ML112371640

Rept of Metallurgical Examination & Stress Evaluation of Recirculation Inlet Nozzle Safe-End Cracking, Rept IE-116

ML112371640
Person / Time
Site:	Duane Arnold
Issue date:	03/16/1979
From:	Weiss S, Dean R, Foley W Parameter
To:
Shared Package
ML112371641	List: ML112371640 ML19269D966
References
CON-NRC-05-77-186, CON-NRC-5-77-186 NUDOCS 7906210333
Download: ML112371640 (87)

Text

(i)

Report No. TE-116 Metallurgical Examination a.nd Stress Evaliation of Recirculation Tnlet N;:.:1e Safe-End Cracking at Duane Arnold Energy Center Iowa Electric Light & Power Co.

Dr. S. Wiss R . S. Dean PARAMETER, Inc.

V. Pasupathi G. P. Smith D. R. Farmelo J.. S. Perrin Battelle Columbus Laboratories Prepared for U. S. Nuclear Regulatory Commission 7 906210333

(ii)

.4 NOTICE This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Govern ment nor any agency thereof, or any of their employees, makes any warranty, expressed or implied, or assumes any legal liability or responsibility for any third party's use, or the results of such use, of any information, apparatus, product or process disclosed in this report, or represents that its use by such third party would not infringe privately owned rights.

The views expressed in this report are not necessarily those of the U. S. Nuclear Regulatory Commission.

Available from U. S. Nuclear Regulatory Commission Washington, D. C. 20555

(iii)

IE-116 Distribution Copie s Nuclear Regulatory Comnission Technical Assistance Branch Office of Inspection and Enforcement 40 PARAMETER, Inc.

File Copy 1 S. Weiss 1

Report of Metallurgical Examination and Stress Evaluation of Recirculation Inlet Nozzle Safe-End Cracking at Duane Arnold Energy Center Iowa Electric Power Company Report No. IE-116, March 16, 1979 Prepared for: United States Nuclear Regulatory Commission Office of Inspection & Enforcement NRC Contract 05-77-186 PAR: NRC-IE-78/79, Task 03 By:

Robpyrt S. Dean, P.E.

Walter J. F> ley, P.E.

PARAMETER, Inc.

Consulting Engineers Elm Grove, Wisconsin

Page 2 SECT TON I Abstract Presented is a report of a metallurgical investigation and stress analysis study of a cracked safe end from a recirculation inlet nozzle from the reactor at Duane Arnold Energy Center. The work was coordinated by PARAMETER, Inc. at the request of the Nuclear Regulatory Commission. The metallurgical investigation was per formed by Battelle Columbus Laboratories, and the stress analysis study was made by PARAMETER staff members. Results correlate well with a parallel investigation of a companion nozzle safe end directed by the Iowa Electric Light and Power Company at Southwest Research Institute. The cracking is identified as intergranular stress cor rosion cracking. The crack originated in the weld heat-affected zone at the tight crevice formed by the safe end and thermal sleeve joint, and progressed radially outward from 30 to 80 percent through the Inconel safe end wall. The crevice provided a location in which contaminant build up was able to occur, as well as a region in which high localized stresses were concentrated. Review of the replacement safe end design shows eliminationof the tight crevice and reduction of stresses to levelsithought to be low enough to avoid recurrence of the cracking.

Page 3 CONTENTS Section Pag'e Abstract 2 Contents 3 IT Introduction 4 Summary of F' indiiins 5 Conclusions 6 ITV Review and Concluisi-ons from 7 Battelle etallurgical Report Dr. S. Weiss, Consultant to PARANMETER V Comments on Meta llargical Findings 11 at Battelle and Southwest Rkesearch R. S. Dean, R1ZAIRMETER, Inc.

VI Review of Stress Levels-Original Design 14 VT1 Review of Stress Levels-Replacement Design 1.7 VII References 18 IX Exhibits and Attachments 19 Exhibit A: Battelle Columbus Laboratories Report BCL-585-9 January 1979 Examination of Inconel Safe End from Duane Arnold by V. Pasupathi, et al Attachment 1: Stress Tabulations, CalculAtions, and Comparisons by R. S. Dean, PARAMETER, Inc.

Page 4 SECTION II Introduction Task Order #3 on Parameter NRC Contract #05-77-186, issued by The Office of Inspection and Enforcement of NRC, requested assistance and consultation services for a metallurgical evaluation of a nozzle safe end from the reactor vessel at Duane Arnold Energy Center. This safe end was from a companion nozzle to one that had cracked-through, and was being investigated under the jurisdiction of the Licensee, Iowa Electric Light and Power Company. The main intent of the parallel Parameter study was to verify findings of the Licensee investigation.

In addition, the task order requested assistance in review and evalu ation of the Licensee's failure analysis of the cracked-through safe end, and assistance in review of the Licensee's design basis analysis of the recirculation inlet nozzle.

The Duane Arnold Energy Center is a nuclear power generation plant with a boiling water reactor, operated by Iowa Electric Light and Power Company, and located at Palo, Iowa, near the City of Cedar Rapids.

General Electric wasIthe reactor designer, and Chicago Bridge & Iron, CBI Nuclear, was the' fabricator. The reactor was assembled on site by CBI.

The subject safe ends are from the recirculation inlet nozzles of the reactor, of which there are eight, all of which had indications of cracking as found by non-destructive inspection at the site after leak age had been detected from the cracked-through safe end. The nozzle designation for the Parameter investigation is N2E.

The N2E safe end as cut from the vessel was shipped without decontamin ation to Battelle Columbus Laboratories for the metallurgical studies.

This was done with coordination at the site by Mr. Raymond Sutphin, Quality Assurance Engineer, of Parameter.

Radiographs of the nozzle N2E safe end, taken on site prior to cutting out the section, were carried to Parameter by Mr. Sutphin. These radio graphs were reviewed at Parameter by Mr. Kenneth Ristau, Consultant for non-destructive testing, and Dr. S. Weiss, Consultant for metallurgy, and during this review, locations were selected for sectioning for metallurgical study.

The metallurgical investigation at Battelle was conducted under the direction of Parameter Consultant, Dr. Stanley Weiss. The Battelle report is Exhibit A of this report. Dr. Weiss' evaluation, conclusions, and comments are Section IV of this report.

The review of the original and replacement design analyses of the re circulation inlet nozzle were made by Parameter staff engineers. The discussion is included herein asSection VI and VII. Stress tabulations, calculations, and comparisons are included as Attachment #1.

A summary of the more significant findings and conclusions are given following this introduction.

Page 5 SECTION III Summary of Findings

1. Metallurgical Analysis 1.1 The mechanism and mode of cracking is identified as inter granular stress corrosion cracking.

1.2 The crack originated at the crevice formed by the safe end and thermal sleeve, and extended a full 3600 circumferentially around the safe end.

1.3 Sulfur was observed on the fracture surfaces and adjoining crevices.

1.4 The crack originated in and propagated from 30 to 80 percent through the safe end wall from the weld heat-affected zone of the safe end/thermal sleeve weld joint.

1.5 The heat-affected zone exhibits partial re-solutionizing of the grain boundaries, wlhich exhibit significant sensitization in the Inconel base material.

1.6 No corrosion pitting attack or multiple cracking was observed.

Only a single crack was found.

2. Review of Stress Analysis, Original Design 2.1 The original stress analysis identified the safe end/thermal sleeve joint as a high stress location, due to the stress concentration effect of the crevice.

2.2 Fatigue analysis at this location per Code with added conser vatism showed that the design was acceptable to the Code.

2.3 Normal operating loading conditions cycle stresses to over yield strength of the material at the crack location, due to the stress concentration effect.

2.4 The original design was acceptable from a stress and fatigue analysis standpoint for the extreme sudden recirculation pump startup. This transient was never experienced in actual plant operation.

3. Review of Stress Analysis, Replacement Design 3.1 Safe end operational loading stress levels are reduced to approximately 50% of those of the original design.

3.2 Residual stress from the safe end/thermal sleeve weld is reduced to below yield stress level.

3.3 Normal operation loading stress, including stress concent ration, cycles within the yield strength range of the material.

3.4 Extreme transient stress range, including stress concentration, only slightly exceeds yield strength range. Fatigue evalu ation shows this to be well within Code acceptable limits.

Page 6 SECTION III - Continued Conclusions

1. The metallurgical investigation of the N2E safe end at Battelle Columbus Laboratories under Parameter direction are in close agree ment with the findings of the Licensee directed investigation of the cracked-through N2A safe end conducted at Southwest Research Institute.

2. Main factors contributing to the intergranular stress corrosion cracking of the original safe end design appear to be:

2.1 Presence-of the tight crevice, causing:

2.1.1 Stress concentration 2.1.2 Contaminant build-up 2.2 Relatively high stresses at cracking location, although not over code allowable.

2.3 Susceptibil'ty of the Inconel to cracking due to the sensi tization by heat treatment and welding.

3. The proximity of the repair welding on the O.D. of the original safe end appears to have had no effect on the cracking.

4. The importance of residual stress from welding is an unresolved question. It is a possibility that residual stress could have been a major source of stress to initiate the cracking, but the stress analysis review shows that pressure and thermal load stresses were high enough for intergranular stress corrosion cracking to occur without residual stress being present initially.

5. The replacement design accomplishes objectives in:

5.1 Reduction of stresses from operaLing loads.

5.2 Reduction of residual stresses from the safe end/thermal sleeve weld.

5.3 Elimination, or at least drastic reduction, of the objection able effects of a tight crevice by replacement of the crevice with an annulus with controlled dimensions.

Page 7 SECTION IV Review and Conclusions f rom Battelle Metallurgical Report by Dr. S. Weiss, P.E.

Metallurgical Consul tant to PARAMETER, Inc.

Page 8 Review and Conclusions from Battelle Metallurgical Report by Dr. S. Weiss In accordance with the request of the Nuclear Regulatory Commission, a metallurgical examination has been performed on safe end N2E at the Battelle Laboratory under the direction of PARAMETER, Inc. The ob jectives of this study were to:

a. Conduct an independent study to compare with a parallel study performed by the Licensee on a companion safe end section containing a through wall crack.

b. Determine the mechanism and mode of failure in safe end N2E.

c. Provide assistance in the review and evaluation of the Licensee's failure analysis of a through wall cracked safe end reigoved from a companion BWR Nozzle.

and d. Provide assistance in the review of the Licensee's design analysis of the recirculation inlet nozzle.

The work at Battelle is now completed and has been reported on in the final January 1979 Battelle Data Report entitled "Examination of Inconel Safe End From Duane Arnold" by V. Pasupathi, et al.(Exhibit A)

Additionally, considerable information and numerous inputs into the understanding of the problem were gained by interim visits to Battelle Laboratory during the course of the investigation, visiting the South West Research Institute (SwRI) to review their preliminary findings on behalf of the Licensee and attending numerous meetings between the Licensee, General Electric, NRC, Battelle, SwRI and PARAMETER, Inc.

The most relevant findings from the Battelle study are:

1. The mechanism and mode of cracking in Nozzle N2E is identified as intergranular stress corrosion cracking.

2. The cracks observed originate and propagate from a relatively tight crevice which ranges in length from 0.2 to 0.3 in. and ranges in gap size from 0.002 to 0.005 in.

3. Sulfur was observed on the fracture surfaces and adjoining crevices. Concentration profiles show that the intensity of sulfur was found to in crease towards the crack tip. The level of in tensity of sulfur observed in these studies re vealed that the sulfur found was present as a contaminant and not as an inherent constituent of the base materials.

Page 9 Review and Conclusions from Battelle Metallurgical Report - continued

4. The cracks originate in and propagate from the weld heat-affected zone of the weld joining the Inconel 600 thermal sleeve to the Inconel 600 safe end. The heat-affected zones of these welds exhibit only partial re-solutionizing of the grain boundaries,.which exhibited signifi cant sensitization in the initial safe end base material prior to welding.

5. No corrosion pitting attack or multiple cracking was observed in any of the specimens analyzed.

In each specimen only a single crack was found to originate in the heat-affected zone.

6. No apparent evidence of fatigue or cyclic load ing was observed. However, this does not pre clude the possibility of crack propagation by low cycle corrosion fatigue in addition to that attri butable to ISCC, since these two modes of fracture may have similar fractographic appearances in this application.

From these findings it is concluded that the major factors contributing to cracking and failure of these nozzles appears to be:

A. The presence of a tight crevice within which localized chemical reactions and conditions are occurring.

B. The presence of high localized applied stresses of yield strength magnitude at the safe end weld.

C. The presence of high localized residual welding stresses at the same location resulting from field welding of the thermal sleeve to the safe end.

D. The presence of an adverse sulfur rich chemical environment which is known to promote stress cor rosion cracking in high nickel base alloys. Exa mination of the performance history of the Duane Arnold Reactor may indicate the potential source of this contaminant.

These findings appear to agree with the preliminary report of studies performed at SwRI on behalf of the Licensee. The final report from SwRI was not reviewed and thus a direct comparison with the Battelle final report has not been made. However, it is believed that the findings of each of the studies essentially corroborate one another.

The stress relieved repair welds introduced to the safe ends prior to final field fabrication did not appear to be a contributing factor to the cracking problem experienced. The thermal sleeve and safe end

Page 10 Review and Conclusions from Battelle Metallurgical Report - continued were both sensitized exhibiting discrete precipitated carbides at the grain boundaries. Although these base materials exhibited sensitized conditions, no stress corrosion cracking or corrosion pitting was ob served in either material at the faying interfaces within the crevices adjacent to the weld heat-affected zones which contained the crack.

The findings of the study confirmed that re-design concepts must take the following recommendations into consideration:

- Minimize applied and residual welding stresses to a safe level by revising the structural design.

- Eliminate crevices and crevice conditions in welded joints.

Study of the cracking problem experienced in the Inconel 600 safe ends at Duane Arnold has raised numerous important questions which can be answered only by formalized resedrch and developmental programs. Among those factors and questions which appear to be of vital importance are:

- The role of both the applied stresses and residual welding stresses in causing these failures.

- What geometry constitutes a crevice?

- What is the influence of specific BWR environments that can potentially cause IGSCC in the presence of a crevice, weld, and stress condition?

- Which non-destructive test methods are most reliable for detecting initiation and propagation of cracks such as were observed in these studies?

Page 11 SECTION V Comments on Metallurgical Findings at Battelle and Southwest Research by R. S. Dean, P.E.

Staff Engineer PARAMETER, Inc.

Page 12 Comments on Metallurgical Findings by R. S. Dean, P.E.

PARAMETER Staff Engineer Cause of Failure Ample evidence was found that intergranular stress corrosion cracking (IGSCC) of the inconel safe end material was responsible for the cracks in the Duane Arnold recirculation inlet piping. This is well presented and stated in the Southwest Research Institute interim report (Ref.l,pg.

40) and is substantiated by the Battelle Columbus Laboratory report (Ref.2,pg.43).

Conditions Required for IGSCC Reference (1) infers (pg.40) and Reference (3)(pg.2-1) states that three conditions are required to initiate IGSCC: 1) a susceptible material,

2) an aggressive environment, and 3) stress. It seems agreed that yield stress level is required locally to the zone of crack initiation (Ref.1, pg.40; Ref.3,pg.5-1).

Conditions Causing IGSCC at Duane Arnold:

Material Susceptibility Reference (1) sites published data supporting susceptibility of Inconel 600 to IGSCC in high purity water environments (Ref.1,pg.40).

Analyses of the chemical composition of the safe end material verify that it is Inconel 600 (Ref.l,pg.40; Ref.2,pg.43). Cracking occurred in both sensitized and re-solution treated zones in the SwRI study (Ref.

1,pg.40), and only in the partially re-solution treated zone in the Battelle findings (Ref.2,pg.43).

Crevice and Aggressive Environment The presence of the crevice is agreed by References (1) and (2) to be a major contributing factor to the cracking. All cracking initiated near the tip of the crevice formed by the safe end/thermal sleeve joint, with only one initiation location at any given radial section. No incipient intergranular attack, pitting, or other evidence of significant chemical attack was found present along the surface of the crevice by either in vestigation (Ref.1,pg.39, Ref.2,pg.43).

Reference (1) sites the crevice as an entrapment location for concen trations of the necessary corrosive environment for IGSCC (Ref.1,pg.40).

Stress Both the SwRI and Battelle reports describe the safe end cracking as relatively regular in depth circumferentially a full 3600, and that crack progression through the base material shows no evidence of being step-wise in nature (Ref.1,pg.40; Ref.2,pg.43). Initial cracking pro gressed along a single line for some small percentage of depth, and then some intergranular branching occurred in the deeper sections, possibly

Page 13 Comments on Metallurgical Findings - continued Stress - continued indicating relief of the initial very large tensile stresses by the first straight length of crack. Reference (1) expresses the opinion that this initial stressed condition is due primarily to the residual stresses caused by the un-stress-relieved welding of the thermal sleeve to the safe end (Ref.l,pg.42). After this stress was relieved by the initial crack, normal load stresses caused by pressure, temperature, and other mechanical loads present, would contribute stresses high enough at the tip of the crack to continue the progression. Even though primary and secondary load stresses could be relatively low, peak stress, due to the stress concentration by the presence of the crack, could be expected to exceed elastic limits and continue to propagate the crack.

Page 14 SECTION VI Review of Stress Levels, Original Design and SECTION VII Review of Stress Levels, Replacement Design by R. S. Dean, P.E.

Staff Engineer PARAMETER, Inc.

Page 15 SECTION VI Review of Stress Levels, Original Design by R. S. Dean, P.E.

(A sketch of the safe end/nozzle geometry is shown on Page 3 of Attachment 1.)

An adequate stress analysis per code was made at the time of design by Chicago Bridge & Iron in Reference (4). In this analysis, the section through the safe end at the tip of the crevice was identified as the highest stressed section of the nozzle assembly. A fatigue analysis was made at point 13 (Ref.4,F8-14 thru 30) on the safe end at the tip of the crevice. A theoretical stress concentration factor of 4 was used for the analysis at this point, and a usage factor of .515 was calculated.

In the code (Ref.5,pph.N-415.3,pg.28 and Ref.6,pph.NB-3222.4-e(2),pg.64) evaluation of stresses at structural discontinuities, the statement is made, "Except for the case of crack-like defects, no fatigue strength reduction factor greater than five need be used." On the basis that the crevice is quite crack-like, it is felt that a factor of at least five should have been used, since no experimental or other basis for using four is given.

In order to make a comparison using factors of both 4 and 5, Attachment 1 presents a fatigue analysis made strictly per the ASME Code procedure (Ref.5,pph.N-415.2). The original analysis used an elastic-plastic method per Reference (7) which modifies the alternating stress for use with Code design fatigue curves (figs.N-415(a),(b)), and arrives at more conservative results.

Also, it appears that the design fatigue curve N-415(b) as in the 1968 edition of the Code (Ref.5) was used in the .original analysis. This curve was corrected in the summer 1968 Code addenda to agree with the curve in the 1965 edition of the Code, which also agrees with the current (1977) edition. However, the corrected curve is less conservative than the one used (1968). Attachment 1 uses the corrected curve.

The results of the fatigue analysis of Attachment 1 show that the cumula tive usage factor is quite low; 0.09 for k = 4, 0.190 for k = 5. Both results are less than the .515 with k = 4 calculated in the original report, and much less than the Code allowable of 1.0. This clearly makes the original design acceptable on the basis of fatigue analysis per Code.

Also presented in Attachment 1 are comparisons to yield strength of point 13 axial stress and stress intensity range of the three cyclic conditions used for fatigue analysis. According to the operation history of Duane Arnold (Ref.9, Attachment B), the sudden recirculation pump startup (Transient 1) was never experienced, so the very high stresses (strains) of this transient never occurred. Axial stress from hydrotest is less than yield strength, and stress intensity range from hydrotest is well within the 3 Sm Code allowable. However, add to this the residual stress expected as a result of the sleeve-to-safe end weld, which is of the order of yield stress (see Att.1,pg.14), and the stress at point 13 is cycled through a shakedown at first pressurization. This might initiate crack ing, but normally it would not be expected to do so because of the ductile nature of the Inconel.

Page 16 Review of Stress Levels, Original Design - Continued A plot of cycling axial stress (without peak stress) at point 13 is shown on Page 14 of Attachment 1, assuming an initial cyclic history of the nozzle very roughly based on reported history (Ref.9, Att.B).

The plot indicates that after an initial shakedown of stress, from residual through two hydrotests and shutdowns, cycling axial stress (without peak stress) of the milder transients and normal startups and shutdowns would fluctuate thereafter with tensile yield stress as a maximum. One cycle of the extreme cold pump startup is shown to show the shift in mean stress that it would cause for subsequent milder cycling. This transient did not occur in the actual history, as men tioned earlier, but is of interest for comparison with the analysis of the replacement desiqn.

The presence of the crevice, however, creates stress concentration peak stresses as tabulated for the various conditions on page 10 of Attach ment 1 (also Ref.4,pg.F8-26). The tabulation shows that even normal startup and shutdown transients cause point 13 psuedo-elastic stress to exceed yield strength, and cause local plastic strains to occur. The more severe cool-down-warmup trahnsient #2 causes greater strains, of course. These strains are local on the safe end I.D., and do not cause distortion of the full cross-section.

Stress cycling as described in the previous paragraph is normal Code design practice for high peak stress locations in ductile materials, keeping the cumulative usage factor less than 1.0. However, this local strain cycling, along with steady state (sustained) stress condition at yield stress level, in the presence of susceptible material and corro sive environment, provides the conditions which promote stress corrosion cracking.

It could be concluded from the plots of cyclic stress that the presence of the initial residual stress from welding is immaterial. The stress mechanism for stress corrosion cracking is present whether or not there is initial residual stress. Normally, with a ductile material, it is unlikely that the residual stress would be severe enough to initiate the cracking immediately. Also, the corrosive conditions required for inter granular stress corrosion cracking would not have been present initially when the weld was made, so it is doubtful that the residual stress from welding initiated the crack.

The relative uniformity of the crack around the entire periphery of the safe end, as found by Refs. 1 & 2, leads these references to the con clusion that the residual stress from welding was the principal contri butor to cracking, because of the relatively axisymmetric nature of re sidual stress~from the sleeve to safe end weld. The main load stresses are from pressure and temperature, and also are axisymmetric in nature.

The non-axisymmetric stresses from piping loads produce stress of some lesser magnitude, and, in addition, although it is not analyzed specifi cally, the direction of the piping loads would be expected to vary during temperature and pressure cycling, causing principal stress directions to vary. There would be a resulting tendency for this also to contribute to the cracking all around the periphery, rather than in one circumferen tial location only, as might be thought would result from just piping load stresses.

Page 17 SECTION VII Review of Stress Levels, Replacement Design by R. S. Dean, P.E.

On page 11. of Attachment 1, comparisons are made between stresses at the point on the inside wall of the safe end (point 130) and corres ponding point 13 of the original design. The replacement design avoids the sharp crevice by substituting an annulus with controlled radii at the tip, keeping the stress concentration much lower, and the safe end wall is thicker at this section (.96 in. vs. .57 inches on the original).

These changes are responsible for reducing) the stress levels by nearlV 50 percent. Fatigue analysis results in a very low cumulative usage factor of .002, compared to .09 for the original design, and to maximum Code allowable of 1.0.

The replacement design also reduces the level of residual stress in the safe end wall by removing the weld from direct attachment to the wall, as with the original design. The flexibility of the joint to the thermal sleeve allows elastic displacement to accommodate radial weld shrinkage, and residual stress is kept below yield stress (pg.11 of Att.1).

For a comparison with cyclic axiai stresses of the original design, use is made of the superpositioned total of axial stress from the various loadings calculated in Ref. 8 for the replacement design (see Att. 1, pg. 12). These are taken at a critical time during the extreme sudden pump startup transient. Primary plus secondary axial stress is plotted, page 15, using stress without thermal component for normal startup and shutdown cycles. The plot shows stress shakes down to the range between yield stress limits, whereas, for the.original design, the last cycle plotted on page 14 shows the primary plus secondary stress (psuedo-elastic) far exceeds the yield stress.

Total stress, including peak stress, using the extreme pump startup transient, only slightly exceeds yield stress (Att.1,pg.13). For the normal controlled startup and shutdown cycles, total stress remains within the yield stress range, which it did not do in the original design.

It is reasonably certain that total stress for the mild heatup and cool down transients also would cycle within the elastic range.

Page 18 SECTION VIII References

1. Burghard, H.C., Jr., Metallurgical Investigation of Cracking in a Reactor Vessel Nozzle Safe-End, Interim Report, SwRI Project 02-5389-001, Southwest Research Institute, October 20, 1978.

2. Pasupathi, V., Smith, G.P.,Farmelo, D.R., and Perrin, J.S.,

Examination of Inconel Safe End from Duane Arnold, Battelle Columbus Laboratories,. Final Report, Project BCL-585-9, January, 1979.

3. Lemaire, J.C., Ranganeth, S., Prevention of Stress Corrosion by Limitation of Applied Static Loads in BWR Piping and Components, General Electric, DRF6880052, Nuclear Energy Engineering Division, San Jose, California, September, 1978.

4. Stress Analysis Report of Original Recirculation Inlet Nozzle N2 (including Safe-end Stresses), Sections T8 (Thermal Analysis, S8 (Stress Analysis), F8 (Fatigue Analysis Chicago Bridge & Iron, Rev. 1, November, 1971.

5. ASME Boiler and Pressure Vessel Code,Section III, Nuclear Vessels, 1968. with Summer 1968 Addenda.

6. ASME Boiler and Pressure Vessel Code,Section III, Division 1, Nuclear Power Plant Components, Sub-section NB, Class 1 Components, 1977.

7. Tagart, S. W., Plastic Fatigue Analysis of Pressure Components, ASME Paper No. 68-PVP-3, American Society of Mechanical Engineers.

8. Stress Report, Recire Inlet Nozzle Safe End Replacement, Duane Arnold Nuclear Plant, CBI Nuclear Company, August 10, 1978.

9. Responses to five questions concerning The Duane Arnold Energy Center Recirculation Inlet Nozzle Analysis, letter to J. G. Keppler, U.S. NRC, Region III, from Lee Liu, Sr. V.P., Engineering, Iowa Electric Light and Power Company, July 26, 1978, IE-78-1137.

10. Recirculation Inlet Safe End Repair Program, Duane Arnold Energy Center, Iowa Electric Light and Power Company, IE-78-1782, December 8, 1978.

11. Drawing, Original Design, Thermal Sleeves for Recirculation Inlet Nozzles Mark N2 A/H, Chicago Bridge and Iron Company, Contract 68-2967, Drawing No. 32, Sheet 7, Purchaser's No. 205-1289.

'12. Drawings,Replacement Design:

1. Reactor Vessel, General Electric Drawing No. 794E904, Sheet 2, Revision 0.

2. Safe End, General Electric Drawing No. 112D2504, Rev. 0.

3. Adapter, Thermal Sleeve, General Electric Drawing No.

137C7284, Rev. 0.

Page 19 SECTION IX Exhibits and Attachments Exhibit A:

BattelleColumbus Laboratories Report BCL-585-9 January 1979 Examination of Inconel Safe End from Duane Arnold by: V. Pasupathi G. P. Smith D. R. Farmelo J. S. Perrin Attachment No. 1:

Stress Tabulations, Calculations, and Comparisons for Safe End/Thermal Sleeve Joint Location in Reactor Nozzle N2 at Duane Arnold Energy Center by: R. S. Dean and W. J. Foley PARAMETER, Inc.

Page 1 of 15 Attachment No. 1 to Report IE-116 Stress Tabulations, Calculations and Comparisons for Safe End/Thermal Sleeve Joint Location in Reactor Nozzle N2 at Duane Arnold Energy Center Iowa Electric Light and Power Company Prepared for: U. S. Nuclear Regulatory Commission Office of Inspection & Enforcement NRC Contract 05-77-186 PAR: NRC-IE-78/79, Task 03 by: PARAMETER, Inc.

Consulting Engineers Elm Grove, Wisconsin S ROBERT S. '

DEAN T E8826t E rS ELM GROVE, .

S0\ WIS. IF N ** Robert S. Dean, P.E.

Checked by: -

Walter J. Foley, P.E.

Page 2 Introduction Sketches of original and replacement safe end design are shown on page 3.

For the original design, a re-calculation of fatigue analysis at point 13 is presented, following 1968 code procedure, and using the code fatigue curve of the summer 1968 addenda. Strength reduction factors of both 4 and 5 are used for comparison of resultant usage factors.

Longitudinal stresses are compared with yield strength for each signi ficant cyclic load condition. Cycling stress intensity range is com pared to 3 Sm limit and is roughly plotted to illustrate shakedown and the effect of adding in residual stress estimate.

Stresses from the replacement design report for the comparable location on the new safe end are tabulated and plotted for comparison to the original design.

Summary and Discussion See Sections III, VI, and VII of report IE-116.

References See Section VIII of report IE-116.

Comments on Safe End Replacement Design A comparison of the sketches of the two designs for the safe end show that the replacement design accomplishes the following:

1. The wall thickness is increased in the tapered section at the thermal sleeve attachment location which reduces stresses from operating loads.

2. The sleeve-to-safe-end weld joint is separated from the pressure boundary wall of the safe end, preventing any possible cracking in the weld heat affected zone from propagating through the pres sure boundary wall. Weld location on the relatively flexible projection from the safe end accommodates circumferential weld shrinkage, reducing residual stresses from welding.

3. The crack-like crevice is eliminated, and is replaced with an annulus of controlled dimensions. This reduces the stress con centration by over 50%, and substantially reduces the tendency to trap corrosive contaminants at this location.

'S ere. N0 Of

/

-- I 0- /

"1" '-4

/Wro*E, .5a r EV 5#7

/1 57 L 112

- /NCO'V.L 7/EP7eAL C0Eet/Ce U i47~

_~4gg~~IT (eep~ /~

SAFE toL II 70

• O 14.

13.1? .p io'*z7#

Id&'#Wt. T~fA1AL AZ- SAF= EAID el//d4 L AAJ CEPLACt [- r SCIAt, - 4Pweo'i 4

1r m 4

& Anod P/-c /V 044,' V-'h~

n/F' EAA/, A s-7Inle'&fl

/ Fo 4 - 19 4 A/L~) t2 ~

L

///e k e t OO / Sd-~//b (A'.' v L/,~T~O ~y (AREv $ 4 b) ,4 .57;F-

£i'v. .L./ -$Wj-TV  : Ca ./4 /so 0 s 2$o .J EL~ri 5Tw UU6'z3*

Jie.L~ A £4&J4 .4,E6 TA4.^J fIhe, A F 4 ,,47- ,4/4t-4S/4 L 0OAl PO-" A ge-r 4), FO -, 7ru, z).

,*RESS A/04l/BAlt47vJ45F, (4) 7EA/4/o) 0-(7'- OA/4,7 d,4 xIAt) .aMe-ss O- "A/ .).743ss srvrC-t&&/F-7 (I4'2 S. A..Sra~st4

-S $es Aa/.s r

/ 0 "

,4 o' C

$ z /4red't A/,#ral 77 .#4kS 1474 A4 t-r A 2 0?A(CULAr71dAj 0OC FAr7-/ u 6 x~'vasa ,4r /l/or /3 NCOttA / 4}0(t1 4teoed Coo f 40/4 J CA/(*rf VEOF 74E JA#/79E.

'4 A0d ^JO 0 5wea/4rfl ac /vrro A) FA'010 O eaF 4o7;

(,gA A.> S 14e r4dED C;C r 4W.0 A L 0 4-7-Lt447 M4E - r4C7Vid. 5 -7) 9 4 v$ ^/4A7e 7*/lt4% R SAdF i C4<:*ZtV 7 <;',:Lm7 4A,47> /

A'o L) bb E4 / P4.5/( ^h???#4/T CL .44 C~aldinOAAJ , fli^

A- ~ r J( CT kVJ/~ ~

/

~

('(A,)

/ ,'- n-A; u~ '3z Ci 1 ~ ~

1'~~ ~

~5. k/,3A~d)(k~ ~?C QQ,,4~JAJ --

/ / e P 77C 76 AT i/ZA #i- F-:/L 74 - /ors, ) 4st, ,r sd 94E ,4E 4.'75S q7/ v' rvsef 4/774,oddY

,s 77,t ,40a 7 4 04 m 4fA4 Lo'-as, el 78 (7 /7o 7 2S33) 41 / 7 Z3 aOr~/.- orem OP CAo.~/C, estm:

SV = f49 r3z I r04

C A- I~ sunjar t - --------------------- Swar? No.. ... o ...

ZFnKo. -- Joe N O. .

-- oa ----.--

A T..... ........ ------------

//DA'. Cre -aS 7 -tdy k 4. th;/f, 45

/)7/:4,4'#7n/4 &ri £ 3 (.& 5) F/iA 1e)

<' <-E r e-,o JI 20 (4/od5  ? 4foevj P-3 FeLL4a I . /e Y &

b oc

,4unwast$Le CA/ctde: =; lo'o cc

'I

.oo4 Ne 2 e - 9z 2, / De 4c444.5,EV47,447&P

/',f S7 e7r "e" 4, p4, F e- za 4 7( PL T) + < Cr) {

9 :%530 - -S41 = 23770

-7q 7 / r.4

• 13// fs1 - / A , ,.

p( vHE L6Ao 7 'e -? 7454 -: 7S[3 Ps .

TOTAL 19071 4 79,f,3 s./ d0 6 -41, I4- re =7 Sc - *A/,war#A9 f 75 r ex r 23; -;i4

//

1%

, eG O

4 - ~~ ~ t ~ v~ fj c -- ---

Ssu---.. ------ --- -.-....--.---.-.-..-...

.... PuseT #e. .7 .0.or..

4~- --------------------- -om F Ej AM T/A/A 4.t- Adithd&4

3. 77 Av ~J2o.4/

Fory F 4, P~ Fd -14 d /7 9-?.  :

(P-t0 7) 1 (vi 7) 37 4 oo -A/45 1472o + )3&s

= 37456 Ps. 1610sP51 PIPE 1-040 7 3 3PS

-ro rk L J- 8 735 7 -75

- 4 to b6 redP 3=k 4 - ~ As A *4 4- ( y-7 ) /754--7 Z .3 4 Z A o7e: C LC C Tewas

,zr z S - 0 Y7 5-4 7 z UA4 2 /

4t -1 !r 2603 9S2 /

A/ /N- 4 26boo c c le$s /Z o r- e rA)o

.OSS ,I32 ,7

,_5q (ULi UL.. Ar 1V 4 Fe-l?

j1 z

/1z ,Al 74

-,o 4 .cIa ,o t SetI ( ESULT's Av'E 4&veLL *J CAh' /i-c CUIZVle A/-4 fyr(ej O I4eCt4S T674&/ TA7,C 1=,/c eL50 L&& OF Y* (LC,(e4/ o o.0

mt4C-FT NOX JOR NO -.

J14 . ~. .A r^ -

-J ,4 rur 45.

t I ~ ~'- ~ ~/

.1, I 2~/~ ji A a) P Lut.- 0-oiQ~ 'I /aye p o, t4 C ,4 , Sw . Z i I JTET'tTI (A p +) + Po PCA4 S57C43ss Gr~~- =2671{o 2235 -(*56Psi , S'tq;- SS, b.s ) o9s 4 -(s )

=)0 7 q s4 10%t > -3Sw, 6- to Pas.

E.*Coo d-.De eAss )mie 2

2. rA4* 4:o *-. >4 28/oo Ps .

t. Sr [4ts6E -(- 87sJ - 0 47 5W4 3 PJS< 25 4~ ~ ,* -+4 4 , < / ( 7/d+J).:v3 P+ LT AkJT

b. S -z2 -(ft - 32 " -"

3&44/a

Y *,-- DATEc - 71 SUBJECT-...... SHEFT NO-.9 OF

.C1b NO

' ... 1 0 '\ 1& && E12 7 I l1 A1 AT c- 5cflo o7 r 1 - 14  :

WAL T ' :4 7 Wt Z' :- o5P 0 t. J.2.75 kf4tt 734 t a A/& 1  : 1 Ru L st*' g 00 po ;" P' ce T) SA Y j,/ PPa; FO A.

FRM e& 5, Pts(a 42 Foit /l eP e ZLcya t 4 fFoo 26 / #- .cot 4Z7.3 <o. AT C75 EF db70o P!41.

So yeeloo, = , 47 ( 867oc)

- 4076o cf OK eso.

(4Z rees G7 4cooQ Pa.A Ib (n" Ps&P).

te7, /3 Z ( rV Ree.

oQ z SOaO P4) AT Pt

DATE.- 'It ECT ------ SHEET NO. / O

-t/J~'.

- - - - JOe NO. -

pi4n To / DoedTor777

%7Z33 1,/750 Lac PPooLL 7A3 f

70TA l4 A td

1. CrJ

~~ (r~  ; z) /j0 4z 4 q q 1 3Z COOL t4 ,'.2 374oo 45263 1l3Z (2~~.oi ~ lPa 907/ f  ?~ La3 Toro 1 I IA I .o' PA, .44.e 3 3 -z- 713 3 A.// 7 A tL No' ..

4. Cos r& i. i e e) kv4.e,,'oi44 7W 4,?elay'  :;r,4 77 7, 70 z6,3 784C-#45433 Iz74:1 2 3 Id 3- 4 4S245 Z

74(eS7cM3. 4 352& 3 I' y A:; swra~40 -To 0 -r-er6 a 1/

/2 I

• ~Qe~/ 7iJ7.M'47-/

COAT~Dt~r'Q~.S~ae~4 ~ ~Z~53~

a. T 46P1A 4-b<63

/3 14

f. c/ < 6GAT9 A

S----JECT . - SHEET NO / --. OF JO3 NQ0 Sw4e Ev. 4afrmeJr AmiJ 4 , 4 A S6) A/Aled 2 dtt 5 0 .J04 _0 6o  ? cod 0

toe IT41 L. AT 7' E A/O. #4 7// C/E CE Od' W d'~tLQ; 4 rr1? A?/e Co veno rA c0 (cicet me nod £4rve) 4r ,. /  :

A =/ 7at cS, ty.4 o#FA v 4 A) oei, as v

-4 U (:t 4 _1 e OF0Cf(.(el 1)<O D&

.* 84V5, P,10 1,0

-Z.ti5 P, 4P q 4O4 td 4f" 4/,D-Ar IGo t .(I

&_9_D& L4 8232a 0q8-So 9 o 3 Fee IA__J

- - - - -vs

BY AQ&.k. DATE &79 SUBJECT... -- SHEET NO... Z.. OF C1+KD. ElY . /'o DATE sa JOS

.. NO.

/4'A 11. ,W<

/4 e A ( - ;7P"' 4z7Ae s s ,4 Yr- j/ o'/4./r /3a ( etc.

(o, 4M/' a~ To rg6t r 4Tr '43 O

/0, 9 -6 'jA4 4. C6 -

F Ao' / rr6. /< .3o v F e. 6  ;

'K tJSo -r_ Z&4 /a Ps i1. wr / .stoacr (pr, /.%,

AJ~I' 44

/re# e6/ /zb . '- . Nelin4t*L Jr'es5s , JppeAJ%I 74;erVu?

2 = // $Jo Ps i.

So 2&1 /o 4

= -4 o0 4o /4 I.

,43 4 ce -a0'0 A %1A newte (,C2) Feaa */A?0ez ,-529

,4pjo .Sd - 40 7-11-1a0- 46 ACC- 4 aoneW L.00 A. CA ,44 s/ / I ta*-e-Ade Aewd 4P 7 t.

/000 E4 m4oevr Pso5 ,2,e--svde4

. 4c 4/ 1 73 Z 4

t'oost Pei4 s. 4.5 - 14

,4x,44 "4AO 44 3&7 No.Ste A,,<AL LOAo 45 -7 2z Ji-Esus 5 t LOA0 .4 -3174 TO74. 347Z *;e 404a A /AVC, C P/r ^0dE Fed ,4Ve4t .- '7t'e53 /,5 .SO ,~AE A444 v/Ae7AM/4 I/ rE t'A ex rT4< r70o) F~r* totPrK ev oF A4 //#d, McA 7 rA// .5 A4/re 47Se , ' Ed4006 Od! 1 s/4.5 VQihe1,44-VC, 3 Pyd~Q r$T1 4,$ Tod Aeo T.. c.. 78 tIt4>Ft~'e~C 4~#De c

-- SHEET NO V. O

/ ~ b~VJA 6 S N- -'

~ V

'*1 f, d'4 C& i- *o) F0,40 Pct.

Ti/ /leL e4,A ,4pe JrVA O't r AMje~&4AI'/et; ,r4xr,7S /2d/k'ts

-')(')=3 &s~ op4+' 4r5ZSo ,e vc 74AJ (74 oA 2& '-*,S .r.......

Ii  %* DC0 .7;447P7vPJ (10S (14 -24) obvioustL Y, ThIo

'7QCw 7S'T444 Avo'LcE#AcerA

/Pstoves A At -so

/.5 1.9/4 2(er4, r (AJ s'oJ

's i dy PA. F-6 o 'sod o

, 7>

- /,0rs elF C/ CLi C t4XIL .57e4E 55 o 14 /

sujECT t -. 4 .Q &C.

SHEET NO. OF 1a /4, x /J 7 A't.Lt .OE JoB NCO.

Xv L rES Pe V4Y .ppcod)4?YCV /I ov4 AY1 AIT '5'o)

"I 0

F-A 4St to 5

&:7(ef S5 O0 4o

-2S 76.1 =-2.ks

• ote,, "THiF $UDoocA r/41P ITAtivo UP1 MAd5il&A/r IS SJotwA Foe,, 0t

-SleA , OsJLIJ. TT 4 JAL %Jor eF-9. To P4o. 10 fbot 'TA8LATION.

7 I

. It tb OA FE 3K 2 74 ~,Ij~3JFC $tiEET SHEET NO.P40 / F JOIN NO.-

(

C yY t AJ(; - ~

I feA~'&;4 r 4'.) rA 4A/~')QL "~ ON /,D.A44e^ AJ-D, (ft4our /Jo)

"Il

.~

~

XIA L Swess(

le I to 0

-/0 A----.

-3 a2d1O00

-30 20r.eit To PAizzt II AV#- 127 "e<. TrA6tA'ro*#J.

I

BCL-585-9 FINAL REPORT on EXAMINATION OF INCONEL SAFE END FROM DUANE ARNOLD to PARAMETER INCORPORATED January, 1979 by V. Pasupathi, G.P. Smith, D.R. Farmelo, and J.S. Perrin

-YI

-,fi~~

PF 3 BATTELLE Columbus Laboratories 505 King Avenue Columbus, Ohio 43201

TABLE OF CONTENTS Page INTRODUCTION. . . . . . . . . . . . . . . . .1 EXAMINATIONS AND RESULTS. . . . . . . . . . . . . . . . . . . . . .1 Receipt of Shipment. . . . . . . . . . . . . . . . . . . . . .1 Visual Examination . . . . . . . . . . . . . . . . . . . . . .1 Dimensional Measurements . . . . . . . . . . . . . . . . . . .2 Destructive Examinations . . . . . . . . . . . . . . . . . . .2 Tensile Tests. . . . . . . . . . . . . . . . . . . . . . . . 17 Chemical Analyses. . . . . . . . . . . . . . . . . . . . . . 17 Scanning Electron Microscopy . . . . . . . . . . . . . . . . 23 Electron Microprobe (EMP) Analysis of Duane Arnold Sample No. 2. . . . . . . . . . . . . . . . . 39

SUMMARY

OF 'OBSERVATIONS AND CONCLUSIONS . . . . . . . . . . . . . 39 REFERENCES. . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

LIST OF FIGURES Page FIGURE 1. APPEARANCE OF DUANE ARNOLD INCONEL SAFE END IN THE AS-RECEIVED CONDITION. .........

FIGURE 2. LOCATION OF METALLOGRAPHIC AND SEM SPECIMENS ON DUANE ARNOLD SAFE END N2-E . . . . 5 FIGURE 3. CONDITION OF SAMPLES AFTER BEING CUT FROM SAFE END. . . 6 FIGURE 4. CHANGE IN OUTER CIRCUMFERENCE DURING SECTIONING OF DUANE ARNOLD SAFE END . .

FIGURE 5. MICROGRAPH MONTAGE OF CRACK IN SAMPLE 2 . . .10 FIGURE 6. MICROGRAPH MONTAGE OF CRACK IN SAMPLE 4 11 FIGURE 7. TYPICAL MICROSTRUCTURES AT VARIOUS LOCATIONS OF SAMPLE 3 CROSS SECTION . . . .12 FIGURE 8. MICROGRAPH MONTAGE OF CRACK IN SAMPLE 3 AFTER ETCHING . . . . . . ......... . . . 3 FIGURE 9. MICROGRAPH OF AREA A ON FIGURE 8 FROM ETCHED SAMPLE 3 . . . . . . . . . . . . . . . . . . . .14 FIGURE 10. MICROGRAPHS OF AREAS B AND C FROM FIGURE 8 FROM ETCHED SAMPLE 3 . ...... . . . .15 FIGURE 11. MICROSTRUCTURE OF THE CRACK INITIATION SITE IN SAMPLE 4. . ........... ..... . . . . 16 FIGURE 12. APPEARANCE OF THE GREY PHASE ADJACENT TO A TIGHT CRACK BRANCH. . . ......... . . . 8 FIGURE 13. LOCATION OF THE SECTION FROM WHICH TENSILE TEST SPECIMENS WERE MACHINED. ....... . . . . . 9 FIGURE 14. SCHEMATIC DIAGRAM INDICATING LOCATION OF TENSILE TEST SPECIMENS MACHINED FROM SAFE END ..... . . .20 FIGURE 15. TENSILE TEST SPECIMEN DIMENSIONS..... . .

. . . . .21 FIGURE 16. SEM MICROGRAPH OF GREY PHASE AND CORRESPONDING EDAX ANALYSES OF BASE METAL AND GREY PHASE. . . . . . .25 FIGURE 17. SEM MICROGRAPH AND CORRESPONDING EDAX ANALYSES FROM SAMPLE 4. . . ......... . . .27 FIGURE 18. SEM MICROGRAPH AND CORRESPONDING EDAX ANALYSIS 0 F IRON RICH MATERIAL IN CREVICE OF SAMPLE 4 . . . . . .28 FIGURE 19. SEM MICROGRAPH AND CORRESPONDING EDAX ANALYSIS OF TITANIUM INCLUSION FROM SAMPLE 4 . ..... . . . .29 FIGURE 20. SCHEMATIC DIAGRAM OF SAFE END SAMPLE 5 INDICATING FRACTURE SURFACE EXAMINED BY SEM . . .30 FIGURE 21. PHOTOMACROGRAPH OF FRACTURE SURFACE FROM SAMPLE 5 . . .31

List of Figures (Continued)

Page FIGURE 22. SEM MICROGRAPH MONTAGES OF SAMPLE 5 FRACTURE SURFACE. . . . . . . . . . . . . . . . . . . .32 FIGURE 23. SEM FRACTOGRAPHS OF SAMPLE 5 NEAR CREVICE . . . . . . .33 FIGURE 24. SEM FRACTOGRAPHS OF SAMPLE 5 AT MID FRACTURE. . . . .

. 34 FIGURE 25. SEM FRACTOGRAPHS OF SAMPLE 5 AT CRACK TIP . . . . . . .35 FIGURE 26. SEM FRACTOGRAPHS OF SAMPLE 5 NEAR CRACK TIP . . . . . .37 FIGURE 27. DUANE ARNOLD SAFE END SULFUR PROFILE ON FRACTURE SAMPLE 1........... . . . . . .. 38 FIGURE 28. SAMPLE 2 MICROPROBE RESULTS......... . . . . .. 42

LIST OF TABLES

-Page TABLE 1. RESULTS OF WALL THICKNESS MEASUREMENTS..... . . . .4 TABLE 2. RESULTS OF ROOM TEMPERATURE TENSILE TESTS FOR DUANE ARNOLD INCONEL SAFE END MATERIAL..... . . .. 22 TABLE 3. CHEMICAL ANALYSIS OF INCONEL SAFE END BULK METAL MATERIAL. .......... . . . . . . .. 24 TABLE 4. ELECTRON MICROPROBE RESULTS OF 2e SCANS IN BASE METAL, GREY PHASE AND WELD METAL...... . . .. 40 TABLE 5. ELECTRON MICROPROBE ANALYTICAL RESULTS FOR Ni, Cr, Fe FROM AREA/POINT COUNTING IN BASE METAL, GREY PHASE (POINT), AND WELD METAL......... . . . . . . .. 41

1.0 INTRODUCTION

Recently Iowa Electric Light and Power Company found cracks in all of the eight Inconel safe end sections of the recirculation inlet nozzle in the Duane Arnold Plant. One of the safe ends was found to contain a visible throughwall crack. Cracks in other safe ends were detected by a combination of radiography and ultrasonic techniques. In order to determine the cause(s) of cracking Iowa Electric Light and Power Company initiated examination of the safe end section containing the throughwall crack. In parallel with this effort, examination of another safe end from Duane Arnold was initiated at Battelle's Columbus Laboratories (BCL). This parallel effort had the objective of obtaining an independent evaluation of the nature and extent of cracking.

The safe end designated as N2E was shipped to BCL and subjected to detailed nondestructive and destructive examinations including optical metal lography, scanning electron microscopy, electron microprobe analysis, chemical analysis and mechanical property evaluation. This document is a final report of the data obtained in this investigation.

2.0 EXAMINATIONS AND RESULTS 2.1 Receipt of Shipment The Inconel safe end section was received at the BCL Hot Laboratory during September, 1978. Upon opening the shipping container, the internal activity was found to be rather high, -500-700 mRem/hr at or near the specimen.

In addition the specimen was found to be highly contaminated, with smearable activity being 900,000 dpm.

2.2 Visual Examination A visual examination of the sample was made using a magnifying glass.

Care was taken not to disturb the deposits on the specimen surface. The outer surface of the specimen was relatively clean with azimuthal orientation marks 0-20 around the circumference of the safe end. These markings had been made with a felt tip marker and corresponded to the locations of radiography films.

2 In addition to these marks, the piece also contained a hose clamp containing the specimen identification number plate showing N2E. The location of the repair weld could be clearly seen on the outer surface. The inner surface of the specimen had a rust colored coating of loose powder. This powder could be easily scraped off. Careful examination of the inner surface failed to reveal any cracks. The as-received condition of the specimen was documented by photography in detail. Figures la and lb show the appearance of the specimen.

2.3 Dimensional Measurements Dimensional measurements made on the specimen consisted of wall thickness and diameter measurements. Wall thickness measurements were made using a micrometer. Location of the measurements and the results are shown in Table 1. Diameter measurements were made from photographs taken during visual, examination. The outer diameter at the large end was 14.00 in. and that at the smaller end was 11.3 in.

2.4 Destructive Examinations 2.4.1 Specimen Sectioning. The specimen was marked for sectioning according to the cutting diagram supplied by Parameter Incorporated. Five thin samples were cut with a band saw. Locations of the samples are shown in Figure 2. Figure 3 shows the samples cut from the safe end.

Sample No. 1 - At radiography location 5 Sample No. 2 - Between radiography location 4 and 5 Sample No. 3 - Between radiography location 3 and 4 Sample No. 4 - Between radiography location 14 and 15 Sample No. 5 - Between samples 1 and 2.

3 I

(a)

I I

I (b)

FIGURE 1. APPEARANCE OF DUANE ARNOLD INCONEL SAFE END IN THE AS-RECEIVED CONDITION

4 TABLE 1. RESULTS OF WALL THICKNESS MEASUREMENTS Sector Locations (Orientation) 1 2 3 4 0 (00) 1.4688" 1.3065" 0.8642" 0.7515" 1.4850" 1.3030" 0.8610" 0.7450" 3 (450) 1.2988" 0.8592" 0.7380" 5 (900 ) 1.4862" 8 (1350) 1.4802" 1.3018" 0.8650" 0.7215" 10 (1800) 1.5112" 1.3015" 0.8565" 0.7142" 13 (2250) 1.4715" 1.3038" 0.8420" 0.7190" 15 (2700) 1.4608" 1.3085" 0.8452" 0.7238" 18 (3150) 1.4565" 1.3090" 0.8650" 0.7322" SCHEMATIC OF SAFE END CROSS SECTION INDICATING LOCATIONS OF WALL THICKNESS MEASUREMENTS Safe End to Sleeve Weld Outer Repair Weld

5 FIGURE 2. LOCATION OF METALLOGRAPHIC AND SEM SPECIMENS ON DUANE ARNOLD SAFE END N2-E

6 33 I FIGURE 3. CONDITION OF SAMPLES AFTER BEING CUT FROM SAFE END I

(b)

37 Prior to making the first cut, small indentations were made with a center punch on either side of the first cut location. The distances between these marks were measured. Three such pairs of indentations were made and the distances measured.

While the first cut for Sample No. 1 was being made, it was found that the cut closed tightly and the band saw blade could not be pulled out.

The frame was removed and the blade was left in the cut. A new blade was inserted and the second cut made. After completing this cut the sample had to be pried loose. The distance between the centerpunch marks was measured after the sample was cut out. The measurements obtained before and after are shown in Figure 4. Also shown are the locations of the marks.

2.4.2 Metallographic Examination. Samples 2, 3, and 4 were mounted in epoxy resin and prepared for metallographic examination. The samples were milled to obtain a flat surface and ground with silicon carbide papers of grit 120 through 600. They were then polished with a slurry of Linde A alumina. Samples 2 and 4 were examined in the as-polished condition and Sample 3 was etched electrolytically with a 10 percent solution of oxalic acid.

All three samples contained cracks. The cracks were all inter granular in nature. In all samples, the initiation site of cracks was in the region of tight crevice between the sleeve and the safe end and radiated outward. The length of the tight crevice was found to be in the range 0.2-0.3 in. and the width was 0.002-0.005 in. No cracks were observed in the sleeve.

In Samples 2 and 4 some grey areas were observed adjacent to tight branches of the cracks. These areas appeared to be of different composition. Similar areas were also observed along "tunnels" radiating from the cracks. The source of this grey phase area is not known.

The crack did not penetrate through the cross section of the safe end in any of the samples examined. In Samples 2 and 3 the crack had pene trated approximately 80% of the wall. In Sample 4, which was obtained

8 Location of Center Punch Marks 14.0 in.

&~ I 136,

,2 IA2.7 Repair Weld 11.3 in.

Distance Between Punch Marks Inches Before Cut After Cut Change No. I 0.9230 0.8754 0.0476 No 2 1.0220 0.9250 0.0970 No 3 0.9650 0.8800 0.085 FIGURE 4. CHANGE IN OUTER CIRCUMFERENCE DURING SECTIONING OF DUANE ARNOLD SAFE END

9 from the opposite quadrant, the crack penetration was about 30%. The crack characteristics were documented by photography. Figures 5 and 6 show the results from Samples 2 and 4, respectively.

Sample 3 was examined in detail to characterize the microstructure of various parts of the sample. Figure 7 shows the results. The photo micrographs show that no major abnormalities are apparent in the micro structure of the safe end with the exception of sensitization near the sleeve to safe end weld. However, it appears that the thermal sleeve is sensitized (as evident from carbides precipitated at grain boundaries) rather uniformly even away from the weld.

Sample 3 was subsequently reprepared and etched with a mixture of 20 ml H2 0, 20 ml HNO 3, and 80 ml HC1. The purpose of this procedure was to clearly identify the location of the crack with respect to the weld and heat affected region. With this technique, the crack initiation site was found to be in the re-solution treated region of the heat affected zone. Figures 8 through 10 show details of the crack location and crack characteristics.

In order to further characterize the extent of sensitization of the safe end in the vicinity of the crack, sample 4 was repolished and electro lytically etched with 10% Nital. This procedure is expected to show grain structure in the material regardless of the degree of sensitization. After examination, the sample was repolished and etched electrolytically with a 15% solution of phosphoric acid. This etching process preferentially attacks carbides in the grain boundaries and the matrix. Figure 11 shows a comparison of the microstructure at the crack initiation site with the two etching procedures. Examination of the photomicrographs shows that the sensitized region in the safe end is rather narrow and that the crack initiation site is located in the re-solution treated region adjacent to the weld. The thermal sleeve also appears to be sensitized to a greater extent in comparison to the safe end.

Additional experiments were carried out on sample 4 using modified glyceregia as the etchant. The etching solution consisted of 10 ml HNO 3 '

10 ml acetic acid, 20 ml HC1 and 30 ml glycerine. This process was expected to delineate chromium depleted regions along grain boundaries.( 1 )

(1) References at end of text.

mMm -mMma m m m"m 170 Crevice FIGURE 5. MICROGRAPH MONTAGE OF CRACK IN SAMPLE 2

Crack Region Analyzed By EDAX (See Figure 17)

Crevice FIGURE 6. MICROGRAPH MONTAGE OF CRACK IN SAMPLE 4

12 Area 10 Area 9 Area I2 Area 7 Area 13 Area 6 Area 15, Area 1 Area 3 Area 4 FIGURE 7. TYPICAL MICROSTRUCTURES AT VARIOUS LOCATIONS OF SAMPLE 3 CROSS SECTION

MWW- - W---Wamm -m

'V.

I AV.

-I B

FIGURE 8. MICROGRAPH MONTAGE OF CRACK IN SAMPLE 3 AFTER ETCHING Details of areas labeled A, B,and C are presented at higher magnification in Figures 9 and 10.

14 IUUX FIGURE 9. MICROGRAPH OF AREA A ON FIGURE 8 FROM ETCHED SAMPLE 3

15 1OOX (B)

I UUX (C)

FIGURE 10. MICROGRAPHS OF AREAS B AND C FROM FIGURE 8 FROM ETCHED SAMPLE 3

16 19% Nital Etch (a)

Phosphoric Acid Etch (b)

FIGURE 11. MICROSTRUCTURE OF THE CRACK INITIATION SITE IN SAMPLE 4

17 Examination of the specimens showed extremely narrow and barely discernible regions (probably chromium depleted areas) in the relief adjacent to the grain boundaries in the heat affected region. Initially, it was planned to analyze this region using the scanning electron microscope to determine if the composition was different from the grain matrix. This plan was abandoned because of the difficulty in discerning these regions. The etching procedure, however, was found to define more clearly areas previously identified as grey phase. Figure 12 shows one such area in the sample.

2.5 Tensile Tests From the remaining parts of the Inconel safe end a large piece was sectioned so that specimens for tensile tests could be machined. The large piece was cut between radiography locations 0 and 4 as shown in Figure 13. The cut piece was decontaminated and ultrasonically cleaned and five tensile test specimens were machined. Figure 14 shows schematically the location of the specimens in the safe end and Figure 15 shows the dimensions of the tensile specimens used.

The specimens were tested at room temperature. The results obtained are shown in Table 2. From the results presented in Table 2, it is clear that no degradation of the Inconel safe end tensile properties was observed.

2.6 Chemical Analyses 2.6.1 Tests of pH in Crevice. Tests to determine the pH of the residue in the crevice were carried out using deionized water and litmus paper.

Although litmus changes indicated pH to be in the 4-6 range, such reactions were variable and not sufficiently positive to enable conclusive determination of the acidity of corrosion products.

2.6.2 Liquid Samples From Crevice. The section cut from the pipe for tensile test samples was used to obtain samples for chemical analysis. Prior to decontamination the crevice area was rinsed with distilled water and the rinse solution was collected and analyzed by emission spectroscopy. This procedure involves evaporating the water and analyzing the residue. The major element in the residue was found to be Na. Trace amounts of Mn, Si, Cu, Ni, Cr, Ti, Al, B, Fe, Mg, K, Ca, Ba, and Sr were also detected.

18 I

I I

I I

I I

3 ,

500X FIGURE 12. APPEARANCE OF THE GREY PHASE ADJACENT TO A TIGHT CRACK BRANCH Sample was etched with aqua regia glycerine and acetic acid mixture to reveal chromium depleted regions.

I I

I I

I I

19 Radiography Locations Section For Tensile

'Test Specimens O0 Sample 3, End Thermal 900 FIGURE 13. LOCATION OF THE SECTION FROM WHICH TENSILE TEST SPECIMENS WERE MACHINED

mmmmm m - m m m m m mm m Crevice aD Safe End to Sleeve Weld Sleeve FIGURE 14. SCHEMATIC DIAGRAM INDICATING LOCATION OF TENSILE TEST SPECIMENS MACHINED FROM SAFE END

-MW m -- m m mm M 0.4375 -14 UNC-2A A .005 R Both ends

[_0.375 +/-

0.003+

5/32 R 0.50 0.75 min

+/- 0.20 Reduced section 2.060 +/- 0.050 Notes: 1. D= 0.160 +/- 0.001 diameter at center of reduced section. D'= actual D+0.002 to 0.003 at ends of reduced section tapering to D at center.

2. Grind reduced section and radii to 32V radii to be tangent to reduced section with no circular tool marks at point of tangency or within reduced section. Point of tangency shall not lie within reduced section.

FIGURE 15. TENSILE TEST SPECIMEN DIMENSIONS

22 TABLE 2. RESULTS OF ROOM TEMPERATURE TENSILE TESTS FOR DUANE ARNOLD INCONEL SAFE END MATERIAL Test Yield Percent Total Percent Specimen Strength UTS Elongation Reduction No. (ksi) (ksi) (in 0.640 in.) in Area DA-1 47.5 107.0 41.9 57.8 DA-2 48.0 106.0 40.15 57.5 DA-3 46.8 105.7 39.5 58.2 DA-4 48.8 104.5 38.3 61.0 DA-5 49.3 107.0 40.6 58.7 Precharacterized*

Material 44.0 100.0 38.0 53.0

• Results of Precharacterized Testing obtained from Chicago Bridge and Iron Company Nozzle Certified Test Reports. Specimen gage length for Precharacterized Testing - 2.0 in.

23 2.6.3 Corrosion Deposits on the Inner Surface of Specimens. The red powdery deposit observed on the inner surface was scraped off and collected. This sample was analyzed using an X-ray diffraction technique.

The major portion of the scrapings (60-70%) was found to be hematite (Fe 0 )'

2 3 The remainder could not be identified.

2.6.4 Bulk Metal Analysis. Chemical analysis of a bulk metal sample (from safe end) was carried out. The samples consisted of a metal chunk and fine chips. The results, shown in Table 3, indicate that the composition of the material was within the limits of specification.

2.7 Scanning Electron Microscopy Four safe end specimens, each containing a partial thruwall crack, were examined by scanning electron microscopy (SEM). Two samples (Sample Nos.

2 and 4) were mounted and metallographically polished prior to SEM examination as stated in Section 2.4.2. The remaining SEM samples, identified as Sample Nos. 1 and 5 were examined along the fracture surface. Figure 2 identifies the location and the surfaces examined of each safe end sample.

2.7.1 Polished Samples, SEM Examinations. SEM examination and Energy Dispersive X-Ray Analysis (EDAX) of the fractures and fracture areas of Samples 2 and 4 provided several interesting results. As can be seen from Figures 5 and 6, the cracks observed in both samples contain numerous branches or tributaries. In addition, the cracks appear to originate in the weld heat affected zone at the crevice between the safe end and the thermal sleeve.

SEM examination of many of the tight crack tributaries in Samples 2 and 4 indicated that a phase different from the Inconel base metal was present directly adjacent to the crack. A typical SEM micrograph of this phase with its corresponding X-ray -spectrum is presented in Figure 16. EDAX analysis indicated that this grey phase was chromium rich. The chromium enhancement of the grey phase was determined to be approximately 60-80% relative to the base metal material. In addition to the composition variation of the grey phase, numerous instances, of what appears to be transgranular tunneling (shown in Figure 16) were observed in areas where grey phase was found.

24 TABLE 3. CHEMICAL ANALYSIS OF INCONEL SAFE END BULK METAL MATERIAL Element Percent Ni 73.6 Cr 15.6 Fe 7.8 Al 0.4 Ti 0.3 Mn 0.2 Si 0.2 Cu 0.1 Mo 0.1 S Q .002 Si 0.3 P 0.03

25 Base Metal Grey Phase 3000X FIGURE 16. SEM MICROGRAPH OF GREY PHASE AND CORRESPONDING EDAX ANALYSES OF BASE METAL AND GREY PHASE

26 EDAX analysis of material observed within the crack both adjacent to the grey phase and in areas where no grey phase is found indicate an iron rich material relative to base metal compositions. Figures 17 and 18 present SEM micrographs and EDAX spot analysis performed well within the crack and at the crevice on Sample No. 4. As can be seen in Figure 17 the EDAX analysis of material within the crack indicates excessive iron relative to base metal analysis. However, no iron depletion can be observed in the material adjacent to the crack. EDAX analysis performed on material within the crevice (Figure 18) at a location near the crack origin indicate similar iron rich material. Comparison of the two relative iron contents from EDAX analysis at the crevice with analysis in the crack indicate a significantly higher iron content of material well within the crack.

Numerous titanium inclusions were observed throughout the base metal on Samples 2 and 4. A typical titanium inclusion observed in Sample 4 is illustrated by SEM micrograph in Figure 19 in conjunction with its X-ray spectrum.

2.7.2 SEM Fractography. As stated previously, safe end Sample No. 5 was examined by SEM. Figure 20 presents a schematic of Sample No. 5 showing the fracture surface examined. In order to expose the fracture surface for examination the specimen was mechanically fractured. Figure 21 presents a photomacrograph of the fracture surface examined. The bright material to the left is the region of uncracked safe end wall thickness which was mechanically fractured. Examination of this region following fracture indicated the material to be ductile.

Detailed SEM examination was performed on the entire fracture surface.

Figure 22 presents SEM fractograph montages at the beginning (near crevice) middle, and tip of the crack along the fracture surface. The mode of cracking, as indicated by the SEM fractographs, was intergranular fracture.

Numerous EDAX analyses were performed on several areas of the fracture and crevice. Figures 23, 24, and 25 present SEM fractographs illustrating typical areas chosen for EDAX analysis (i.e., Figures 23b, 24a, and 25b present actual areas examined by X-ray near the crevice, in the middle and at the tip of the fracture, respectively). In addition, Figure 24b presents a SEM micrograph of a grain facet (from Figure 24a) showing a network of needle-like corrosion products on which a spot EDAX analysis was performed.

-m - WWm=mWmm m m = m m-m-m Fe Rich FIGURE 17. SEM MICROGRAPH AND CORRESPONDING EDAX ANALYSES FROM SAMPLE 4 Area shown is located within the rectangular region highlighted in Figure 6

m m mmmmm m m m m mmmm FIGURE 18. SEM MICROGRAPH AND CORRESPONDING EDAX ANALYSIS OF IRON RICH MATERIAL IN CREVICE OF SAMPLE 4

---

mmm -mm -m no - m mm IbOOX FIGURE 19. SEM MICROGRAPH AND CORRESPONDING EDAX ANALYSIS OF TITANIUM INCLUSION FROM SAMPLE 4

mm mMM m mmmm-mmmmmm a

Crevice Safe End to Sleeve Weld Sleeve FIGURE 20. SCHEMATIC DIAGRAM OF SAFE END SAMPLE 5 INDICATING FRACTURE SURFACE EXAMINED BY SEM

AlA FIGURE 21. PHOTOMACROGRAPH OF FRACTURE SURFACE FROM SAMPLE 5

1 32 I

I I

I I

(a) Crack Tip (c) Near Crevice FIGURE 22. SEM MICROGRAPH MONTAGES OF SAMPLE 5 FRACTURE SURFACE (a) Crack Tip, (b) Mid Fracture (b) Mid Fracture, (c) Near Crevice

33 (a) Typical area on fracture surface near crevice (b) Actual area examined by EDAX analysis FIGURE 23. SEM FRACTOGRAPHS OF SAMPLE 5 NEAR CREVICE

34 (a) Actual area examined by EDAX analysis JA (b) Needle network of corrosion products on grain facet. Spot EDAX analysis indicated high sulfur contert.

FIGURE 24. SEM FRACTOGRAPHS OF SAMPLE 5 AT MID FRACTURE

35 ZUA (a) Typical area on fracture surface at crack tip.

200X (b) Actual area examined by EDAX analysis.

FIGURE 25. SEM FRACTOGRAPHS OF SAMPLE 5 AT CRACK TIP

36 Results from EDAX analysis of Figures 23, 24, and 25 and other areas not shown indicate no significant variations from base metal composition.

However, small amounts of sulfur were detected in almost all of the fracture and crevice EDAX analyses. Figure 26 presents SEM micrographs taken on the fracture surface near the crack tip. The crystalline material, magnified to 2000X in Figure 26b, was subjected to spot EDAX analysis. Results indicate high sulfur contents (-8 times greater than the highest sulfur content detected in area EDAX scans on the fracture surface). Similar crystalline structures were observed and analyzed on the crevice surface, and they too indicated relatively high sulfur content.

In order to characterize the sulfur distribution along a fracture surface and determine the source of the sulfur contaminants Sample 1 was mechanically fractured in the same manner as Sample 5 and examined by SEM.

EDAX analyses were performed along the fracture surface at designated intervals to quantify local sulfur concentrations. A total of 11 EDAX scans were per formed along a radial axis from the crevice to the tip of the crack. In addition, two EDAX analyses were performed on the mechanically broken ductile material. Each EDAX scan examined in surface area of approximately 1.6 x 10-2 cm2 .

The results of the EDAX analyses are presented as Figure 27. The relative sulfur concentration data presented in Figure 27 are plotted versus wall thickness.

Based on this data an increase in relative sulfur concentration was observed with increasing crack penetration (with the exception noted directly at the crack tip).

The relative sulfur concentrations measured in the ductile material beyond the crack tip indicate a significant concentration deviation from the fracture surface*.

Based on the difference in relative sulfur concentrations on the fracture surface and the ductile material and the appearance of the sulfur rich particles and the high concentration of sulfur in them, it is believed that the source of sulfur in the cracks was external to the base metal material.

• It should be noted that sulfur concentration shown for the base material are intended only for comparison with those in the fracture surface. The values should not be taken as representative of actual sulfur concentration in the base metal. (See Table 3 for sulfur content in base metal.)

37 Wll (a) Cluster of crystalline material near crack tip.

2000X (b) Crystalline material subject to EDAX spot analysis. Results indicate high sulfur content (>20%).

FIGURE 26. SEM FRACTOGRAPHS OF SAMPLE 5 NEAR CRACK TIP

w mwm mMmmm m w mm mm m 4-C Q) Crack Safe End O2.0 Base Metal CL

/

A C

0 /

/

C

0) / Y

/ Safe End

0) /

0.0 L.

/ 0. D.

(A.

Q.)

0'a 0)-

/

0-1 /

.1 01

- -O...~ %% /

I I I I I

I 10 2U0 30 40 I 50 60 70 80 90 100 Percent Wall Thickness FIGURE 27. DUANE ARNOLD SAFE END SULFUR PROFILE ON FRACTURE SAMPLE No. 1

39 2.8 Electron Microprobe (EMP) Analysis of Duane Arnold Sample No. 2 Sample No. 2 was examined in detail with the electron microprobe.

Examination consisted of 2e scans, X-ray mapping and area or point counting.

Areas analyzed included base metal, grey phase, and weld metal. In preparation for semiquantitative analysis the sample was mounted in a stainless steel ring with epoxy, ground with SiC papers and polished with A12 03 powder.

Two theta scans were obtained for each area by simultaneously scanning with LiF, PET, and KAP crystals to analyze elements from 1 1 Na through of Ti and Si 94Pu. In addition to the major elements Ni, Cr, and Fe, traces were noted in the base metal and grey phase-crack areas. The weld metal indicated a minor amount of Mn. Results are tabulated in Table 4.

Fixed time (30 sec) area counts were then performed on the same three areas for Ni, Cr, and Fe to obtain semiquantitative analyses. The results are shown in Table 5. By comparison with pure standards, results on a first-approximation basis (e.g., no corrections for atomic number, absorption or fluorescence) were obtained. Base metal results are shown to agree reasonably well with the nominal and analytical chemistry results. Of main importance is the Cr increase in the grey phase by -50% (relative). This analysis supports the qualitative results obtained in the scanning electron microscope (SEM).

X-ray mapping, as shown in Figure 28, compares the Ni, Cr, and Fe distribution in a grey phase-crack location. For orientation purposes the X-ray images must be compared with the electron backscatter (EBS) image, which in turn can be compared with the adjacent photomicrographs. The Ni X-ray map shows a decrease in Ni content in the Cr rich phase. The Cr image appears to indicate a slight intensity increase. Iron appears to follow the general area topography with relatively uniform distribution.

3.0

SUMMARY

OF OBSERVATIONS AND CONCLUSIONS An examination of the data presented in Section 2.0 of this report leads to the following observations and/or conclusions.

• All samples taken from the safe end and examined either by optical metallographic or SEM techniques contained part-wall cracks. No cracks were observed to penetrate the repair weld on the outer surface.

40 TABLE 4. ELECTRON MICROPROBE RESULTS OF 2e SCANS IN BASE METAL, GREY PHASE AND WELD METAL Base Metal Grey Phase Base Metal Grey Phase Weld Metal Area 3 Area 1 Area 4 Elements Detected Listed in Order of Decreasing Intensity*

Major (>300 cps)

Ni Ni Ni Cr Cr Cr Fe Fe Fe Minor (50-300 cps)

Mn Trace (<50 cps)

Mg Mg Mg Ti Si Ti Si Ti Sr Si Nb

• Elements 11 Na through 92 U.

I

41 TABLE 5. ELECTRON MICROPROBE ANALYTICAL RESULTS FOR Ni, Cr, Fe FROM AREA/POINT COUNTING IN BASE METAL, GREY PHASE (POINT), AND WELD METAL*

Base Metal Grey Phase Weld Metal Inconel -600 Chem ical Nominal Analysis Area 3 Area 1 Area 4 (ASM) (Base)

Percent Ni 77 63 80 76.0 73.6 Percent Cr 15 24 17 15.5 15.6 Percent Fe 7 9 5 8.0 7.8 Total 99 96 102 99.5 97.0

• First approximation results do not include atomic number, absorption, or fluorescence corrections.

42 GREY EBS PHASE IMAGE AREA 500X 1 OX 500X GREY PHASE Ni X-RAY ly IMAGE CRACK 500X 30X Cr X-RAY t

IMAGE 500X MACRO 5X Fe X-RAY IMAGE 500X FIGURE 28. SAMPLE #2 - MICROPROBE RESULTS

43

• In all samples, the crack originated in the crevice between the sleeve and the safe end and radiated outward.

No cracks were observed in the sleeve.

• The location of the cracks in all cases were in the re-solution treated region of the heat affected zone of the weld joining the safe end to the sleeve.

See Figures 9 and 11.

• In the three samples examined metallographically, only one defect, namely the crack, was observed. No other crack precursors such as pitting were evident in the crevice region of any of the samples.

a All the cracks observed by metallography and fractography exhibited typical characteristics of intergranular stress corrosion cracks observed in nickel base alloys.(2-5) o The depth of crack penetration in four of the five samples was about 80% of the safe end wall thickness.

In the fifth sample which was taken from the opposite quadrant, the depth of crack was about 30%.

o From the location of the cracks, it is believed that the cause of crack initiation and subsequent propagation is not related to the repair weld on the outer surface of the safe end.

• The chemical composition of the safe end was within specification limits.

• Results of tensile tests on specimens from the safe end show no abnormalities.

• In all of the metallographic samples, small grey areas were observed around many tight branches of the cracks.

These areas were found to contain relatively higher amounts of chromium. The source of the grey phase or its relationship to the cracking mechanism is not known.

44

• Small amounts of sulfur were detected in almost all of the fracture surface and crevice EDAX analyses. Relatively high sulfur content was detected in a crystalline appearing material near the crack tip on the fracture surface. No sulfur was detected on the metallographically prepared samples. From the appearance of the sulfur rich particles, the amount of sulfur in these particles and the concentration profile of sulfur on the fracture surface, it is believed that the sulfur was entrapped from the environment. However, it is not known if the presence of sulfur as a contaminant contributed to the cause of cracking.

45 REFERENCES

1. C. S. Tedmon, Jr., and D. A. Vermilyea, "Carbide Sensitization and Intergranular Corrosion of Nickel Base Alloys", Corrosion-Nace, 27, No. 9, pp 376-381, September (1971).

2. H. Coriou, L. Grall, P. Olivier, and H. Willermoz, "Influence of Carbon and Nickel Content on Stress Cracking of Austenitic Stainless Alloys in Pure or Chlorinated Water at 350 C", Proceedings of Conference Fundamental Aspects of Stress Corrosion Cracking, Columbus, 1967, Ohio State University, NACE (1969).

3. H. R. Copson and S. W. Dean, "Effect of Contaminant on Resistant to Stress Corrosion Cracking of Ni-Cr Alloy 600 in Pressurized Water, Corrosion, 22, pp 280-290 (1966).

4. H. A. Domian, R. H. Emanuelson, L. W. Sarver, G. J. Theus, and L. Katz, "Effect of Microstructure on Stress Corrosion Cracking of Alloy 600 in High Purity Water", Corrosion-Nace, 33, No. 1, pp 26-37, January (1977).

5. J. Blanchet, H. Corious, L. Grall, C. Mahieu, C. Otter, and G. Turluer, "Historical Review of the Principal Research Concerning the Phenomena of Cracking of Nickel Base Austenitic Alloys", Proceedings of Fifith European Congress of Corrosion, Paris, September (1973).