy and scanning pnology. In contrast, examination of the failures in the AtSI electron microscopy (SEM), for (1) indications of SCC on the 316 NG SS revealed that the failures had occurred by a com-tracture surface and (2) surf ace cracking along the specimen bination of. shallow, multiple cracking and ductile failure gage length. These results,in addition io the measured time to (Figure 3). The cracks were thumbnail shaped and were gen-f ailure, elongation, reduction in area (RA), and uttimate tensile erafty less than 0.65 mm deep. The SCC was transgranular, l strength (UTS), were used to characterize the SCC susceptibil-and markings on the fracture surf ace indicated that the cracks had propagated discontinuously by a series of small crack ity of the specimens. jumps. The crack initiation site was always evident because of Results these crack front marl (ings and the thumbnail nature of the The results of all of the SSRTs performed are given in

• Registered trade narna.

CORROSION-NACE

/ TABLE 3 - Test Results in a Simulated Resin Intrusion Environment"> e Time t > Weld Specimen Elongation RA Failure UTS Metal Type (*/.) (%) (h) (MPa) Type of Failure I 72 A 17.0 76.1 365 490 Ductile in A:St 316 NG l72 A 17.8 68.7 385 496 Ductile in AISI 316 NG 1-72 B 18 6 43.3 420 564 Ductile in weld metat 172 B 20 0 53.4 431 558 Ductile in fusion line* SCCG n AISI 316 NG R-127 A 21.6 45.4 439 488 i G n weld metal R 127 A 19.0 54.1 429 481 SCC i G n weld metal R-127 B 11.6 29.0 299 422 SCC i SCCG n weld metal R-127 B 13.3 39 6 330 445 i SCCG n AISI 316 NG R 135 A 18.1 55.1 367 454 i R-135 A 22.2 72.6 429 473 SCCG in AISI 316 NG R-135 B 25.7 47.8 488 577 Ductile in fusion line* R 135 B 29.1 41.6 527 573 Ductile in weld metal

• 288 C. 7.93 MPa, pH 4.6,6 ppm 0, and 2 x 10-7 s-2 l
• Failures in the fusion line propagated through both the AISI 316 NG SS and the weld metal.

1

• Mixture of SCC and ductile failure.

h.. N k s 1

1.=

y e:. [ ~ 7 l ...y;,gpc.4 h )~ gd-l.5mm j J g.j iOOum l a b ik j N' ig,.,ei r f?;T*..A*T*f y. a,. J.

• .7h s.

y o e q U4"2 lOOum C -mm ammi FIGURE 2 - Fracture surf ace of the R 127 weld metal SSRT specimen tested in the simulated resin intrusion environment; high magnification view of (b) the SCC region and (c) ductile region. Vol. 47. No.11. November 1986 685

l ./ ,c p ? h m

m.2 7

-.+. y3- .",, i ^ 2g.n %p m

a..;

w y%:hs...~h,..'. % [W } 4g4h ]h. .P.. [Q,Q. -t e x, e hd M kha* y WQ..W. '#v Q h _:. / dQ,.y.,Q, p g3g ' ..l,p - s'x* (v T.Qq...,/J r' " y .a r . &;$,&W {[ 2 5 Cum a /.,'/.o $e & ffbN.0 h. % s.n,< ,f

;t -

+ .y 1j'p. 4 n s W M,%y.,, to Q ,um. .f ) s w?Q, e :n N.q..a ~ w. c ll % L .Q >h %% ~~r n f f ? d ': n-f _,, 7 b Y;f.: 5 f. Qf 4 ~w *~ -%W,... r' s t &- Q R s 2M I . h. h, f ( I p kl p -=% } [f, y.:,p ' 3.j g y w ..~. % i, c. 4 d.-2r-50um - = ~ = ' ' b FIGURE 4 - Transgranular stress corrosion crack l FIGURE 3 - Fracture surf ace of AISI 316 NG SS tested formed in AISI 316 NG SS during SSRTs in a simulated in a simulated resin intrusion environment: (b) high res n intrusion environment:(b) energy dispersive x ray magnification view of the TGSCC region. spectrum of the deposit. there was a sman amount of de-scsit constsimg ct angular the R-127 and AIS! 316 NG SS material became very sensitive crystathtes and extending no more than 0 05 mm on the crack to oxygen content. becoming more negative as the oxygen f racture surtace (Figure 4) It is unakely that the crystals had le.el dropped. Of tre two materials. the R-127 weld metal dis-been formed af ter f adure. as they were found only iccahy at played the ! east notte pctential The pcientials of the 1-72 and the matiation s tes Energy disperswe x-ray anatysis (EDXA) cf R-135 weld rnetais were ve y semdar betow 0 3 ppm 0;. and this depos.t mdicated a sman amount of su:f ur contammant although thetr potentials also dropped with decreasmg oxygen (Figure 4L Trns suggests that the cracks e:ther initiated at le.el. they were not as sensitive to oxygen content. At an oxy-manganese su:hde mc'us.ons m the steel or that sufur gen level of 0 2 ppm the electrode pctennals of a:1 of the test species f rcm the resm ntrus4cn erwircnment nad cc"centrated aHoys were m the range O to - 50 mV SHE. However at low cx-ar.d set un scme part.cular icca! e ectrochemical conitions ygen levels to C5 ppmt the electrode potentials of the A-135 that caused crack mit;abon Smce manganese was not de-and 172 we!d meta!s were -150 to 200 mV mcre ncble than tected in the x ray analys,s. this would imply the latter process those of the AISI 316 NG SS and R 127 materials. of m:tiahon resultmg from the a;gressaeness of the test solu-in summary, this senes of S3RTs has demonstrated that 3 f1) under the test cond.tions used. the 172 and R-135 we!d tion Figure 5 shows the results of the test to measure the e;ec-meta!s ^ere immune to crachng in both the normal BWR wa:er pironment and the resm mtrusicn enorenment: f 2i the trode potennais of the test aGoys m the sirnulated resm mtru. sicn enwrcnment cwer an or pgen range of 0.05 to 8 porn. It is R-127 f c r metai was susce:. tit;le to crachng in the resm miru. si n e'C'onm&nt. but was irnmune m the normal BWR water 1 C seen that above 0.3 p;.m 0 the potentials of the three weld 7 meta:s ar.d AISI 3t6 NG SS are au <ery simnar. encept abcve 3 e%rcoment: and (3) AiS! 316 NG SS was +cund to be irymure ppm 07 wrere the pctennal of AISI 316 NG SS is shgntly more to crachng in the normal BWR water chemistry. but was pos:tue than that ct tne werd metals The SSRTs in the simu-suscephtHe to crcchng m the resm mtrusion enwronment, ex-j lated resm intrusion environment were conducted at an ex-CP;t *h" he!ded eth I 72. n which case et was irnmune ygen le<el of 6 ppm, under these condmons, the potennats of the weld metals au teH m the range 31 to BS mV SHE (standard DISCUSSION hydrogen e:ectrodel whereas that of AISI 316 NG SS was The resuits of this work have demonstrated that with the shgnity rnore positive at 122 to 127 mV SME. Between 0.3 and excephon of R 127, the h gh. chromium weld metals (R-135 and 1.5 ppm O,,, the e.ectrode potent >ats of au cf the test aucys I-72) were higNy resistant to SCC in both pure and impure we'e very simdar and almost mcepencent cf the oxygen con. simu!ated Bi'.R envirornents under the test cond;tions used tent. However, below 0 3 ppm 0 the e!ectrode pctentials of Prenous SSRTs. howe <er have shown that both A 135 and 7 enR mnmON-N ACE e e

s 9 300 ,,,3g ,,3 g x Type 316 NG 200 A l-72 ~ O R-127 x O R-135 100 x g g - B i o 8N O S I 0 D B = { 9 4 . -100 m E t O -200 O - 300 -400 '.1 0 1 10 O.01 Dissolved Oxygen Content. ppm FIGURE 5 - Electrochemical potential of AISI 316 NG SS and the I-72, R-127, and R 135 weld metals in a simulated resin intrusion environment (1 ppm H SO at 288 C) as a function of 2 4 dissofved oxygen content. R127 were immune to SCC in'both the creviced and uncreviced with 02 ppm 02 at 289 C), but that TGSCC occurs in an conditions when welded to Alloy 690 in an identical simulated impurity-containing environment (oxygenated water with 0.2 resin intrusion environment.7 Thus, the demonstration of SCC pom O and 0.1 ppm 50/ at 289 C). The TGSCC occurred 2 susceptibility of the R 127 weld metalin the resin intrusion en-over a strain rate range of 10-8 to 10-7sd, and the average vironment in these tests was surprising. Comparing the conds-transgranular crack growth rates in AISI 316 NG SS were ap-tions for the two test series, the cause of the SCC susceptibil-proximately a factor three lower than the intergranular crack ity is not obvious. Both sets of SSRTs were conducted in ident-growth rates in AISI 316 SS. In a more aggressive environment ical environments at the same strain rate; the only difference (8 ppm O2 + 0.5 ppm Cl-), one order of magnitude dif ference was that in one case, Alloy 690 was used as the base plate ma-was observed? Ljungberg, et at.12 have also tested AISI 316 terial, and in the other case, AISI 316 NG SS was the base plate NG SS in a flowing test loop installed at the Swedish material However, since the rneasured electrode potentials of Ringhals-1 BWR and found from SSRTs that the material was AISI 316 NG SS and Alloy 690 were very similar in this environ-immune to SCC in a normal BWR environment at a strain rate ment (122 and 154 rnV SHE, respectively),' It is improbable that of 5 x 10-a gu. However, Clarke" has shown more recently galvanic effects were the cause of the variable SCC suscepti-that AISI 316 NG SS is susceptible to TGSCC even in pure bility. Also, the electrode potential of the R 127 weld metal water environments. Clarke believes that the TGSCC is a was identical to that of the other two weld metals at the high phenomenon associated with SSRTs, and since it has never oxygen level of the SSRT, and the other weld metals were im-been observed in service,it is not presently considered to be of mune to cracking. Also, there was no possible effect of heat-engineering significance. to-heat variability, since in this work Ge R 135 welding elec-The present work has shown that AIS1316 NG SS was not troce was made from the same heat of material as the R 127 susceptible to TGSCC in the simulated resin intrusion environ-filler metal. ment when welded with the 1-72 weld metat. This effect is .l The strong SCC resistance displayed by the 1-72 weld unlikely to result from test-to-test variability since duplicate metal is consistent with the results of previous U-bend tests tests demonstrated the same ef fect. The I-72 weld metal, performed on this material,8 which demonstrated a strong ef. which had a much higher chromium content than R-127 and fect of chromium content on the SCC behavior of weld metals. R 135 (42.8 compared to 29.3 wt%), appears to reduce the SCC in those tests, welds containing less than 24% chromium susceptibility of AISI 316 NG SS (16.9 wt% Cr), which tends to cracked easily, while welds containing more than 24% Cr were support the strong effect of chromium content on SCC resist-significantly more resistant. As can be seen from Table 1,I 72 ance.8 Thus, it would appear that the galvanic or metallurgical had the highest chromium content (42.8%) of the three weld conditions (or both) formed by the AISI 316 NG/l 72 couple rnetats tested; therefore, it would be expected to be highly reduce SCC susceptibility. However, the electrode potentials SCC resistant in these environments. of all of the weld onetals in the test environment (oxygenated The results of this program have also shown that AISI 316 resin intrusion environment,6 ppm O and 1 ppm H SO,) were 2 2 NG SS is immune to SCC in a simulated pure BWR environ-identical (see Figure 5); therefore, this implies that it is the ment, out susceptible to TGSCC in an impure BWR environ. metallurgical conditions established at the AISI 316 NGil-72 rnent. These results are consistent with previously published couple that act to reduce the SCC susceptibility of AIS1316 NG data.532 Maiya and Shacks have shown from SSRTs that no 8 SS. Although Nelson has postulated that local partitioning of SCC occurs in a high purity environment (oxygenated water elements in collaboration with chromium depletion can affect Vol. 42. No.11. November 1986 687

-~ l the SCC susceptibility of these nickel-base weldments, de-decrease in potential below 0.3 ppm 0, and its potential was 2 -100 mV at 0.1 ppm 0. Although the reasons for the slightly tailed microstructural examinations are clearly required to ac-2 count properly for the differences in SCC susceptibility. more positive potentials for AISI 316 NG SS are believed to be Wcth respect to theimmunity of the AISI 316 NG SS when the same as those previously described for the weld metals,it welded withl-72 and the unexpected SCC susceptibihty of the should be noted that very positive potentials have been meas. t2 ured in actual BWR reactor environments. Ljungberg meas-B 127 weld metal,it is possible that the observed behavior may ured potentials of + 110 to ~+ 190 mV SHE for AISI 316 NG SS result from strain rate effects. The stram rate m a composite weld metal-base metal spectmen can vary as a result of dif. in the Ringhals 1 BWR reactor environment containing 0.27 to ferences in the mechanical properties of the two materials, 0.405 ppm 0 - 2 which can cause the strain to localize. Since SCC is a strain From the practical standpoint, the,results imply that any of the weid metals investigated could be,used to weld AtSI 316 L rate-sensitive phenomenon, then apparent immunity or sus- . NG SS in service if the purity of the BWB reactor environment L ceptib41 sty in these composste welded specimens could result could be guaranteed. However,in environments characteristic f rom strain rate ef f ects, and testing at slower or f aster nomina of resin intrusions that can be encountered in service. SCC im-strain rates may produce opposite behavior. Very little data are available that show the ref ationship be-munity for the complete weidment was provided only by 172, tween the electrochemical potentials of nickel-base weld the high-chromium weld metal.Thus, based on these results, [ [ metals and oxygen content in impure (S0,2-) simulated BWR the 172 filler metal of f ers the best alternative to presently used filler metals when welding AISI 316 NG SS for BWR service. l environments at 288 C. However, the data generated in this However, whether I 72 is supenor to AIS1308L weld metal, or a I program can be compared to data generated for Alloy 600 in high-purity water. Staehle has demonstrated a sigmoidal matchmg AISI 316 weld metat, has yet to be determined. d relationship between potential and oxygen, similar to that Conclusions ( which has been well documented for AtSI 304 SS." Above The following conclusions can be drawn from the results 0.2 ppm O. the potentialis not strongly influenced by oxygen 2 content, and it increases linearly from 0 mV SHE at 0.2 ppm to of this investigation:

1. In oxygenated pure water at 288 C, the AISl 316 NG SS, i

200 mv SHE at 8 ppm of dissolved oxygen. Below 0.2 ppm 0, l 72 R-127, and R-135 weld metals are allimmune to SCC at a 2 the electrode potential of Alloy 600 is mucn more strongly in. strain rate of 2 x 10-7 s-'. fluenced by oxygen content, dropping to - 200 mV at 0.1 ppm

2. In an oxygenated simulated resin intrusion environ-j O and -250 mV at 0.05 ppm 07 ment (6 ppm O and 1 ppm H SO ) at 288 C, AISI 316 NG SS ex-2 The potentials measured for the nickel-base weld metats hibited TGSCC except when welded with the 172 filler metal.

g 2 4 are in very good agreement with the data of Staehle above 0.3 The R 127 filler metal also exhibited IGSCC in this environ-ppm, However. although the we!d metals did exhibit increased ment, whereas the 172 and R-135 weld metals were f ound to be sensitivity to the oxygen content below 0.3 ppm (with the ex. immune. ception of R 127), the potentials did not drop as sharply as the

3. The immunity of AISI 316 NG SS in the impure environ-data of Staente would precict. However, there are two possible ment when welded with I-72 probably did not result from any t

I reasons why potentials measured in this program are - 150 to galvanic ef fect, since the electrode potentials of all of the weld 200 mV more noble than those measured by Staehle. First. metals were identical in the test environment. Staehle found that steady state potentials were not reached

4. Local enrichment of sulfur species from the environ-for 24 h in pure water. Although this tirne period is consider.

ment appeared to be associated with the initiation of cracks in t ably less in impure, acidifed water,"it is possib!e that in the AISI 316 NG SS. present experiment, steady-state potentials were not properly j established at the lower and intermediate oxygen levels, since Acknowledgments l the oxygen level was reduced by constant deaeration. At the This work was supported by the Electric Power Research i lowest oxygen levels. the potentiats were allowed to institute under Contract No. T305-6. The author thanks J. L equilibrate for 2 to 3 h, wnereas Tay!or" has shown that it can Netson for many valuable discussions. take up to 5 h to reach steady' state potentials in an acidified i (H;SO,) water environment containing 0.1 M Na2SO. Second. 4 the electrode potential of Alloy 600 responds to pH changes References according to the Nernst equation, with a decrease in pH result-

1. R. W. Weeks, Proc. Int. Symp. Environmental Degradation

[ ing in the potential becoming more positive. Since Staehle's of Materials in Nuclear Power Systems-Water Reactors, data were for pure water and the present expenment was con-National Association of Corrosion Engineers, Houston, ducted in acidified 11 ppm SO,2-) pure water with a room Texas.p. 69,1984. temperature pH value of 4 6 to 4.8. then the more noble poten-

2. J. C. Danko, Proc. Int. Symp. Environmental Degradation tials measured in the present experiment would be expected to of Materials in Nuclear Power Systems-Water Reactors, result solely from the pH eff ect. Taylor and Silvermanta have Nat onal Association of Corrosion Engineers, Houston, shown that decreasing the pH value from 7.2 to 4.6 in a 0.1 M Texas,p. 209,1984.

Na SO, acueous environment at 288 C can increase the poten-

3. W. J. Shack, et al.," Environmentally Assisted Cracking in z

tial of Alloy 600 by 180 and 140 mV for oxygen contents of 0.02 Light Water Reactors " ANL Report NUREG/CR.3292, and 8 ppm. respectively. October 1981-September 1982 Annual Report, Argonne The potential-oxygen relationship measured for AISI 316 National Laboratory, Argonne, Illinois,1983. NG SS can be compared to those that have been generated for

4. J. E. Alexander, et al.," Alternative A!!oys for BWR Piping l

AISI 304 SS in pure water.*5 Again, AISI 304 SS exhibits a Applications," Final Report NP.2671-LD, General Electric sigmoidal relationship, but the steep position of the curve, Co., Schenectady, New York, October 1932. where the potential changes rapidly for slight changes in oxy-

5. P. S. Maiya. W. J. Shack, Corrosion, Vol 41, No.11, p. 630, 1965.

gen content, has been found to occur over different oxygen level ranges by different investigators, which indicates that a

6. H. C. Burghard, A. J. Burste, Final Report, SwRI Project f

certain amount of variability exists in these expenments. In-02 5839-001, Sputhwest Research Institute, San Antonso, dig 6 found that the steep position of the curve occurs below Texas, December 1978. O.05 ppm O and that the potential of AtSI 304 is between

7. R. A. Page. A. McMinn Final Report, SwRI Project 06-

-100 and -200 mV at 0.1 pdm O. Staehle,'S on the other 5721-001, Southwest Research institute, San Antonio, 2 2 hand, measured a sharp decrease in potential below 1 ppm O Texas. December 1985. I and a potential of - 380 mV at 0.1 ppm O. In comparison, the

8. A. McMinn, R. A. Page, " Stress Corrosion Cracking of I

2 l 2 inconel Weidments in a Simuf ated BWR Environment," AISI 316 NG SS tested in this program exhibited a significant j CORROSION-NACE RAR

Proc. 2nd int. Symp. Environmental Degradation of Mater-

12. L G. Ljungberg, D. Cubicciotti, M. Tr olle, CORROSIOW85, ials in Nuclear Power Systems-Water Reactors, National Paper No.100, National Association of Corrosion Engi-Association of Corrosion Engineers, Houston, Texas, neers, Houston, Texas,1985.

13. P. S. Maiya, W. J. Shack, NUREG-CR 3689 Vol. 4, ANL,

p.108 1966 e i {

9. J. L. Nelson, S. Floreen, "An Evaluation of the SCC Be-85 Vol. 4. Quarterly Progress Report, Argonne National havior of inconel Alloy 690 Weldments in a Simulated Laboratory, Argonne, Illinois. October December 1983.

I BWR Environment," Proc. 2nd Int. Symp. Environmental

14. W. t Clarke, Private Communicatiori, General Electric f

Degradation of Materials in Nuclear Power Systems-Co., Pleasanton California,1986. Water Reactors, National Association of Corrosion Engi-

15. A. W. Staehle, et al., EPRI Project RP 31'.1, Final Summary neers. Heuston, Texas. p. 4.1956.

Report, Electric Power Research insitute, Palo Alto, California January 1975 to December 1977.

10. F. F. Lyle Jr., E. B. Norris, Stress Corrosion Cracking-

16. M. E. Indig. A. R. McIlree, Corrosion, Vol. 35, No. 7, p. 288, i

The Slow Stram Rate Technioue, ASTM STP 655, A. M. 1979. l Ugiansky, J. H. Payer Eds., ASTM, Philadelphia, Pennsyl-

17. D. F. Taylor, C. A. Caramihas, J. Electrochem. Soc., Vol.

vania, p. 388,1979. 129, No.11, p. 2458,1982.

11. I. J. Magar. P. E. Morns. Corrosion, Vol. J2, No. 9. p. 374,

18. D. F. Taylor, M. Silverman, Corrosion, Vo. 36. No.10, p.

1976. 544,1980. r l Probing Microbiologically Induced Corrosion

• S. M. GERCHAKOV,' B. J. LITTLE ' and P. WAGNER' Abstract The apparatus described in this paper, consisting of a two-compartment cell in which the compartments are biologically isolated and electrolytically continuous, is ideally suited for the study of microbiologically induced corrosion (MIC). The electrochemicalimpact of three bacterial isolates on three metal substrata has been Quantified using this apparatus, in addi-tion, three proposed mechamsms for MIC have been evaluated.

Introducilon the dual compartment system, these species were evaluated Failures in tanks and piping systems made of steel, copper, on metal electrodes in electrolytes similar to those from which aluminum, and nickel alloys exposed to aqueous environ' they were isolated. ments have been attributed to microbiologically induced cor-r rosion (MIC)." In all instances, microbial slimes or deposits Materials and Methods have been associated with the failures, cnd in some cases, specific microorganisms have been isolated. However,in the The Measurement Cell absence of analytical techniques to venty and quantify the The measurement cell is described schematically in electrochemical impact of the microorganisms, it is impossi. Figure 1, and the letters used in the text refer to the figuce. ble to establish direct cause and effect relationships. The cell comprises two halves, which are mirror images of This paper describes a system that can be used to eval-each other, separated by a 0.1-pm cellulose acetate /ceflutone uate the electrochemicalimpact of microbiological species un nitrate membrane. The membrane (C) is secured between tr e metal electrodes, it consists of two compartments (electro. two half cells (A) with a clamp that also holds the two half-lytically continuous but biologically isolated) externally con-cells together.The medium inlet (D) is at the bottom of the ce I nected to a zero resistance ammeter (ZHA). When either of the to ensure proper mixing, which is also aided by the gas flow. e compartments is perturbed by microorganisms, anodic and The position of the medium outlet (E) determines the liquid cathodic currents are established. An anodic current indicates level maintained in the cell. The cell cover (B), secured to the that the microorganisms are responsible for oxidation reac. cell body with a clamp, holds the cell's accessories.The gas l tions. A cathodic current indicates reduction.The extent of the inlet (K) is situated in such a way as to mix the contents in the microbial mediation is reflected in the magnitude of the chamber adjacent to the membrane. The cover also holds the observed current. metal specimen holder (H), Luggin capillary (1), Pt electrode (G), The microorganisms discussed in this paper were iso. lated f rom corroding copper, nickel, and mild steel surf aces, in gas outlet (J), and a, utility port (F). The utility port is useful for j 1 rnedium sampling or inoculation with bacteria. Inlets and out-lets (llouid or gas) are appropriately protected from bacteria ' Submitted for publication August 1985; revised April 1986. migration.

• Deceased; formerly with Dept, of Microbiology and Immu-nology, University of Miami, Miami, Florida 33101.

" Naval Ocean Research and Development Activity, NSTL, Bacteria Mississippt 39529-5004. The etiect of a marine pseudomonad, an obligate thermo- ) 0010 9312/86/000205/$3.00/0 m m urs 11. November 1M6 wn,.w twation of Corro+ don Fnomeets 689

J' ./ i. SCC-Resistant Welding Alloy for BWRs Nickel-base and stainless steel weld metals are commonly used in BWRs to join safe-ends to low-alloy Figure 2 steel reactor pressure vessel nozzles 12-c0 o and to carbon steel pipes, pumps, and valves. Nickel-base weld met-o _ so-als are also used for many reactor 31 internal applications, including y vessel-attachment welds. Intergran-g ,o_ ular stress corrosion cracking a QGSCC)has occurred in nickel-base alloy 182 weld metalin several op-o erating BWRs. Laboratory and j 2c-plant crack growth data show that SCC in alloy 182 can occur under o normal water chemistry conditions ,o 2o "a'o at a rate far exceeding that of sensi-e is tized Type 304 stainless steel cr concentration (wt %) 309L/3$Lst Effect of chromium concentration on stress corrosion cracking s steel eld t-als has been used when safe-end nicke!-base weldmg alloys. replacement has been an option. Al-though the IGSCC behavior of BWR service. Crack growth rates normal BWR water (200 parts per these low-carbon austenitic stain-have been measured in the labora-billion 02), alloys 182 and 308L ex-less steels has been excellent to tory using fracture-mechanics type date, results from laboratory tests specimens in normal and highly hibit measurable crack growth at a raise concern regarding their be-faulted BWR environments. Weld stress intensity of approximately 40 havior in off-chemistry conditions. metals 182 and 308L were also ksi Vin., whereas alloy 72 did not This concern, together with the de-tested and used as a basis of com-exhibit any measurable crack sire forimproved mechanical com-Parison. The results indicate that in grmvth until a stress intensity of at patibility with carbon-or ahoy-steel (continuedon pageII) base metals, has led to a search for an ICSCC-resistant, nickel-base weld metal for BWR applications. Figure 2 Several years ago, laboratory ~ investigations sponsored by EPRI Weld Metal Crack Growth Data revealed that in normal and off-crack gm.e m rae Faufted Chemistry normal BWR environments,IGSCC 33 g.7 %*C susceptibility of nickel-base weld-m ing alloys was related to the alloy's 3.00E-8 U chromium content (see Figure 1). a These studies clearly showed that o io se G c for nickel-base welding alloys, ag o g 0 chromium levels greater than ap-3 cog., 00 oo o proximately 27% of weight were re-t.ocE-9 'O g qtured for maxunum resistance to crack initiation. Welding alloys sez aut 72 3 ME-to. O o 6,i such as alloy 182 and 82, with typ' i-cal chronnum levels of 18-20% by ~ ,,xg,33 weight, were far more susceptible o 2o e so so to IGSCC than weld alloy 72 with stress intensity (ksiG) 40-43% by weight chromium. Re-cent EPRI studies have focused on Effect ofstress m. iensity on crackgrowth ratefor severalweld alloys m BM enmronment. qualifying welding alloy 72 for 10 Nuclear Notes. First issue 1993 - ~~

SCC-Resistant Alloy ALWR (continued from page 10) (continued from page 9) least 50 ksi Vin. was reached. The gas tungsten arc welding (GTAW) containment would create a natural results from similar tests carded out and submerged metal arc welding in a highly faulted BWR em' iron-use, respectively. Subcommittees evaporative cooling film on the containment's outer steel surface. ment are shown in Figure 2. The su-have approved these welding al-perior resistance of alloy 72 as com-loys, but they are still in the AWS However, analyses prove that the external evaporative film is not ab-pared to alloys 182 and 308L is approval chain. It may be another solutelv necessary. Tests demon-dearly demonstrated in this high-year before AWS final approval is strate tilat with the absence of any sulfate, high-oxygen environment. obtained. Then acceptance by operator action, the AP600's natural While high-chromium welding ASME is basically automatic, al-alloys 52,152, and 72 have been though further time is required. safety systems would safely re-shown to be superior to alloys 82 The second approach has been move the heat and protect the pub-lic and the reactor. Tests and analy-and 182 in IGSCC resistance, none an ASME Code case submitted by ses have demonstrated no series of of these alloys are currently ap-Westinghouse fordirect acceptance proved for use under Section II of on behalf of South Carolina Electdc postulated circumstances that would breech AP600 contamment the Amedcan Society of Mechanical & Gas Co., which wishes to use al-Engineers (ASME) Code. Two ap-loys 52 and 152 in the manufacture integrity. proaches have been taken to obtain of replacement steam generators The final step in safety testing approval. The first approach in-with alloy 690 tubing. The code will be a series of integrated tests volved submission to the appropd-case has been approved and fabd-to examine the behavior of all pas-ate American Welding Society cation of the steam generators may sive safety systems during a wide range of postulated accidents. One-(AWS) committees for approval. start in February 1993. fourth-scale tests will be conducted This was initiated two years ago in Weldability studies have shown a request for indusion of alloy 72 that all three of these alloys are ac-at Oregon State University, fol-in AWS's A5.14 Subcommittee. ceptable. Increased relative cost of lowed by full-height, full-pressure When International Nickel Com-alloy 72 makes alloy 52 a mor tests at the SPES facility in Italy. The panyInc. decided to make alloys 52 ufole choice for uAW use. SPES tests are part of a cooperative and 152 available as commercial For additional information, contact

1. fort with the NRC and represent an ther major milestone for the products, the request was broad-hplie On7ds, (415) 855-2058, or L2rry ened to indude these two allovs for, Nelson,(415) 855-2825.

AP6p0 program. 4 l With the success of the SPES k te'sts, Westinghouse will have the W Gr NRC's agreement that passive eMgEq: logy @fo& %@-CLTOMMMQt safety systems will meet fulllicens-eTennmo rMana glgmgw& plants jq r qmckreference by e$g&M e ing requirements independently of %@$h f4 YAs 2nostqoperadng $ iff GMAWE expensive backup systems. ngmeers approEch the" midpoint of theid and plant personnel Use of the. With completion of the full- ~ height, full-pressure tests at SPES 'currenElideh55d term /fiotentiaria terminofogy'is recommended by ^ ~inisun~derstihding ~of ?aginfMEPRIyucleifMandg~eme~nnnd-in 1993, Westinghouse is confident Tdegiddatioir'in:systefidfstruc-GRes6uihis:Coun'cillind.th'e'NRC - the AP600 will meet the utilities' goal of having a licensed, standard- - tures,"iiid Tosp6nentsYoccugyto fifiprosre the~imdFrstifidif bf~ $ing-'froiiEriproperuseEfthe tei-dQging pHEiiiIehi7facilitR@the3 ized, mid-size passive 1.WR avail-able by the mid-1990s. &minorofdrkscribin siEihagment-Xfin'g' man &ginD"reportmjQE~reTefift}l5EtTail$ For further information, contact ^ agem umdEta;anicianfy'tIieEdgerpreR Chuck Welty,(415) 812-2322. 5fs rkey'eleEIentof udinie ~ 7 ~ ia'tioiE6f15IEdards del reguld4 Tidifift5fdfe5Ea i @plhiiElifessFrenekar;.' ima6EgEnEEEirid eip'ufmen teffirfaEciTe~filiitioiE}$1ng1TIiel aregEner2f ksalifi6TtfoE?!EPRI desteropdif allyEidiiIt2EEGitEliifdlar'fsi9 3 and has piiblishildefinis'6Es fo~ EFn'undI6giriised bfthnIn's'tififtEof7 1 J oveiihuridiffteimsTflated tigNucleir'PoWiiOpeiEtioiiffii~thef i OaghigEi~fisil"heporE ~ NucleifElfiit'Reliabili ata' 1 $ock@et-sife gIffsEy', BR-IO1PowerHaniC5?dGidn A&~ Syi~tQQ"rmahon$o$ W $ \\ T ? holo TIq00844,*anc _ For,nore. nticU GeEIg75 fills (475)SI2-2BI2^$$ % M P,r*f;;U w Gt%4e WW % Q.%.m. ~. ~ w-M-3m

• Nuc' ear Notes, First issue 1993 11

4 41. H. Kaesche. Metanic Corrosion (English translation of Die

45. L Rathe, W. Gruht, Werkstoffe und Korrosion, Vol. 31, p. 768.

' Korrusion der Metaffe), 2nd ed., transtated by R. A. Rapp.

1980, t

National Association of Corrosion Engineers,1985. 46. J. C. M. U, R. A. Oriani. L S. Darken. Z. Phys. Chem. (NF), Vol. I 42.' E. Brauns. H. Temes, Werkstoffe und Korrosion, Vol.19, p.1, 49,p.271,1966. 1968. 47. H. Vosskuhler, Werkstoffe und Korrosion, Vol.1, p. 357,1950. ( 43. H. L Logan, J. Research Nat. Bur. Stand.. Vo!. 61, p. 503. 48. N. Ruym K. Baardseth. J. Inst. Met., Vol. 96, p. 92,1968. f 1958. 49. R. N. T. Unwin, G. C. Smith, J. Inst. Met., Vol. 97, p. 299,1969. 44. J. A. Feeny, M. J. Blackbum, Theory of Stress Corrosion j Crack ng in Anoys, J. C. Scup, Ed., Brussels. Belgium, p. 355. j y 1971. P t k t Stress Corrosion Cracking of Inconel Alloys and Weidments in High-Temperature Water - The Effect of q Sulfuric Acid Addition

• q A McMINN~ and R. A PAGE" Abstract i

The stress corrosion cracking (SCC) susceptibilities of Alloys 600 and 690. A!Sf" 316 NG stainless steel (SS), ASTM

• A508 carbon steel, and a number of compatible weld metals have been evaluated at 288 C in pure water and in pure water containing sulfunc acid additions. The sulfuric acid was added to simulate the effects of a resin release from the demineralizer system of a boiling water reactor (BWR).

A combination of creviced and noncreviced slow strain rate, constant load, and crack growth rate tests -i were used in the evaluation. The results indicated that a!! of the alloys tested in the uncreviced condition ] were immune to cracking in the pure water environment. The presence of crevices in the pure water l environment produced a susceptibdity to SCC in Alloy 600,in inconel 1-82 and 1-182 weld metals, and ASTM A508 steel, but not in Alloy 690. Cracking was enhanced by the addition of 1 ppm H,SO. in slow strain rate tests (SSRTs) and constant load tests, but crack growth rates were not enhanced. AH of the j alloys tested in the resin intrusion environment were susceptible to cracking, except for the high chromium weld metals R_-135 and inconel 1-72. i 4 Introduction no deliberate additions are made to the water, water chemistry

{

Over the past years, there have been a couple of incidents of transients, brought about by resin releases frorn the dominera52er i in-service cracking of boiling water reactor (BWR) inconel Alloy 600 system or by cooling-water leaks through conderiser tubing, do i safe-end forgings. The occurrence of intergranular stress corrosion occasionally introduce impuritier. into the pnmary coolant system. j cracking (IGSCC) in Alloy 600 and the adjacent inconel 1-182 wetd Demineralizer resins rapidly decompose at normal operating tem-rnetal has raised senous questions abotit the long-term suitability of peratures, yielding su! fate ions from the decomposition of the cationic these materials for use in BWR environments. As a result of these resin and tnmethylamine and ar9monia from the anionic resin." Since l questions, a number of experimental studies have been performed to the alkaline ammonia and trimethylamine are volatile at BWR i provide further information conceming the susceptibility of Alloy 600 operating temperatures, the net result of a mixed resin intrusion is an j and its proposed replacement, inconel Alloy 690, under BWR increase in the suffate ion concentration and conductrvity, and a operating conditions.b8 Included in these tests were a number of decrease in the pH in the primary coolant system. On the other hand, weld metals that are compatible with A!!oys 600 and 690, cooling water leaks generally lead to chloride ion intrusions. For a The typical primary coolant in a BWR is neutral pH, high punty resin intrusion, the changes in conductivity and pH are related to the water containing an exygen concentration of -200 ppb, which is rate and duration of the resin retease, the rate of resin decomposition, present as a result of radiolysis of water in the reactor core. Although and the rate of water cleanup. Resin intrusions, yielding room temperature water conductivities as high as 55 pS/cm, have been observed in service.s tRegistered trade name. " Submitted for publication December 1986; revised May 1987. Of the two common cooling water upsets, resin intrusions are

• Failure Analysis Associates,1100 S. Washington St., Alexandna, thought to affect the IGSCC behavior of nickel-base alloys, such as All y 600, far more adversety than upsets resulting from cooling Virginia 22314 wa a g ace ge atas sdate aMWo

" Southwest Research Institute, P. O. Box 28510,6220 Culebra Rd., pure water have on IGSCC has been demonstrated for both Alloy San Antonio, Texas 78284 600' and a number of weld metals that are compatible with Alloys 600 ("Amencan iron and Steel Institute (AIS!), Washington, DC. and 690.7The purpose of this paper is to examine more closety the

• American Society for Testing and Materials (ASTM), Philadelphia.

' effect of a water chemistry excursion, brought about by a resin Pennsylvania. Intrusion, on IGSCC of a number of NiFeCr and steel-base snetals

~ l 4 and weld metals using test results obtained from a combination of conductivity of the inf vent water ranged from 0.06710 0.077 Sicm. slow strain rate, constant load, and crack growth rate tests. The conducevity of the effluent was -0.2 gS/cm during the uncre-viced SSRTs and -1 pS/cm dunng the creviced tests, influent and i Experimental emmt oxygen levels differed by no more than 20% wrth the effluent l l level always slightly less than the influent level. Material As noted in the introduction, the primary effect of a resin Four base metals. Alloys 600,690. A!SI 316 NG stainless steel s n s an ycr ase in me sudatgdon and coMuM_ (SS). and ASTM A508 carbon steel, were incluced in the test matnx. and a decJege in the pH of the pnmary coohng water 7Thus, the e nment chosen to simdate a resin Msbn b Ws wM The composttrons of the four base metais are listed in Table 1. Welcments made with commercially available inconel I-82,1-182, onsisted of high punty water with a deliberate addition of 1 ppm of_ 4 5 sutfuric acid. The addition of 1 ppm of suirunc acic resulted in a 1 ppm and I 72 and with expenmental alloys R-t27 and R-135, were also sulfate concentration, a conductivity of ~8 nS/cm, and a pH of ~4.8 included in the matnx. The nominal chromium content of these weld -At73m T6Y5dmTJrE" These values are all weii within the r$ 3 ~ metals ranged from 15 to 43% Actual wetd chemistries are listed in ygg Table 2. The welding and heat treatment procedures and the simulated resin intrusion environment were run with a dissolved I correspondin microstructures have been fully descnbed elsewhere. For the sake of brevity, they will not be repeated oxygen level of ~7 ppm to accelerate cracking, f Test Matrix TABLE 1 - Chemical Composition of Test Alloys The test matnx consisted of a senes of slow strain rate, crack [ growth, and sustained load tests. The SSRTs were performed in a compo.moa twry multispecimen test facility incorporated in an autoc! ave, in which six Ei.m.ni Ahoy 600 Alicy 69o Alst Ji6 fvG ASTM A50s specimens were tested in parallel. This test facility has been fully desenbed elsewhere." Each of the specimens was galvanically c o.oe c.012402 c.ois 0.22 isolated from the autoclave, loading frame, load train, and other Q$ specimens using zirconia and Teflon' inserts. Cylindncal, one-piece j c Fe e.so a.73-e.sa tw tai tensile specimens with cross-sectional diameters of 6.35 mm and "M gage lengths of 25.4 rnm were used. For each specimen, the applied J c. D.is o.11412 027 load and the potential vs an Ag/AgCl reference electrode of a design As o 43 c2242s o oc1 similar to that described by Magar and Moms" were monitored and ~ h 2 recorded by a digital microcomputer. The crosshead displacement I s c.oc7 0.00s 4.005 c.oos o.013 (related to the total strain experienced by each specimen) also was h j [ $a measured via a displacement transducer. Displacement was con- [ v c.cas o.c21 trolled by a screw-driven machine. A strain rate of 2 x 10" s~' was i used for all of the SSRTs. a rn. composson rance msmo r oresems ine e-wnum no mamm conneo sor Crevices on the SSRT specimens consisted of an inner layer of the inme hesis of Anoy eso useo m ine anysm. These composmonas vananons d d graphite cloth and an outer layer of nickel toil. Creviced specimens noi r.saii m any ascemos emance e sc0 bensar-were exposed for 168 h in the autociave prior to the initiation of 3 straining. This provided adequate time for the crevice chemistry to TABLE 2 - Chemacal Composit6on of As-Deposited Weld Metals develop,ir Straining of uncreviced specimens was initiated as soon as the autoctave temperature and oxygen level stabilized, con =' moat m. m n mc.a. oram The crack growth rate tests were conducted on 24.13-mm-thick, o.m .-- m me Tina ny 4 bolt-loaded wedge opening loading (WOL) specimens. Side grooves, c o 007 o o4 o os a c:s o on 1.21-mm deep, were used to maintain a planar crack geometry and i y y y y y to increase the degree of constraint present at the surface and thus ci F. on to 2s to ao to or o os minimize the amount of crack tunneling. The WOL specimens were 7 o "3 fatigue precracked in accordance with ASTM Standard E-399 and L c. e ct 0 07 o oi c o4 o ot were bolt loaded to the desired stress intensity prior to exposure in the autociave. Specimen potentials vs an Ag/AgCI reference elec-co n os a re o ci o nt trode and a platinum wire were monitored throughout the test. The l 8 ' "5 8 "' crack growth specimens were removed and optically examined for crack growth once every 1000 to 1500 h. At the end of each test, the "*rwe composamn or wxv. nr r wes.ma en aswow.o composaun final load and crack length were determined for each specimen in order to calculate the final stress intensity. } Environment The sustained load tests were performed by stressing 3.18-mm-7 To simulate BWR envronments, the stress correston testing diameter, smooth tensile specimens in individual loading blocks. l was conducted in a recrculating pressurized water test loop. A Load was applied by tightening nuts on the specimen ends to schematic of the recirculating test loop can be found in Reference 1. produce a precetermined strain corresponding to the desired stress All heated metallic portions of the system were constructed of level. Loads of 1.25 and 1.5 times the 288 C yield stress were used, t titanium to permit close control of the dissolved oxygen content of the Special Belleville-type washers made of high-strength Alloy X-750 1 4-water. For all of the tests reported in this paper, the recirculating test were used to minmze load relaxation dunng testing. Each of the loop was operated at a flow rate of 11.4 L/h which resulted in specimen leading block assemblies was galvanically Isolated using i replacement of the fluid in each autoclave apprcximately once each ceramic inserts. Specimen potentials vs an intemal Ag/AgCl refer-j

hour, ence electrode and a platinum wire were monitored throughout the Tests were conducted at 288 C and a pressure of 7.34 MPa in j

test a high punty water environment and a su!! uric acid-dosed, high purity water environment to simulate " normal" primary coolant chemistry Results l and a resin intrusion environment, respectivety, in the vadosed high 4 g punty environment, creviced slow strain rate test (SSRT) specimens The results of all of the SSRTs performed are summarized in were run in water containing 16 pom of dissolved oxygen t d accelerate cracking. UM SSRT specimens were run with erther 8 ppm pf dissolve _d crygen to accelerate cracking or with 200 any cracking. Complete data from the SSRTs, which include material ppo or dissorved oxygen to simulate actual BWR leverif~TrE ' Registered trade name, j

TABLE 3 - Slow Strain Rate Test Results Surface Condition: Pure water Environment Ream intrusion Environmens Orygon Concentrauon: uneww c,e,.cw unewee c cea Potentiat* 200 poo o, a ppm o, is ppm o, 7 psen o, 7 ppm o, umries (-o_2e se -6.37 v ) (-e.os to -e.ie v,, ) g-e.c2 i -o.is v,,,) (c.1s is o.os v.) to.is se o v, ) s e Amoy 600 no SCC nosCC surface cracks surrace anca.s surface cracks Asoy 690 se SCC no SCC nosCC estaca cracas surface cracts AAoy 600rincanei 0a2 no SCC no SCC sursace cracas SCC or kcones ba2 surface es,m Aaoy 600/incones > 182 no SCC SCCs.ortace cracas SCC or mconer etE2 SCC of incones &ia2 AAoy 690nncunes M2 no SCC no SCC Amoy 6904ncones 0182 no SCC no SCC SCC of incones &is2 mcorus ktS2 sCOsune e crack.a SCC SCC mcones M2 surface cracks sunace cracas surtace cract.s Amoy 6fs0/R-i27 nosCC no SCC Amoy 690rR 135 no SCC no SCC Amoy 600/b182rAsTM A506 TG surface'* TGsCC of ASTM A505 SCC or armones &tt2 SCC of mconal 6182 nack.s in ASTM A506 SCC of Amoy 600 Alloy 600tinconel M2/ ASTM A508 SCC of incomt 6 82 SCC of Aaoy 600 surface cracas en sanace cracks a Amoy 6(E & inconal 8 82 Amoy 600 & incunes M2 Aany 690/kconel 9182/ ASTM A508 SCC of incones FIS2 SCC of kuxines F182 l Amoy &WQanconel1-82/ ASTM A506 auttaca aacis a swiace cracks ans i sconssi M2 SCC *' en incunet M2 j AISI 316 NGMcones 072 no SCC

• no SCC AdSi 316 NG4b127 no SCC
• SCC of R-127 j

+ TGsCC of Assi 316 NG l AISI 316 NC.'R.t3$ no SCC"' TGsCC of Als! 314 NG i "' stress corrossun cracksng was energraradar unless carerwene nr.aaed. ~

1. % of featured soecamen ponenhast "One out of leur sommmens.

d ]

• • 'Poseread. a.C3 in 0 Da v.

I

• • ore cui of esgris specament.
• too opo o,.

I c b) < ~ .d*k' 4 g M.,,,," 7 .r .:::.7f 4.- 4 w .~ a~ M* ne%wfi:rr i y-;.. - ~ .~ ~ x, 11 ~ kD.. ~7. ' I Y ~~ .. w 4 k~f@s kT .k ) i u% h. 5 o y . 4-tm u---- 1 i 200um 2 2mm L . y l a b FIGURE 1 - (a) Fracture surface of creviced inconel1-182 weld metal specimm slow strain rete I tested in high purity water at 288 C;(b) high magnification view of IGSCC region. 1 l heat treatment, time to failure, percent elonganon and reduction of 0.2 and 8 ppm. However, the presence of a crevice and a higher I area, and ultimate tensile strength have been presented oxygen concentration (16 ppm) did result in SCC in the pure water i elsewhere.s.2 si in Table 3. materiais that have surface cracking environment. The matenal that exhibited the greatest SCC suscep-indicated are considered to have exhibited a borcertine susceptibility tibility was the Inconel L182 weld metal, with six of the eleven to SCC, Le., one that falls between immunity and gross SCC. Sur' ace spectmens tested exhibiting stress corrosion cracks or shallow l cracks were always less than 100- m deep. Spewnens that were surface cracks. The, SCC was intergranular and a typical SCC j deemed resistant to SCC failed in a totally ductile manner and no fracture surface of the inconell-182 is shown in Figure 1. TypiCalload surface cracks were observed. Specimens that were deemed sus-vs time curves for the cracking and no cracking conditions are shown ceptible to SCC exhibited large stress corrosion cracks, which were in Figure 2. Other matenals that exhibited a susceptibility to SCC in intergranular, unless stated otherwise, the creviced condition were Alloy 600 and inconel 1-82. These i The results show that all of the nickel-base alloys and weld matenals exhibited shallow intergranular surface cracks less than metals tested were immune to SCC in pure water in the uncreviced 100-pm deep (Figure 3). A!!oy 690 was the only material that was I condition. This immunity to SCC was observed over a range of immune to SCC in the pure water environment in either the creviced specimen potentials at two widely diffenng oxygen concentrations. or uncreviced cond:tions.

susceotible to SCC in the resin intrusion environment; whereas, only e 4 3000 Inconel 1-182 was susceptible in the pure water environment. These - weld metals exhibited large intergranular cracks on their fracture 2530-M CC surfaces. Similarty, Alloys 600 and 690 exhibited enhanced suscep-tibdity, as shallow surface cracks were observed in these matenals 200C tested in the resan intrusion environment. The AISI 316 NG SS, which I was immune to cracking in the pure water environment. underwent p 1500-extensive transgranular stress corrosion cracking (TGSCC) in the 3 resin intrusion environment (Agure 4). The only two matenals that 1900-were immune to cracking rn the resin intrusion environment were the high Cr weld metals. R-135, and inconel 1-72. 51 There was little effect of crevice conditions on the SCC susceptibility of the alloys tested in the resin intrusion environment. O Matenats that were res:stant to SCC in the uncreviced condition 0 100 200 300 400 500 600 700 remained resistant and those that were susceptible exhibited very Tim h similar degrees of susceptibility. In companson to creviced tests conducted in pure water, the sulfunc acid addition essentially only FIGURE 2 - Typical load vs time curves for SSRTs at 2 x affected the susceptibility of the A!!oy 690. In summary, the ef'ect of 10" s". the 1 ppm sulfunc acid addition was to increase the occurrence of IGSCC in inconel 1-182 and 1-82, to increase the amount of surface 73 e.umsw cracking in Attoys 600 and 690, and to cause TGSCC of the AISI 316 i gWig yg NG SS. M I o M l ~~ ' 3 I .b f -lh fb& Y* / k. P [\\ p 1 l M .I,? m_p sp*= or*.gn.. n, }

• • 9 l

'} -)) t 50um .~ . g.. u- .r m .y.. ~. 1 p RGURE 4 - TGSCC in AISI 316 NG SS slow strain rate i MM t.".%-M(62,; Nth.. tested in a resin intrusion environment at 288 C. j . ; y m ~, q, A Q M Q W t \\ ?' $?%

• Crack Growth Tests I

4 MNd 3. The results of the crack growth tests performed are summanzed h[ in Table 4. The crack growth rates given in Table 4 were determined ~ C by dividing the rneasured crack length by the test time. Since this '" ^ - y,,,4 g - ', p p:4g? procedure includes the time to initiate the stress corrosion crack from

• ? CW:04*:

the fatigue precrack, the actual crack growth rate will be greater than - - W.-4%W4$ @= ~ 200pm this. In this type of testing, there is also the possibility that crack M:6mppgW g70,:n,1op, onc, g,,,, no, 3,,n,,,cn,e,,nien og,in n,3 in,

,.,,g,m pd

,2.9-eG a N ef'ect of underestimating crack growth rate. However, monitoring g

h crack growth through frequent visualinsoection minimizes this effect. Welded WOL specimens were used for tests conducted in the pure water environment.The specimens had been machined so that RGURE 3 - Typical surface cracks in Alloy 600 slow the crack would propagate in the weld heat affected zone (HAZ).The strain rate tested in creviced condition in pure water and specimens exhibited no crack growth after 1500 h at a stress intensity in resin intrusion environment; (a) SEM micrograph, (b) (K) level of 33 MPad.,The K level was then increased to 49 metallographic cross section. MPad and the specimens were exposed for a further 6700 h. Crack growth occurred in the Alloy 600 welded specimen, but not the The results of the SSRTs conducted in the resin intrusion Alloy 690 welded specimens. These observations were confirmed by environment are also shown in Table 3. Specimen potentials for this breaiang open the specimens after the test. The fracture surface of environment were in the range of 0 to 0M6 Vsse. For uncreviced the welded Alloy 600 specimen is shown in Agure 5. Fractographic speomens, the effect of the sulfunc acid additdn was to lower the and metallographic examinations of this specimen revealed that the I SCC resistance of all the materials for which a cornparison could be fatigue precrack was located in the inconel 1182 weld metal rather mace. The weld metals inconel 1-182, l-B2, and R.127 were all than the HAZ. as planned. Initial IGSCC occurred in the inconel i.182 1 [

-. - -. -. ~ _ - - -... l ! i TADI.E 4 - Crack Growth Data for Specimens Tested in Pure Water and Resin intrusion Environments l pwewetw twimnmem no m emn...o t.,vtronmone 5 ress Cseca Exposwo craca Growin strese craca Expoewe Crack Grmistn entenalty m W Growth Time Rene imensrly Crowits Time Rate f Matwtet (MPs yh (mm) (n) (mm e-') (MPs V m) (mm) (h) (mm e * *) { Aaoy600 49 4 (43 41 48 4123 32 m 10" Alby 690 50 0 3048 0 l 60 0 "D*8 0 l 70 0.24 3rea 2.2 a 10-* AGoy 60Gancones 1182 33 0 1500 0 49 (31) 145 ' 6730 5 s 10 " 49 4 (46) 2f" 4129 1.7 a10"' d i l l Anoy 6tvuncones 6-82 49 4 (4s 2) 4 0'*' 4129 2.7 a 10-' l Anoy s90ancone. Ma2 23 0 1500 0 g 49 0 6700 0 l l l "has seass runsay snown m parencem Scraca growe m Amoy soo. Sceaca growe e incmer Ha2. l

• • • Crack growm n inconal i.82.

between 60 and 70 MPad. K sce for Altoy 600, inconel 1-182, and l w -g} t g,,, l-82, on the other hand, is below 49 MPaVE l '.4 The above results indicate that the resin intrusion environment l ' = enhanced cracking, since only Alloy 600 cracked in the pure water l .. ~'a environment; whereas, alt of the alloys cracked in the resin intrusion l i * -A enwronment. However, where a direct companson of growth rates ?-- could be made for cracking in the Alloy 600, the growth rate in the l y ,.g pure water environment was slightly accelerated compared to that in l the resin intrusion environment, i.e.,6 x 10-7 rnm s-' compared to t 3.2 x 10-' mm s-', respectively. L ) ~.,r.- j l . C '. n ' Sustained Load Tests l W '.p-The results obtained from the sustained load tests are grven in j Table 5. In the pure water environment, both the Alloy 600 and j l -Q*Q ** a inconel 1-182 were susceptible to cracking; whereas, Alloy 690 was .c j A. " " resistant to cracking. There were differences in the degree and type l N C.

• of attack between the Alloy 600/inconel1-182 and the Alloy 690/Incone!

l l ,.F.Y M l-182 welded specimens. In the former specimens, localized corro-i '$ ;,7 sive attack had occurred preferentally at machining marks, and l -Wy' Smm

• • • • 't 9Phio 5'eti "5 th'o"9" th' SP'oim*"5 ''*'*d 5"8'to" t

cracks up to 25-m deep propagating from the base of these grooves. Since the cracking was only associated with the grooves l FIGURE S - Fracture surface of Alloy 600!!nconel i 182 c.d not the smooth specimen surface, this was considered to be I ewdence of SCC susceptibility in the presence of a crevice. Surface j crack growth specimen. Fatigue precrack (A), IGSCC (B), l grooving and a typical surface crack are dlustrated in Figure 7. In the and fast fracture in air (C) are shown. Alloy 690/inconel I-182 welded soecimens, shallow cracks were i observed in the 1-182, but no localized attack or cracking of the Alicy at an angle of between 45 and 90 degrees to the procrack plane 690 was observed. The results indicate that a stress greater than 1.5 l because of the elongated weld grain structure (Figure 6). In-plane r, and/or time in excess of 8200 h would be required to initiate SCC q I cracking could only occur by transgranular cracking of such grains. in Alloy 690 in a pure water environment. I After less than 1 mm of growth, the crack grew into the Alloy 600, in the res'a intrusion environment, inconel 1-182 was again [ after which, it gradually retumed to a plane normal to the loading b M N e h6 l directon. Thus, although inconel 1-182 and Alloy 600 were both environment accelerated crack initiation compared to the pure water i susceptble to IGSCC, the crack growth data generated was onfy environment, since deeper cracks (100 m) were obtained in a much applicable to the Alloy 600 base metal. shorter test exposure, 3000 vs to 8200 h in the pure water The crack growth data for specimens tested in the resin environment. However, cracking was not observed in the Alloy 600 t intrusion environment are also given in Table 4. Crack growth and 690 base metals or in the R-127 and R-135 weld metals. I occurred in each of the alloys tested, and ADoy 600, Inconell 82, and i 1-182 exhibited similar growth rates, ranging from 1.7 x 10-' mm s-' for inconet 1-182 and 3.2 x 10-' mm s-' for Ancy 600. It is believed DiSCUSSIOrt j that the columnar grains, which were perpendicular to the plane of Companson of the results in the simulated resin intrusion the faDgue precrack, hindered the IGSCC crack growth in both weld environment with those obtained in pure water indicates that the j metals. Higher crack growth rates would be expected when the addition of 1 ppm H SO, generally enhanced cracking. In the 2 j columnar grains are onented parallel to the preferred crack growth uncreviced SSRTs, the SCC resistance of the majonty of alloys j g tested was lowered by the addition of sulfuric acid, as evidenced by directon. Of the three Anoy 690 specimens, onty the specimen exposed SCC cracks in Alloy 600, inconel 1-82,1-182 R-127, and AISI 316 I at the highest stress intensity (70 MPaVrE) experienced any crack NG, and by surface cracking in Alloy 690. Onfy inconel I-72 and j j growttL The measured growth rate was much lowe.r than the rates R-135 were immune to cracking in both environments. In the creviced i l measured in the other apoys, which were tested at lower stress SSRTs, there were no apparent differences in the SCC resistance of j intcasses. The absence of cracking at 50 and 60 MPaVrn indicates the a!!oys tested in the pure water and resin intrusion environrnents, j that the K,sce for Alloy 690 in the resin intrusion environment lies except for Alloy 690, which exhibited surface cracking in the resin } b

i ~ im. ym:~ v.?w % H%.iq,t pg.%.C ~ a ^ ' r V' W %~~. 5l O 5 W.'CEE G.* Q &j " 4 ',

ri.:.yk A.,.uM..C.3 - s : -

.. fy,... by,': .74+#i. ** ".1;='- n- ...q, l .&.. M. w-

• Q~ $-i.l.i-4*,.'

~ ~ t w ..*9i.'M8 ** 1..517tWf" AW ] .y - ' ' , Q.L^y,p ?f. .r. ", y -' ~,,: - s. .,W ,y ] <gg[D g >.w '^ .v., W, 5

fhg,

, Tec. g,ft 7', 3 -w I -g 3 + __W ) __m-_ lmm same-1 wm,h a t l i i f e rw .y r~.,1 I If ~ :q y: 3 t e v, i. u h D. ) b - ' '. g[' l 4, % 'f - ~, i A ) .'f,s }~- _,,,M. u i l T ~~ I ~< ) .~ . ~ - 1 l ~' .u-%. n l, -a t b l - s. ) [ ~ ~

s,

i [ 2 5 Oum l I l t i FIGURE 6 - Cross section through fatigue precrack and IGSCC of Al;oy 600/inconell-182 WOL Crack propagation was from left to right. The fatigue precrack (horizontal surface at left) was j located in the inconel1-182 weld. initialIGSCC was through inconel1-182 at an angle of 45 to 90 degrees. Crack moved back toward 0 degrees in the Alloy 600, as can be seen in (a). i l TABLE 5 - Results of Sustained-Load Tests Metena!"' Pure Water Enytronment* Resan eritrusion Erwwonment l l Ahoy 60G&anet k182 Locataed anaca 10Nm cracas m encxvisi h182 l 25-,m cracks n bom ( AAoy 600 and incones vit2 i May 69Mnconet F152 25 am cratas m inconei F182 ho arm or cracka m Ahoy Ei>3 a May 690/R.127 No anaca w crwu Aaoy 69G4125 No anaa or cracks i "%. amens insied n uncrevc.ed and creviced cxndnons e p.re aa:er at 13 are 1.5 o, svens nevois. Onry creviced specur= ens were insied an the resen snuuseon on wonment 288 C. O a 8 pom, t = 2200 FL l n 8'284 C. O, 7 poet t = 3000 h.

"A 55 gS/cm. The rabonale for using the high conductvity environment yfL h k i 0 ? f b (55 SS/cm. 6.5 ppm H,SO.) was that water conductvi:y excursions to 55 S/cm at room temperature have been observed in BWRs.' I 21 h, g-11 p%$is.IflM Andresen determined that the time to failure (or time to crack l h N h E. initiation) in uncreviced constant load tests was decreased by 2 to 3 .5 { h j f *" ' g orders of magnitude in impi tre water compared to pure water and that d ~^ 7d. l U ; tailure occurred at stresses below the matenal yield strength. The i .g65 ' !* @b h Y.pTN e*fects of impunty concentration were studied, and, at 10 pS/cm, a h marked decrease in the seventy of cracking and a marked increase 5' f. 4 i Tir - g Jg h y in the time to failure, compared to the results at 55 pS/cm, were 8 observed. At 5 S/cm, no grain boundary cracking occurred. These l 7. [J'- 6N j g / ~ results are in good agreement with those obtained in the present 4 'J, 68@M } Mh j B-S/cm water (resin intrusion e donment) or in pure water if I work. which showcd that shallow cracks could be obtained in Y M-s '6 o 'EJ 2 crevices were present. These rr: Ats indicate that the threshold T, E %[ conductvity for cracking in water ofth sulfuric acid additions is in tha bfg@g IjM j] .F range of 5 to 10 S/cm. At conduct:vities much lower than this i jg threshold (pure water). cracking would not be expected unless ylF' J. L crevices were present, since crevices may produce a local environ-3..C 4. {.3 a h [ jI h i ment above the threshold. 7 -E Several other investigators have studied the SCC behavior of x ]s.R v. w.- g 100um a inconei alioys in born aqueous suifate saiutions and surfuric acid l ~ environments.** Vermilyea performed straining electrode exper-iments at controlled potentia!s on Alloy 600 at 283 C in both an FIGURE 7 - SEM micrograph of cracks which initiated at aqueous 0.01 M Na SO. solution (neutral pH) and a pH 2.5, H SO. 2 surface grooves of constant load specimens. solution. Cracking was only obtained in sensitized Alloy 600 and not in the annealed or cold-worked conditon. In the neutral sulfate intrusion environment but was immune in pure water. This suggests solution, cracking occurred at 0.25 Vsse., but the cracks only that an aggressive environment may have been established in the occurred at crevices. Cracking in the pH 2.5, H,SO solution crevices of the pure water tests. Based on the limited data for the occurred at 0.25 and 0 Vs,.,e in the uncreviced condition. Later work Alloy 600/inconel 1182 WOL specimens. it may also explain why by Vermilyea'* showed that Alloy 690 was immune to cracking in the crack grwh 'ates were not increased by the resin intrusion pH 2.5, H SO. solutions at 0 V34, even in the sensitzed condition, 2 environment in companson to growth rates in pure water, since the and that cracking of Alloy 600 could be prevented by potentals rnore fatigue precrack in the WOL. specimens acts like a crevice. Also, in negative than -0.2 Vsys. the sustained load tests, the effect of the sulfunc acid environment Jacko'7 has evaluated the susceptbility of A!!cys 600 and 690 was to produce deeper surface cracks in a shorter test penod as a functon of solution pH in 8% Na,SO.. Tests were conducted for t compared to the pure water environment. The acceleration of 5000 h on C-nng spec: mens ioaded to 150% of the matenal yield l cracking througn HyO. additions has also been coserved by a strength. As the pH decreased in the range of 10 to 2, the SCC number of other investgators and their results are discussed below. susceptbility of Alloy 600 was increased with cracks up to 1000- in Copson and Economy'8 were among the first to observe the deep being observed. However, the susceptibility of Ancy 690 acceleration of cracking in Alloy 600 in pressunzed water at 316 C. remained low and constant, with only shallow cracks (<50 Sm) being They tested double U-bend and double bent beam specimens in observed. Jacko also measured average crack proccgation rates for desonized water that had either sulfunc acid added to adjust the pH mill-annealed Alloy 600 in the approximate range of 5 x 10-' to 1.4 value to 4 or ammonium hydrcxide added to adjust the pH to 10. At x 10-8 mm s" in a pH 3, 8% Na SO. sclution. Growth rates for r a pH value of 10, the tests lasted 18 weeks. but, at a pH value of 4, thermally treated Alloy 600 and for Alloy 690 were approximately an the tests were completed after 2 weeks. and cracking was detected order of magnitude slower. This environment was more aggressive as earty as ene week. Cracks in creviced areas were up to 375-m

• Mn Du'e water, but less aggressrve than a 10% NaOH environment.

deep after only 2 weeks of testing at a pH value of 4 terwoodss has also investgated the behavior of Alloy 600 in Floreen and Nelson and Floreen have also evaluated a senes 290 L acid sulfate solutons, using SSRTs uncer potentostatic 7 3 control. Tests were conducted in sulfate, bisulfate, and mixed of Alicy 690 weldments in both a pure water and a simulated resin intrusion environment. The weldments were made witn the filler solutions having sodium-to-sulfate ratios of 1:1,3:2, and 2:1. The rnetals inconel 1-182.1-82,1-112. I-625. R 135, R-127, inconel I-72, susceptibility to SCC was found to increase as the sodium-to-sulfate and R 128, which had chromium contents in the range of 16 to a4%. ratio increased (pH increased). The SCC susceptibility was also a l Single and double (creviced) U-bend spec: mens were onginally functon of sulfate concentration and electrochemical potencal. The l exposed for 18 weeks in a pH 6. 6 ppm oxygen, pure water susceptibility increased with increasing overpotential from the free 4 environment at 316 C. None of the weldment specimens cracked in corrosion potental. this environment. The same set of specimens ure then tested in a One of the reasons for the enhanced cracking in the resin resin intrusion environment with pH value of 4.6 (adjusted by suff:me intrusion environment observed in the present work may result from acid),6 ppm 0. and 316 C. Faded soec: mens were detected amer 12 the su!func acid shifung the electrode potential of the alloys tested 2 weeks, and the majonty of the specimens had failed aher a 40-week into a potential regime, wherein, they are susceptble to IGSC_C. In exposure. A strong ef'ect of weld metal chromium content on the the uncreviced pure water tests, the potential of the alloys tested fell , propensity to failure was observed, with stress corrosion resistance in the range -0.08 to -0.37 Vsms, and the alloys were immune to increasing with increasing chromium content. The recults of Floreen cracking. In the resin intrusion environment, the potential of the alloys and Nelson are in good agreement with those of Copson and tested were in the raoge 0 to 0.16 Vsse. Both Vermilyea and Indig Econorny'8and with those of this program which demonstrates that and Andresen' have performed polanzation tests on Alloy 600 in different test methods can be used equally to determine the sulfunc acid solutions at a high tempertture (289 C). Both investi-aggressiveness of test solutons and rank the SCC resistance of gators obtained very similar polanzaton curves, which indicated that l vanous alloys, the alloy was actve at the open ctreuit potential (-0.4 Vsg) and Andresen*" used constant load and SSRTs to determine the passivated readily upon polanzation. Passrvity was obtained over the e*fect of sulfunc acid additions on the IGSCC suscept:bility of Alloy range of C to -0.2 Vswe. Unlike stainless steel tested in the same 600 in 288 C water containing 0.2 ppm oxygen The range of suffunc . soluton, Alloy 600 did not exhibit secondary passivity, but corroded acid additions used resulted in water conduchvity in the range of 5 to at increasing rates at increasing potentials. Vermilyea suggested

. ~- ), ? presence of sodium sulfate wouac not be applicable to the lower. - that removal of chromium from the passive wrf ace film by the sulfunc acid at high-temperature results in a matenal that is not readily conductivity, sulfate-free environment. l _. passivated, and hegh active currents are therefore octained. Com-Metal dissoluton and/or hydrogen embnttlement are generally panson with the potanzation curves in dilute acid indicates that the Delieved to be responsible for IGSCC in nickel-base aHoys, although i specimens tested irt the resin intruson environment were at poten-the exact mechanism is neither well understood nor unanimously taals above the passive regon for Alloy 600, as determined by agreed upon. Even with the many expenmental studies performed in Andresert this area, a fundamental understanding of all of the factors that The importance of electrochemical pcter.ual on the cracking influence 1GSCC remains elusive. Although the mechanism of behavior of Anoy 600 in deaerated causte solutions at hign temper. IGSCC is generally classified as esther anodic SCC or hydrogen ature has been well Jocumented.' but very lit!!e data in sulfunc acid ementtiement, it should be remembered that these two mechanisms solutions have been generated. However, a number of factors are not mutually exclustw, and the possibihty exists that both regarding the SCC susceptibility of Alloy 600 are aoparent from the contribute to crack propagation at the same time. It is beyond the 2 literature. Hubner, et ai ' identified two potential regimes in which scope of this paper to discuss in detail IGSCC mechanisms that have cracksng occurred in aerated. 300 C water containing 20 ppm 50.'- been desenbed elsewnere,352*-25 a!! hough an outline of the The first regime was a strongly cathodic regon (-0.56 Vsme), and the proposed mechanisms will be desenbed here. i second regime was just below the transpassrve regon (0.63 Vsue). Some studies have indicated a possible role of hydrogen i No cracking was observed at 0.5 and 0.3 Vs# On the other hand, embnttlement as a mechanism for IGSCC in nckel. base alloys.'" Vermitvea has idenufied cracking at 0 and 0.25 Vsg n a pH 2.5 Airey" has shown that SCC initiaton times in pure water were i H SO. soluton at 289 C, but no cracking at potentials more negative reduced when hydrogen was present in the environment, it was 2 i than -0.2 Vssa. The results obtained from the SSRTs in this program postulated that the hydroger, had affected the composition and are consistent with the aoove observatens of Vermdyea. structure of the oxide film, resulting in a film lower in chromium. Although the importance of electrochemical potential on crack-Economy, et alz' have cited evidence that A!!oy t@ specimens ing benavior has been discussed, the results from the pure water exposed to hydrogen-containing water at 363 C containei P to 14 SSRTs also demonstrate the importance of crevees in promoting ppm hydrogen. However, the work of Economy, et al failed to show enhanced susceptibility to SCC. The measured electrode potentials a relationship between IGSCC initiation times and hydrogen partial of the crevced and uncrevced SSRTs were essentially identcat. and pressure. In fact, IGSCC was absent in the environment containing i yet, the alloys tested were immune to SCC if uncreviced and were the highest hydrogen partial pressure (6.300 kPa), implying that i susceouble to SCC if creviced. It is probabie that the potential within hydrogen embnt!!ement does not play a significant role in nickel-base f ~ the crevice was significantly different than the bulk pctential, since in alloys. In general, it would be expected that hydrogen embnttlement I the SSRTs, only the bulk potential was measured. The results imply would be more pronounced at lower temperatures where hydrogen that the funcien of the crevice may have Deen to establish an acide recombination is slow and diffusivity such that hydrogen concen-g crevce environment with an atWndant shift of potential into the trates near the crack tip. Garud and McIlree'* believe that the i cracking regime; the crevce environment would then act like the operable mechanisms of IGSCC is anodic dissolution, but that I acide resin intrusen environment in facistating stress corrosion crack hydrogen may affect the electrochemical activity, as well as the local anitiation. However. Taylor'8 haa performed crevce expenments on detormation response of the metal. thereby influencing the net rate of Alloy 600 at 2S8 C in both pure water and sodium sulfate (0.1 M IGSCC without attenng the basic mechanism. Na SO.) environments. He found that in pure water, the crevice pH in the coinion of the authors, most of the existing data. including l r lay on the basic side of neutral. but that the degree of alkalinity was that from this study, support a film rupture, metal dissolution. p self-timrting, thereby preventing very caustic solutions from develop-repassivaten mechanism for cracking. Briant" has shown that in eng in a pure water / Alloy 600 system. Taylor found that acidic oxygenated high-temperature water, IGSCC can be correlated with l conditaons could only develop in a crevice in the presence of ionic chromium depleton. These grain boundary, chromium-depleted impunties. The exposure to air-saturated water of a crevce initially regens are electrochemically different compared to the matnx containing 0.t M Na,SO soluten generated a strongty acidic matenal and thus provide preferential sites for crack initiation and 22 environment (pH 3). However. later work by Taylor and Caramshas favorabie paths for crack propagation. Also, Lee and Vermityea'* showed that the acidificaten of Alloy 600 crevces at hign tempera-have shown that the deformation charactenstics of Alloy 600 are such tures was tr.nsitory and that a neutral pH was attained after that strain localizes in the grain boundary region. This suggests a apprcrimatety a one-week penod. localized film rupture mechanism of the grain boundary region. Once Understanding of eiectrochemical conditons within a crack or the film is ruptured, then metal dissoluton at the grain boundary 3 crevice has improved because of expenments such as those causes localized grain boundary attack. Repassivaten will prevent conducted by Taylor and others. However, most of the work has been further attack until the next rupture event. However, once the depth l l conducted on static crevices" or cracks. Investigators have not of the antergranular attack is sufficient to produce an acid crevce i addressed the influence of dynamic straining, as exists at a growing soluton, a propagating crack would be formed. 1 crack tro. on the chemistry of the crack tip environment. A number of The mechanism by which the sulfuric acid solution enhances mathemaical rnodeis used to predict SCC rates ass 9me that the SCC suscept:bility is thought to result from the impure environment chemistry inside a propagating stress corrosen crack is similar to that being more aggressive (increased corrosion rate), and therefore a i 23 i within a static crevce. However, recent work by the authors has greater crack advance per film rupture event occurs. Vermilyea and shown that this is not always a vand assumption, and the work offers Indig have shown that matenal depleted in chromium dissolves at an attemate explanation for the effect of the resin intrusion environ-high rates an acid solutions. In addition to the acid nature of the resin g ment. These expenments on AISI 304 SS found that crack tip strain impunty environment. crevice / crack growth processes can result in q affected both the potentiat and the pH at a simulated crack tip. Using impunty concentration and metalion hydrolysis, which lead to a more elevated temperature potential-pH diagrams to p!ot the potential and highly concentrated and acidc crack tip environment. The kinetics of pH changes at the crack tip, it was speculated that the conditions at SCO in impure high-temperature water could be controlled by the rate the tip of a dynamcally strained crack may favor the reducten of of film rupture the metal dissolution rate, or the rate of transport of suttate sons to suffite ons and may produce less protectrve sulfates vanous onic impunties to the crack tip. Andresen has provided ar d suffices of nckel and iron instead of the equivalent metal oxides extensrve evidence that supports the hypothesis that mass transport or byttn:rxides. Thus, the ionc soecies present in a crack or crevice is the rate controlhng step in the mechanism of IGSCC of nickel-base are very important in affecting crack tip reactons and the crack tip alloys. j chemistry. These reactons would affect the validity of experiments conducted in whch sodium sulf ate is considered a harmless additive Conclusions and is aoded to the environment to solety increase soluten conduc. The following conclusions can be drawn from the results tsvity and facilitate electrochemcal measurements. The system octained in the present investigation: Chemistry Could be changed so much that data obtained in the

1. In an uncreviced geometry, Alloys 600,690, and AIS1316 NG

~ i. and weld metals inconel I-82,1-182, I-72 R-127. and R 135 were all 9. A. McMinn, Corrosen, Vol. 42, No.11, p. 682,1986. immune to SCC at 288 C in oxygenated pure water. Transgranular 10. F. F. Lyle, Jr., E. B. Noms, " Stress Corrosion Cracking - The surface cracking of ASTM A508 was observed in pure water, s;ow. Strain-Rate Technique." A. M. Ugiansky, J. H. Payer,

2. The additen of 1 ppm HAO. to oxygen containing pure water Eds., ASTM STP 655. ASTM, Philadelphia. Pennsytvania, pp.

resulted in a marked reducnon in SCC resistance of all the alloys 388 398,1979. tested in the uncrevced conditon, except for the high chromium weld 11.

l. J. Magar, P. E. Moms. Corrosion, Vol 32, p. 374,19'76.

i metals, R-135 and inconel 1-72, which were smmune to cracking. 12. D. F. Taylor, Corrosion, Vol. 35. No.12, p. 550,1979,

3. All of the alloys that were suscepuble to SCC exnibited intergranular SCC, except AISI 316 NG and ASTM A508, which H. R. Copson, G. Economy. Corrosion, Vol 24* No* 2, P 55' 13.

g exhibited TGSCC.

4. Ranking of the alloys in terms of their SCC resistance in the P. L Andresen. "A Mechanism for the Effects of lonic impurities 14.

s: mutated resin intrusion environment indcated a correlation with on SCC of Austenitic Iron and Nickel Base Alloys in High Temperature Water, CORROSION /85, Paper No.101, Na-bu!k chromium content, with increased chrornium content providing an increased resistance to SCC. Alloy 600, AISl 316 NG, inconel ti nal Association of Corrosen Engineers, Houston, Texas, O 182. and 1-182 with a chromium content less than 20% exhibited a marked susceptibility; whereas, R-135 and inconel 1-72 with chro-15. D. A. Vermilyea, Corrosen. Vol. 29. No.11, p. 442,1973. miam contents greater than 29% were immune. Alloy 690 and R 127, 16. D. A. Vermilyea. Corrosion, Vol. 31, No.12, p. 421,1975. with chromium contents of 29 and 28%. respechvely, exhibited 17. R. Jacko, " Stress Corrosion Testing of Candidate Steam borderfine susceptbility; Alloy 690 exhibited only surface cracks Generator Tubing Materials" The Electric Power Research rather than deep SCC cracks and crack growth only at high stress Insttute Workshop on Alloy 690 EPRI, Palo Alto, Califomia. t intensity levels (70 MPaVrE), and R-127 exhibited SCC in some 1985. SSRTs and not in others. 18. T. C. Underwood, "The influence of Sulfates and Otner

5. The effect of crevce condibons in the pure water environment impurities on Some Aspects of Corrosion in PWR Stcam was to increase the cracking susceptibility of all of the alloys tested.

Generators." Electric Power Research Insutute Report WA except for Alloy 690. Crevees had little influence on ciacking 157. EPRI, Palo Alto, Califomia,1980. suscepubility in the resin intrusion environment, presumably because 19. D. A. Vermilyea, M. E. Indig. J. Electrochem. Soc., Vol.119, No. the aggressive solution was present initialy. 1,p.39,1972.

6. The resin intrusion environment was found to shift the 20.

E. Serra, " Stress Corrosen Cracking of Alloy 600," Electric potential of the alloys tested ~160 mV in the noble direcuon into a Power Research Insttute Report NP 2114-SR, EPRf, Palo A'to, potential regime where 1GSCC could occur. Califomia November,1981. 21, W. Hubner, B. Johansson, M. Pourbaix, " Studies of the i Acknowledgments Tendency of intergranular Corroson Cracking of Austenitic f This work was funded by the Electnc Power Research Insutute Fe-Cr-Ni Alloys in Hign Punty Water at 300'C." A. G. Atome-(EPRI) and was conducted as part of EPRI Research Proiect 1566-1. nergi Report AE-437, Studsvik, Nykoping, Sweden,1971. The authors would like to thank J. L Nelson. W. J. Childs. and A. R. 22. D. F. Taylor, C. A. Caramshas, J. Electrochem. Soc., Vol.129, McIlree for their support. No.11, p. 2458.1982. 23. R. A. Page, S. J. Hucak, Jr., A. McMinn. Effects of Dynamic References Strain on Crack Tip Chemistry - Volume 2: Tests using an 1. R. A. Page Corrosion, Vol. 39 No.10. p. 409,1983. Intemally Creviced Tensile Speermen," Electnc Power Re-2. R. A. Page. A. McMinn, Met. Trans. A., Vol.17A, p. 877,1986, search lettute Report RD.4649. EPRI, Palo Alto, Califomia. 3. J. L Nelson, S. Floreen. Second Int. Symp. Environmental July 1986. Degradation of Materials in Nuclear Power Systems - Water 24. R. L Cowan, G. M. Gordon, "Intergranular Stress Corrosion Reactors Amencan Nuclear Society, LaGrange Park, Illinois, Cracking and Grain Boundary Composition of Fe-Ni-Cr Alloys," General Elecinc Report NEDO-12399, General Elecinc Co., p.4,1986. 4. R. J. Kurtz, D. W. Shannon, B. Francis,- F. M. Kustas, P. L Schenectady. New York, September.1973. i 25. D. Van Rooyen, Corrosion, Vol. 31, No. 9. p. 327,1975. Koehmstedt. "Evalvauon of BWR Resin Intrusions on Stress 26. Y. S. Garud, A. R. McIlree, "IGSCC Damage Modet An Corrosion Cracking of Reactor Structural Matenais." Electric Power Research Institute Final Report NP-3145 EPRI, Palo Adproach and its Development for Alicy 600 in High Punty Water," CORROSION /BS, Paper No. 88, National Association Alto, Califomia. June 1983. f Corrosion Engineers. Houston, Texas,1985. 5. R. M. Asay, J. Blok. J. H. Holloway. ' Water Quality in Boiling Water Reactors," Electnc Power Research insttute Report & R y, h h Symp. EnWonmed Degrahn d 27. a enals in Mear Nww Systems - Water Reams, j NP-1603, EPRI. Palo Alto. Califomia. November,1980. National Association of Corrosion Engineers, Houston. Texas. 6. P. L Andresen, "The Eftects of Sulfate impuntes in 288 C p.462,1984. Water on IGSCC of inconel 600 in Constant Load and SSRT 28. G. Economy, R. J. Jacko, J. A. Begtey, F. W. Pement, Experiments," CORROSION /84. Paoer No.177, National As- " Influence of Hydrogen Partial Pressure on the IGSCC Behav-sociation of Cor osion Engineers. Houston, Texas,1984. ior of Altoy 600 Tubing in 360 C Water or 400 C Steam," 7. S. Floreen. "Evaluaton of Alley 690 Weidments in Simulated CORROSION /87, Paper No. 92, National Association of Cor-BWR Envitonments," RP 1566-2, Sixth Semi-Annual Report-resion Engineers, Houston, Texas,1987. Inco Alloy Products Company Research Center, March-August 29. C. L Briant, C. S. O'7oole, E. L Hall, Corrosion, Vol. 42. No.1 1982. p.15,1980. 8. R. A. Page, Corrosen, Vol. 41, No. 6, p. 338.1985. 30. D. Lee, D. A. Vermilyea. Met. Trans., Vol. 2, p. 2565.1971. i

WAPD-T-2999 INCONEL FILLER METAL 52 (R 127) DISCUSSION ON MANUFACTURE AND WELDABILITY TESTING l 1 G. K. Mathew, T. W. Hauser, and D. F. Krawiec I NOTICE This report was prepared as an account of work sponsored by the United States Government. Neither the United States, nor the United States Department of Energy, nor any of their employees, nor any of their contractors, subcontractors, or their employees, makes any warranty, expressed or impiled, or assumes any legal liability or responsibility for the accuracy, completeness or usefulness of any information, apparatus, prodtet or process disclosed, or represents that its use would not infringe privately owned rights. l r BETTIS ATOMIC POWER LABORATORY PITTSBURGH, PENNSYLVANIA 15122-0079 Operated for the U. S. Department cf Energy. by WESTINGHOUSE ELECTRIC CORPORATION

WAPD-T-2999 ) ABSTRACT Inconel Filler Metal 52 (1-52), also known as R-127, is a high chromium weld filler material developed by inco Alloys, International (inco) for welding Alloy 690 base materials. A heat of this chromium-nickel-iron filler material was manufactured by inco in the 1970's, but it did not find widespread use and was not commercially marketed. An additional heat of l-52 was melted in 1991 using the Vacuum Induction Melting process and remelted using the Electroslag Remelting process. The chemical composition of this new melt, as well as the melting practice, were chosen so as to yield a product with low concentrations of residual elements. Groove welds made using most of the commonly used weld processes have been tested. Base materials welded included Alloy 690 to Alloy 690, Alloy 600 to Alloy 690, EN82 Buttering to Alloy 690, and Alloy 690 to HY-80. Welding and test results are generally accept-able. No weld cracking has been observed in any of the welds, except in an application i l where it was used to weld stainless steel to Alloy 600, and when it was used with the submerged arc welding procesc. Because of the high chromium content, and the presence of aluminum as one of the constituents, good gas shielding and weld cleaning are necessary. Mechanical properties of Inconel Filler Metal 52 are comparable to that of RN-82. l f

WAPD-T-2999 l ^ Based on availability of Alloy 690 as an alternate base material for reactor plant applications, the acceptability of Alloy 690 as weld filler material has been i evaluated i i Weldability tests performed: e V-Groove weld tests

• Slotted type root cracking tests
• All tests used the manual gas tungsten arc welding process I

i i I

l i l WAPD-T-2999 Chemical Composition of Alloy 690 (Wt%) i i Element Alloy 690' Cr 27-31 Ni 58 min i Fe 7-11 i C 0.05 Mn 0.50 Si 0.50 Al Ti Cu 0.50 S 0.015 P 1.From ASME SB168 a F i

j l WAPD-T-2999 i b i Weldability test results were acceptable. i r o redrawn base material; Inco has 1 reported weld cracking when 690 is used as a filler i' I F t i

1 i ) WAPD-T-2999 inconel' Filler Metal 52, previously called inconel R-127 developed by Inco Alloys International in the 1970's Commercial heat previously produced not marketed widely I t I

i WAPD-T-2999 ~ Chemical Comoosition of Inconel 52 (R-127) Cr 28.0-31.5 Ni Balance Fe 7.0-11.0 ~ C 0.04 max. Mn 1.00 max. 3 Si 0.50 max. Al 1.10 max. Ti 1.00 max. l ~ t Cu 0.30 max. S 0.015 max. P 0.020 max. Mo 0.50 max. Cb 0.10 max. Others 0.50 max. Al + Ti 1.50 max. l The above chemical composition is from inco Alloys, International { I

I ) WAPD-T-2999 Technical Requirements for a new heat of Inconel 52 (R-127) developed

• Specific " aim" for major elements
• Melting practices to be consistent with obtaining the " aim" chemistry, and controlling residual elements New heat melted in 1991
• Vacuum Induction Melting
• Electroslag Remelting t

WAPD-T-2999 Chemical Composition of Different Heats of Inconel 52 (R-127) Element Heat No. Heat No. Heat No. Y50A9SL Y9378K Y9379K Cr 28.81 28.97 28.95 Ni + Co 60.78 60.37 60.37 Fe 8.69 8.98 8.99 C 0.01 0.03 0.03 Mn 0.26 0.23 0.24 Si 0.16 0.17 0.17 Al 0.70 0.63 0.63 Ti 0.49 0.56 0.56 Cu 0.01 <0.01 <0.01 N 0.074 0.007 0.007 B 0.001 0.001 S 0.005 <0.001 <0.001 P 0.009 0.004 0.005 Pb <0.01 <0.001 <0.001 Mo <0.01 <0.01 Cb + Ta <0.01 <0.01 Co 0.01 0.01 Heat No. Y50A9SL is a heat manufactured in the 70's L

1 ~ WAPD-T-2999 l 1 l 1 Weldability Tests Performed on the New Heat of Inconel 52 i e Mechanical testing in accordance with MIL-E-21562 using Alloy 600 and Alloy 690 base materials 1 e V-Groove weld tests with different base materials i e Segmented - groove circular l restraint tests with Alloy 690 base material ^ e Fillet weld test with Alloy 600 / stain-I less steel base material 4 I o -I

3 8" = 3/8" - - 3/8" --- 1/2" AND 3/4" = =- = t i e l i i s I I i i i 1 1 i l l I I i i I I i l i I I I I I i i i l 1 i I f BACKING I i i i \\ / STRIP i /~ i q;;;;;;; ' ~ ~._ _ _ _ _ _ _ _ _ _ - j;;;;;;;;s;i 4, 1 00 ipooi 1 60 I 1 lo.n. 0 0 _. --------- ~~_ioonok i I .- oi-s l l l l N il i I l l l l --1/4" MAX. 1 I I I I I I 1 i i i i TENSILE SPECIMEN / SIDE BEND SPECIMENS 5> T MIL-E-21562 TEST PLATE Z e8

12 1/2" /* = /q ,/ =' u 7= 61/4" / / ./)' ' k, 7 / g,? t o" - 12" TEST PLATE TEST PLATE f f )hk / / ?' sb 450 GROOVE f v s r '"" p// l s mm'.- // \\ \\ / / 1/8" F -- 2" ~ l (TYP.) STRONGBACK -= 1/4" ROOT GAP =, G /4l 4 b TYPICAL BACKING STRIP h 12" = = 8 V-GROOVE TEST PLATE

WAPD-T-2999 Weld Processes Usec for V-Groove We c Tests Manual - Gas Tungsten Arc Welding (M-GTAW) Automatic - Gas Tungsten Arc Welding (A-GTAW) Semi-Automatic - Gas Metal Arc Welding (SA-G M AW) Automatic - Gas Metal Arc Welding (A-G M AW) 1 Submerged Arc Welding (SAW)

j WAPD-T 2999 i i Base Materia s Weldec i e Alloy 690 to Alloy 690 I e Alloy 600 to Alloy 690 i e EN82 Buttering to Alloy 690-j e HY-80 to Alloy 690 j e 304SS to Alloy 600 (fillet weld test) I i i L w en r w ,,-+,n,.

---

w-

Weld Parameters Used (V-Groove Test Plates) i Weld Material Weldino Voltage Travel Wiro Test No. Process Joined Current (volts) Speed Feed (amps) (ipm) (Ipm) 1 A-GTAW 690-690 205-235 12-12.3 4 30-43 2 A-GTAW 690-HY80 205-235 12-12.3 4 3043 3 SA-GMAW 690-690 245 33.5 8.5-10 370 4 SA-GMAW 690-HY80 245 33.5 8.5-10 370 5 A-GMAW 690 490 195-205 30.5 11 370 6 A-GMAW 690-HY80 195-205 30-31.5 8-9.5 300-310 I l t O i h !8 8

WAPD-T-2999 RESULTS e All NDT results generally acceptable to the requirements of Mil-E-21562 some porosity observed initially - attributed to base material used for testing Guided bend test results acceptable e no visible defects Oxide islands on weld surface reduced by good gas shielding, interpass cleaning; can be removed by a light mechanical grind i 4

WAPD-T-2999 RESU _TS(cont'c ) Cracking in fillet weld when welded to stainless steel possibly due to excessive weld metal dilution, similar to when EN82 is used to weld Alloy 600 to stainless steel Cracking when the submerged arc welding process is used due to high heat input, and possibly due to lack of suitable flux

Room Temperature Mechanical Test Results (MIL-E-21562 Test Plates) We!C Material Yield Tens!!e Elongation Test No. Proces t Joined (KSI) (KSI) 1 M-GTAW 690-690 73.2 98.1 32 2 M-GTAW 690-690 60.6 93.6 42 3 M-GTAW 690-690 69.6 89.6 38 4 M-GTAW 690-690 67.8 97.3 34 5 M-GTAW 690-690 64.6 95.5 40 6 SA-GMAW 690-690 47.6 82.5 51 7 SA-GMAW 690-690 49.0 83.5 52 8 SA-GMAW 690-690 49.2 83 51.5 9 SA-GMAW 690-690 46.5 82.2 54.5 10 SA-GMAW 690-690 49.1 82.8 53 6 20 O h @8

Room Temperature Mechanical Test Results. (V-Groove Test Plates) Weld Material Yield Tensile Elongation Reduction Test No. Process Joined (KSI) (KSI) Area (%) 1 A-GTAW 690 490 70.9 96.2 36.5 55.5 2 A-GTAW 690-HY80 68.3 94.2 42.5 58 3 SA-GMAW 690-690 53 83.8 48.5 69.2 4 SA-GMAW 690-HY80 53.4 84.1 48 67.5 5 A GMAW 690-690 53.9 86.2 49 56.1 6 A-GMAW 690-HY80 52.3 83.4 48.5 70.2 I l ~ 1 6f M !8 l l s r +- - - --- - - - - - - - - - - - - - - - - ^- - ' ' - - ' " - - -

^ WAPD-T-2999 i CONCLUSiO \\lS e inconel 52 (R-127) is a suitable weld filler metal to join Alloy 690 and Alloy 600

• All the commonly used weld processes (except SAW process) can be used to weld with Inconel 52 (R-127)
• Good gas shielding is required to minimize oxide formation.

Good interpass cleaning is required; a light mechanical grinding between weld passes to remove surface oxides may be necessary

• Normal controls used for welding nickel based alloys are required r

WAPD-T-2999 CONCLUS ONS (cont'd.)

• Inconel 52 (R-127) can be used to make dissimilar metal welds between Alloy 690 and low alloy steels.

Minimize weld metal dilution to provide margin against crack.ing during welding Further evaluation necessary before using inconel 52 (R-127) for welding inconel to stainless steels J

.c STRESS CORROSION OF ALLOY 600 WELD METAL IN PRIMARY WATER JV Mullen and RJ Parrington i General Electric Company KNOLLS TOMIC POWER LABORATORY A Schenectady, New York Presented at the 1992 EPRI Workshop on PWSCC of Alloy 600 in PWRs Orlando, Florida l December 1-3, 1992 1 -..a

9 NOTICE This report was prepared as an account of work sponsored by the United States Government. Neither the United States, nor the United States Department of Energy, nor any of their employees, nor any of their contractors, subcontractors, or their employees, makes any warranty, expressed or implied, or assumes any legal liability or responsibility for the accuracy, completeness or usefulness of any information, apparatus, product or process disclosed, or represents that its use would not infringe privately owned rights. i f L 2

O E l ABSTRACT A test program consisting of 502 double U-bend and 112 C-Ring specimens was conducted to evaluate SCC initiation of EN62, EN82 and EN82H weld metal. The testing was conducted at temperatures of 600,640 and 680*F for up to 249 weeks. The data was statistically evaluated where possible, although many of the parameters ofinterest did not have sufficient data sets with failures to do this. There is wide variation between heats and even within' heats in SCC initiation time. Therefore, large numbers of specimens are required to characterize SCC behavior. In general, thermal treatment of the weld metal was beneficial to reducing SCC initiation. Increasing carbon content of the weld metal (EN82H) had a smaller benefit, while pickling and base metal microstructure had a minimal or mixed effect on the weld metal SCC. i A small scoping test was performed to evaluate the SCC initiation and stress corrosion crack growth rate (SCCGR) of Type EN82 weld metal in primary water. Testing was performed on three 0.4T precracked compact tension specimens machined from an EN82 weld cradle formed by hot wire tungsten-inert gas welding. Instrumented constant load testing was performed for / 108 days in 680*F primary water at a stress intensity of 35 ksiVin. These tests support a best V estimate SCCGR of 0.80 mils / day for EN82 in 680* primary water. Additional tests were performed to evaluate whether EN82 and EN82H weld metal are susceptible to hydrogen-assisted low temperature crack propagation (LTCP) as observed in other, higher strength nickel based alloys. Rising load tests were performed on 0.4T compact tension specimens at slow load rates in 130 and 200*F hydrogenated primary water and 130*F air, f indicating that both EN82 and EN82H are susceptible to LTCP at a stress intensity of 50 to 60 y i ksiVin. An Alloy 600 pipe weldment SCC test was also conducted. Three inch diameter Schedule 160 Alloy 600 pipe segments were welded with Type EN82, EN82H and EN62 weld metals. The pipe mockup was loaded to approximately 20 ksi (combination of axial and pressure loading); residual stresses in the weld and heat affected zone (HAZ) are estimated at 30 ksi. After 78 [ [ weeks of testing in 680*F primary water, SCC initiation was observed in the weld metal (0.009" maximum) and the HAZ (0.024 maximum). Unless otherwise stated, all testing was conducted in a high purity primary water environment with the following specifications (STP): pH - 10.0 to 10.3,0 < 10 ppb, H - 25 to 45 cc/kg H 0, 2 2 2 and resistivity - 0.012 to 0.058 meg ohm-cm. 3 m

M 5 Stress Corrosion of Alloy 600 Weld Metal in Primary Water Topics to be Covered 1. SCC Initiation Testing 2. Stress Corrosion Crack Growth Rate Testing 3. . Hydrogen Assisted Low Temperature Crack Propagation (LTCP) 4. Alloy 600 Pipe Weld Test 4

i Stress Corrosion of Alloy 600 Weld Metal in Primary Water l l l Testing Environment (STP) i l pH - 10.0 to 10.3 0 - < 10 ppb 2 H - 25 to 45 cc/kg H O 2 2 Resistivity .012 to.058 meg ohm-cm h i 5 f

1. STRESS CORROSION CRACKING INITIATION 502 Double U-bend & 112 C-ring specimens EN62, EN82 weld metal Tested at 600 F,640 F,680 F (315,338,360 C) for up to 249 weeks Four strain levels; 10% and 16% for U-bends .45% and 4% for C-rings 6-

SCC Initiation - Results i Thermal treatment of weld metal (1125 F (607 C) for 7 hours) Beneficial in reducing SCC Statistically - 95% confidence of a significant difference 56% of non-thermal treated specimens failed 21% of thermally treated specimens failed 7

9 a Effect of Thermal Treatment Mill annealed, high strain, EN82, double U-bend specimens { 50 15 40

S S

h 10 s0 il ~ / llf x d d 10 m 10 20 40 60. 80 100 150 200 >200 Weeks Not Treated ThermalTreated O-Not Treated (Cumulative) + ThermalTreated (Cumulative) JVM 10,92

SCC Initiation - Results Carbon Content of weld metal (EN82) Low Carbon - <.03% High Carbon - 2:.03% (EN82H) Small benefit in using high carbon EN82 Statistically - 29% confidence of a significant difference 60% failure rate for low carbon 44% failure rate for high carbon 9

E=ect of EN82 Carbon Content Mill annealed, high strain,680F, double U-bend specimens -O 60 - r, 20 50 J V/ -c 15 E Y e c e / i 10 g 20 p/z i 1 7 3 d 5 7 ~ 0 O 10 20 40 60 80 100 150 200 >200 Weeks Low Carbon liigh Carbon O-. Low carbon (Cumulative) 4 IIigh carbon (Cumulative) svM m92 1

Effect of Strain on mill anneal & cold worked,680F, WDUB specimens 20 60 -C so 15 .0 l. i M l 2' .E

1. ~

J O -0 e 20 5 ~ 10 0 0 10. 20 40 60 80 100 150 .200' >200 Weeks High Strain kw Strain .-O Low Strain (Cumulative) + liigh Strain (Cumulative) JVM 10S2

E=ect o" Tem aerature on EN82 High Strain, Mill Anneal & Cold Worked, WDUB specimens 20 60 O 50 15 m 5 z 3 p P 2 o 10 30 i 3 s$ /# 3 c = /s s 3 a l / ^ m A A A A 0 0 10 20 40 60 80 100 150 200 >100 Weeks 600 F 640 F 680 F --O-. 600 F (Cumulative) + 640 F(Cumulative) + 680 F(Cumulative) JVM 1G92

Ef ~ec: o" Weld Wire Type on High Strain, Mill Anneal & Cold Worked, WDUB specimens at 680F 3a so 25 y , 20 o so I e 10 ~ N 2 2 w Io 20 40 60 so too iso 200 >200 EN62 EN82 o EN62(Cumulative) o EN82(Cumulative) JVM 1092 11

Effect of Base Material Microstructure on HAZ Failures on EN82, IIigh Strain, mill anneal & cold worked,680F, C-Ring specimens 25 60 50 40 i 8 1s o l3 C D ~ ,o i 20 / 5 / to r l 4 4 s l 0 0 v w l Weeks Good Bad -O-Good (Cumulative) 4 Bad (Cumulative) Jvu ta92 - I

t SCC Initiation - Conclusions Thermal treatment (1125 F (607 C),7 hr) reduces SCC susceptibility but does not eliminafe risk of short-time failures Low carbon EN82 is more susceptible than high carbon HAZ SCC resistance is better in base material with good microstructure Increasing strain increases failure rate and decreases failure time Increasing temperature increases failure rate and decreases failure time EN62 is more susceptible than EN82 15 t

2. STRESS CORROSION CRACK GROWTH RATE TESTING 3 .4T precracked compact tension specimens 'EN82 weld metal Constant load - stress intensity of 35 ksiVin (38 MPaVm) 680 F (360 C) primary water for 108 days Results/ Conclusions Specimens showed a range of growth rates of 0.62 to 1.08 mils / day (1.8 x 10-2 to 3.2 x 10-2 m/s) Best estimate SCC growth rate of 0.80 mils / day (2.4 x 10' m/s) SCC initiated in less than 16 days 16

Characteristics of' Weld Metal Used for Crack Growth Rate Specimens Characteristic l Filler Metal EN82 ID 761668S Carbon Level 0.009 w/o Welding Process Hot Wire Gas Tungsten Welding Parameters Shielding gas Argon Current 300 amps Volts 12.5 v Note: 0.4T Compact Tension Specimens were oriented such that the notch is perpendicular to the welding passes. Wold Metal Wrought Metal / / I ./ / \\ M ~ -/ I I 17

COMPARISON OF SCCGRs FOR EN82 WELD METAL SPECIMENS Specimen Test Time Crack Detection Cracking Time Maximum SCCGRW SCCGR A I.D. (days) Time (days) Crack Depth (mils / day) (mils / day) (days) (mils) 7741 108 3.60 104.40 112.3 1.08 0.95 Side 1 7741 108 3.60 104.40 95.9 0.92 0.83 Side 2 0726 108 3.40 104.60 67.8 0.65 0.56 Side 1 0726 108 3.40 104.60 79.8 0.76 0.69 Side 2 7742 108 15.90 92.10 67.8 0.74 0.74 Side 1 7742 108 15.90 92.10 56.8 0.62 0.62 Side 2 Overall Average I = 0.80 I = 0.74 Notes: (1) SCCGR = (Maximum Crack Depth - Crack Depth at Detection)/ Cracking Time For this calculation, the crack depth at detection was assumed to be zero. (2) SCCGR = (Maximum Crack Depth - Crack Depth at Detection)/ Cracking Time For this calculation, the crack depth at detection for specimens 7741 and 0726 were determined by correlating the total SCC area to the LVDT compliance. Due to a broken LVDT wire, a crack depth at detection could not be calculated for 7742 and was assumed to be zero. 18

. __ _ ~ _. t 8 l 1 g . ~.), (%e " ***'. r, y,.,*._. / t

• i D OT17 '

~ f..,.,,', * ' \\

• '".s.

,-~..e s ~g L ' f'.:. ;, >.:i p, e .. g, n% n .y g.j.t. \\.

p

.*. m,,c.f:.

n... nE..,.,.,,,.. $

bhfYk, h g-kkhh.lA$thi?hfh.' i -k k ( l y,.. T'c tg0. p.. Ap nq. hl1,;,V.. : ' ; 4,. ' '.,..w 7',,.3 ..., a g. ..,) 5 ~_s, A....r ' y

a l

l q' {:.*:... f,.'*'y ...yr. a ', I y ..}* y . :'{. ' _,. \\ _ ' s, I .. 'J.,. '. : i. ,[', 1. ' l:p i I

s. L

.t.

.9 3 ..., a., ,s i a) Low magnification (16x). Interdendritic stress corrosion cracking (IDSCC) region j outlined in black. l I l I 4 I i i .+. . oI.- r 7 s - / ; ,.e l l . i,' l \\ \\9 I .h '\\. r l l I i I l b) Enhanced view of IDSCC region (23.5x). j FIGUlW II.1 : Photomicrographs illustrating interdendritic stress corrosion cracking (IDSCC) on the fracture surface of EN82 precracked 0.4T compact tension specimen 7741. i i t A. -1

i i i l l l e ' [ W - ',-, n -

..,r

,3 ,. ),r ; _ ; c,; i - ~ k.: p,. 7,..,,- &j,(- .;,.y .; c '~ .gj, t, i \\g v' .. Et, '. j se ..u ',. .. g. ~ ,\\. e.::. .a .c. c :.::. ;.;;., i 3. j , q w.., 4.. ..,y, ..c., ..,.,c.,.. ..m. o v *d,. ..; y, c _ y 2 7.fi.H'IQ.jl,' ,..,r. I ' i: k, 1.,/,.' .t.I ^- I ' 6 )h :%j^.~f..(Y*+!);/ l { f:O\\ )i jfe' f',i't J. -J N, s i f l c' / i[ g4... .,t.,y ty[ con 2. ~.,. s. j v,. y.y, - ~~ .,-. <; j. - 3r . y s. 2.- 3, -. _ 'y ' 6,.'.i': ,r ? ' ~- '4 ,y i .i s )i 7. t _3 4 i S... l a) Low magnification (16x). Interdendritic stress corrosion cracking (IDSCC) region outlined in black. 1 { l l i 1 t i. / I i e

b i

t b) Enhanced view ofIDSCC region (23.5x). FIGURE II.1: Photomicrographs illustrating interdendritic' stress corrosion cracking (IDSCC) on the fracture surface of EN82 precracked 0.4T compact tension specimen 7741.

D l 3. HYDROGEN-ASSISTED LOW TEMPERATURE CRACK PROPAGATION (LTCP) 7 .4T precracked compact tension specimens 2 - EN82 (low carbon) 5 - EN82H (high carbon) ~ Rising load test l - Test parameters: Ramp rate 10-14 lb/hr (~0.4 ksiVin/hr) (~.4 MPaVm/hr) Temperatures 130 and 200 F (54 C and 93 C) Environments air and hydrogenated water Hydrogen levels 10 to 95 cc/kg H O 2 20

i l 1 .s RESPONSE WIIEN TESTED IN AIR SEVERE ENVIRONMENTAL EFFECT j ON P,,,AND K,,, p P dL aox a O E S S t t P TIME --> TIME ---> l I i SIGNIFICANT ETTECT ON LESS EFFECT ON ) CRACK PRDPACATION CRACK PROPACATION j - P.,x i -P J asx w O fi! S S + t 1/2 P,x 1/2 P (-t ux i t IZSS TRAN 2.0 MINUIES t EQUAL TO OR GREATER THAN 2.0 MINUTES f i TIME---> TINE W i l I Schematie illustrations of risine load testinc { i P ~ 21 f

Hydrogen-assisted LTCP 4 Results/ Conclusions LTCP occurs at 50 - 60 ksiVin (55 - 65 MPaVm) in primary water Results i~n relatively rapid intercolumnar failure Only 4.6 ksiVin (5 MPaVm) difference between EN82 and EN82H > 1.5X environmental degradation in K ,x ym 1.4X increase in K ,x when hydrogen level decreased from 85 to pm ~11 cc/kg Higher temperature (130 F to 200 F) (54 C to 93 C) increases K ,x by a pm factor of 1.2 22

m m l .,., e ~.L. C1 4 ? e .g. _I a C~. O v n C1 ~ C ~, Ne

1, C C LY I

Il V O 30 c.e g +- g - 0; C# g c -J v o ,,3 I t% 1 .a .,..~. A t p,-. N:. J t. s g s, g. W ..... ;. o 4. -. ; ,.~4.. ...-.f, .,..........g....p. ,...;...,....i.~.;n- .. ~ ~ + s. 1

• s....

t. 't. ..s. s. . S ....... - g, . s., 1 . i 2

i..

2 . i : 1 .t ~ t 8 s y==( .. s. f L t t. 1 . a go ~...... - g. .. g.. 4......{.....g... 3 >..... . g.. o ! ~. 4...+.. r .,gou g e.4.... ).....e,- o y (f) { W t-- >== = 1 3 y 1.LJ >== t 6 - y, F-- >W .T g 4 t 4 g a1 L e. Z 1 1 . ~.. ............................. - N t.5 . 2 4....5....,.4....... ...4...4...4...&....;...,;......+....

5.. ;.. 4...

..4...4.........I...... .. 3 .t t .t 1 .. +.......... s .g C 4 .A e e a 6 a 4 a 1 e s i a e a e e a a s a a h ) O O O O C O O O O O O O O'O O O O O O O O O O O O C O O O O O O O O O O O O O O O O O O O O O O O O O I n v nN o c) O tw to O N M N 7. O C) to IN oO N M N n n n n M' - - e - - - n-SE l 'GVOl n M0 h ,_m-I 1

Characteristics of Weld Metal Used for LTCP Specimens Characteristic Specimen 1 Specimen 2 Filler Metal EN82 EN82II ID 761668S NX3651D Carbon Level 0.009 w/o 0.0% w/o Welding Process llot Wire Gas Tungsten Cold Wire Gas Angsten Welding Parameters-Shielding gas Argon Argon Current 300 amps 90-150 amps Volts 12.5 v 9-11 v Note: 0.4T Compact Tension Specimens were oriented such that the notch is perpendicular to the welding passes. Weld Metal Wrought C Metal / / / / i odo / 24

s EN82 and EN82H Weld Metal low Temperature Crack Propagation Test Results Spec teen Test Test H Level Preconditioning Pmax Kpus fine to Coments ID Environment Top cc/kg A,,,, pounas rsion u2 P., 2 (*F) inch (1) Minutes 7743 DN 130 Nom. 60 550 600*F for 4 to 5 Days in 0.409 2000 59.3 605 HVTIG -EN82 35 cc/Kg u, erd 200 lbs 0730 DPW 130 Nom. 95 550-600*F for 4 to 5 Days In 0.390 2002 54.9 100 HVilG - EN82 35 cc/Kg H, ard 200 tbs CB5-3 DPW 130 Nom. 85 550-600*F for 7 Days in 35 0.390 1820 50.3 69 CVTIG - EN82H cc/Kg H and 200 lbs 3 CBS-2 DPW 130 Nom. 85 550-600*F for 7 Days in 35 0.388 1658 45.5(2) NA(2) CuilG - EN82H cc/Kg H2 and 30*F for 5 Deys in 83 cc/Kg N, I CBS-6 Air 130 NA None 0.386 2877(3) 78.4(3) NA(3) CWilG EN82H i Cs5-5 DPW 200 82-86 550-600*F for 7 Days in 35 0.391 2224 61.8 260 CVTIG * [N82H cc/kg H2 and 130*F for 5 l Days in 83 cc/Kg H, CBS-9 DPW 130 10-12 550 600*F for 7 Days in 35 0.390 2466 68.4 620 CWTIG - EN82H l cc/kg H2 and 130*F for 5 Drys in 83 cc/Kg My Specimen Type: 0.4T Fatigue Precracked Compact Tension Specimen, side-grooved 10% Fatigue precracked with the last 2.5% crack extension < 16 Ksi/in a,ni, is the notch length plus the fatigue precrack length Test Ramp Rate: 10 to 14 lb/hr (approximately 0.4 Ksi/in/hr) l (1) Kpmax was calculated using the equation specified in ASTM E813 for side-grooved specimens and includes a seven point internal average fatigue precrack length. l Kpmax - Load x Function (crack length / width)/ /(Thickness x Net Thickness x Width) i (2) Kpmax may not be representative. Abrupt test shutdown to attempt to achieve shorter LTCP crack length may have modified behavior. ~ (3) Specimen LVDT saturated at this stress intensity. 25

l E l l I } l k 'f b C~ ~ 'I u,'. G+ ,k gl' ? ' + y.g f. / / V ~? $(. + ~... l 1 ', 'r h .,i t ' lR ' A 'N l l '4Tt p s, ~~ p-j .? .t- - s.: g4 p- .m. ',(n'. '. p~.p 3.. 1 .c.,. ( :5, n -...1 u p' .\\ '.j \\ 7' 4 \\v, Y - . 7. r c m.s;.' "$ y f ' l 'r '? ;I l {l l W

r. i -{4,.,4 1.;

,,d m

• A l

.. s.., ;, J, ', > - s I 8 - a g,.p'"'}y,4.pf{s9-ig

.9 9-b h,.k l

s ~ / .c i. 1 /. v "v T y SEM PHOTOMICROGRAPH AT 11 ' h

1 a 4 ...T.'....g. .,'g,, ."*\\,' . as..L' 1 Ef hMl, 1 y I :' ]%' } fTf,Q.W((r I ( f. ] ', L. [.,. 4} gf.; l 1.11.l.x))&. y D,'! ~' .~h 4 *19q(0... ?9r..e.t:$[A~rp<< The ~ 'M-df ' i / h) v ~ s. v ~ ;k'. ' '.' ;'.* p n y k> x. v. w: x.,,, t f 4,{;.pn ~[6. t. Jtg gy.?' $..w.').p,? ~,i a..a 4,. 9 ~ 7 '; j,< - s-p y AREA A L.s. u..:' w :.. r, z a

r

" 3 ~. '...n f'- j,,.>.,.; '.y_ t...' ..)2, c 7 f,Q L ~> y .. i -c-a _.. .j > w.a.u;, .,s -. : 3 f t.,,y A,. '5l' ,fr ~ A, . s,E.4% .A ., j -fR* .q r y -- u } w:} 4- ,/ h ' j..r,' ~ M( 4~ I. EI N,, A [. i AIR FRACTURE e, .r ~i. ,TY. ' AFTER TEST . e 'r.5f , Q'; ',fg,' f +

• <n>

.... w., 1 ':, -.., ~ '. ?;. *.

[,

Iy:-, %e ;.;. :. /q.(,.., y.5. ;, 9 i _(-.o_ >-. / ;.. 3. : gg;,o e., ,e i.* cl ;}, }.[. w q:. y - m.,h%.;:. - ~ . ~, l ; y f 4..,. ,-;V y '. f?.:. v, ;....m._. a f.] y. ~ . 1. m;: ~ './' ' ^ Qd',t.b :. " y ~%_,.. SEM PHOTOMICROGRAPH AT 13.5X I,

4. ALLOY 600 PIPE WELDMENT STRESS CORROSION TEST r Axially loaded pipe specimen Nine 3" diameter Schedule 160 Alloy 600 pipe test segments Two heats of piping (good and bad microstructure) Total of 10 welds l 8 EN82 welds (2 low,2 medium,4 high carbon content) l 2 EN62 welds e

h Pipe Cover Test Adapter se0ments j Number 9 8 7 6 5 4 3 2 1 0 iF W iF iF iF iF W W

F W

7 AL AL AL Ak Ak AL AL AL AL AL Endcap EN82 EN82 EN82 EN82 EN62 EN82 i Assembly High Low Medium Hi h Hi h 0 G Carbon Carbon Carbon Carbon Carbon Cross sectional (schematic) view of test pipe assembly Test Parameters Autoclave pressure 2850 psi (20 MPa) Water temperature .680 F (360 C) Hydrogen overpressure 40-60 cc/kg L Axial load plus pressure 20.3 ksi (140 MPa) Axial tensile residual stress 30 ksi (207 MPa) Test duration 78 weeks 29

'W

SUMMARY

OF Post-TEST EVALUATION OF PIPE WELDS' 5 dtd Metatl * * ~ Heat Affected Zone'- J' ease Metal ' Weld. Met! 'PenetranhfestResults .ULOasonic' Test? f V N o.

Sect.

4. ^Results e 1 A 1 ciretnferential linear No indications No cracks 2 cracks No cracks Indication,1.0" tong, root (U) > 0.100" long, 0.015" deep (U) 0.013" deep (U) 1 B 8 tongittMinal linear No indications 1 tongitudinal crack No cracks No cracks Indications: (5) 0.125" Long, 0.020" long, 0.0075" deep (3) 0.063" tong, weld metat 1 C 6 tongitudinal linear No indications No cracks No cracks No cracks indications: (4) 0.031" Iong, (2) 0.125" Long, weld metet 1 0 1 circumferential linear No Indications No cracks 1 crack - < 0.001" deep (U) No cracks indication, 1.0" long; 3 tongitudinal linear indica-tions, 0.125" tong, weld metet i 2 A Lack of fesion, 0.700" long No indientions No cracks 1 crack - > 0.080" long, 0.024" No cracks I (A), scattered porosity deep (U) 2 B Scattered porosity No Indications 2 cracks 1 crack 0.020" long, 0.005" No cracks > 0.060" long, 0.004" deep deep (U) > 0.050" long, 0.006" deep 2 C Scattered porosity No Indications 'l crack

• 0.030" long, 2 cracks No cracks 0.009" deep

> 0.030" tong, 0.002" deep (U) > 0.010" tong, 0.001" deep (U) 2 E Scattered porosity 1 Indicat{on No cracks No cracks No cracks weld metal 3 8 Scattered porosity No indications No cracks No cracks No cracks 3 C Scattered porosity No Indications No cracks No cracks 1 crack - 0.002" deep (A) 3 0 Scattered porosity No indications No cracks 2 cracks 1 crack - 0.002" deep (4), 0.001" deep (A) counterbore 0.007" deep (U) 4 8 Scattered porosity No indications 3 cracks 2 cracks 1 crack - > 0.060" long, > 0.020" tong, 0.003" deep > 0.060" 1ong, 0.001" deep (U) 0.016" deep ( A), > 0.020" tong, 0.002" deep > 0.040" long, 0.001" deep (A) counterbore 0.002" deep i I 30 -., _, _.... --~ ._.-,.--..,m---_ .,. ~ - ...,.c. r-. .-c ..m ~v

SUMMARY

OF Post-TEST EVALUATION OF-PIPE WELDS (CONTINUED) Weld

Met.

PenetrantTestRAsultsi ~ Results1 Uttresonic test' Veld Metal-Heat Affected Zone Base Metal No. ' sect. 4 D Scattered porosity 1 Indication No cracks 2 cracks No cracks weld metal > 0.020" long, 0.001" deep (U) 0.001" deep ( A) 5 B Scattered porosity No indications No cracks I crack - > 0.020" tong, 0.003" No cracks deep (U) 6 A Scattered porosity No indications No cracks 1 crack - 0.001" deep (U) 1 crack - 0.001" deep (A) 4 6 C Scattered porosity 1 Indication 1 crack - 0.0006" deep (A) 1 crack - > 0.04" tong, 0.006" No cracks HAZ deep (U) 6 0 Scattered porosity 1 Indication No cracks I crack - > 0.060" Long 0.007" 1 crack - > 0.080" Long, HAZ deep (U) 0.007" deep ( A), comterbore 7 A Lack of fusion, 4.1" tong (A). No indication No cracks 1 crack - > 0.060" long, 0.009" No cracks Linear porosoty, 2.1" long deep (U) 7 8 5 diagonal linear Indications, No irxfications. 2 cracks I crack - > 0.060" tong, 0.024" No cracks 0.060" 0.120" tong, weld metal 0.002" deep deep (U) 0.002" deep 7 C 1 diagonal linear Indication, 1 Indication 1 crack - 0.001" deep 2 cracks No cracks 0.125" long, weld metet. base metet > 0.060* tong, 0.012" deep (U) Linear porosity, 0.700" Iong > 0.060" tong, 0.007" deep (A) 8 B Scattered porosity 1 Indication 2 cracks 2 cracks 1 crack - > 0.080" long, HAZ 0.003" deep > 0.080" tong, 0.012" deep (U) 0.011" deep (U), 0.001" deep > 0.080" tong, 0.006" deep (A) counterbore 9 A No Indications No Indications No cracks i crack - 0.020" long, 0.001" No cracks de=p ( A) 0 8 No Indications No indications No cracks No cracks No cracks NOTEst track length was established by grinding and polishing incrementally tsitil the crack disappeared. If no crack tength is shown, the crack had disappeared after the first increment (usually 0.020") was removed. Att cracks are circumferential unless otherwise indicated. (A) Side of weld with acceptable microstructure base metal - Nx8908 (U) side of weld with tasacceptable microstructure base metal - N18913 Pipe wall thickness = 0.437" nominal 31 I

t

SUMMARY

OF TEST RESULTS SCC RESULTS (NUMBER / DEPTH - MILS)- Weld No. Weld Type / Insert Weld liAZ Base Metal ' Carbon Carbon 1.& 2 EN62/.047% .030% 4/8.8,8,7.5,3.8 Accept M/S - 0 Accept M/S - 0 Unacc M/S - Unacc M/S - 0 7/23.8.15,12.9,5, 2.1,1.3,0.4 3&4 EN82/.084 .084 3/3,2,2 Accept M/S-Accept M/S - 3/1.4.1,1 3/16,2,1.5 Unacc M/S-3/7,1,1 Unacc M/S - 0 5&6 EN82/.023 .027 1/0.6 Accept M/S - 0 Accept M/S - Unacc M/S - 2/6.5,1 4/6.5,6,3,1 Unacc M/S - 0 7&8 EN82/.010 .015 5/2.8,2,1.8,1.2,.9 Accept M/S - 2/7,6 Accept M/S - 0 Unacc M/S - Unacc M/S - 1/11 4/23.5,12,11.5,8.8 2 9&0 EN82/.084 .027 0 Accept M/S - 1/0.9 Accept M/S - 0 Unacc M/S - 0 Unacc M/S - 0 Notes: 1. Except for 1 mil deep crack reported for weld 6 and 2 mil deep crack in weld 3, all base metal cracks were associated with discontinuity at counterbore. M/S - microstructure rating per Bettis Gallery. 2. Welds 9 & 0 were structural welds attaching the pipe test section to adapters. 32 I

Alloy 600 Pipe Weldment Stress Corrosion Test - Continued Results SCC initiation observed in weld metal and HAZ of each weld Location Maximum Crack Depth (mils /mm) Weld Metal 8.8/.22 HAZ 23.8/.60 Base Metal 16.0/.41 EN62 and low carbon EN82 had highest propensity for weld SCC Worst HAZ SCC occurrgbase metal with bad microstructure HAZ and base metal cracks initiated from stress concentrators 33

Alloy 600 Pipe Weldment Stress Corrosion Test - Continued Conclusions SCC can occur in the weld and base metal of Alloy 600 pipe Good microstructure produces better HAZ SCC resistance Stress concentrators can be preferred SCC initiation sites SCC resistance of EN82 is better than EN62 SCC resistance of high carbon EN82 is better than low carbon EN82 34 ~.

Alloy 600 Pipe Weldment Stress Corrosion Test Lotigitudinal Section of Pipe Weld l OUTER DIAMETER ( l l l l INNER DIAMETER / / ./ \\/ COUNTER 80RE WELD METAL HEAT AFFECTED CRACKS CRACKS ZONE CRACKS l l 35

c .,7 6-General Electric Co. g Knolls Atomic Power Lab Weldability and Properties of EN52 Filler Wire WELDAEILITY AND PROPERTIES OF EN52 FILLER WIRE j D. A. SCHAEFFER AND A. ECHEVARRIA 1992 EPRI WORKSHOP ON PWSCC or ALLOY 600 IN PWRS NCTICE This report was prepared as an account of work sponsored by the Uaited States Government. Neither the United States, nor the United States Department of Erergy, nor any of their employees, nor any of their contractors, subcontractors, or their employees, makes any warranty, expressed or implied, or assumes any. legal liability or-respnsibility for the accuracy, completoness or usefulness of any informttion, apparatus, product or process disclosed, or represents that its use would not in#eitge privately owned rights. l KNOLLS ATOMIC POWER LABORATORY SCHENECTADY, NY Operated for the U.

8. Department of Energy by General Electric Corportion f

I

W e General Electric Co. g Knolls Atomic Power Lab Weldability and Properties of EN52 Filler Wire WELDABILITY AND PROPERTIES OF EN52 FILLER WIRE Abstract Results from work to develop welding procedures for A690 weld filler wire (EN52) are summarized. This work includes weldability tests, mechanical property determinations, ] microstructural characterizations and development of a temperbead welding process. The results to date generally indicate good weldability and accesptable mechanical properties. Further work maybe needed to more closely define acceptable composition limits for EN52. h e

General Electric Co.. g Knolls Atomic Power Lab Weldability and Properties of EN52 Filler Wire 3 OBJECTIVE: To EVALUATE THE WELDABILITY OF MECHANICAL' PROPERTIES A690 (EN52) FILLER WIRE V-GROOVE WELDABILITY TESTS (A690 AND A625 BASE PLATE) y l TEMPERBEAD OVERLAY PROCESS + w I t 9 - 4 . - - ~.. -...-e.- ,,--..~.. ..~r....,--..,...._m., .....-.:-..,~,---...

General Electric Co. g KnoIIs Atomic Power Lab Weldability and Properties of EN52 Filler Wire RESULTS: l EN52 FILLER WIRE WAS SHOWN TO BE COMPATIBLE WITH:

A690, A625 USING AGTAW

SA508, EN82 AND A600 USING TEMPERBEAD PROCESS.

MECHANICAL PROPERTIES OF EN52 ARE COMPARABLE TO WROUGHT 690 t MICROSTRUCT JRE OF THE SA508 HAZ BASE MATERIAL USING THE TEMPERBEAD PROCESS IS TEMPERED MARTENSITE

General Electric Co. g Knolls Atomic Power Lab Weldability and Properties of EN52 Filler Wire i 1 l EN52 WELDABILITY AND PROPERTIES WELD CHARACTERISTICS OXIDATION l FLOW ADHERENCE (WETTING) MICROSTRUCTURE EVALUATION l WELD' DEFECTS l POROSITY l LACK'OF FUSION i HOT CRACKS CONSTITUENT VARIATIONS OBSERVED CARBIDES WELD METAL DILUTION I y,v v t -w,- r w-v%- -erw-d=--

• e 4w

~s--*------d<* ,re,-vw, tm- -*-1

• m4==~--v

~ t u -- * * + * -'- - - ' e-' iw-" * -*-- - =. e ar ^--*e--= w w w e-s e '- w--v'1----- .*w,-

General Electric Co. g Knolls Atomic Power Lab Weldability and Properties of EN52 Filler Wire MECHANICAL PROPERTIES EVALUATED INCLUDED: TENSILE PROPERTIES OF THE AS-WELDED AND POST WELD HEAT TREATED (PWHT) A690 FILLER METAL THE EFFECT OF THE TEMPERBEAD PROCESS ON-THE PROPERTIES OF THE BASE MATERIAL l MICROHARDNESS AND GRAIN SIZE OF THE HAZ AND FUSION ZONE FOR BOTH V-GROOVE AND TEMPERBEAD. WELDS 4 4 ,,__....__._...,.._m .._,...,,,,..-_.__,_..-...__..._.-,..__.s...

General Electric Co. g Knolls Atomic Power Lab Weldability and Properties of EN52 Filler Wire ALLOY CHEMISTRY KEY DIFFERENCE FROM WROUGHT 690 MATERIAL IS THE AL AND TI ADDITIONS AL AND TI ADDED TO INCREASE FLOW AND ELIMINATE HOT CRACKS EN52 WELDABILITY WELDING IS COMPARABLE TO EN82 0XIDATION OF THE WELb METAL HAS BEEN REPORTED AND ATTRIBUTED TO THE AL k EXCELLENT SHIELDING Is REQUIRED To MINIMIZE OXIDATION OF WELD METAL, DUE To AL, TI ADDITIONS ~ --e,- yw-<-upwe. m-e- +t="'- ma^ 9M9* * * --"

• '9

=t-97"TT e'M

• 'M 7*"1 W'9-9'M
• '-'8 Y"'*
• 14-**t:

g-y-b+ ?-,g%-hT

==34 e e r~ p t1*b ?-e ? P su hD-

• -W

-*'N W-W'e e- ?"e-ww'-usW -

~ General Electric Co. g Knolls Atomic Power Lab Weldability and Properties of EN52 Filler Wire _ CHEMICAL COMPOSITION OF EN52 FILLER WIRE ELEMENT HEAT No. Y9378K 4 C 0.03 MN 0.23 FE 8.98 S 0.001 SI 0.17 Cu 0.01 NI 60.37 CR 28.97 AL 0.63 TI 0.56 P 0.004 8 0.001 Ps 0.001 N 0.007-OTHER 0.50 ELEMENTS t sm-r-s- -e r -m---w ee r--, ar e,r H. +~nn-w, ---w- -,w -r --e+--e --m + -mm-w w

~ General Electric Co. g Knolls Atomic Power Lab Weldability and Properties of EN52 Filler Wire V-GROOVE WELDABILITY TESTS A690 MATERIAL WELDED WITH EN52 FILLER WIRE q

== A690 A690 5/8' = I E t A690- - Current =115-125 amps Voltage =12-15 volts - Filler VIre Dia.=0.062' Preheat Temp =60F i Interpass Temp =350F . t_ .--..,5- .-...s- .-.-4. .,c,- --+,w= =s -..---e-#+4-w -- v-ww-w.-m....w-tm ,r.-- , -. ~, -. m..- mmm.-.

~ General Electric Co. g Knolls Atomic Power Lab Weldability and Properties of EN52 Filler Wire V-GROOVE WELDABILITY TESTS A690 PLATES WERE AGTA WELDED AND METALLOGRAPHICALLY EVALUATED TRANSVERSE CROSS-SECTIONS SHOWED NO DEFECTS 1 HARDNESS AND GRAIN SIZE IN THE HAZ WERE FOUND TO BE COMPARABLE WITH WROUGHT 690 MICROPROBE ANALYSIS IS CURRENTLY UNDERWAY. PRELIMINARY RESULTS INDICATE: MC CHROMIUM RICH CARBIDES ARE AT THE GRAIN Bb3 hDARY AND M C CARBIDES IN THE DENDRITIC ARM U l 3 TI (C,N) PARTICLES ARE IN'THE INTERDENDRITIC f REGION o --ww.- ea-- -r, -e m .me- +.~4--wi we< ye,e- .r-a. ye ,~w1-w- w w w -r v. -wa..,--<#.z w A ..w-r

i e i i 9e, h ,. %%r% Y C' s ,,Q 'y *.)y,. 4 ,& 5 o -j,.54, - d -

• g..%

',. N g . gdQ7.x -r.., ..;_. }g ~..f* -(% pung e +1 N q ,p',. 9 ' ' *k. - .~o 's 1 A g. - <O q s,. m. - . r.8 O h i W D4%yd -m-3 90 iE .o e tr" .m. m s w .f 5+ ~ l4 T h $ b ( g 6 't i ._ = n g a +. ~ ^ p

m__,

==* N M 12 C l l '.hf' 3 i d " e ~~{ e m = 1 5N wwme-g. _ @@f ~ w sE E 75%M o sh-r. e aMhmil.PlE o ~ \\ v ~ ~. ~ " I"" .k $(

• p'. h[

h 0 g A' $ ig f W..;g-}% b,5 E o f ~~ A g & 8

E

-G f 8 1 l 65 3 1! J " 3 F o %.: " 5 a-Ou - g 4psiQkhl5,;:# ^ p Iy-m w. .e s.

r m

G d ;

• 4 h<N.y y...

M**.r n:,, n O h"* e uma i ws< ]* hg 6 &~.f$s :.. b$k i S

1. :.?. $.N' U

% 3._ : -15 l nun u n w %p 3, 4*.% a = FI +.- i, W a EQ., ' j q 'm ~,, /...1 U.9 4 [ A s w #- [4 ::aY':?b. ' d.i*N,.[-/.,C $*'b .[ r. g .u.. . = m$$ 1' 'd. / Wf g e L' ; e.; g,T.~4 G s n g g 5.,y

fh Q

..t .,e h ... l$'??$ \\.% hl_,. l'.O i, L la,C[W 'N

• T*

W. .&&z ;y5.:. "b::te N.,:.:,.1&&A5i 3

General Electric Co. g Knolls Atomic Power Lab Weldability and Properties of EN52 Filler Wire A625 MATERIAL WELDED WITH EN52 22 deg 6th 17-20 ID VH62 ,l ID VM62 5th 14-16 _g A625 3rd 0-10 ,a 2nd 5-7 1st 3+4 A690 A690 A690 60 deg F Preheat 300 deg F Interpass 170 Anps 10 +/ Volts 5 Ipn Travel Speed 40 Ipn Vire Feed Speed 35 CFH Argon shield, (18 gas cup 3/32" dia 2% Thortated tungsten electrode, 9/16' stickout 5 I

General Electric Co. g Knolls Atomic Power Lab Weldability and Properties of EN52 Filler Wire MECHANICAL PROPERTIES OBTAINED FROM AGTAW EN52 ON A625 BASEPLATE Post WELD YIELD UTS EL RA -320*F HEAT TREATMENT '(KSI) (KSI) (%) (%) CHARPY 1 (FT-LBS) 1050*F/7H 65.7 89.9 38 79 206.0. 1600*F/4H +- 51.2 84.7 34 67 141.2 l .1050*F/7H 1800*F/4H + 46.2 88.1 44 64 127.5 1050'F/7H 2000*F/1H + 44.4 89.7 44 61 124.0 1600*F/4H + 10507F/7H l TEST SPECIMENS WERE TRANSVERSE TO WELDING DIRECTION l L __,...._,.....-__..~...._..._..._.__._.;.

m ' ~ General Electric Co. g Knolls Atomic Power Lab Weldability and Properties of EN52 Filler Wire EFFECT OF PWHT-ON-EN52 FILLER METAL PROPERTIES _ .3 ROOM TEMPERATURE TENSILE PROPERTIES FtM THE [ 1050*F PWHT ARE EQUIVALENT TO THE AS-WELDED L FILLER WIRE A REDUCTION IN THE YIELD STRENGTH WAS OBSERVED FOR THE PWHT IN EXCESS OF 1600'F. THE ULTIMATE TENSILE STRENGTH WAS NOT SIGNIFICANTLY AFFECTED BY THE PWHT ALL FRACTURE SURFACES SHOWED A DUCTILE FRACTURE (CVN ENERGIES OF 100 + FT-LBS WERE OBTAINED AT -320"F) t 't m. ... ~. _m-2,, mm m.

t General Electric Co. - g Knolls Atomic Power Lab Weldability and Properties of EN52 Filler Wire t 6 TEMPERBEAD WELDING. PROCESS 3RD LAYER N 1st EN52 OVERLAY + l S A 508 Clo.s s 2 ~ A 600~ 5 y / 9" EN 82 WelCl MetO.l Current =115-125 amps Process requires controlled heat Voltage =12-15 volts input Fron successive layers to Filler Vire Dia.=0.062, obtain tempered nortensite in the Preheat Temp =60F 3A 20 heat nFFected zone WAD. .Interpass. Temp =350F I e -+ + e- --+-c w r

l General Electric Co. g Knolls Atomic Power Lab l Weldability and Properties of EN52 Filler Wire i l WELD TEMPERBEAD PROCESS l SA 508 HEAT AFFECTED ZONE (HAZ) l CHARPY V-NOTCH (CVN) DATA O y300 a m -+- 250 l v g200-E = {150= 100-o. _E 50- $0 e O -350 -300 -250 -200' -150 -100 -50 0 50 100 Temperature (deg. F) CVN DATA SHOWED HIGH TOUGHNESS IN THE.HAZ REGION (NDT AT 35 FT-LBS-IS -152'F) v.- . v. - %,4- +w + +-.,--4w-wm .,r..-.i,,-i--~-- v, _,e.. -.. -, - - _ _, _ _ _ _ - -

~ l General Electric Co. g Knolls Atomic Power Lab l Weldability and Properties of EN52 Filler Wire l TEMPERING EFFECT OPTIMUM TEMPERING ACHIEVED AFTER THREE LAYERS. THE MICROSTRUCTURE OF THE HAZ IN THE SA 508 BASE MATERIAL-WAS TEMPERED MARTENSITE 4: TEMPERING OF THE HAZ MICROSTRUCTURE IS OBTAINED WITHOUT POSTWELD HEAT TREATMENT. L 8 .-,.m,- + - - - - 4 v-- ,.--,,,.,,e--. .~.,,,-4 m- --

General Electric Co. g Knolls Atomic Power Lab Weldability and Properties of EN52 Filler Wire TEMPERING EFFECT OF EN52 m. 45' y Ph jH~ i

,9*==ig l

G *iff~ N 630V ' i\\ ... in,o toye, M \\ \\ + S""*'

• 2 ~<

N 9 Q . n,,i toy., J_ + n,,is.oo 20: 1 15~ ^ ^ io / I v w 5' ? O' .000.010.020.030.040.050.060.070.080.090.100.110.120.130.140.150 DISTANCE from weld fusion line (in) ,~-=*-nm--- - + - -. .s n-- r m v'- = ' - e-war 4+' y 4'-e m-qw-v-r -wm- - - - * - -ee-r- +, =+ -- =i -+ = w +.e--- -,.w-, .,er- - m~-

General Electric Co. g Knolls Atomic Power Lab Weldability and Properties of EN52 Filler Wire TEMPERBEAD EVALUATION RESULTS METALLOGRAPHIC EVALUATIONS REPORTED NO DEFECTS IN THE FUSION 2DNE OR INTERFACE BETWEEN THE EN52 FILLER METAL AND BASE PLATE. SIDE BEND TESTS OF EN52 FILLER METAL SHOWED NO CRACKS AFTER 20% STRAIN THE DILUTION OF THE FIRST WELD LAYER OVER SA 508 WAS DETERMINED BY MICROPROBE ANALYSIS WEIGHT % CR NI AL TI FE-SI A 690 27-BAL 0.5-0.2-7-11 0.05-SPEC 31.5 APP'. 56 1.1 0.7 0.30 FIRST 12.80 24.48 0.23 0.03 61.8 0.28 LAYER-(Ave) l l vw,a e.-.<

,wa w

...c.w..

• e

---

w m.

.a..... i.-,-

---

wra

>-,,-e .1r- ++a.-.w.-,- vs... -,--..,--,._w-..~,--w...i e .e-w-m--ee.e- - e - ne-o ..%-6-e

General Electric Co. g Knolls Atomic Power Lab ~ Weldability and Properties of EN52 Filler Wire ~ TEMPERBEAD MECHANICAL PROPERTIES i l Material Orientation Temp 2% Y.8. UTS Failure

• F Elongation kai kai Location 4

EN52 & axial 76 38 58.5 86.8 A600 A600 EN52 E axial 76 36 59.5 86.9 A600 A600 EN52 transverse 76 40 57.9 84.3 Weld EN52 transverse 76 43 59.5 89.9 Wald EN52 transverse 76 42 64.4 88.1 Wald m A490 Smte Bend Specsnen , pe tch centered se Aant h spa *wa Tremwree M apecswo p-Lse or R-tey 1 s e P 4 .__,-e.- ww --4,-,-,4-,~,---,.ww-r. m ,,n-3.--w-w--- ,y-- e - ..,_r~.. .v i. e. w ..m.---=e.w...e a, - -,. + -. = -.--e..,. -,--_-e--*= .~ u --. -m---

General Electric Co. g Knolls Atomic Power Lab Weldability and Properties of EN52 Filler Wire -CONCLUSIONS EN52 FILLER WIRE HAS BEEN SHOWN TO BE COMPATIBLE WITH . WELDING A625 AND A690 USING AGTAW AND WITH.SA-508, EN82 AND A600 USING TEMPERBEAD-PROCESS MECHANICAL PROPERTIES OF THE EN52 WELD METAL ARE COMPARABLE TO WROUGHT 690 NOTCH TOUGHNESS (CVN) OF THE EN52 WELD METAL IS EXCELLENT GOOD MECHANICAL PROPERTIES ARE OBTAINED USING A TEMPERBEAD PROCESS O 4 ..--,-,.-,--.---,,e~.+.. -v wn-..-- ---.,e + -m, e +,,... - ---n---

General Electric Co. g Knolls Atomic Power Lab Weldability and Properties of EN52 Filler Wire i CONCLUSIONS (CONT'D) THE OPTIMUM AMOUNT OF TEMPERING IS ACHIEVED IN THE -THIRD LAYER MICROSTRUCTURE OF THE SA 508 HAZ BASE MATERIAL USING THE TEMPERBEAD' PROCESS (CONTROLLED HEAT-INPUT) IS TEMPERED MARTENSITE s (AltNCifA 6rdE7 N 36% LA-16o 'fJ 44' k' PM)Ol b r }{Wtf Cfhte 4 O Thf WWWK EVLM N etM Ltmtr Ns C tl.ntKs% No 3PGt It a c r G t V ed 5 C lb6 f ft 5 A P S ns, Q % [LM& L tro L44W 4ef. & .t% L i ~. ,...,... _. =. _.. ~..~....

3 Q;{ \\ I tE 33 -1$ dff. Ii 4g Proceedings 4) of the { Second International Symposium s 1il on .I.M ENVIRONMENTAL DEGRADATION OF w ly MATERIALS IN NUCLEAR POWER i& SYSTEMS-WATER REACTORS or 3i !b

a d b.

Monterey, California rv ty September 9-12,1985 b $' <a ) IA

y j

TMS 'f Sponsored by AlME ag'qp American Nuclear Society i y; Metallurgical Society of the AIME ANS NACE lh National Association of Corrosion Engineers i i elw[L l l-iiE( q \\ h'O D Me 086i h S >N A / l: Published by the 3/77g.g y df/ 9 m a };( American Nuclear Society ),7 La Grange Park, Illinois 60525 USA tt 7 I ll ~

's .Q i

• e.
• s.

l AN EVALUATION OF THE SCC BEHAVIOR OF INCONEL ALLOY 690 c a WELOMENTS IN A SIMULATED BWR ENVIRONMENT 'e** b' S. FLOREEN, Knolls Atomic Power Laboratory ./ J. L. NELSON, Electric Power Research Institute Box 1072 3412 Hillview Ave. Schenectady, NY 12301 o Palo Alto, CA 94303 (518) 393-6611 +#*, la (415) 855-2825 h 4 9... In practical terms, by use of higher chromium ABSTRACT weldment chemistries and gas tungsten arc weld-ing, the susceptibility of welds to failure j The purpose of this program was to evaluate the could be considerably reduced.

• ap stress corrosion behavice of alloy 690 weldments

,4,l in an accelerated BWR environment. Four differ-INTRODUCTION ent weld chemistries corresponding to nominal +,,e alloy 600, 690, 625, and 671 type compositions Stress corrosion cracking (SCC) is of were prepared as shielded metal arc welds and concern in various boiling water reactor (BWR) g[ gas tungsten arc welds. Ease plate specimens of components and considerable efforts have been alloy 690 and alloy 600 also were prepared. expended to determine the causes of these prob-f,,f lems and to minimize failures in service. Thus g*4# Single and double U-bend specimens were exposed far, relatively little attention has been paid 40 weeks in pH 4.5, 6 ppm oxygen water at to the behavior of weldments in these systems.

• [k 316*C. The pH of the environment was adjusted The incidence of stress corrosion failures in to 4.6 by the addition of sulfuric acid.

welds has been small compared to problems in g+#+h wrought structures, but the potential for Af ter the exposure, the majority of the weld-trouble exists and deserves attention, ',I. W ments had failed. Cracking was confined to the k weld metal. There was a strong effect of the In the present study an examination has chromium content of the weld on the propensity been made of the SCC susceptibility of sixteen >{O p," to failure. In welds with less than 24t Cr, 31 different welds madg with either low carbon or d out of 32 samples failed. At higher Cr con-high carbon INCONEL alloy 690 base plates. Tv tents, only 8 out of 32 specimens failed. In additional alloy 600 welds plus additional bast 4 these latter welds, 7 of the 8 failures occurred plate and control specimens were also exam-6, C ' L in single U-bend specimens of shielded metal arc The aim of the study was to characteriz' ined. ^ wel ds. Only one gas tungsten arc weld specimen the behavior of different types of weld chemis ^

• a i

f ailed. None of the base plate 690 or 600 tries and two welding processes in an acceler-specimens failed. Thus welds appear much more ated BWR environment.

7,

-' : 4 prone to cracking. EXPERIMENTAL PROCEDURE Micrnprobe examinations of the weldments did not -a 1 /

• • • l reveal measurable microsegregation of chro-Base Plate Preparation

~V / Based on studies of other alloys in the mium. [ / literature, some finer scale microsegregation 6f Two types of alloy 690 base plates contai chromitn would be expected, but it is argued ing different carbon contents were used. The that simply segregation of chromium per se would low carbon base plate was taken from a 31,700 ' + In H not explain the present results, heat made at Huntington Alloys as 12.7 mm thic [ plate that was mill annealed at approximately This material is typici 1010*C for 0.5 hours. f fl commercial material. a ' t I This work was carried out while the authors a !j were with the Inco Alley Products Co. Research birademark of the Into family of companies. Center, Sterling Forest, Suffern, N.Y. d} r u, i

f i i l i No commercially prepared high carbon plate were observed by radiography. However, areas of l stock of alloy 690 was available. Therefore slight undercut were detected by the liquid-several 45 kg laboratory vacuum induction heats penetrant inspection. These defects were not were melted and hot worked at 1250*C to 19 mm considered significant and were removed during l thick plate stock. These plates were then sample preparstion. j solution annealed for 0.5 hours at 1175*C, air cooled, cold rolled to 14 mm thickness and then Cross-sections of the weldments were exam-annealed 0.5 hours at 1010*C and air cooled to ined optically and by side-bend testing. All l simulate the mill anneal used comercially. sections of the weldments from which the test specimens were to be machined were found to be i The composition of'the alloy 690 materials sound. The bend tests revealed no defects and l tested are given in Table I. Also listed are all weldments exhibited excellent ductility. the compositions of a commercial heat of l Type 304 stainless steel and a low-carbon labo-Sample Preparation and Testing ratory heat of alloy 600. Both were added to j-the tests as controls. While it would have been Tensile specimens were machined perpendicu-more desirable to compare commercially prepared lar to the weld beads so that the weld metal was high and low carbon alloy 690, earlier stress located in the centers of the test sections. l, corrosion studies carried out during the The tensile specimens were 6.3 m diameter with development of this alloy had shown no signifi-a 31.7 m cylindrical gage length. Two speci-cant differences in SCC behavior between lab and mens were machined from each weldment. commercial heats with comparable chemistry. e

{

Furthermore, as discussed below, SCC in this Stress corrosion samples were 12.7 m wide, ii study was confined entirely to the weld metal, 3.05 m thick single and double U-bend speci-so variations in processing history of the base mens. These samples were machined perpendicular lA ; i E plates should not be crucial. to the weld beads such that the tension face was normal to the weld direction. The double U-bend [ Weldment Preparation configuration was designed to locate the weld metal beneath an outer U-bend of alloy 690 so s t Four different types of weld chemistries that a crevice condition existed at the tensile Ri were evaluated, corresponding to the nominal surface of the weld. Two single and two double W compositions of INCONEL alloys 600, 690, 671 and U-bends were tested from each weldment and also i i l 625. For each type of alloy chemistry a gas from the various base plate materials. tungsten arc weld (GTAW) and a shielded metal a f arc weld (SMAW) were evaluated. Table 11 gives Weld metal chemistries were determined from j , a summary of welding materials. It will be drillings taken in the central portions of the % noted that five of these materials are com-welds. These results are shown in Table IV. y mercial products, while the remaining three are The chemistries are generally what would be ~ f ~ experimental materials. Experimental in this expected for the various consumables, and no 5 case means that these products were developed in significant dilution effects from the base r iearlier laboratory studies, but they have not plates are evident. 'been sold comercially. Stress corrosion tests were conducted in a i i 1 For control purposes two additional weld-two-gallon alloy 600 autoclave. Tests were run ents were prepared with alloy 600 as a base at 316*C in an accelerated BWR aqueous environ-plate. Only two welds were prepared with this ment containing 6 ppm oxygen and acidified with H 50a to a pH of 4.6. As discussed by Andresen base plate, one with INCONEL Filler Metal 82 and 2 0,ne with INCONEL Welding Electrode 182. and indig(1), this type of chemistry is repre-sentative of the contamination produced by a T '" Table III sumnarizes the procedures used breakdown of the ion exchange resin beds in a Q r making the weldments. The base plates were BWR. The samples were tested for a total expo-chined to 12.7 mm thickness with an 80* sure of 40 weeks, with periodic shut-downs to _luded angle joint. The welds were made at examine the specimens for f ailures. At the travel speeds and any oxide layer on the completion of the 40-week exposure, the unbroken ._facefoftheweldbeadswasremovedbetween specimens were mechanically pinched to open up . asses. - No preheating was employed and the any

• nonvisible" cracks. As discussed below, a

_imum' interpass temperature was 90*C. The large number of cracks were detected after ase ' plates were restrained by six straps during pinching. lding. The welding parameters were kept .nstant during preparation of the various In addition to weldment and base-plate ,t s. specimens, additional general control samples i were added to test the efficiency of this envi- .' eld soundness was verified by radiographic ronment. These s'amples are sumarized in W quid-penetrant inspection. No defects Table V. They include an alloy 600 base plate s 'S i w

-.m.. . _5Y/8thi; W _3 6sJL1rM e ~@ -

• eurm -

i.o ) ~ in welds containing more than 24 wt.% > di welded with welding electrode 182 in which the chromium only 8 out of 32 specimens failed. alloy 600 plate was given a low temperature heat Seven of these failures occurred in shielded ( o treatment of 200 hours at 400*C af ter mill metal arc welds, and all of these seven speci-annealing. Type 304 stainless steel specimens mens were single U-bends. Thus both weld 44 with two dif ferent low temperature heat treat-process and specimen configuration played a 5 i ments were also included in the test. significant role at higher chromium contents. l The greater propensity for f ailure in the SMAW I RESULTS ANO DISCUSSION samples compared to the GTAW ones is not too surprising in view of the coarser solidification Mechanical Properties structure generally produced by the SMA pro-The significant distinction between the cess. The results of the tensile tests of the single and double U-bends is noteworthy, sug-base plates and weldments are summarized in gesting that the oxygen content was a critical fable VI. The values are the averages for variable in the behavior of the higher chromium The individual values in 3 duplicate specimens. SMA welds, g each pair of samples were generally in close The properties for Of the higher chromium content GTA welds, agreement with each other. both the weldments and base plates are in good only 1 sample out of 16 was cracked after j This particular sample was a double agreement with handbook values for these materi-40 weeks. Quite clearly the alloy 671 The tensile strengths of the weld metal U-bend specimen. als. were usually somewhat less than those of the and alloy 690 type chemistries made by GTA l base plate and the tensile failures were com-welding appear to offer significantly improved The ductili-SCC resistance in this type of environment. monly located in the weld metal. ties in all cases appeared satisfactory. A number of the failed specimens were Stress Corrosion Results_ sectioned and examined metallographically to study the fracture behavior. Cracking in all The results of the stress corrosion tests cases was within the weld metal, and no cracking 7 are summarized in Table VII. As expected the was detected at the fusion line or in the base Many samples showed multiple small Type 304 stainless steel specimens failed i j plates. cracks throughout the weld metal and there was No cracking was rapidly in this environment. observed, however, in any of the alloy 600 or no indication of any preference for cracking alloy 690 base plate samples after the 40-week near to or remote from the fusion line region. Thus local chemistry change due to dilution from

exposure, the base plate does not appear important to the Numerous cracks were observed in the weld cracking propensity.

samples, particularly after pinching the speci-mens at the end of the 40-week test period. Figure 2 shows an example of typical stress One striking result in the weldment data was the corrosion cracks. The cracks in most cases ]f strong effect of the chromium content in the appeared to follow an intergranular or inter-weld metal on SCC resistance. Figure 1 shows cellular path in the weld metal. Crack branch-the data for the welds made with alloy 690 base i ing was occasionally seen but was not comon. d plate in terms of the number of failed specimens Figure 3 is a scanning electron micrograph of E The intergranular or after 40-weeks versus the chromium concentration typical fracture surfaces. of the weld metal. Of the 32 welds containing intercellular nature of the stress corrosion 3 less than 24 wt.% chromium, 31 failed. path is again evident. 4 y It should also be noted that the two welds Electron microprobe scans across the weld t metal in several samples did not reveal any made with alley 600 base plate showed 100% f ailure, i.e., all 4 specimens f ailed.. The localized enrichment or depletions in either the [ chromium contents'of these welds were also less principle alloying elements or suspected m - m Thus the local segregation g L than 24 wt.%. Thus, regardless of the base h impurity elements. plate chemistry, whether single or double in the -cast structures is on a finer scale thanj U-bends were tested, or GTA versus SMA welding could be resolved by the 1 to 2 um microprobe C Localized partitioning on a finer scale 2 processes, welds containing less than 24 wt.t chromium were extramely prone to failure, would be expected, and probably could be g-4s beam. g i resolved by scanning transmission electron l These results may be contrasted with the microscopy (STEM). J absence of cracking in the alloy 600 base plate z,,-o n The present heat of alloy 600 con-A STEM study by Brooks, et al..(2) of 1. i. samples. tained 15.6 wt.% Cr. Thus, if a mininum c5co-Fe-21Cr-)4Ni GTA welds showed local variation y mium content is required for cracking rests-in Cr content on the order of 12% between the h Simila d tance, it is evidently much higher in the weld-cell boundaries and cell interiors. y f ments than in the base plate, 5 6 ? d

l 1 studies in Ni-base GTA weldments evidently have associated with the greater microsegregation not been performed, but judging from the solidi-tendencies of this type of weld structure, fication diagrams for the Fe-Ni-Cr system it Further work clearly is required, however, to seems unlikely that the local variations in establish the reasons for the present weldment chromium content would be of significantly chemistry effects. greater magnitude. In the practical sense the message seems If this conclusion is correct, the strong quite clear. Lower chromium fillers are effect of chromium content in the present extremely sensitive to SCC in the present envi-results cannot be explained solely on the basis ronmen t. Although this environment is not of localized depletion of chromium. That is, typical of normal BWR operation, it is charac-the alloy 600 base-plate results showed that an teristic of an excursion that could be encoun-alloy with a uniform chromium content of tered in service. Quite clearly it would be 15.6 wt.1 resisted cracking. Thus if crack *ng wise to use higher chromium fillers and GTA is attributed solely to chrcmf um depletion, the welding when making welds for BWR service. critical local level of chromium must be less than 15.6 wt.% and possibly much lower than this ACKNOWLEDGMENTS value. However, based on the STEM results dis-cussed above, it seems unlikely that the filler This work was funded by the Electric Power metal 625 GTA welds containing an average chro-Research Institute under contract RP1F66-2, mium concentration of 23.5 wt.% would contain Project Manager Dr. W. J. Childs. The con-local areas where the chromium was depleted to siderable assistance of E. P. Sadowski in pre-less than 15.6 wt.%. paring the welds and of I. J. Magar in carrying out the autoclave t?sts is greatly appreciated. Thus the strong effect of chromium shown in Figure 1 appears due to something beyond or in REFERENCES addition to just local depletion of chromium. Local partitioning of other elements in colla-1. P. L. 'Andresen and M. E. Indig, Corrosion, a boration with chromium depletion may be respon-Vol. 38, No.10, October 1982, pp. 531-541. sible for the cracking susceptibility of the + lower chromium weldments. In the weldments with

2..

J. A. Brooks, J. C. Williams and A. W. h more than 24 wt.1 chromium the greater cracking Thompson, Met. Trans. A. Vol.14A, January g propensity of the SMA welds is also very likely 1983, pp. 23-31. 2 1 ) tao s ! TaDie !! (se gnt Perce ts i, l SCO 3C? ~ C an Fe tr ' $1 al Tf 5 ei y Hea t to. C at tov 690 reial 0.020 0.14 8.4 26.6 0.30 0.27 0.30 0.002 Bal IE E la g trod 182 Coated lutrode 3-125 Emp. Filter Altog 690 Erperimental C( M4a C-a11ov 690 R-135 Esp. Costed Electrode $abora tory 0.051 0.18 9.5 29.8 0.13 0.45 0.28 0.004 Bal !<0NEL Filler stetal 72 Filler Allog 671 Commercial c""'""'- 5.;

= ::n ::' !!:: :n ::n ::! ifJ :

O IEONCL Ff11er Itetal 525 Filler Allog 625 Cosmekiel g ~ ICONEL Welding Electrode 112 ' Coated Electrode ,e tory C.016 0.17 7.3 15.4 0.16 0.25 0.25 ' 0.003 841 Type 30s staWess steel etial

0. 04 $ 1.58 Bai 19.07 0.07 0.002

<0.01 0.012 8.55 i; li s. k N. x-N 7 2

%i Table III WELDING PARAMETERS +

1. q1 Gas Tungsten Arc INCONEL Welcing Electrodes 182 Shielded Metal Arc

? l INCONEL Filler Metals and 112; R-128 Exp. and R-135 Exp. 82, 625, 72 and R-127

• 9 Consumables a

3.2 m (0.125") } 1.6 m (0.062") Consumables Dia. DCSP DCSP )* Polarity 24 I *, } 10 85 Voltage 220 I 95 i Amperage 1st ps 230 13 plus sealing bead i Amperage Others 13 l 3 No. of Passes 3.2 m (0.125") Tungsten Dia. 35 l.*'- Shielding Gas (cfh) None None Preheat 90 90 ax. Interpass Temp. (*C) 6 U-straps I 6 U-straps Restraint 40 Travel Speed (m/ min)(minimum) (@"*. c Wire Feed (mm/ min)(minimum) 0,45 m (0.020") Root Space 450 80* Included Angle; Joint Design ts{. 0.70 m (0.031") Root Face 12.7 x 152 x 305 mm 12.7 x 152 x 305 m p (0.5 x 6 x 12 in) (0.5 x 6 x 12 in) Plate Size Table IV I*Jy CHEMICAL COMPOSITION OF VARIOUS WELO DEPOSITS y 1 Composition. Wt.% (Ni Balance) Alloy Filler Metal Process Cr Mn_ Si Fe Ti Al CD Mo C S i Weldment . Welding i 0.035 0.006 'i~ 690A 182 SMAW 16.5 6.4 0.50 8.7 0.42 1.6 0.034 0.006 1.6 690B 182 16.1 6.5 0.53 8.8 0.29 1.6 0.031 0.006 600 182 15.4 6.4 0.54 8.8 0.31 0.012 0.002 2.2 4 690A 82 GTAW 20.9 2.9 0.19 3.0 0.31 0.015 0.001 2.1 6908 82 21.3 2.8 0.15 3.3 0.29 0.008 0.001 2.3 19.0 3.0 0.16 2.8 0.29 600 82 _. SMAW 22.0 0.30 0.50 4.7 0.11 3.1 8.0 0.035 0.007 6908 112 21.5 0.33 0.47 4.6 0.08 3.4 8.4 0.044 0.008 690A 112 690A 625 GTAW 23.5 0.13 0.16 5.3 0.33 0.31 2.7 7.4 0.021 0.001 23.3 0.14 0.13 5.4 0.31 0.29 2.9 7.8 0.022 0.002 0.075 0.006 6908 625 640A. 128 .SMAW 43.9 0.09 0.91 0.40 0.49 0.19 0.064 0.007 43.7 0.09 0.89 0.50 0.39 0.18 0.024 0.001 i 6908 128 690A 72 GTAW 42.1 <0.05 0.19 1.0 0.42 0.18 0.029 0.002 42.6 <0.05 0.16 1.0 0.43 0.17 0.040 0.005 r 6908 72 640A 135 SMAW 26.2 5.1 0.39 7.1 0.05 <0.10 2.1 l 6908 135 25.8 5.2 0.36 7.1 0.05 <0.10 2.1 0.042 0.005 690A 127 GTAW 28.5 0.26 0.21 6.9 0.33 0.61 0.013 0.002 0.018 0.003 690B 127 28.9 0.29 0.17 6.9 0.34 0.61 ,,g NOTE: 690A = Alloy 690, 0.02 wt.%C Ja 6903 = Alloy 690, 0.06 wt.%C 600 = Alloy 600, 0.02 wt.%C 3 q g@, m I@ \\ w 2 4 p[. 8 1 r,

M I Table V CONTROL SPECIMENS Materi al Heat Treatment i Type 304 Stainless Steel 1120"C/1 h/WQ + 677"C/2 h/AC I + 1120*C/1 h/WQ + 500*C/24 h/AC INCONEL Alloy 600 MA + 400*C/200 h/FC i INCONEL Welding Electrode 182/ MA + 400*C/200 h/FC I Alloy 600 Base Plate 'Jeldment l NOTE: MA = Mill Annealed. l Table VI TENSILE PROPERTIES - AVERAGE VALUES OF TWO TESTS h5 iCU Filler Metal or Weld Yield Strength U.T.S. El. R.A. T Electrode Process Base Plate MPa (ksi). MPa (ksi) (%) (%) Weldment Properties 1 Filler Metal 82 GTAW Low C Alloy 690 438 (63.5) 687 (99.6) 27.5 59.0 Welding Electrode 182 SMAW 456 (66.2) 665 96.6) 15.5 43.5 ~ R-127 Exp. GTAW 432 (62.7) 655 (95.0) 25.5 57.2 R-135 Exp. SMAW 472 (68.5) 736 (106.7) 22.5 44.5 . Filler Metal M GTAW 485 (70.4) 718 (104.2) 25.0 49.5 ' =_r 20.5 39.0 i R-128 Exp. SMAW 512 (74.3) 739(a) (107.2) 753 (109.3) 32.0 50.2 4 Filler Metal 625 1 TAW 471 (68.3) 'f. Welding Electrode 112 SMAW 505 (73.3) 801 (116.3) 27.0 46.0 l U 7; Filler Metal 82 GTAW High C Alloy 690 417 (60.5 686 (99.5 31.0 54.5

Welding Electrode 182 SMAW 418 (60.6

'687 -(99.6 25.5 45.0 C* R-127 Exp. GTAW 447 (64.9) 686 (99.5) 24.o.45.7 % R-135 Exp. SMAW -431 52.6) 718 (104.1) 30.0 48.2 $1 Filler Metal 72 GTAW 441 63.9) 695 (100.9) 28.0 48.5 SMAW .'421 (61.0) 735 (106.6) 34.0 42.2 ",'s.R-128 Exp.. .,GTAW .. ~ 469 68.0) 764(b) (110.9) ^ 35.5 45.2 ller Metal 625 ' 443 64.3) 776 (112.5) 34.0 62.5 ding. Electrode 112 .SMAW Th'!.d5 .2 Low C Alloy 600 348 (50.5) 623(b) (90.3) 41.0 78.0 Filler. Metal'82 GTAW 290 (42.1) 611 .(88.6) 40.0 77.7 lding Electrode 182 - SMAW ' .: ? ~ --), ,, $,. ; s, f, Base Plate Properties E'N$$'di.- 3 ;.

Lod C Alley 690 357. (51.8) 748 (108.5) 42.5.64.5 High C Alloy 690 486 (70.5) 818 (118.7) 35.0 60.5

+b. ,--J' f ,-g. ,{' Low C A"oy 600 170 (24.6) 582 (84.4) 55.0 77.2 [f* One sample failed in heat affected zone. b Samples failed in base plate. 3 3 0%.61+i 1.;( 9 ?.<

4 4 ' i D ^^ 100 l o o suAW woutments a h,I Table VII a GTAW we dments STRESS CORROSION RESULTS - 75

• ,V !

h [p% - Failure Time dolding Mat ed el_ Rase #1ste __ _ U + nd Type (Weers) _ gM C un ee We1Mmt Results. y o d 2 25 Filler Metal R2 tow C 690 Single 33,40 O Double 40.40 l l l_ l l 1 High C 690 Single 33,40 Double 40.40 g 600 Single 40,40 16 to 24 28~ 32 36 40 44 b

• ee Double 40,40 f,hromium Concentrahon (wt %)

f( Electrode 132 Low C 690 $1ngle 12.27 Effect of chrtraium concentration and welding Double 27.40+ Fig. 1 procedure on the 5CT 1>ehavior of alloy 690 4 Hign C 690 single 40,40 4 Double 27.40 welc'.ments. k., 600 single 40.40 ,e Double 12,12 l R-127 Exp. Low C 690 51ngle 40+,40+ Double 40*.40+ High C 690 Single 40+.40+ , 100 um, P +e. Double 40+.40+

1. A*.

j ~& R-135 Eno. Low C 690 Single 33.40 -% ( ! Douole 40+.40+ Hign C 690 $1ngle 40,40 5 [

• 3 -

Double 40+.a0+ $p* h w tal 72 t w C 690 Single 40+,40+ Double 40+.40+ e A Filler e ,j .a C 690 Stagle 40+.40+ g Double 40,40+ V, ( R-129 E sp. Low C 690 $ ingle 40.40+ Double 40+.40+ l Hig% C 690 Single 33,40 ] Double 40+.40+ 3 0 ) Filler Metal 625 Low C 690 Single 40.40 o Double 33.40 High C 690 . Single ' 33.33 Double 40,40 } W ~M~ pil,ler N tal,BI seld.. --.3, -, n Electrode 112 Low C 690 Single 33.40

. 7 %

'y' High C 690 single 40.40 "-s - . Double 40.4w yt Double 40.40 s. 7 y ~-iOOvb fH O 4 g (Y b1D ~I-D. b *

• Base Plate 4 Control Samples Results

'N'.~.*k- " iY I Y ' ' '.'-r n

r. ; e b,P.~7y Low C 690 Stagle 40+ 40+

C,+1.\\. i -m ~ Double 40+.40+ .. ~ '.; ym y. 'E$ 6 $ d.' il j

• e:

. f,(/, ~ 1 f. s ..-y i Nigh C 690 Single 40+,40+ Double 40+.40+ ~ ' l j 411og600 singte 40+.40+ ' ' J *' W 't. e Double 40+.40+ ,( Type 304 N.T.1 ' Double . 3.3 L Double 12.12 F Type 304 4.T.2 l 600-H.T.3 r,ouble 40+.40+ 5' - ~- '2. J f ~ ~.' S ~ ll 1R2 Welo - M.T.3 Single 16.16.28,28+ q, '.;,- ~- ' d,

• D N.?

Alloy 600/ Electrode Double 10.16 2~ . ~. ,.g 3 - ' g-ig g M r. (M;.; #; - before welding -2t;G '.34 C y,$;*..y R. p h Electrode 182 Weld 3[4 l !U N'1TF: 's -;9 H.T.1 = 1120*C/1 h/WQ + 677'C/2 h/4C. f N.T.2

• 1120*C/1 h/Wa + 500*C/24h/AC.

, ; g c.- ~ N.T.3 Mill Anneal + 400*C/200 h/FC. Meta 11ogra{ Examples of cracks in failed welds. cross-sections etched with 10 percent osalic ec Tig. 2 1 1 ~. t f ti 10 i t m

-~ 4 ) I y-P "4. VA -- - 1; E," 'fkIN[f ' l L .e a ae. s. rer i l Filler Metal 82 Weld i ~.. 4 f. f 'y. ' 5-f ,p j l 4A g.

r. t h/

3..

J.w ( .f~ r((.EE - :.> ,e 1 - ~ ' ~' ~ n. J 6.,. p

37 r s i

_.pq.,.. p y.- seu ' Q[ tis t** l-g3,ggwg, ggg y,3g - j ,Qsu i r b.s,s I mfig.:3 Scanning election micrographs of f ailed weld h,;, $Petimens, f tac .,E.9 se m. $f *r"? d:) $U : y? IU 8 %.. u.

• ~

, u;>.- l II. D ] Gli-C L ; _. 71 1.: $a 3?) $3 ?; e ~. p 4ug l Y h. l S: TJnu a.w i i YV .11 ..t ..J i v wi -}}