ML19347C671
| ML19347C671 | |
| Person / Time | |
|---|---|
| Site: | Quad Cities  |
| Issue date: | 07/31/1980 |
| From: | Diercks D, Dragel G ARGONNE NATIONAL LABORATORY |
| To: | |
| Shared Package | |
| ML19347C669 | List: |
| References | |
| NUDOCS 8101050070 | |
| Download: ML19347C671 (46) | |
Text
{{#Wiki_filter:em J FAILURE ANALYSIS OF CRACKED CORE SPRAY PIPING FROM Tile QUAD CITIES UNIT 2 BOILING WATER REACTOR
- by D.R. Diercks and G.M. Dragel Materials Science Division Argonne National Laboratory Argonne, Illinois 60439 July 1980
- NOTICE: This report was prepared as an account of the work perfor.ned for the Commonwealth Edison Company under Ack. No. 85050. It is not to be distributed, referenced, excerpted, quoted, or reproduced in any form without the permission of Argonne National Laboratory and Commonwealth Edison Company.
ib D 0 5<b 0\ Y
o Please make the following correction in the report:
" Failure Analysis of Cracked Core Spray Piping from the Quad Cities Unit 2 Boiling Water Reactor" by D. R. Diercks and G. M. Dragel Page 23, Figure 7 change; --elbow-t o-we dge--
to read; elbow 'to-pipe L.
% f 4
TABLE OF CONTENTS Page List of Figures . . . . . . . . . . . . .. . . . . . . . . . . . . . Iv List of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . vi INTRODUCTION . . . . . . . . . . . . .. . . . . . . . . . . . .. . 1 COMPONENT REMOVAL, SHIPPING, AND DECONTAMINATION . . . . . . . . . . 2 INITIAL SECTIONING . . . . . . . . . . . . . . . .. . . . . . . . . 2 DYE-PENETRANT EXAMINATION . . . . . . . . . . . . . . . . . . . . . . 3 X-RADIOGRAPHY . . . . . . . . . . . . . . . .. . . . . . . . . . . . 4 CHEMICAL ANALYSES . . . . . . . . . . .. . . . . . . . .. . . . . . 5 METALLOGRAPHY . . . . . . . . . . . . .. . . . . . . . .. . . . . . 5 DELTA-FERRITE DETERMINATION . . . . . ... . . . . . . . .. . . . . 9 MICR0 HARDNESS MEASUREMENTS . . . . . . .. . . . . . . . . . . . . . 10 TENSILE TESTS . . . . . . . . . . . . . . . . . . . .. . .. . . . . 10 SENSITIZATION DETERMINATION . . . . . .. . . . . ... . . . . . . . 11 DISCUSSION . . . . . . . . . . . . . . . ... . . . . . . . . . . . 11
SUMMARY
. . . . . . . . . . . . . . . .... . . . . . . . . . . . . 13 ACKNOWLEDGMENTS . . . . . . . . . . . . .. . . . . . . . . . . . . . 15 REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 16 i
f 1 l l t lii
List of Figures No. Title Page
- 1. Schematic Representation of the Core Spray Injection Line in the .
Quad Cities Unit 2 Boiling Water Reactor . . . . . . . .. . . . 19
- 2. View of Decontaminated Elbow and Wedge Assembly Before Initial Sectioning . . . . . . . . . .. . . . . . . . . . . . ... . . 20 Alternate View of Decontaminated Elbow and Wedge Assembly . . . 20 3.
- 4. Diagram of Elbow and Wedge Assembly with Initial Sections Indi-cated by Dashed Lines . . . .. . . . . . . . . . .. . . . . . 21
- 5. Appearance of ID near Elbow-to-Wedge and Wedge-to-Pipe Welds in Region from 6 o' clock (RHS) Through 9 o' clock to 12 o' clock (LHS) after Dye-penetrant Examination . . . . . . .. . . . . . 22
- 6. Same as Fig. 5 but in the Region from 12 o' clock (RHS) Through 3 o' clock to Approximately 5:30 (LHS) . . . . . .. . . . . . . 22
- 7. Appearance of ID near Elbow-to-Wedge Weld in Region from 6 o' clock (RHS) Through 9 o' clock to 12 o' clock (LHS) after Dye-penetrant 23 Examination . . . . . . . .. . . . . . . .. . ... . . . . .
- 8. Same as Fig. 7 but in the Region from 12 o' clock (RHS) Through 23 3 o' clock to 6 o' clock (LHS) . . . . . . . . . . . . . . .. . .
- 9. Circumferential Cracks Revealed by Dye Penetrant at 6 o' clock PosiLion on Elbow Side of Wedge-to-Elbow Weld . . . . . . . . . 24
- 10. View of OD Near Wedge-to-Elbow Weld Showing Dye-penetrant Indi-cation (Arrow) on Wedge Side of Weld, at Approximately 6:30 24 Position . . . . . . . . . ... . . . . . .. . ... . .. . .
- 11. View of OD Near Wedge-to-Elbow Weld Showing Dye-penetrant Indi-cation (Arrow) on Elbow Side of Weld at Approximately 8 o' clock 25 Position . . . . . . . . . . .. . . . . . . . . . ... .. . .
- 12. View of OD Near Wedge-to-Elbow Weld Showing Dye-penetrant Indi-cation (Arrow) on Elbow Side of Weld at Approximately 10 o' clock 25 Position . . . . . . . . . . .. . . . . . .. . . . . ... . .
Radiograph of Central Portion of Specimen Shown in Fig. 5 . . . 26 13.
~
- 14. Diagram Indicating How- Specimen Shown in Fig. 5 was Sectioned for Metallography . . . . ... . . .... . . ... .. . . . 27 28
- 15. Intergranular Cracking observed in Metallographic Section A-A. .
- 16. Penetration of Cracking in Section A-A into Weld Metal at Wedge-28 to-Elbow Weld . . . . ..... . . .. . .. . . . . . . . . .
iv
List of Figures (contd.) No. Title Page Low-magnification View of Larger Circumferential Crack Observed
~
17. in Section C-C . . . . . . . . . . . . . . . . . . . . . .. . . 29
- 18. Higher-magnification View of the Crack Shown in Fig. 17 .. . . 30
- 19. Low-magnification View of Smaller Circumferential Crack Present in Section C-C . . . . . . . . . . . . . . . . . . . . . . . . . 30
- 20. Low-magnification View of Axial Crack Present in Section D-D . . 31
- 21. Axial Crack Present in Section E-E Approximately 26 mils below the Reference Plane . . . . . . . . . . . . . . . . . . . . . . 32
- 22. Axial Crack Present in Section E-E Approximately 61 mils Below the Reference Plane . . . . . . . . . . . . . . . . . . . .. . 33
- 23. Higher-magnification View of Cracking Present at Wedge Material-Weld Metal Interface in Same Section Shown in Fig. 22 . . .. . 34
- 24. Higher-magnification View of Cracking Present Near Pipe Ma-terial-Weld Metal Interface in Same Section Shown in Fig. 22 . . 34
- 25. Axial Crack Present in Section E-E Approximately 100 mils Below the Reference Plane . . . . . . . . . . . . . . . .. . . 35
- 26. Higher-magnification View of Secondary Cracking Present in Weld Metal in Same Section Shown in Fig. 25 . . . . . . . . . ... . 36
- 27. Axial Crack Present in Section E-E Approximately 210 mils Below 37 the Reference Plane . . . . . . . . . . . . . . . . . .. ...
- 28. Subsurface Crack Present in Weld Metal in Section F-F . .. .. 38
- 29. High-magnification View of Typical Delta-ferrite Distribution Present in Weld Metc1 of Section E-E (Same Section Shown in Fig. 27) . . . . . . . . .. . . . . . . . . . . . . . . ... . 38 i
! 30. Heavily Attacked Microstructure Observed in Wedge Material Near Wedge-to-Pipe Weld af ter Etching According to ASTM Procedure ' A 262, Practice A . . . . . . . . . . . . . . . . . . . . . . . 39
~
- 31. Lightly Attacked Microstructure Observed in Elbow Material Away from Heat-affected Zone after Etching According to ASTM Pro-cedure A 262, Practice A . . . . . . . . . . . . . .. ... . . 39 I
l V l I
List of Tables . No 3 Title Page I. Compositions (wt. %) of Pipe, Wedge, and Elbow Samples - Compared with Specifications for Type 304 Stainless Steel . . . 17 II. Delta-ferrite Levels Observed in Section E-E in Regions Near and Away from Through-weld Crack . . . . . . . . . . . . . . . 17 III. Microhardness Data Obtained on Elbow Material at and away from the Weld Prep Area, as a Function of Distance from the ID Surface . . . . . . . . . . . . . . . . . . . . . . . . . . 18 IV. Summary of Tensile Test Data Obtained on Pipe and Elbow Material . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
+
e O vi
. INTRODUCTION Commonwealth Edison Company's Quad Cities Nuclear Power Plant, located near Cordova, Illinois, consists of two essentially identical but independent 800-MWe boiling water reactor (BWR) units and turbine-generator systems lo-cated in a common building. Both BWRs were supplied by Caneral Electric Company, and both units began commercial operation in 1972. On March 8, 1980, during a pressure test of the Unit 2 reactor vessel, water was observed to be leaking from the loop A core spray injection piping in the region near the junction af this piping with the reactor pressure vessel. An ultrasonic in-spection of the entire core spray piping system located within the primary containment was subsequently conducted. This inspection indicated the pres-ence of several cracks near welds at the leaks in loop A, and suggested that cracks were also present at similar locations in loop B.
The loop A core spray injection line is shown schematically in Fig. 1. The piping is 10-inch Schedule 80 ASTM A 312, grade TP 304 (Type 304 stain-less steel) seamless pipe, and the elbows are A 403, grade WP 304 (Type 304 stainless steel) schedule 80 forgings. Type 308 stainless steel filler metal was used in the welds. Of particular interest to the present investigation are the 90 clbow and adjacent 15* wedge located between the reactor pressure vessel and the first manual isolation valve. Ultrasonic inspection revealed crack indications at all three welds joining these forgings to each other and to the adjacent piping. Crack indications were also detected ultrasonically at the welds joining the first manual isolation valve to another 90* elbow immediately adjacent to it. The layout of the loop B core spray injection line is not greatly dif-ferent from that shown in Fig. 1 for loop A. However, in loop B the po-sitions of the 90* cibow and 15* wedge between the pressure vessel wall and the manual isolation valve are reversed in that the wedge is nearer the vessel. Cracks were detected ultrasonically in loop B at the same four welds I . as in loop A. During reactor operations, the manual isolation valve and testable check valve shown in Fig. 1 are normally open, and the core spray isolation valve is normally closed. Since the core spray inlets to the pressure vessel are below the level of the reactor coolant water, that portion of the core spray lines up to the core spray isolation valve is normally filled with more or 1
less stagnant coolant water. At the cracked elbow and wedge welds, the water tem;,erature is estimated to range from 400 to 500 F under normal operating 'onditions, and the pressure is about 1000 psi. COMPONENT REMOVAL, SHIPPING, AND DECONTAMINATION The piping regions from loops A and B including the 90* elbow and 15* wedge pieces were selected for detailed failure analysis. These two 105* elbow assemblies were therefore removed from the core spray line by making circumferential cuts approximately 8 inches away from the pipe-to-elbow and pipe-to-wedge welds, respectively. The elbow assembly from loop A was partly decontaminated at the plant before being shipped to Argonne National La-boratory, but the loop B assembly was shipped in the non-decontaminated con-dition. The cracked weld region between the manual isolation valve and the adjacent 90* elbow from both loops was not sent for examination. Upon receipt at Argonne, the non-decontaminated loop B elbow assembly was placed into storage without decontamination. The loop A elbow assembly was subjected to a thorough decontamination procedure to facilitate subse-
- t quent handling and examinations. The component was first immersed in a caustic permanganate solution at 200 F to oxidize the surface residue, after which this residue was dissolved in o phosphoric acid solution. However, this technique was found to be only partly successful in removing the surface contamination. A " bright dip" solution of acetic, nitric, phosphoric, and hydrochloric acid was then swabbed over the surface in order to clean and passivate the surface region. This approach was found to be more successful, and the radioactivity of the elbow assembly was finally reduced to about 100 mR/h as compared to levels in excess of 1 R/h in the as-received con-dition. Two views of the decontaminated loop A cibow assembly are shown in Figs. 2 and 3.
INITIAL SECTIONING . After decontamination, the core spray elbow assembly from loop A was sectioned to permit a more detailed examination of the cracking present at the ID in the vicinity of the weld. The 15* wedge was first removed from the assembly by making a pair of circumferential cuts through the elbow and 2
pipe material, respectively, approximately 2-1/2 to 3 inches away from the circumferential welds to the wedge. A similar pair of cuts was made at the other end of the assembly to each side of the circumferential elbow-to-pipe weld. The resulting circular rings were then split asially through the 6 and 12 o' clock positions to produce the four semicircular pieces on which the ensuing examinations were performed. (The clock positions around the cir-cumference of these semicircular pieces are defined looking toward the reactor pressure vessel, with the 12 o' clock position originally oriented upward in the installed position in the plant.) The locations of the above cuts are shown as the dashed lines in Fig. 4. DYE-PENETRANT EXAMINATION A macroscopic examination of the veld region at the 1D surfaces of the four pieces described above was performed using Spotcheck type SKC-NF dye penetrant (Magnaflux Corp., Chicago). The surfaces were first cleaned with alcohol, acetone, and a special dye-penetrant cleaner. The dye penetrant was then applied and the excess removed, and a developer was subsequently , applied to reveal the presence of cracks through the bleed-out of retained dye penetrant. Figures 5 through 8 show the appearances of the four ID surfaces after this procedure. Numerous cracks are indicated in the regions near the welds, and in almost every case the cracks are oriented in the axial direction. The only exceptions are (1) an irregular crack on the elbow side of the wedge-to-elbow weld at approximately the 8 o' clock position (see Fig. 5), and (2) a pair of circumferential c. racks at the ID surface on the elbow side of the wedge-to-elbow weld at the 6 o' clock position. These latter cracks are not readily visible in the photograph shown in Fig. 5, but are shown with more clarity in Fig. 9. It may be seen that the dye penetrant revealed no cracks in the weld metal itself or on the pipe side of the welds. The cracks tend to be concentrated to either side of the wedge-to-elbow weld. . It should be noted that an axial strip was cut from the semicircular sample shown in Fig. 6 before the dye-penetrant inspection was performed. . This strip, which included approximately the 5:30 to 6 o' clock positions, was removed during initial sectioning to permit a preliminary metallographic examination of an axial crack on the wedge side of the wedge-to-elbow weld. Thus, the semicircular piece shown in Fig. 6 includes only the 12 o' clock to 3
approximately 5:30 positions. . A dye-penetrant examination was also performed at the OD of the wedge and adjacent welds in an attempt to locate the leaking cracks. Unfortunately, the . relatively rough surface at the exterior of the pipe, particularly at the interface between the weld head and the adjacent base metal, greatly hindered the inspection. Relatively faint surface indications were found at several locations along the weld metal-base metal interface at the OD; at least three of these indications may correspond to leaking cracks. Figure 10 shows one of these OD indications at approximately the 6:30 position on the wedge side of the weld, probably corresponding to the darker of the two ID cracks seen in this region in Fig. 5. A second possible leaking crack is shown in Fig. 11; this indication apparently corresponds to the non-axial crack to the elbow side of the weld near the 8 o' clock position in Fig. 5. The third indication (Fig. 12) appears to match up with an elbow-side crack at approximately the 10 o' clock position in Fig. 5. It should be noted that the OD indications , shown in Figs. 10 and 12 agree reasonably well with the positions of the leaks observed in the field (see previous discussion), but the indication at the . 8 o' clock position shown in Fig, 11 does not correspond to the location of a reported leak. Sinilarly, a reported leak near the 10 o' clock position on the wedge side of the weld does not appear to have been confirmed here. X-RADIOGRAPllY After the dye-penetrant examination was completed, the ecmicircular piece containing the apparent leaking cracks at the wedge-to-elbow weld (i.e., the piece shown in Fig. 5) was inspected using x-radiography. Several additional possible flaws or extensions of cracks previously observed using dye penetrant were detected. The most significant of these is shown in Fig. 13 as the crack at the bottom center of the picture. This crack corre-i sponds to that seen, using dye penetrant, to the wedge side of the wedge-to-pipe weld at approximately the 9 o' clock position in Fig. 5. However, Fig. 5 , suggested that the crack stops at the weld fusion line, whereas Fig. 13 indi-cates that this crack penetrates the weld metal and extends into the pipe . material on the other side. The remaining suspected flaws identified by radiography were not as clear-cut as the crack shown in Fig.13, and did not reproduce well in the 4
photographic prints. These additional possible flaws are shown in the sketch of Fig. 14 in red. Three of them represent extensions of cracks al-ready detected using dye penetrant. Two of these cracks may extend through the weld metal, as indicated in Fig. 14; one is seen in the wedge-to-clbow weld at about the 11 o' clock position and the other in the wedge-to-pipe weld at approximately the 11:30 position. A faint indication of a possible crack on the pipe side of the wedge-to-pipe weld at about the 11 o' clock position was also detected in the radiograpbs. CHEMICAL ANALYSES After the above examinations and procedures were completed, chemical analysis sacples were taken from the pipe, elbow, and wedge material. The results of these analyses are presented in Table I. The compositions of all three samples may be seen to fall withru the specified values for Type 304 stainless steel for all of the elements analyzed. Additional emission-spectroscopy analyses were performed to check for the possible pres-ence of such elements as Ti, Nb, V, Mo, Zr, and Pb. None of these elements was found to be present in significant quantity. An additional chemical analysis for carbon only was performed on a weld metal sample taken from the wedge-to-elbow weld at approximately the 9 o clock position. This sample was taken from the center of the weld pass at the OD i surface to reduce the likelihood of dilution by the surrounding base metal. The carbon content of this sample was found to be 0.064 wt. %, which is below the maximum allowable for Type 308 stainless steel. l METALLOGRAPHY l l t The semicircular piece on which radiography was performed was subse-quently sectioned for metallographic examination. The dashed lines in Fig. 14 Indicate where the cuts were made, and the metallographic sections examined are shown in green and denoted by the letters A-A through F-F. The observations j made on these sections are discussed below. The first metallographic sections examined were sections A-A and B-B through an axial crack on the wedge side of the wedge-to-elbow weld between about the 5:30 and 6 o' clock positions. These two sections were taken in 5
the radial direction through the pipe wall at two locations along the crack length. The circumferential weld parallel to the two sections has a V-groove configuration tapering outward toward the OD, and section A-A is sufficiently close to the weld that it intersects the weld metal approximately midway through the thickness of the section. Thus the ID surface of section A-A consists of wedge base metal very near the wedge-to-elbow weld, and the OD surface consist of weld metal. Section B-B cuts entirely through base metal % 0.5 in, away from the weld at the 1D, but is presumably still within the heat-affected zone (HAZ). Two cracks were detected in section A-A, both of which initiated at the ID and penetrated approxit2tely halfway through the wall thickness at this point. The base-metal portions of the two cracks exhibited the branched intergranular propagation mode commonly associated with intergranular stress-corrosion cracking (IGSCC), and some oxide was present in portions of the cracks (see Fig. 15). Both cracks penetrated % 60 to 80 mils into the weld . metal before stopping, as shown in Fig. 16. It should be noted that the crack seen in Fig. 16 appears to be discontinuous in the weld metal; however, the crack segments seen probably link up above and below the plane of the polish. One of the two cracks seen in section A-A was also observed in section B-B, where its appearance was similar to that shown in Fig. 15. The depth of the crack seen in section B-B was about 60% of the wall thickness. The two circumferential cracks detected in the elbow material immediately above the wedge-to-elbow weld in the 6 o' clock position were examined metallo-graphically in section C-C. The larger of these two cracks, which was closer to the weld, is shown in Fig. 17. The crack is again intergranular and highly branched, but follows a much more tortuous path than the axial cracks observed in sections A-A and B-B. In addition, large voids are observed where entire grains have apparently fallen out during sample preparation. A higher-magnification view of a portion of this crack (Fig. 18) reveals that corrosion product, probably oxides, is present in the crack as well as in some of the voids. The crack seen in Fig. 17 extends about one-third of the way through . the wall thickness from the ID, and apparently stops at the weld metal in this section. , The second circumferential crack seen in section C-C is shown in Fig. 19. This crack extends for about 130 mils, or about 20% of the way through the vall thickness. and again displays a highly complex nature. Both of these 6
circumferential cracks, like the axial cracks seen in sections A-A and B-B, appear to have been caused by ICSCC. The circumferential section D-D shown in Fig. 14 was also examined met-allographically. This section was taken to the pipe side of the adjacent wedge-to-pipe weld, and is near enough to the tapering weld-metal region that it intersects the weld metal approximately halfway through the section thick-ness. Thus, the ID surface in this section is pipe material and the OD sur-face is weld metal. Section D-D was chosen for examination because the dye-penetrant check did not reveal any flaws in this location, but the radio-graphic examination indicated that a surface crack to the wedge side of the weld, which was seen with dye penetrant, extended into the pipe material in this region. Figure 20 shows a low-magnification photomicrograph of the crack found in section D-D. This micrograph clearly indicates that the crack intersects
. the ID surface, even though it was not seen with dye penetrant. The crack path is intergranular, and some branching is evident. The overall crack depth in this section is about 400 mils, or about two-thirds of the wall thickness, and the crack penetrates approximately 110 mils into the weld metal before stopping. The crack again appears to be the result of IGSCC.
The radiographic results suggest that the present crack may have com-pletely penetrated the circumferential wedge-to-pipe weld. However, this i. not been confirmed here. In addition, a smaller second crack on the pipe i side of the weld, to the right of the present crack (see Fig. 14), was indi-cated by radiography. An attempt was made to include this crack in section D-D, but it could not be found in the prepared metallographic section. It is l possible that this smaller crack, if actually present, was lost in one of the saw cuts. The crack examined metallographically in section E-E is of particular interest. The radiograph shown in Fig. 13 indicates rather clearly that this crack completely penetrates the wedge-to-pipe weld, even though the dye-penetrant examination revealed it to be present at the surface on the wedge side only. The metallurgical sections to be discussed here were obtained
~
by mounting the specimen with the ID surface containing the crack parallel to the face of the mount, so that the plane of the polish in the prepared section is apploximately parallel to the ID surface of the wedge and pipe. The mounted specimen was then ground down to a level at which the weld bead 7
was exposed across the entire plane of the polish, but the base metal to each . side was still slightly below this plane. This level was arbitrarily defined to be zero depth into the specimen, and the various metallographic sections discussed below were taken at several depths below this reference plane. Figure 21 shows a metallographic section obtained at a depth of 26 mils below the reference plane. An intergranelar crack is visible in the wedge material only, and no cracking is seen in either the weld metal or the pipe base metal. However, as Fig. 21 indicates, the wedge-metal crack terminates abruptly in this section where it intersects a region of lack of fusion at the wedge material / weld metal interface. This region of lack of fusion ex-tends % 0.4 in, along the interface. Figure 22 shows the metallographic section obtained at a depth of 61 mils below the reference plane. The lack of fusion at the wedge / weld-metal inter-face is no longer visible, and the crack in the wedge material now extends somewhat into the adjacent veld metal. An intergranular crack on the pipe . side of the veld is also now visible, though it does not link up with the crack in the wedge in this plane. Higher-magnification views of the cracked ~ regions at each side of the weld are shown in Figs. 23 and 24. An interesting nicrostructural feature visible in the wedge material in Fig. 22 is the presence of delta-ferrite stringers extending along the axial direction. These stringers parallel the flow lines in the forged micro-structure, and are not present in the pipe material, which was not forged. The stringers also parallel the axial cracks present in the wedge and elbow, and their possible influence on this cracking will be discussed below. The metallographic section taken at a depth of 100 mils reveals one continuous crack extending from the wedge material across the weld metal and into the pipe material on the other side, as shown in Fig. 25. The crack path in the base metal on both sides of the weld is again intergranular. In the weld metal, the crack path appears to be influenced by the alignment of the columnar structure along the directions of heat flow that occurred during solidification. A higher-magnification view of a secondary crack in the weld . metal (Fig. 26) suggests that the cracking process here is one of alternate crack extension through the austenite matrix between the delta-ferrite par- , ticles and subsequent propagation along the interface between these particles and the austenite. 8
A final metallographic section at this location was taken at a depth of 210 mils, and is shown in Fig. 27. The weld-metal region is larger here, since the V-groove weld taper;. outward as one approaches the OD. In addition, the cracking in the weld metal is wider and somewhat straighter in this sec-tion, and is not as discontinuous. The features of the crack are otherwise similar to those seen in the previous sections. The final metallographic section examined was section F-F of Fig. 14 This section was taken normal to the pipe wall down the weld centerline between adjacent cracks in the wedge and elbow material. The purpose of this section was to determine if these adjacent cracks were interconnected through the weld metal, as had been suggested by radiographic results. Figure 28 shows a portion of the section obtained. The small, discontinuous, subsurface crack seen is located N 100 mils below the specimen surface, and was clearly visible in both the unetched and etched conditions. It thus appears that the
- adjacent cracks are probably interconnected through the weld metal.
DELTA-FE" RITE DETERMINATION In view of the extensive weld-metal cracking observed in metallographic section E-E in particular, quantitative determinations of the amount of delta ferrite present in the Type 308 weld metal in this section were performed. Regions both near (within 10 mils) and away from (more than 100 mils) the through-weld crack were examined, and all determinations were made at se-lected locations on the section at the 210-mil depth (see Fig. 23). A special l etching technique using Kalling's reagent was used to accentuate the delta-ferrite regions in the weld metal. A Quantimet 720 image analyzer system was used to measure delta ferrite from photomicrographs taken at a magnification of 700X; one of these is shown in Fig. 29. Five randomly selected locations on each photomicrograph were analyzed, and the average of these five values is reported here as the delta-ferrite content for the region shown in that i
- photomicrograph.
The results of the delta-ferrite determinations for the five photo- ! - micrographs analyzed are presented in Table II. No significant cifferences were observed in the levels present near and away from the through-weld crack. In all cases, the volume percentage of delta ferrite was within the
% 4-8% range typically recommended to avoid susceptibility to weld cracking.
9
MICR0 HARDNESS MEASUREMENTS . Before the core spray piping and elbow sections were welded together, the ID surfaces of these pieces were machined near the weld regions. As a result of this machining operation, slight deformation could be visually ob-served in the first 1-2 grain layers from the prepared surfaces. A series of microhardness readings was therefore taken to determine the extent and depth of this surface deformation. The readings were taken on elbow material near the 6 o' clock position adjacent to the elbow-to-wedge weld, with hard-aess measured as a function of distance from the machined ID surface. For comparison purposes, a similar series of readings was taken on this same specimen near the ID surface but awiy from the machined region. All readings were taken using a 50-gram weight and a diamond pyramid (Vicker's) indenter. The l'ardness data obtained are summarized in Table III. Only a modest increase in hardr.-ss is observed near the machined surface as compared with the ID surface region away from the weld. The depth of the affected region appears to be of the order of 10 to 20 mils. TENSILE TESTS Roon-temperature tensile tests were conducted on specimens of pipe and elbow material oriented both parallel to the pipe axis (axial specimens) and perpendicular to the axis (transverse specimens). The test specimens were four inches long and approximately 100 mils thick with a one-inch gage length. The specimen geometry and testing procedure conformed to the requirements of ASTM Standard E-8. The tensile-test results are summarized in Table IV. A relatively modest difference in tensile properties is seen between the axial and trans-verse pipe specimens. In addition, the axial specimen from the forged elbow exhibits about the same properties as the two pipe specimens. However, the transverse specimen from the elbow is noticeably stronger than the axial . specimen, indicating a possible directionality in properties in the forged microstructure. It is somewhat surprising, however, that the room-temperature tensile strength is greater in the transverse direction, even though virtually all of the IGSCC observed in the failed component occurred more or less normal to this direction. It thus appears that either (a) the 10
. service plus residual stresses present in the failed core spray line were substantially greater in the transverse (hoop) direction than in the axial direction, or (b) the susceptibility to IGSCC in a givca direction cannot be directly related to room-temperature tensile properties.
SENSITIZATION DETERMINATION A simple qualitative test for the presence of sensitization in the austenitic stainless steel pipe, elbow, and wedge material was performed using ASTM Procedure A 262, Practice A. Under this procedure, the sample to be tested is electroctched in a solution of 10% oxalic acid in water at a current density of 1 A/cm for a period of 90 seconds. The degree of sensitization in the sample is revealed by the extent to which it is attacked by this etching procedure. Samples in and away from the weld HAZ were tested. as expected, the samples from the veld HAZ were found to be heavily sensitized, but little sensitization was indicated in the samples away from the weld regions. Figure 30 shows an etched sample of wedge material from the HAZ of the wedge-to-pipe weld. The heavily attacked " ditch" structure along the grain boundaries is apparent. In contrast, relatively little attack was noted, for example, in the sample of elbow material taken away from the weld region (Fig. 31). DISCUSSION Based upon the above observations, it seems reasonable to conclude that the cracking observed in the Quad Cities Unit 2 elbow assembly is due to IGSCC. The corrosive species was probably dissolved oxygen in the semi-stagnant reactor coolant water present at the ID of the core spray line. The fact that cracks were present only at the weld regions indicates that the sensitized microstructure and probable residual welding stresses present in these regions contributed to the failure. Any welding residual stresses present would, of course, be supplemented during normal operation by a tensile hoop stress of about 8000 psi due to the nominal 1000-psi internal pressure contained by the piping. Failure of this piping due to ICSCC is not in.itself a unique occurrence; at least 133 incidents of IGSCC in US and foreign BWRs have been reported 11 I
through January 1979. However, two aspects of the present failure dis- . tinguish it from the sorts of ICSCC failures typically observed in BWR piping systems. The first of these is the overwhelmingly axial orientation of the cracking observed. Approximately 40 cracks were detected by various means at the ID of the elbow assembly, and all of these cracks were axial with the exception of the two circumferential cracks seen in Fig. 9 and the mixed cir-cumfercatial and oblique crack seen in Fig. 5. The second distinguishing feature of the present failure is the fact that, in at least one and probably two instances, a crack has completely penetrated a weld metal region. These two aspects of the present failure will be addressed briefly here. In the IGSCC failures seen to date in BWR recirculation piping, cracking is typically observed to take place in a weld HAZ and to propagate circumferentially around the pipe in this zone at a more or less constant dis-tance from the veld fusion line. However, axial cracks have been seen in, for example, the Type 304 stainless steel steam-to-isolation condenser piping from the Dresden Unit 1 BWR. In the case of the IGSCC failures in BWR core spray lines, both axial and circumferential cracking have been found. . Cracks seen in the safe-end reducer and spool piece from the Peach Bottom Unit 3 BWR and in a furnace-sensitized safe end from the Nine Mile Point reactor were almost entirely circumferential in character. However, in the Vermont Yankee A and B loop core spray lines, both axial and circum-ferential cracks were found near several welds located between the pressure vessel and the first manual isolation valve. Mixed axial and circum-ferential cracks were also found near welds in the 10-inch core spray lines of the Dresden Unit 2 BWR.( } In this case, the cracks tended to be concen-trated in the furnace-sensitized Type 316 stainless steel safe ends and safe-end extensions (dutchmen), all of which were forgings. A single axial crack has also been detected in the safe end of one of the core spray lines in the Tsuruga BWR in Japan. However, the authors are not aware of any instances in which the axial cracking of a BWR component was as pronounced as in the f present case. There is at least some circumstantial evidence to suggest thau .xial j cracking seen here is related to the banded microstructure present in the . forged elbow and wedge. First of all, the cracks observed are heavily con-centrated in the elbow and wedge material, and all the cracks detected in the wrought pipe material appear to have initiated in the adjacent elbow and 12
1 1
. wedge. Similarly, the axial cracks seen in the Dresden Unit 2 core spray line were located primarily in the forged safe ends and safe-end extensions.
Secondly, the axial cracks in the elbow and wedge are seen to closely parallel the banding present (see Fig. 22). The presence of delta-ferrite stringers implies that the forging temperature may have been somewhat in excess of the maximum 2100 to 2300 F commonly recommended, llowever, austenit ic stain-less steels containing delta ferrite have been found to passess increased resistance to stress-corrosion cracking in aqueous solutions,(ll) and so the l mere presence of delta ferrite in the forged microstructure does not explain the cracking seen here. The final point to be considered here is the propagation of at least one stress-corrosion crack completely through the Type 308 weld metal. Devine ' has analyzed the problem of inte rgranular cracking in duplex austenite plus delta-ferrite. weld-metal structures, and has developed quanti-
- tative criteria for assessing the susceptibility of these alloys to such cracking. He assumes that cracking in duplex microstructure is contingent
. upon sensitization of the austenite-austenite grain boundaries, and that a critical amount and distribution of delta ferrite is required to prevent this sensitization by causing preferential carbide precipitation at the austenite/ delta-ferrite phase boundaries. The key parameter in Devine's criteria is S , ythe amount of austenite/ delta-ferrite boundary area per unit volume of alloy. For Type 308 weld metal, Devine predicts that complete
~1 immunity to sensitization should be attained at an Sy value of 2900 cm for an alloy containing 5 percent delta ferrite and having a carbon content of 0.064 wt. %. Alloys with S ylevels somewhat below these critical valuca should be highly resistant but not totally immune to sensitization. The Sy
-1 value of about 2250 cm for the cracked Type 308 weld metal shown in Fig. 27 (carbon content = 0.064 wt. %) places it somewhat below Devine's critical level for complete immunity to sensitization, although considerable I
resistance to sensitization would still be expected. SUM 1ARY An elbow assembly and adjacent piping from the loop A core spray in-jection line of Commonwealth Edison Company's Quad Cities Unit 2 Boiling Water Reactor have been examined in order to determine the nature and cause 13 v _. . __ __ _
of coolant leakage detected in this region during hydrostatic testing. The . results of this investigation may be summarized as follows:
- 1. The coolant leakage was found to be due to cracks at the ID of the elbow assembly, at least three of which apparently penetrated to the OD.
- 2. A total of approximately 40 cracks were detected at the ID of the loop A elbow assembly, and all of these cracks were located in the vicielty of the circumferential welds joining the elbow, wedge, and pipe pieces.
- 3. All of the cracks detected were predominantly axial in direction with the exception of two circumferential cracks and one mixed circumfercatial and oblique crack.
- 4. Virtually all of the cracks were located in the forged elbow and wedge components, and in all cases, cracks found in the .
pipe material appeared to have initiated in the wedge or elbow ma t e r ia.'. . ,
- 5. The base-metal cracks were intergranular and more or less branched in all cases, and are apparently due to intergranular stress-corrosion cracking.
- 6. In one case, a crack was found to initiate in the wedge material at the ID surface and eubsequently grow below the surface com-pletely across the weld metal and into the adjacent piping material. In a second case, a crack through the weld metal apparently grew below the surface to join two ID surface cracks in the wedge and elbow material, respectively,
- 7. The chemical compositions of selected base-metal samples from the elbow, wedge, and pipe were found to conform to the specifications for Type 304 stainless steel. The carbon level in a weld-metal sample from the wedge-to-elbow weld was found to fall within the requirements for Type 308 stainless steel. .
- 8. Room-temperature tensile tests were conducted on axial and transverse base-metal specimens from the pipe and elbow material. .
These tests indicated a modest difference in properties among the axial and transverse pipe specimens and the axial elbow specimen. Ilowever, the transverse elbow specimen displayed noticeably higher 14
tensile and yield strengths than the remaining three specimens.
- 9. The delta-ferrite content in a section of weld metal containing a through-weld crack was found to be between 5.2 and 6.3 percent, which is within the normal range.
- 10. A qualitative test for sensitization using ASTM Procedure A 262, Practice A determined that the elbow, wedge, and pipe materials in the weld 11A2 were heavily sensitized, but relatively little sensitization was indicated in samples away from the weld,
- 11. Microhardness measurements detected the presence of a slight to moderate increase in hardness near the machined ID surface in the vicinity of the veld.
- 12. The forged elbow and wedge material exhibited a pronounced banded microstructure containing a significant amount of delta ferrite strung out in the axial direction. The axial cracks were gen-
- erally parallel to the banded features of the microstructure, and the orientatioas of these cracks may have been influenced by . this banding, ACKNOWLEDGMENTS The authors gratefully acknowledge the cooperation of R. Gaitonde of Commonwealth Edison Company in completing this investigation. The follcwing persons at Argonne National Laboratory also contributed to this work: F.M. Basso, C.W. Benedict, W.B. Conn, K.O. Canner, T.R. Corbett, S.M. Gehl, M.D. Gorman, J.C. Ilaugen, R.P. Isenberg, K.J. Jensen, G.G. Ketchmark, and W.K. Soppet. The final manuscript was prepared by E.L. Hartig and edited by E.M. Stefanski. e 15
REFERENCES
- 1. " Standard Methods of Tension Testing of Metallic Materials," Standard E 8, Annual Book of ASTM Standards, Part 10, American Society for Testing and Materials, Philadelphia (1974). -
- 2. " Detecting Susceptibility to Intergranular Attack in Stainless Steels,"
Standard A 262, Practice A, Annual Book of ASTM Standards, Part 3, American Society for Testing and Materials, Philadelphia (1974).
- 3. " Investigation and Evaluation of Stress-Corrosion Cracking in Piping of Light Water Reactor Plants," NUREG 0531, U.S. Nuclear Regulatory Commission, Washington, D.C. (February 1979), pp. 2.1-2.2.
- 4. " Operational Analysis Department Report on Leaks from Isolation Con-densor Piping, Dresden Station, Unit 1," M-1504-74, Commonwealth Edison Co. (July 16, 1974).
- 5. U.S. Nuclear Regulatory Commission Docket No. 50278-1078 (April 15, 1977).
- 6. U.S. Atomic Energy Commission Docket No. 50220-37 (April 3, 1970).
- 7. U.S. Nuclear Regulatory Commission Docket Nos. 50271-1018 (August 26, 1977) and 50271-1019 (September 8, 1977).
- 8. J.Y. Park, S. :fanyluk, R.B. Poeppel, and C.F. Cheng, " Metallurgical -
Examination of fracks in the Dresden-2 BWR Emergency Core-Spray System
.'.0-inch Diameter Piping," Unpublished Report to Commonwealth Edison Co., Ack. No. 5661 (April 1976).
- 9. H.H. Klepfer et al., " Investigation of Cause of Cracking in Austenitic Stainless Steel Piping," Vol. 1, NEDO-21000-1, General Electric Co.
(July 1975), pp. 7-2 to 7-3.
- 10. The Making, Shaping, and Heat Treating of Steel, Eighth ed., U.S. Steel Corp. , Pittsburgh, PA (1964) , p.1120.
- 11. R.M. Latanision and R.W. Stachle, " Stress Corrosion Cracking of Iron-Nickel-Chromium Alloys," in Proc. of Conf. on Fundamental Aspects of Stress Corrosion Cracking, National Association of Corrosion Engineers (1969), pp. 214-307,
- 12. T.M. Devine, " Mechanism of Intergranular Corrosion and Pitting Corrosion of Austenitic and Duplex 308 Stainless Steel," J. Electrochem. Soc. 126 (1979), pp. 374-385. .
- 13. " Evaluation of Near-Term BWR Piping Remedies," Vol. 2, EPRI NP-1222, Electric Power Research Institute (Nov. 1979), pp. 6-45 to 6-61.
16
. TABLE I. Compositions (wt. %) of Pipe, Wedge, and Elbow Samples Compared with Specifications for Type 304 Stainless Steel Pipe Wedge Elbow Type 304 Sample Sample Sample Stainless Steel C 0.046 0.074 0.058 0.08 max Cr 18.96 18.62 18.67 18 - 20 N1 10.36 10.34 10.46 8 - 12 Mn 1.62 1.78 1.80 2.0 max Si 0.46 0.41 0.43 1.0 max TABLE II. Delta-ferrite Levels Observed in Section E-E in Regions Near and Away from Through-weld Crack Distance from Delta-ferrite Content Photomicrograph Through-weld Crack (mils) (vol. %) 1 < 10 5.8 4 > 100 6.3 5 > 100 5.9 8 < 10 5.2 9 < 10 5.9 m 17
TABLE III. Microhardness Data obtained on Elbow Material At and Away from the Weld Prep Area, as a Function of Distance from the ID Surface At Weld Prep Area Away from Weld Prep Area - Distance from Vicker's Distance from Vicker's ID Surface liardness ID Surface Hardnegs (mils) (kg/mm2 ) (mils) (kg/m') 4.0 147 2.4 124 4.6 138 2.9 113 5.1 135 3.7 113 6.2 135 7.4 112 7.1 124 8.5 95 7.5 124 10.0 113 8.1 124 17.3 107 9.2 107 20.0 102 10.0 112 22.3 104 17.1 122 17.3 118 17.6 101 . TABLE IV. Summary of Tensile Test Data Obtained on Pipe and Elbow Material Ultimate Yield Tensile Total Strength Strength Elongation (psi) (psi) (%) Pipe, Axial 39,200 82,000 54.0 Pipe, Transverse 34,300 80,600 66.0 Elbow, Axial 36,700 83,200 56.5 Elbow, Transverse 44,000 93,400 57.9 18
?
1 [ VESSEL E REACTORWALL PRESSURE N0ZZLE- =* < MANUAL ISOLATION SAFE END- VALVE ELB0W-- I WEDGE 1 TESTABLE r CHECK VALVE g ;= -REACTOR CONTAINMENT WALL DCORE SPRAY ISOLATION VALVE e
~
Fig. 1. Schematic Representation of the Core Spray Injection Line in the Quad Cities Unit 2 Boiling Water Reactor. The cross-hatched regions indicate the locations of welds where crack indications were detected by ultrasonic testing techniques. 19 L.
4 I I i l 1
'r 's Q;; 'f 4 - M 7' [)'*Y i
4 ,r.,, t [p,1;: . , . . u -A J t* - w : f P h' .j
'4
! k9 \
- .: 1 y
x > =~:w - - __
, j.
i
, Il FS
.A _~ _16! M i
Fig. 2. View of Decontaminated Elbow and Wedge Fig. 3. Alternate View of Decontaminated Elbow W' b Assembly Before Initial Sectioning. and Wedge Assembly. i
i PIPE i .
/ i I
l I l I r-- -J i ' I i i l j
. ELB0W
= TO 1
REACTOR L M _ __ q 1 WEDGE- = { t
TO PIPE = VALVE Fig. 4. Diagram of Elbow and Wedge Assembly with Initial Sections Indi-cated by Dashed Lines. The 12 o' clock position is oriented upward. ;
l I l i-
N
'/
[ g; , - 4 --;
'~
- a , 4
} , I. . ,
1 3 7.:a _ m: g j i- I-e 25is
+ > : W ,. , ^ .
fj 9 j, g 4 kW * 'A - : w
- y- -
;≫;f
?.h, . . ,. , ,
' f.
( s , . - t,
~
~ q --(",:Q:)
- f. . .. a ; Q
- i. - . , - .
a;
.s j @h
, , '~%'-
U Fig. 5. Appearance of ID near Elbow-to-Wedge and Wedge-to-Pipe Welds in Region from 6 o' clock (RHS) Through 9 o' clock to 12 o' clock (LHS) after Dye-penetrant Examination. The elbow material is at the , l top of the figure. l
; , g _ - .-
I J
' \ }ll
^
~
j ; J. 7 l 1.4 ~4 3 l, 'A [; y 'b)4
>-, , ,,1 (g:
j [,1 -- _. - ,. , ~. t ] a i Fig. 6 Same as Fig. 5 but in the Region from 12 o' clock (RHS) Through 3 o' clock to Approximately 5:30 (LHS). l t l 22
~
D *
- lD *D 3'l $
w o n o L JL Idi% 3,_
. ~
m-my,
~
g' p :, ...;.
=
s
~
w +
, ,-j
- g. % , . . o .
Y, [ '
/
. . gc ;( 2 0 -
,; j . $]
l - . % P- w~
.,~-. ,
- z. .
9{
,,l fygge , m + ma.,g.9,yg
. .: a,x, t .. f . -
1, 1
, ,b "- -d -
(
) . -. -.
.a .
f
..,m u .
Fig. 7. Appearance of ID near Elbow-to-Wedge Weld in Region from 6 o' clock (RHS) through 9 o' clock to 12 o' clock (LHS) after Dye-penetrant - Examination. The pipe material is at the top of the figure, m...~y..._y_-. ,- - 7,.,_,.-,,._,._.., ,,.
* ~
. , ~
'QY g. ,
- h. . ,,-
* = .- ~ n ,, . ,w.c' ; -
~w -vn,.
1 - L t , c.q[. '" %w s , ,,
, j.k)j -
p u, . . , ~ S. > ge .
,; w -
-: y ( '
j ,j
- . 4(p ,:' N" . -
, > s.
%i
- t. 3 i7
*3
(. { : l 4 : $ {~ _ .,:_.__, .-r il _s , .q
- b. .c:.-m 2- , c.-- m. . s._=..
. -,: _ m.L d:.MMJ Fig. 8. Same as Fig. 7 but in the Region from 12 o' clock (RHS) Through 3 o' clock to 6 o' clock.
23
l l 1-
< , 9 J ,.,'
w% .
;a gs '
e- .. ..
- ! k!i ! ! @!'
n ; 3)fl fkN
- . ,[ 1 ..? :..
.Y ~; X :
.. m.$ l . Cb; hjh $7- '&
Fig. 9. Circumferential Cracks Revealed by Dye Penetrant at 6 o' clock Position on Elbow Side of Wedge-to-Elbow Weld.
~ , --
e_ - ,,;. y.;m, - t 4 )d n ,,1p'
- w. ,,~.
's ys, I
t: 4 ab3r w
;gd Llp i .
@ ?;de ;iF.1$ ff s hh ,
Fig.10. View of OD Near Wedge-to-Elbow Weld Showing Dye-penetrant Indi-l cation (Arrow) on Wedge Side of Weld, at Approximately 6:30 Position. l l 24 i t
ry,c cz; ,% y+ n wrtg y , q,:'., ? , , l
., ,., w. ,
s - l
. . ,^
se 'j g)-. 1 F f tg i,n :n . e i s '7w _ m~ ___ .. :' .- l ht~ s I _ . . , -'O l c rp 7g... ng -{, g l1;1, w7 3: -s f ~ $. ..
'(QQl'ygcf - y:gsy;xp,
- 3,3'cy)-n.iQ A y *%%T)% ., .%,-w;e,5~O.;
.cA.r Mi.S.WicridMrastb' 7;;caryg' Ohe' WML Pig,11. View of OD Near Wedge-to-Elbow Weld Showing Dye-penetrant Indi-cation (Arrow) on Elbow Side of Weld at Approximately 8 o' clock . Position.
1 1[ f.'Tf Y ' W.f t
., [
m t].,k;iM 3 ,9
.,:. y. ' > ;
L,. - y.. . ,, s.>ry3
..;;cg
- y. .~.,t
~
[ -4
~
i;.
. ., q.,
I
< i
,,,re s _ ., n . <- _ jd4 ML%
l 8 f ;j Fig. 12. View of OD Near Wedge-to-Elbow Weld Showing Dye-penetrant Indi-cation (Arrow) on Elbow Side of Weld at Approximately 10 o' clock Position. 25
. _ - . . _ _ _ _ _ _ _ . - . - . ~ . . . -.-. .. -- --_ . _ _
- i. O li I
t
- g. ,%
\ ~ L. s
.i k
4 l
- 4 ,
l 4 I l i i 4 i l .
;f.
1
- 1
( 3
)
\ w ,~,. ) '- ! i j + ,
- x. s
', 4 l . v I
l 2 I
/
- I
' ' - 2 U d... -
1 I Fig. 13. Radiograph of Central Portion of Specimen Shown in Fig. 5. f 26 , i I
6 12 6 0' CLOCK 0' CLOCK O, CLOCK t g / t ~ O' CLOCK f I i ELBOW
-- k t u __'~~ ~ ~'~-i l---
% Li WELD L
/ I, i fifi I
'~x -'l' b \ - 2 A
Ff--lF - WELD L A - L
\
[ k [31 l I g a;. __ __ _ _ - i
'f :I q
ijC/ f B
' m_ B I
D I Fig. 14 Diagram Indicating
--- -- - ., _ _ ___ l.7 ___ -
Ibw Specimen Shown in { Fig. 5 was Sectioned for l Metallography. The red
.-% l lines indicate cracks de-tected by radiography
%_ ,,_ _ _ , _ ] (. } , #j j that were not seen using f
lk - I gfELD L - f dye penetrant, and the green lines show the lo-i Dn --60 ( i
) / cations of the various
( L.- i _ ._ _ - - - - L E "1 metallographic sections. Also shown is the strip l DIDC 8 IF L of naterial between I approximately the 5:30 I and 6 o' clock positions from which sections A-A _,,_,_ and B-B were taken.
OP @ A"y 9 , dW- o , . La
.~ ) N -' ~ \
/
\
.p .
1 k* - I I
. .,- j l ) .
, i-3
\
l 't r i 4
- , f A
t ( s, , V
.i .i . ,
l 10 0 m 5 Fig. 15. Intergranular Cracking observed in Meta 11ographic Section A-A. .
. .. : . m w. o,,
,. .,,.,,.,n..:,'gs..,,,,,, ~ ' ' ' '
E s
. $ja . .. Erf; ,. ~ .
- ; & 'A;?l. ' f'!'{
',: f.o, *fpe,h'1.{r3."
! [.Q ','. , e< . -s:'4,k ;'*&".*f.?':.:v'f:
Y ;a >
?
- .s ..w'-.... ~-: -
,1;i,;;. . . <<. :. w 1; ' s,.+1 a ,. s :-;,>
? . 1c: L.E, ;',.n* ..; :; > ;',{.
- Rj. % e: ;~.* .*t49
- l' -
-r"-
~
i .- r . r .
- l . ?.,"' G' 1,
, .;;. ,9 . ; <' .o
' v.e:.ac, w., ,..< >c
> j . . y47 ;r _ .
* ' ;;f 0.: fA :-2 .
;; >* . ' ,jn'j.5:,. .g f 9., ,
i , ~ O. y- n cr: $.1;;,y, w .<,, ,. .c. >- .,. . ,-
~
m ,,.. . ' ':. ' y k. 'e . 5' M.#,.
. /
s L'.
,53 .g. ..n,.., { -Z57'
'~*
,~
.: ,9
..:< ?..
.) c i
.r-
+
~,.1
. , ,, . E -
l s
.' 3 .
; 500p.m ,
.9 ,i ,
J. 1; < - Fig. 16. Penetration of Cracking in Section A-A into Weld Metal at Wedge-to-Elbow Weld. 28 l 1 l- __
I I I l l l l q: . .' - . 4
..4 ; .
^ '
. y y < ,. .
~
Jg,: 7 u,% 500. pm. ,
. . , .. I
.s r o s~4 *
*Qs
,: *' 5.
.,/ , . g '
- Ni y
. ~ . .
u .. 1' v, Np- .f5
* ~
$3*: N, ~ ,
i,
. ' , ' : -w' l r] m. ? * '
* +
Q - 4." af @l'*;;. e - yT .M- , .
~ ' '
y , s
- .w;nh , xe .t ,.
= + .
.- 50 29..l?,u.L.s.-Rg 7 . . ; *c' \ . . .
w
.w
, j, ,
~
.Dl^
[ 3 [M . I/ ,.
~) A
;) q '
*- f ',
[, ., ,,
' R p ', ,z. f iv/ T., }}.n ? [ v J.'A W 4'\ . .- '
', k, 6
)
,g e- '. a. < s q ,
'V ' ja,.-4 -)* ,
^ p *:-.', T . \ '
.,1 s '. .
' i
, ,. * > ; ij s, ..,t -- -
.O -
, .; G i)
, . !h . ;;-!
')
- h- 'A ) U'a 's i-k ;ty ,' ' +
, ~' .Q*l'
- r. 7,..
.db$if f J J.1['l'.Y '
"O 'A
.s %w-f, ' ' ' : e ,; [m.'., .
k.Q.
, : :c.~.- a. .
2 ,
. ,s , . , . . ,.~~
- 1. . . - , . .D?.:w.,% G , ' z'i . .~ 1 1 Q' ,
Fig. 17. Low-magnification View of Larger Circumferential Crack Otserved b cd in Section C-C. The intergranular nature of the crack is evident. V4 2EED b i l i
l
! s - T ;-
s - N . loopm. !
,'3,ge p'rrs-b. O.,f
'/' gj_
l
,T'Q ,':=-{. -
g' il' ~ ,, .[?ff N.'
;K \f; s ,. s p- m.
.I
*N.
, - } {f.{ ,'.u. ,
~
\
-f - -
,,3 d - \g
\ .9 ,
~l f
s- .
. .b / *,
s.2 --' a c , /1 . .. 64 n i (s %.- . , J
^
l i - '_ .f i. d_ '
/\ " -' '
'~
,d , t'
, _' 5$ a I ,
. ,~
* ,:q,
)/ ,. ,
. / a n 'l 1, n. : /. i v..l .
- a. p :7yr . . .
( Fig. 18. Higher-maFnification View of the Crack Shown in Fig.17. l I l l
- l W' s; - r-4 l , . ,
e 't .. . ff. ' , ,
$1
'j Y , ,
y i 1 i, m
.' 'j
' ' i,. .. ,.)'
'f/ ;
s i
.. , 3 4 * *t.,* 'te; ? ;
i , ,'$f, f4% < b4;f*q , N Q;~",,- Q,e - ;;
~*- 'L
,1 >;M ' ,
i >. * . . , ' 1 , l t ,
, d l, .
500 , m . Fig. 19. Low-magnification View of Smaller Circumferential Crack Present in Section C-C. 30
= . . . . ,
J I g myyyy y n~ * - , r w: -. l s e.s.or ./ t n';. ,rg : j,- .-
- c. y,, .,
i 2 r..b~,' i,L . '., 2
~
g%,\;C* y) ~ f,; '
., i
- g. m.4 % ,
g,M:. , '!,1 3 j
., t . i) y lh,yy ? W j:n 'qf
'3
- t s~ 5. k. . . ~ ;.7 p .- --
%'e-1, 3s.$d!E%dt(,,'
4 s , '
. k.:r'. n~yIe. N: , .. -
i .
'A' >a *
'.j;, y" .:.; {,4
- t' Q k:..- r .. ae;,w,p _
r+c@j',
,~
t
\ yg ?.J
.,uxui, inh,%- . '
~ -
. t J
,)
500pm
. . Q i Q
.c -
Fig. 20. Low-magnification view of Axial Crack Present in Section D-D. EEED 1 b PEJ i s . . 4 ' u i i. l
i l I f ' l
.a I ,.
i o
.a r
.i
.$ f, .. . ,
'( . f csy *
%q,?;:
ff []Ib%* . , j'r 3:'c cm pj,[ 9,Q: .y
+ ,
j h% Nv ' ff ., s f,fiQpe -
- 2.m..s .
i 4 ;w ,; v ' p4}Q;M
.g ' lmm l ip.((,N: ,J '
- l l l
g,f 3 ,
%. i. - .. ~ l t ;; ,? ; . 'c ' i. .
l 1- . . , ,~ l s' 3:\ ;l,,lm1 g . . - i-l *
-. * .* i' 1
La .- ;'t , . . :.1 Q- j*w. l I t Fig. 21. Axial Crack Present in f yy;7' , ':b 3 k. Section E-E Approximately i 6*?b'.i' S t.?,0 ; .' l "' ~ ,'i . 26 mils Below the Reference ' Plane. The crack is visible
dM
~ ..
'l Iyf'Q' [.T M _ c.. ,q[i to the upper right, and l ff* ' - q 2 ~' ' terminates abruptly at a
- pc region of lack of fusion at 5/- .
the wedge material / weld-metal (p ~ , interface. l ll '
')
*
- y' l
.= .
f % r i
~
=
+,,
o .
#~ 9
,,/;M.* 'g f.' s* egg;+M~ ; -..
.wa \ .
S, }, e . .
! D . , . .. .. ... .
C; > I. [ ' 'g .<, t r
! . 's p; . 4 *
j, s d, .
. l
#.2 .# . I j
l
%' ^ N. y a ud . . 's l
32
l i I i i l . l l I i i I o--- I
, . . x._e, ., . , y, s-3 ; 7..w.<
- y. ' je j
.j ( s.
! *m= ,],.
e g it j i - . 6
., n /,
1
.s . .-
\
' ' ~ , :. .y ],
. s,. w,
- k. ..
- : 'a i ..( - " ' .; ,.
l : .,/~ > by, , - ' , a
. ~.
- 9. . _ e.
',......,b
- i i - -
-k ' .p -Y',.3*A [ .% M42 i
i.
,,.. J
,.+3, ;
N,,, L. . .. we ev.g e..w ,,,,;f**f,s - . e mr,w t,m..3.. . i i , .y- 2_4. . -n ,
*-~.
,, .~v; -
; c;. -
, . p - l g f' ,-- , #m' ,, ;'_
~
s , i l Q f.,c' .y.4
... . s.
. ... '., g, 4 . _,
,e,
~ g . ,@,3- ..,
...a.
, ' . e-j .: , . . ' R %,., .g . , , . _ ,
. l u
w
* ^ *
- m ~
q , . _' .' :" . [y * { p,: gNd,,.s,.h r.,,
. .... -' W" $,Q .
4 , w ?
' , *_2 -
%QM , s. . -
i
, , , , . ,w -
.r 7 .- >
^-
. . w 6.,.. . ~.
%g' ..., .. ,
j
-Q -p 'f ; .[-"g7 " _ , %Q c.*w--J.gfprmweq we : :
g ' 3
, ^ ~ *w'r,%=.4 m.*W5; MsN ;. 3. e,.,% b t
l l l Axial Crack Present in Section E-E Approximately 61 mils Below the Reference Plane, Fig. 22. The wedge material is located to the right of the weld region and the pipe material M i M, to the lef t of the weld region in the photograph. , , I EsD' b t i l
f D ; 9 ,8,
" If h: s.
}W , ,..[ r.d,25 '
i iit t ' , -
%g'%
Fig. 23. liigher-magnification View of Cracking Present at Wedge Material / Weld-Metal Interface in Same Section Shown in Fig. 22, c: 'a ',
- 1S41
" gy p.y :-
- : ' }7 5
-A .y, . .
.a , Ejy."cc-ii " f4,l';. i l i;; .a : ; g$$fr -
?'I?l-l'l.2..)=
./,ml')
, ;}.)cl, *$.cQW
.O j g..
ung : r - y;]L c
.,7 A .
s., _- .
,2.:si .
c -u ;c a; t y 7.y y W,))r 93(,c.4;i. ~ft -1 m j; .,[ ~ . rw,! . ]y,. ., ,k : . S . 1 ,r . , = . 500 p.m .l.4', ' ,\. -' . ._
; ; , 2
. ...1- . . . .
Fig. 24 liigher-magnification View of Cracking Present Near Pii..- Material / Weld-Metal Interf ace in Same Section Shown in Fig. 2 2, 34
y..#. sv ,,,w
*r g T 7 - '- ~
-.1 - ^ ,. Ig C sid[s-j_Q + *
' - (%,,.gg ,,
u.
~,
.aw%
y,. q;:- .ym ,- .
; 2-- ..
f $ W^ Q~;((g
./ > g - -
,., ',:) $ l".:f.
W i.bj{W'n q
...p O p4
'. n
- ;- v.. . .;. ,
1 1
. y 3. 'g c- y, > .. . '. x ,
,%w, .
, t %.
y . 7, _ 2 - ,
?
m, 7 Clh.
,s - 5 n, r . i c,x. s
, ,, r;, s. t 1
.$: >V, , ~~ v' gf"8L
- _- .
t *' s.c g. cv,..a.4 $g..' . g...
...g4...-s u '
s&fw'? 3, . ' . ._,. . Dy' .. . .
.r-- a. ' ,
,f Ap~.
. %.. ..?; ! g,
* ~ ' , c,y.c. .
- ..~
, . y y ,*q, , ,;
%,..., 3- ;.s
?- ..+3
.. , , . 4
'
- Q,. . .
. _, } , r L. .
m.-u . , . .
' ~~ 3.
...,r y - , Qj Q-z%. > . . .w,
.,x .,
ym Ye.
. . . m_
~~. v : s' _V
~Y ., .w
. . - v
, :lh{ E'. (.p J
Sh Fig. 25. Axial Crack Present in Section E-E Approximately 100 mils Below the Reference Plane. (g The wedge material is located to the right of the weld region and the pipe material to the left of the weld region in the photograph. wo 2 2Ea c=='
=
7n,.7 e,r ._y7,,y y ,. p, 73 7.- 1. m , 7. .gc_o,7.
.n.y,.
. . q, pg
., h l l b f.g;93 +,[ I O
. _ l) re 4, p*
.s. - -
, %. .; W j . . . .
.. y Q- , .
,e- , =. . ..
,z-%, ,
4 ~,. ~, . _ n :, g- .a, t. 4_. 7-
*g- .r . .
>: . ..j t
E p c ; ';.; ; e
- .. .. j
- w. -ge ,+c
-- , :. (g ' t '
Y : ,^;,l- ). .
^
t ' VR' p :_ . -
"n' ' . );;. g . &,;.,;,.. w-.;;
ise A:t -
. g --
.. -- . . 3
-f r.gL, , . < .
. ,J . . , '
t ,' , yw . ...q 3 4
,l , . fl ' c'{~ LQ q _ . -
-C
( p-a. y_ , __s:, .
..b,,
>.9 s , , - - -
y ,,1.- e .
.. ; , [ ..
, 2 - .
o ~i . . ., .' ~- -/- g_ ; ;;".. ,y - . .a - s . . , o. ,: (l/p,j.1 R__ p i i 25 p.m i n?- G1-$_ 3{, l W Lo e, %g ., .Y) v e i p . y .. .
,, j -
. :/ g- vo Li;a~.mana/AL.iw ;f;l; l k %;h:3sj l Fig. 26. Higher-magnification View of Secondary Cracking Present in Weld
! Metal in Same Section Shcun in Fig. 25. f t \ l l . 1 l l ( i 36
v . ~
, , s.p. .m, T 7,
s ,' s c
"$[-. " . ; ' , f&[if"- .V, ,,
'., ;-, ;.: c m .
- r. .
. . s, *g "+<rr~, .?*~~~~,.
'.)
2y. y;I'hhNI ' ' A.....=W Qg'.,lt*;; 2
- f. -
t d-.:?;% ~ . ,y -r< v nm
--s y"w'
. - .- ~ ??' ' :
,3 3. y fi rjy % * .y
-- 3: s' %
!E $ ik. . $ .,,c ,
x 1 ~ ..tu. -
' wa 21Jy.;:f:d Q, - -O .? ." ',-
,cA- -s',
,f '.
;. ;xw L;,3v w ,, m
< , c i ~--. :
3 u- mau ~4- 4.,r; ~~ .. > . -
'Q m :;M. 1. - * - ~ - +
g} yxr;% rs
' ' i : y, ~.
y r e..2; m;. .:. .
~
r-m + _ .- 7~ ~ -
- s . ., _., _ _. <
v a p% s %,,. ,
.A a,
%A, -
_;. ' .. sW p we A.,,.s . ,, s- - - , . . .
"Lu ..
- a. .n23 _ ,_ imm p< .g .
_;,, , s., D ,.r _. . .a. % <
,, y~
~ ~;; ,~- ~,a 3 Fig. 27. Axial Crack Present in Section E-E Approximately 210 mils Below the Y Reference Plane. The wedge material is located to the right of the C weld region and the pipe material to the left of the weld region in Q the photograph. g
$E
~
s Y
*D N h
Dm "l I s A JM D
=
.!*.n.i. . < r .z 4. r.
. . .; ~ . .... < . .
e ,J- . : s .- . .,. h(, w. :.. .. - ). :e,*
.;...,., . . e ?.-v, ... 5,: . j s.4*, g
.:s'..n. ,. ,;} s. y h.:w.:!.%...t,. w.,-
y:?- .
. 3:.r. .\w.*.
- ~ *.
.A
. , , ye.,g K, y. y m e . .
p,r .g
.>, <., a.~. .y..~,. ..;. 1 ) f4. s , . -
u , "
'.. y c.
.n: s.Q,g..f.. .. a n :
.I .y,.< ra>. ..s ~..,.s.r...\.q .Ya .,,. ,.gR
't L . + . 2
- e. . . .
9 g
. ,21 .g .
. / 's.( .
$ ,?. . ? r' >,.
- . ;a 'p,q:.. . ,r o + .; . .,./. p. P.
. . 9.,l i
.i . . * .
* . <+
,3 . .-.h l '. ~ . - p. .f.
. : . ., ')* <
.,,. : . - , . -.. .., s. ... .g, 3.# ': . %,. -
.. .n. . . , . . . ~ .. ,. .. .
a : m. , . .,, . .,:. .. ~ . y.
...". . < . .;.... . . . :e . ; .,. .. ...;f.
. . . ,- .. .r,- .
. .. . ., m :s,. n. . s , ...
. .: .. ~.e.x - - ~ ;. , , ,;.
, ... . . . . . ., . 7 .y . r. .w.
- y-: - . p. .; -
-w7 .
. . i_ ..
.i n a
~. , m y .,.. : . e . .
'... n' v.. . . . :.x. i. , ~, .i
.. :g. ..
4 .. s. . .. r. -v..9 .r a n.~. s ~ y: s 6 , r ,'r:.* v'. , , .q.,.
- . ..2 2 7. ...,..,.~..,e.,... ..
,,e 100 pm e
. f.t,ey . a) .1.a - . - ~..., .c . . . . .w 5
, - . , e e
- y) ) ..* E
' ', .s e 1 e? '. v , . 6 ! ... .=*
..y.%'. l ~ , 41 - . ' , , y.
- Fig. 28. Subsurface Crack Present in Weld Metal in Section F-F.
m q.. - g p , , - g ~ , , , , . 7 %g 7
&R
, , [ Aj
~
,_ , \ % klj ef*.
,y.
Y.
- Y b " % #. , . sp 4 N. ; ' , g/ , ;8'll .? g$}l,1
~< s s
tuay-(' h' 1.y'{qj[R - .; p%y.1,,'g' )b /
- n )f.;M
, \
3 y :,, . % A.
- s. . . .+
c 4, y.t ::.,
,f % pe; Q )*[ e%
p
.,7 t
k x ,o ,;, , s ~--
.v , Q.n
_, y
'?Lp q f_Y;.L
,' f h.* l !.
g '.' * * %* lJy ~l ,C/ 'p M. NQl Q e;i'l W ';qlr,= 4 ;*
. ',f^
A .h;,s ,
&i$l?i : ,*) . --f 'y o; ~ * :* ? t
.; 'LQ * * / _ 'k , ,
.,.e 0 .s ,....; -(~ g; - , . .
t ,.. . s, . ., n' 1. &:. .s 1s'.. ,_ , ,e. t y',,<,. lGi
.~,
w\ Qf - l ,, P^^ _ ,' +
;[ k :\p _;y e a y },% , g}<.y %
kg
.,c. kw N
nt* .: M* > s.4* e p, * .e
. , . * /f e
*~
h ,- a % ;A...T#, w %*.y.,z.,z h,*., Ma ,. J .m N'ip':% . ,' v* Q'a g5 Mm h
.
- i #;.M...,. . y ;M Q . y >.t q ._ _ _
g' , _ *
., , h, .
t.
?*
. ae ..e.
.. . a : ". . Nic
.*'L,.
A, am a p i d' . " J 43r*7d, Fig. 29. High-magnification View of Typical Delta-ferrite Distribution Present in Weld Metal of Section E-E (Same Section Shown in Fig. 27). 38
om o Q J w o Ju o Ju _ . (TU m 1
'{1 h
h e. C
& y
,250 pm ,
# 1 :Xsa9*V11 Fig. 30. Heavily Attacked Microstructure Observed in Wedge Material Near Wedge-to-Pipe Weld af ter Etching According to ASTM Procedure
. A 262, Practice A.
4- 4 .= v r7 m y(; . v3 13 L _7
/ / .s
'I
- r. ',
. .fj -
j( ,
%@&1 -
. N
$LpL 9['14h.A ?.; ,gy
[ I Yb k'%'jrMb
'rie ; % 4n ><1 .N .-~ c, '3 1 s 25 Fm . 'd Fig. 31. Lightly Attacked Microstructure Observed in Elbow Material Away from Heat-af fected Zor:c after Etching According to ASTM Procedure A 262, Practice A.
39}}