Citation
Effect of Neutron Irradiation and Thermal Aging on Cast Austenitic Stainless Steel and Stainless Steel Weld Phase Stability

Material Information

Title:
Effect of Neutron Irradiation and Thermal Aging on Cast Austenitic Stainless Steel and Stainless Steel Weld Phase Stability
Creator:
Li, Zhangbo
Place of Publication:
[Gainesville, Fla.]
Florida
Publisher:
University of Florida
Publication Date:
Language:
english
Physical Description:
1 online resource (153 p.)

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Materials Science and Engineering
Committee Chair:
YANG,YONG
Committee Co-Chair:
FUCHS,GERHARD E
Committee Members:
MYERS,MICHELE V
PATTERSON,BURTON ROE
CHEN,YOUPING

Subjects

Subjects / Keywords:
cass -- degradation -- lwr
Materials Science and Engineering -- Dissertations, Academic -- UF
Genre:
bibliography ( marcgt )
theses ( marcgt )
government publication (state, provincial, terriorial, dependent) ( marcgt )
born-digital ( sobekcm )
Electronic Thesis or Dissertation
Materials Science and Engineering thesis, Ph.D.

Notes

Abstract:
Cast austenitic stainless steels and welds used in light water reactors usually adopt a duplex structure consisting of austenite and ferrite. Their compounds are widely used in LWRs as primary pressure boundary and reactor vessel internal components. The common cast stainless steels in service include CF-3 and CF-8 series with ferrite content up to 30%. For the welds, ferrite contents are limited to less than 20%. Potential embrittlement can occur due to microstructure evolution during LWR environmental degradation. Fully understanding of such phase decomposition-related embrittlement is crucial to the accurate evaluation and assessment of the LWR sustainability program. Therefore, systematic experiments and analysis have been performed in this study to investigate the effect of aging and irradiation on the phase stability of ferrite in CASS/weld. The stability of ferrite in CF-3 under long-term aging and irradiation was examined. Low dose neutron irradiation can enhance the decomposition of ferrite synergistically with thermal aging owing to irradiation-enhanced diffusion. At high irradiation doses, the ferrite phase exhibited a steady evolution of phase transformation with respect of spinodal decomposition and G phase precipitation coarsening. For 308L weld, both long-term aging and low dose neutron irradiation induced spinodal decomposition, with large Cr concentration fluctuation wavelength and amplitude and minor G phase precipitate. Such decomposition could be the major reason for the ferrite phase nano-hardness increase, while little change in the hardness of austenite was observed. The periodic strain field combined with the Cr rich zone lattice friction augmentation would both oppose movement of dislocation and strengthen the ferrite phase. ( en )
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis:
Thesis (Ph.D.)--University of Florida, 2017.
Local:
Adviser: YANG,YONG.
Local:
Co-adviser: FUCHS,GERHARD E.
Statement of Responsibility:
by Zhangbo Li.

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
LD1780 2017 ( lcc )

Downloads

This item has the following downloads:


Full Text

PAGE 1

EFFECT OF NEUTRON IRRADIATION AND THERMAL AGING ON CAST AUSTENITIC STAINLESS STEEL AND STAINLESS STEEL WELD PHASE STABILITY By ZHANGBO LI A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2017

PAGE 2

2017 Zhangbo Li

PAGE 3

To my family and friends

PAGE 4

4 ACKNOWLEDGMENTS I would like to give my most sincere thanks to my ad visor and my committee chair Dr. Yong Yang, Associate Professor of Nuclear Engineering Program, Department of Materials Science and Engineering, Uni versity of Florida for his support and guidance on my research. The consistent establishment of this work benefits greatly from his outstanding vision and extraordinary insight toward s the materials aging and degradation program I am deeply thankfu l for my supervisory committee Prof. Michele Manuel, Prof. Gerhard Fuchs, Prof. Pat Patterson and Prof Youping Chen for their valuable guidance and encouragement to the success of this rese arch. It is such a great honor to have them in my committee. My colleagues and friends have provided tremendous amount of help support and encouragement during the last few years They are my colleagues, friends and collaborators Dr. Wei Y ang Lo, Dr. Yue dong Wu, Nicholas Silva, Haode Yang, Kookhyun Jeong, Chi Xu, Nils Strombom, Dr. Soumitra Sulekar, Dr. Yaqiao Wu and Dr. Yiren Chen Thank you all for your help support and dedication Finally, I would like to give my special thanks to my family for their love, company and trust on me for thi s journey.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 8 LIST OF FIGURES ................................ ................................ ................................ .......... 9 LIST OF ABBREVIATIONS ................................ ................................ ........................... 12 ABSTRA CT ................................ ................................ ................................ ................... 14 CHAPTER 1 BACKGROUND AND INTRODUCTION ................................ ................................ 16 1.1 Basics of CASSs/Welds ................................ ................................ .................... 16 1.1.1 Ferrite and Austenite Structure s and Properties ................................ ...... 16 1.1.2 Alloy Elements in Stainless Steel Alloys ................................ .................. 18 1.1.3 Metallurgy in CASS and Weld ................................ ................................ 22 1.1.4 Structure/Property Relationships in CASS/Welds ................................ ... 24 1.2 Application of CASS/Weld in LWRs ................................ ................................ .. 26 1.3 Aging Degradation Study of CASS/Weld in LWRs ................................ ............ 28 1.3.1 Aging Embrittlement Study Methodology ................................ ................. 28 1.3.2 Microstructure Evolution of Thermally Aged Ferrite ................................ 30 1.3.3 Activation Energy Extraction of Aging Embrittlement ............................... 36 1.4 Motivation ................................ ................................ ................................ ......... 37 1.4.1 Nuclear Energy and LWRs Sustainability Program ................................ 37 1.4.2 Challenges of Neutron Irradiation Enhanced Aging Degradation ............ 38 1.4.3 Comparison of Thermal Aging and Neutron Irradiation Validation ........... 41 1.4.4 Synergistic Effect of Irradiation and Thermal Aging on Ferrite Decomposition ................................ ................................ .............................. 42 1.4.5 Micro Scratch Test for Fracture Toughness Extraction ............................ 43 2 EXPERIMENTAL DETAILS ................................ ................................ .................... 45 2.1 Materials and Treatment ................................ ................................ ................... 45 2.2 OM & SEM Metallography ................................ ................................ ................ 47 2.3 TEM Crystallography and Characterization ................................ ...................... 48 2.4 Atom Probe Microstructural Characterization & Analysis ................................ .. 49 2.4.1 Frequency Distribution ................................ ................................ ............. 51 2.4.2 Radial Distribution Function ................................ ................................ ..... 52 2.4.3 Proxigram of Cr Map ................................ ................................ ............... 54 2.4.4 Cluster Analysis ................................ ................................ ....................... 55 2.5 Mechanical Property Testing ................................ ................................ ............ 58 2.5.1 Nano indentation Test ................................ ................................ ............. 58

PAGE 6

6 2.5.2 Miniature Tensile Test ................................ ................................ ............. 62 2.5.3 Micro Scratch Test ................................ ................................ ................... 63 3 RESULTS ................................ ................................ ................................ ............... 68 3.1 Metallography and Chemical Composition ................................ ........................ 68 3.1.1 CF 3 ................................ ................................ ................................ ........ 68 3.1.2 308L weld ................................ ................................ ................................ 70 3.2 Crystallography and Microstructure Characterization ................................ ....... 73 3.2.1 Spinodal Decomposition ................................ ................................ .......... 73 3.2.2 G Phase Precipitate ................................ ................................ ................. 79 3.3 Mechanical Testing Results ................................ ................................ .............. 84 3.3.1 Nano indentation Results on Welds ................................ ........................ 84 3.3.2 Tensile Test on CF 3 ................................ ................................ ............... 86 3.3.3 Micro Scratch Test on Welds ................................ ................................ ... 88 4 DISCUSSION ................................ ................................ ................................ ......... 90 4.1 Macrostructure and Metallurgical Analysis ................................ ........................ 90 4.1.1 CF 3 ................................ ................................ ................................ ........ 90 4.1.2 308L Weld ................................ ................................ ............................... 93 4.2 Microstr ucture Evolution in Ferrite Spinodal Decomposition ............................. 94 4.2.1 Synergistic Effect of Spinodal Decomposition ................................ ......... 94 4.2.2 Irradiation Enhancement Estimation ................................ ........................ 95 4.2.3 Spinodal Decomposition at High Dose Irradiation ................................ ... 97 4.2.4 Spinodal Decomposition under TEM ................................ ..................... 103 4.3 Microstructure Evolution in Ferrite G Phase ................................ ................... 105 4.3.1 G Phase Formation under Irradiation and Aging in CF 3/308L weld ..... 105 4.3.2 G Phase Coarsening under High Dose Irradiation in CF 3 .................... 108 4.3.3 TEM investigation and APT validation ................................ ................... 110 4.4 Effect of Thermal Aging and Neutron Irradiation Validation ............................ 113 4.5 Mechanical Testing Analysis ................................ ................................ ........... 115 4.5.1 Nano indentation ................................ ................................ ................... 115 4.5.2 Tensile Test ................................ ................................ ........................... 116 4.5.3 Mi cro Scratch Test ................................ ................................ ................. 119 4.6 Structure Property Relationship Validation on CASS/weld ............................. 120 4.6.1 Spinodal Decomposition Strengthening ................................ ................. 120 4.6.2 Particle Strengthening by G Phase ................................ ....................... 123 4.6.3 Other Strengthening Mechanism ................................ ........................... 124 4.6.4 Compact Tension Test and Impact Test of CF 3 ................................ ... 128 5 CONCLUSIONS ................................ ................................ ................................ ... 130 APPENDIX A ATOM PROBE BASICS AND TIP FABRICATION ................................ ................ 133 B MICRO SCRATCH FRACTURE TOUGHNESS ................................ ................... 136

PAGE 7

7 C THE PHYSICS OF NEUTRON IRRADIATION DAMAGE ................................ ..... 139 D SUPPLEMENT RESULTS ................................ ................................ .................... 142 LIST OF REFERENCES ................................ ................................ ............................. 144 BIOGRAPH ICAL SKETCH ................................ ................................ .......................... 153

PAGE 8

8 LIST OF TABLES Table page 1 1 Comparison between ferrite and austenite. ................................ ........................ 17 1 2 Properties of austenitic and ferritic stainless steels. ................................ ........... 18 1 3 Operating pressure and temperature in primary pressure boundary of PWR. .... 28 1 4 Component serviced in PWRs with CASSs and weld. ................................ ........ 41 2 1 Chemical composition of the cast stainle ss steel CF3. ................................ ....... 45 2 2 Chemical composition of 308L weld fusion zone ................................ ............... 47 3 1 APT measured chemical composition of the ferrite at difference. conditions. .... 70 3 2 Average composition of G phase precipitate in CF 3 ................................ ......... 80 3 3 Summary of G phase precipitation. ................................ ................................ .... 82 3 4 Fracture toughness of 308L weld measured from micro sc ratch test. ................ 88 A 1 LEAP Tip Annular Milling Steps ................................ ................................ ........ 135

PAGE 9

9 LIST OF FIGURES Figure page 1 1 Pure iron cooling curve. ................................ ................................ ...................... 16 1 2 Cr Fe binary system ................................ ................................ ........................... 19 1 3 Impact of alloy element on the loop. ................................ ................................ 20 1 4 Phase diagram of Fe Ni binary. ................................ ................................ .......... 21 1 5 Phase diagram calculated using JMatPro for DSS. ................................ ............ 22 1 6 Influence of composite volume fraction on mechanical properties. ..................... 25 1 7 Application of CASS/weld in PWR. ................................ ................................ ..... 27 1 8 Comparison of spinodal decomposition and nucleation & growth. ...................... 32 1 9 Ostwarld step theory of precipitation behavior. ................................ ................... 34 1 10 Flux of element during G phase formation and spinodal decomposition of ferrite. ................................ ................................ ................................ ................. 35 1 11 Full size Charpy V notch impact energy of CF 3. ................................ ............... 3 9 1 12 Application o f CASS/weld in reactor core internals. ................................ ............ 40 2 1 SEM images of a typical TEM sample on the grid fabricated using FIB. ............ 49 2 2 Cr frequency distribution of ferrite in DSS with varied aging time. ...................... 52 2 3 Cr map in ferrite phase in DSS with varied aging times. ................................ ..... 52 2 4 Typical RDF plot of Cr Cr of ferrite in aged CF 3. ................................ ............... 53 2 5 Typical proxigram of Cr of spinodally decomposed ferrite. ................................ 54 2 6 Mn Mn nearest neighbor distributions for orders from 1 to 10 ............................ 56 2 7 Gaussian peak deconvolution for the 4th order nearest neighbor distribution and determination of d max and d e ................................ ................................ ........ 57 2 8 Cluster analysis parameters selection of the N min in MSM. ................................ 57 2 9 Cluster analysis results of ferrite in 308L weld aged for 2,226 h at 400 C. ........ 58 2 10 Optical image of the sample surface of electropolished 308L weld at 500X. ...... 59

PAGE 10

10 2 11 Loading and unloading curve for the nano indentation. ................................ ...... 60 2 12 Tip scanned image of 308L weld with indentations. ................................ ........... 61 2 13 Sample dimension for the sub size ten sile test of CF 3. ................................ ..... 62 2 14 Schematic of the micro scratch test. ................................ ................................ ... 63 2 15 Indenter shape function calibration with fitted curve. ................................ .......... 65 3 1 Metallographic image of CF 3 ................................ ................................ ............. 68 3 2 Metallographic image of unaged CF 3 using SEM. ................................ ............ 69 3 3 Metallographic image of 308L weld ................................ ................................ .... 71 3 4 Metallographic image of 308L weld using SEM. ................................ ................. 72 3 5 Cr distribution in ferrites upon different treatment conditions of CF 3. ................ 74 3 6 Fe Cr elemental frequency distributions in ferrites for different conditions. ........ 74 3 7 Frequency distribution of Fe and Cr in ferrite of irradiated CF 3. ........................ 76 3 8 Wavelength vs irradiation dose of neutron irradiated CF 3. ................................ 77 3 9 TEM images of spinodal decomposition with mottled structure in ferrite of 308L weld The small figure in the lower left corner is the APT reconstruction image. ................................ ................................ ................................ ................. 78 3 10 G phase precipitates in aged, irradiated, and aged irradiated ferrites. The images are sized to have an identical scale, and the isovalue thresholds of Mn, Ni and Si are 2, 10 and 6 %, respectively. ................................ ................... 80 3 11 Cluster size distribution of neutron irradiated CF 3. ................................ ............ 81 3 12 The ionic concentration of Ni, Si and Mn in G phase precipitate in ferrite of irradiated CF 3. ................................ ................................ ................................ ... 81 3 13 Diffraction pattern of ferrite matrix and G phase precipitate of CF 3 irradiated at 20 dpa. ................................ ................................ ................................ ........... 83 3 14 TEM image of G phase precipitate in ferrite of CF 3 irradiated at 20 dpa. .......... 84 3 15 Nano hardness of the austenite and ferrite + austenite phase of 308L weld aged at 0 h and 2226 h. ................................ ................................ ...................... 85 3 16 Stress strain curves of CF 3 with 24% ferrite tested at room temperature. ........ 86

PAGE 11

11 3 17 SEM fractography of CF 3 with 24% ferrite ................................ ........................ 87 3 18 SEM images of the scratch of 308L weld, images on the right are the red box area in images on the left. ................................ ................................ .................. 89 4 1 70 wt.% Fe Cr Ni pseudo binary phase diagram. ................................ ............... 91 4 2 Variation in solidification mode in weld. ................................ .............................. 94 4 3 Schematics of spinodal decomposition. ................................ .............................. 98 4 4 Cr proxigrams in ferrite of irradiated CF 3. ................................ ....................... 100 4 5 Iso surfaces of Mn Ni Si clus ters and interfaces of Cr phases ...... 107 4 6 G phase precipitate in ferrite of CF 3 irradiated at 20 dpa reconstructed in IVAS. ................................ ................................ ................................ ................ 111 4 7 Relative position of the G phase cluster from Ni iso surface and the cylinder. 113 4 8 G phase precipitate in ferrite of CF 3 irradiated at 20 dpa reconstructed in IVAS. ................................ ................................ ................................ ................ 113 4 9 Decoupled stress strain curve of ferrite, austenite and CF 3 in exaggerated illustration. ................................ ................................ ................................ ........ 118 4 10 Strength and hardness of unaged Fe Cr alloys. ................................ ............... 125 A 1 LEAP Tip fabrication steps in FIB SEM procedure. ................................ .......... 135 B 1 Relationship of fracture toughness and sample thickness. ............................... 136 C 1 Schematic of neutron irradiation damage in solid. ................................ ............ 140 D 1 Maximum size, mean size and volumetric number density of G phase precipitates of CF 3. ................................ ................................ ......................... 142 D 2 As weld 308L OM image of the base metal and fusion zone. ........................... 142

PAGE 12

12 LIST OF ABBREVIATION S ANL Argonne National Lab APT Atom probe tomography ASTM American standard testing method AWS American welding society BWR Boiling water reactor CAES Center for Advanced Energy Studies CASS Cast austenitic stainless steel CT Compact tension DSS Duplex stainless steel EBSD Electron back scattering diffraction E DS E nergy dispersive X ray spectroscopy FIB Focused ion beam IVAS Integrated visualization analysis software LEAP Local field atom probe INL Idaho National Lab LWR Light water reactor MaCS Microscopy and Characterization Suite MSM Maximum separation method ORNL Oak Ridge National Lab PKA Primary knock on atom PWR Pressurized water reactor RDF Radial distribution function

PAGE 13

13 SEM Scanning electron microscopy TEM Transmission electron microscopy UTS Ultimately tensile strength YS Yield strength UE Uniform elongation

PAGE 14

14 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy EFFECT OF NEUTRON IR RADIATION AND THERMA L AGING ON CAST AUSTENITIC STAINLESS STEEL AND STAINLESS STEEL WELD PHAS E STABILITY By Zhangbo Li Dec ember 2017 Chair: Yong Yang Major: Material s Science and Engineering Cast austenitic stainless steels ( CASS s ) and welds used in light water reactors ( LWRs ) usually adopt a duplex structure consisting of austenite and ferrite Their compounds are widely used in LWRs as primary pressure boundary and reactor vessel internal components. The common cast stainless steel s in service include CF 3 and CF 8 series with ferrite content up to 30%. F or the welds ferrite contents are limited to less than 20%. P otential embrittlement can occur due to microstructure evolution during LWR environmental degradation Fully understanding of such phase decomposition related embrittlement is crucial to the accurate evaluati on and assessment of the LWR sus ta ina bility program Therefore, s ystematic experiment s and analysis have been performed in this study to investigate the effect of aging and irradiation on the phase stability of ferrite in CASS/weld. The stability of ferrite in CF 3 under long term aging and irradiation w as examine d L ow dose neutron irradiation can enhance the de composition of ferrite synergistically with thermal aging owing to irradiation enhanced diffusion At high

PAGE 15

15 irradiation dose s the ferrite phase exhibited a steady evolution of phase transformation with respect of spinodal decompositio n and G phase precipitation coarsening For 308L weld b oth long term aging and l ow do se neutron irradiation induced spinodal decomposition with large Cr concentration fluctuation wavelength and amplitude and minor G phase precipitate Such decomposition could be the major reason for the ferrite phase nano hardness increase w hile little c hange in the hardness of austenite was observed The periodic strain field combined with the Cr rich zone lattice friction augmentation would both oppose movement of dislocation and strengthen the ferrite phase.

PAGE 16

16 CHAPTER 1 BACKGROUND AND INTRODUCTION 1 .1 Ba sics of CASSs/Welds 1.1.1 Ferrite and Austenite Structure s and Propert ies Shown in Figure 1 1, t he pure iron cooling curve illustrate s the transformation sequence of iron from liquid to ferrite > austenite ferrite a t different temperature range s ferrite both have a body centered cubic (B.C.C.) st ructure with lattice constant of 2 85 2 88 In comparison, the austenite phase has a face centered cubic (F.C.C.) crystal struc ture with lattice constant of 3 598 [1] Figure 1 1. Pure iron cooling curve. ( Source: Ref. [2] )

PAGE 17

17 For the F.C.C. structure the close packed plane has the smallest lattice spacing. Thus, less activation energy is required for slip to occur compared with other planes. In crystallography, there are f our sets of independent close packe d planes in the F.C.C. structure which are {111} type planes. Each plane contains three <110> slip directions. There are 12 slip systems for F.C.C. crystal s in total All of them are identical with the same activation energy for dislocation slip. In comparison, B.C.C. crystals do not have close packed planes, and s lip can occur along the plane with the smallest Burgers vector Plane types identifi ed are six {110} types, twelve {112} types and twelve {123} types. There are two <111> direction s for each {110} plane and one for the {112} and {123} planes. In total, t here are 48 slip systems in B.C.C. crystals. Table 1 1 shows the details of ferrite an d aust enite slip plane configuration s based on their crystallography [3] Table 1 1. Comparison between ferrite and austenite. (Source: Ref. [1] ) Phase Crystal structure Lattice parameter ( ) Slip plane, direction & systems vector ( ) Ferrite B.C.C. 2.865 {110},{112},{123} <111> (48) 2.48 Austenite F.C.C. 3.598 {111} <110> (12) 2.54 Usually, for a duplex stainless steel (DSS) the ferrite phase is less ductile and harder compared with austenite phase at room temperature [4] Ferrite can become brittle at extremely low temperature and exhibits a characteristic ductile to brittle transition phenomenon In contrast temperature drop has limited impact on the ductility of austenite. An implicit explanation is that the dislocation movement in ferrite requires greater activation energy. Above the ducti le brittle transition temperature, the kinetic energy of thermal vibration is sufficient to facilitate the dislocation movement. This helps enable the ductile behavior of ferrite. Below the ductile brittle transition temperature,

PAGE 18

18 dislocation becomes less m obile due to decreased thermal vibration s of the atoms Greater external stress is required to enable dislocation movement If the requi red stress i s greater than that for the crack formation, deformation happens via crack ing instead of dislocation movement Ferrite becomes brittle under such conditions The brittle cleavage usually occurs along the [001] plane, as it is not the close packed plane. In general, ferrite phase has greater strength, hardness and stress corrosion cracking resistance whil e austenite phase has excellent performance in toughness, ductility weldability and castability Table 1 2. Properties of austenitic and ferritic stainless steels. Properties Austenitic Ferritic Toughness Very high Moderate Ductility Very high Moderate Weldability Good Limited Thermal expansion High Moderate Stress corrosion cracking resistance Lo w Very high Magnetic properties Non magnetic Ferro magnetic 1 .1.2 Alloy Elements in S tainless S teel A lloys The alloy elements play a very important role in the austenite/ferrite formation in stainless steel alloys Elements that tend to promote austenite formation are termed as austenite former s or stabilizer s On the contrary, elements that favor the stable fe rrite phase are commonly referred to as ferrite former s Austenite stabilizers inc lude carbon, nickel, manganese and nitrogen. Manganese and nitrogen are usually added to replace some nickel as an austenite stabilizer. Withou t austenite stabilizer, the ret en tion of is almost impossible at room temperature. Figure 1 2 show s the phase diagram of Fe Cr where a loop can be seen. The loop indicates that the composition and temperature range for stable austenite is quite limited for Fe Cr binary alloy s T he addition of an

PAGE 19

19 austenite former, such as nickel, can expand the loop This could help with the control of austenite content Ferrite stabilizers include chromium, silicon and molybdenum. The addition of these elements will retard the formati on of austen ite while favoring the ferrite phase The impact of those alloy elements on the loop is shown in Figure 1 3. Figure 1 2. Cr Fe binary system. ( Source: Ref [1,5] )

PAGE 20

20 Figure 1 3. Impact of alloy element on the loop. ( Source: Ref. [6] ) Carbon has been known to be the most common interstitial element for steel alloys. Its ato mic s ize is only about 1/30 of an iron atom. Carbon helps increase the tensile strength and hardness due to solid solution strengthening. Carbon is a very strong austenite former Its concentration is typically kept less than 0.05% to prohibit the formation of carbide at grain boundaries in austenitic stainless steels Chromium is the essential element in stainless steel to enable the stainless property. When containing greater than 10.5% chromium, an oxide thin film will form on the surface of steel when exp ose d to air as passivation. This thin oxide layer prevents further oxidation of stainless steel. Chromium is a strong ferrite stabilizer. The ato mic radii of chromium and iron are very close with iron atoms at 0.126 nm and chromium

PAGE 21

21 atoms at 0.127 nm [7] As chromium exists in the B.C.C. structure at room temperature, it greatly favors ferrite formation by substitution for iron atom s Nickel is a strong and expensive austenite former widely used in various kind s of stainless steels. Ni tends to partition to the austenite phase during the alloy solidification process owing to its low solubility in ferrite Pure Ni exhibits an F.C.C. crystal structure shown in the b inary Ni Fe phase diagram in Figure 1 4 When alloyed with Fe, Ni can perfectly dissolve in F. C.C. austenite owing to the negligible lattice misfit. When Ni content is less than 20 at.% in Fe, no intermetallic phase s with ordered structure can be formed as indicated in the phase diagram. Nickel can also help improve the fracture toughness, high te mperature yield strength, formability and we ldability. Figure 1 4. Phase diagram of Fe Ni binary (Source: Ref. [8] ) Molybdenum primarily increases the pitting corrosion resistance of stainless steels and high temperature strength in austenitic and DSSs [9] Molybdenum containing grades of stain less steels are generally more corrosion resistant than

PAGE 22

22 molybdenum free grades. In austenitic stainless steels between two to seven percent of molybdenum is usually added, whereas in DSSs between three to five percent is added The addition of one or two percent molybdenum to ferritic stainless steels also significantly increases the corrosion resistance and the elevated temperature strength through solid solution hardening. Application of this effect in industrial alloys can be found in heat exchangers a nd other elevated temperature equipment such as automotive exhaust systems. 1.1. 3 Metallurgy in CASS and Weld CASS and w eld typically have a structure of ferrite phase surrounded by the retained austenite because of the incomplete phase transformation from ferrite to austenite during the subsequent rapid cooling. This is very common in the DSS cooling process as well A phase diagram by Li [10] calculated using JMatPro of slow cooling of DSS is shown in Figure 1 5 In this case, the c ooling rate plays a major role in controlling the final volume fraction of ferrite phase Figure 1 5 Phase diagram calculated using JMatPro for DSS. (Source: Ref. [11] )

PAGE 23

23 Because of incomplete ferrite to austenite transformation s illustrated above CASS c an contain ferrite with volume fraction s up to 30%. The fusion zone s of the stainless steel welds normally have a smaller ferrite content. They both share a duplex structure of branch like ferrite surrounded by aus tenite. Such a composite structure allows for excellent performance of CASS/welds in power plants, oil and gas industr ies wher e oxidation resistance and good strength are required. The determination of ferrite formation in CASSs and welds has been well studied in the past decades [12] As the microstructure de termines the mechanical properties in application s many experiments and theories have been developed to understand the ferrite format ion process. Several American Standard Testing M ethod ( ASTM ) / American Welding S ociety ( AWS ) standards for CASSs/welds have been established. In addition, considerable research has been carried out concerning the ferrite content determination. For CASS AS TM A800 covers grades CF 3, CF 3A, CF 8 and CF 8A I n addition, AWS A4.2 covers most of the welds Normally the ferrite number is used to indicate the ferrite volume fraction in those stainless steel s The quantity of ferrite contained in the stainless st eel is fundamentally a function of the chemical composition and its thermal treatment history. Primarily three different methods can be used for CASSs introduced as described below. The first method is the chemical composition determin ation method. The to tal ferrite promoting element s and total austenite promoting elements are calculated as chromium equivalents and nickel equivalents respectively, usin g empirical equations. The values are then used to compare with the Shaeffler diagram via corresponding

PAGE 24

24 c hromium and nickel equivalents The precision of the ferrite content estimated from the chemical composition depends on the accuracy of t he chemical analysis procedure. The second method is the magnetic induction method which has been widely applied A probe is us ed to measure the magnetic content of a specimen that interact s with its built in magnetic field. Such interaction can induce a change in vo ltage inside the coil which can be measured to estimate the overall content of ferrite. This method is th e most accurate method for casting alloys [13] The third method is metallographic examination. An e tching technique is needed to reveal the ferrite phase Ferrite content can be esti mated from the ratio of ferrite area over the total area based on the recorded micrographs as explained in test method E562. The m ethod for the determination of ferrite content in austeni tic stainless steel weld in AWS A4.2 is very similar to the magnetic response method described in ASTM A800. Because of its facile procedure and good accuracy, the magnetic response method was used for the determination of ferrite content of CASSs/welds in this study. 1.1 4 Structure/Property Relationship s in CASS/W elds CF series steels can contain delta ferrite phase with volume fraction up to 30%. The percentage of ferrite c an be measured by the magnetic method Such a duplex structure of secondary phase surrounded by the matrix is quite similar to the structure of fiber reinforced composite material s Based on the composite theory by Campbell [10] the ferrite content and structural configuration can impart excellent mechanical property enhancement to the bulk Content greater than 30% of ferrite could easily allow for the connected and continuous structure shown in Figure 1 6 This could enable a much greater impact on strength and modulus compared with random short configuration in

PAGE 25

25 structure Thus for typical application s of CASS, the ferrite content is best mai nt ained at around 15% 25% to allow for a sufficient amount o f toughness. In comparison welds usually have much less ferrite content. Figure 1 6 Influence of composite volume fraction on mechanical properties (Source: Ref. [10] ) Another widely a nalyzed structure/property relationship in CASS/welds is the impact of ferrite morphology. It has been reported by Kamiya and Karlsson [14,15] that ferrite in continuous film like form can cause a certain amount of fracture toughness we akening of the weld. In comparison the globular or discon tinuous vermicular ferrite shows little fracture toughness shift. As further explain ed by Ch opra [16] a globular ferrite morphology will have a higher fracture toughness. In comparison, a lacy network of ferrite with con tent greater than 30% could act as a continuous fracture path for crack

PAGE 26

26 propagation. At very low temperature, ferrite turns brittle while the austenite stays ductile. The continuous thin film morphology allows crack s to form and propagate along the failed ferrite network. Meanwhile in the discontinuous ferrite morphology configuration, crack propagation could be slowed or stopped by the austenite matrix However, Bonnet [13] proposed that the ferrite skeleton c ould remain continuous even with content less than 5%. He selectively dissolved the austenite phase by an electro chemical method. While the metallographic image indicated that the ferrite phase was fully isolated the ferrite s keleton structure was continuous. The ferrite morphology in CASS, such as CF 3 and CF 8, usually exhibits a partially connected short range acicular morphology [17,18] Such a morphology can maintain excellent strength enhancement without harming the ductile and tough austenite ma trix. In general t he ferrite/austenite duplex structure in CASS/weld provides excellent combination of toughness, weld ability and castability with high strength. 1.2 Application of CASS /Weld in LWRs T he CF 3 and CF 8 alloys have been used in LWRs in primary pressure boundaries and core internal components [19] A general description of the types of steels used in pressurized water reactors ( PWRs ) i s illustrated in Figure 1 7 For the primary pressure boundary, CF 8A and CF 8M have been primarily used for reactor coolant pipes, pump casings, reactor coolant valve bodies and cover s. CF 3 later replaced CF 8 in the construction of those components. Type 308/308L weld metals are often used for welding the austenitic steel parts, e.g. ci rcumferential and vertical welds in boiling water reactor ( BWR ) core shroud s, in vessel weld s for the core control drives, and in jet pump assembly welds. These weld ing alloys are also used in the overlay cladding of the reactor pressure vessels. To prevent hot cracking, stainless steel welds

PAGE 27

27 ferrite up to 10% in the austenite matrix. Table 1 4 listed the operation temperature and pressure for the primary and secondary loop. Generally those components operate i n the temperature range of 275 C 315 C under high pre ssure at 2250 psi and in corrosive environment s Figure 1 7. Application of CASS/weld in PWR. (Source: Ref. [20,21] ) T he ferrite phase can help improve the strength, corrosion resistance and weldability of the casting /welds However the duplex structure stainless steel

PAGE 28

28 cast/welds are prone to ther mal aging embrittlement With a design ed service lifetime of 40 years aging can induce various phase decompositions in ferrite. This c an dramatically affect its m echanical properties Based on the experimental data phase, phase carbide and Cr rich phase all will contribute to the embrittlement of the casting under various condition s With the stabili phase formation temperature was pushed up to 950 C but t o avoid formation, casting must be cooled past 850 C in phase was typically formed above 550 C and h as rarely been detected when aged below 500 C T hus Cr rich phase plays the major role in cast embrittlement Table 1 3 Operating pressure and temperature in primary pressure boundary of PWR. (Source : Ref. [22] ) Loop Pressure T (hot leg) T (cold leg) Primary 15.5 MPa (2250 psi) Secondary 6.2 MPa (900 psi) 1 3 Aging Degradation Study of CASS/Weld in LWRs 1.3.1 Aging Embrittlement S tudy Methodology The earliest research interest on CASS c an date back to t he aging degradation of CASS/weld program initiated in 1985 at Argonne National Lab ( ANL ) Its goal wa s to characterize and correlate the microstructure of in service reactor components and laboratory aged material embrittlement and to ident ify the mechanism of embrittlement The basic methodology i s summarized in the following paragraph s To evaluate the aging related embrittlement kinetics, the aging process was assumed to be thermally activated The activation energy can be estimated by ex amining the onset of embrittlement ( C harpy V notch impact toughness measure d at room temperature) through aging experiment s at varied time and temperature. Based

PAGE 29

29 on the initial data tested by George Fisher of Switzerland on CF casting in the temperature range of 300 C 400 C, an activation o f 100.4 kJ/mol was extracted. The predict ion of the long term aging embrittleme nt of steel casting can be achieved through the Arrhenius e xtrapolation shown in Eq. 1 1 in which E a is the activation energ y and R is the gas constant. For a service component at aging temperature T and time frame t, the equivalent aging parameter P c an be calculated with respect to the degree of embrittlement at 400 C. (1 1) The application essence of those material s in LWRs was the focus of a field engineering study Depending on the source of the material obtained, the heat treatment history can be quite different For samples obtained from a decommissioned power plant, i t is difficult to trace the ir original metallog raphy and heat treatment history. Because of this some researcher s have performed aging on CAS S samples without consideration of the heat treatment [24] In other cases, c asting s in small dimensions were solution annealed at 1040 C 1060 C [25] Recently, Li [11] carried out solution treatment at 1080 C for varied time length s on DSSs Combined with subsequent TEM a nd tensile testing on the later aged samples, the study showed that the solution annealing has little effect on ferrite compositions with insignificant impact on the thermal aging kinetics. It is trustworthy that the grain size and the grain orien tation effect of the casting were not considered as key parameter s in th e aging degradation studies described above The austenite dendrite grain in CASSs can reach a few m ill i m eters Within each

PAGE 30

30 austenite grain, hundreds of scale ferrite grains reside at the dendrite core of austenite Ferrite acts as a reinforcement inside the ductile austenite matrix. This duplex structure plays a critical role in maintaining the high strength and excellent ductility of CASS Such a comp osite like structure has been confirmed by Schwarm [26,27] via finite element modeling (FEM) Schwarm investigated CF 3 and CF 8 in non aged and aged condition via tensile testing and nano indentation. Further electron backscattering diffraction and FEM were also performed to simulate the process of mechanical testing. Ferrite tends to be more stress sensitive while austenite is more strain sensitive. Compared with austenite grain boundaries, the ferrite/austenite heter o phase boundaries play a significant role in plastic deformation. Furthermore, Wa ng [28] investigated the plastic deformation of ferrite/austenite duplex structure steel s using t ensile testing. Scanning electron microscopy ( SEM ) was used to record the specimen surface deformation pattern after a certain strain was reached. The result indicate d that the ferrite /austenite phase boundary plays a major role in strengthen ing the materi al after long period s of thermal exposure. Thus to facilitate the study and simplify the discussion, only the ferrite/austenite duplex structure w as cons idered as the key factor in the present study. 1.3.2 Microstructure Evolution of Thermally Aged Ferrit e It has been known for over 60 years that f errit e containing stainless steels are susceptible to aging embrittlement when serviced in the reactor operational temperature range [17] Identified phase transformation s in ferrite include spinodal decomposition and Ni Si rich G phase in the temperatur e range of 400 C. The Fe and Cr binary system has been well studie d from a metallurgical perspective Similar to other binary systems such as Cu Ti Ai Ag and Ni Au, th ey all

PAGE 31

31 exhibit a miscibility gap in a certain temperature range. The phase se paration evolution inside the miscibility gap can sometimes go without nucleation and growth leading to formation of an interconnected morphology known as spinodal decomposition Spinodal decomposition of ferrite occur s when the ferrite composition falls inside the spinodal region of the phas e diagram of Fe and Cr (Figure 1 2) Fe and Cr atoms demix from each other through uphill diffusion shown in Figure 1 8 T h is process eventually results in a very fine modulating morphology of Cr rich A s the Cr concent ration fluctuates the ex tent of decomposition can be measur ed by the wavelength of the Cr conc entration profile can lead to hardening of the ferrite phase Historically the characterization and measurement of spinodal decomposition in Fe Cr binary system has mainly relied on the Mossbauer spectroscopy [29 33] Nowadays, spectroscopy with atomic resolution tools are normally used for the characterization and quantification. T o explain the spinodal decomposition in the perspective of free energy, the Cahn Hilliard equation was introd uced as shown in Eq 1 1 to Eq. 1 3 [34] In these equations, G denotes the Gibbs free energy f(c) indicates the Gibbs free energy as a function of solute concentration c The term is the in crease in free energy due to concentration fluctuation. If we assume the binary solution to be incompressible and isotropic with c negative if

PAGE 32

32 inside the spinoda l region. Such a driving force can enable the binary system to demix into tw o phases with different composition s via uphill diffusion. Depending on the lattice parameters of the two phases, the misfit value can be calculated. For the Fe Cr binary system, is quite sm all i.e., far less than 5%. This can help explain why the two phases are highly coherent with little distinction in X ray diffraction ( XRD ) and transmission electron microscopy ( TEM ) diffraction pattern s Figure 1 8 C omparison of (a) spinodal decomposition and (b) nucleation & growth. (Source: Ref. [35] )

PAGE 33

33 An early study performed by Miller [36] on Fe Cr binary alloy used a field ion atom probe ( FIAP ) to identify the Cr rich was sometimes mistakenly interpret ed as a precipitate during the 1990s. Chopr a [16] studied CF 3 and CF 8 but reported no discovery of spinodal decomposition in the ferrite phase due to the precipitation interpretation Many recent experiment s performed on CASSs/welds have found spinodal decomposition in ferrite phase using atom probe tomography ( APT ) due to its atomic resol ution For a nucleation and growth process, the driving force is usually the Gibbs free energy shown in Eq. 1 4 (1 4) The first term , denotes the difference in chemical potential when a nucleus with radius of r is formed from the matrix. The term indicates the the induced strain energy of the nucleus and denotes the corresponding surface free energy increase Traditionally, precipitate forms via the sequence of G.P. zone > > > [37] The G.P. zone normally has exac tly t he same microstructure as the matrix, but slightly diff erent chemical composition. Th exactly the same chemical composition but slightly different crystal structure s with respect to the coherence level with the parent phase. Overall the composition difference between the precipitate and surrounding matrix is achieved via long range solid diffusion within the parent phase [38] A more detailed phase transformation i s illustrated step by step in Figure 1 9 The final phase is different matrix in both chemical composition and microstructure Instead of the phase nucleati ng directly from the th a large

PAGE 34

34 activation barrier, precursors can be formed as a result of intermediate process es with much smaller formation barrier s Figure 1 9 Ostwarld step theory of precipitation behavior (Source: Ref. [3 5 ] ) In principle, the precipitate tends to nucleate along the grain boundary of the parent phase where int erfacial and strain energy can be well accommodated [40] The name G phase was given as it wa s a new phase discovered at the grain boundary. This normally takes a long time as both the crystal structure and chemical c omposition of the precipitate are quite different from the matrix. Thus, it was normally discovered in austenit e when aged at high temperature G phase precipitation is an intermetallic silicide It has a n F.C.C. structure with lattice parameter no greater than 1.14 nm

PAGE 35

35 depending on chemical composition [1,39] Its c ommon clustering elements include Ni, Mn, T i and Nb. As a secondary phase found in CASS/weld materials when aged at reactor operating temperature [25,41] G phase ca n strength en the ferrite phase via precipitate hardening For CASS/weld ferrite, G phase is known to have a longer incubation time compared with spinodal decomposition when aging temperature is lower than 400 C [39,42] Thus, G phase format ion is ty pically considered to rely on the spinodal decomposition process. Li [11] investigated DSS aged at 400 C for 3000 h using TEM. He report ed that G phase was not observed as there was only one set of diffr action pattern s in the ferrite matrix It was further explained that the incubation time for G phase could be eve n longer than 3000 h at 400 C. Figure 1 10. Flux of element during G phase formation and spinodal decomposition of ferrite. (Source: Ref. [39] )

PAGE 36

36 Spinodal decomposition is known to be a pivotal process for G phase precipitation. In fact, Mateo [39] proposed the correlation of G phase formation on spinodal decomposition in a model depicted in Figure 1 10 F or Cr rich Ni i s expelled towards the phase boundary owing to its limited solubility. Likewise, Si is expelled from Fe The combined effect is that more and more Si and Ni atom s gather and accumulate at the interface region of an To minimize the Gibbs free energy, a more stable G phase starts to nucleate out of those clusters. This assumption was further enhanced by Takeuchi [43] in his study on stainless steel welds 1.3.3 Activation Energy Extraction of A ging Embrittlement Many mechanical tests have been performed to investigate the aging embrittlement CASSs/welds. Pumphery [25] pe rformed aging experiment s on CF 3 in the followed by the C harpy impact test. With an activation energy of 169 kJ mol 1 he concluded that aging at s sufficient to simulate the end of 3. Chemical composition may have a direct influence on the spinodal reac tion. However, there wa s in suf ficient evidence showing that chemical composition can also influence the aging activation energy. Chung [19] further conducted both aging experiment s a nd mechanical property test s over a wide temperature range on CF 3, CF 8 and CF 8M. It accuracy was primarily determined by the agin g activation energy, which can be affect ed by the chemical composition. In the meantime, Pareige [44] performed long term aging experiment s by mechanical property

PAGE 37

37 test s B y using the equivalent time model Experiments to extract the weld aging embrittlement activation energy were performed during the early 1990s by Grobner [9,45] Trautwein and Gysel [46] with inconclusive results ranging from 100 kJ mol 1 to 230 kJ mol 1 Tavassoli [47] attempted to determine the aging activation energy of spinodal decomposition in the temperature range of 490 and reported an activation energy of 67 kJ mol 1 Vitek [41] further conducted aging experiments on 308L welds at an elevated temperature range from 4 results indicate d that the aging kinetics of sp s the most rapid [29] group performed aging experiment on 304 and 316 They reported that aging at higher temperature tended to yield greater activation ener gy value s compared with lower temperature range aging 1.4 Motivation 1.4 .1 Nuclear Energy and LWRs Sustainability Program The N uclear R egulatory Commission has licensed 99 LWRs which produc e about 20% of the electricity in the United States over the past two decades T he LWRs contribute to more than 70% of the non greenhouse gas emitting electric power generation Over 75% of those LWRs are generation II reactors reaching the end of their 60 year operating licenses In the meantime, the dom estic electricity demand is expected to grow at around 1% annually. If those nuclear power plants do not continue to operate beyond 60 years, more new power plants need to be built to meet the energy demand To build a new advanced nuclear fleet can cost hundreds of billions of dollars. T raditional fossil plants are cheaper but they will increase greenhouse gas emission

PAGE 38

38 dramatically In comparison, extending the current license lifetime of 60 years is a low risk option and cost effective To ensure that t he current LWRs can be operat ed safe ly reliably and economically t he O ffice of N uclear E nergy has identified four R&D pathway s within the LWR sustainability frame. They are material aging and degradation, advanced LWR nuclear fuel, advanced instrumentati on information and control system technologies and risk informed safety margin characterization. The goal of the material aging and degradation pathway is to develop a scientific basis for better understanding and more accurately predicting the long term environmental d egradation behavior of the primary system, structure and component s in LWRs The program will prov ide data and methods to assess LWR performance which is e ssential to the ir safe and reliable operation 1.4 .2 Challenges of Neutron Irradiation E nhanced Aging D egradation CASSs and welds in LWRs are susc eptible to aging embrittlement in 32 has been well established in the past two decades as discussed in Section 1.3 S pecifically, for the low carbon steel CF 3 when aged at elevated temperature, an obvious characteristic ductile brittle transition phenomenon of its impact energy can be observed as shown in Figure 1 11. Both a decrease in upper shelf energy and an increas e of ductile brittle transition temperature indicate the embrittlement tendency. Fortunately, if only consider ing the designed 40 year lifetime at reactor operation temperature, the corresponding impact e nergy of CF 3 i s still quite high to guarantee its safe and reliable functionality.

PAGE 39

39 Figure 1 11. Full size Charpy V notch impact energy of CF 3 (Source: Ref. [17] ) As s hown in Figure 1 1 2 s ome CASS component s such as core support columns and mixing devices, may be exposed to neutron irradiation with fluence level o n the order of 10 20 n/cm 2 (E > 1 MeV ) when considering an extended service life time of 80 years For the current 60 years of service, the neutron irradiation dose on those CF series steels can range from 0.01 displacement per atom ( dpa ) up to more than 10 dpa. Thus it is quite important to establish the corresponding irradiation response s of those servic e component s As pointed out by Chopra [48] current data on neutron irradiation damage is more focused on dos e level between 1 dpa 5 dpa. More d ata points are needed in the dose level range of less than 0.1 dpa to greater than 5 dpa.

PAGE 40

40 Figure 1 1 2 Application of CASS/weld in r eactor core internals. (Source: MRP 276) The stainless steel welds in the LWRs are also exposed to neutron irradiation to a relatively lower dose as compared with reactor core components. A s noted by the ERPI technical reports MRP 276 summarized in Table 1 4 the welds can be in the high fluence regions of the PWR internals (e.g. core barrel welds and flux thimble tube plugs) T he a nticipated neutron fluence can be o n the order of 10 22 n/cm 2 .There are

PAGE 41

41 several studies regarding the neutron irradiation eff ect on the fracture property of stainless steel welds [49,50] TEM examination s conducted by Lee et al [51] could not reveal the nature of fine defects even though micro hardness tests show ed a significant ferrite phase after irradiation. Takeuchi [43] observ ed the neutron irradiated microstructures in a stainless steel electro slag weld overlay irradiated to ~ around 1 dpa at 290C. A slight progression of Cr spinodal decomposition and an increase in the fluctuation of the Si, Ni, and Mn concentrations were observed in the ferrite phase Nevertheless, related microstructural studies, particularl y at LWR relevant condition s are very limited Table 1 4 Component serviced in PWRs with CASSs and weld. (Source: MRP 276) Components Grade Irradiation fluence ( n/cm 2 E>1.0 MeV) Potential concerns CRGT assembly spacer castings CF 3M 10 20 Thermal embrittlement CSS assembly cast outlet nozzles CF 8 10 17 Thermal embrittlement CSS assembly vent valve discs CF 8 10 17 Thermal embrittlement IMI guide tube assembly spiders CF 8 10 20 Thermal and irradiation embrittlement Core barrel ASA with 308L 10 20 Irradiation embrittlement Core support shield ASA or AMA with 308L 10 18 Irradiation embrittlement Lower grid STIG with 308L 10 20 Irradiation embrittlement 1.4.3 Comparison of Thermal Aging and Neutron Irradiation Validation H igh energy neutron irradiation can introduce point defects in materials at va ried rates and densities depending on the irradiation parameters Sub sequent thermal diffusion of the se point defects can lead to a healing process. This occurs when the

PAGE 42

42 interstitials and vacancies recombine with e ach other or sinks in the materials and become annihilated. Eventually, the point defects reach a dynamic equilibrium level owing to the continuous neutron irr adiation. Such irradiation induced defects will have a wide va riety of effects on the materials w ith respect t o microphysics and microchemistry such as therm odynamic equilibrium, transport kinetics and local disordering. B ased on the theory and experiment s by Was and Rothman [52,53] the aforementioned equilibrium density of vacancy tends to dramatical ly outgrow that of interstitial by a few orders of magnitude at low irradiati on temperature s Such high density vacancy can enhance the diffusion kinetics in many ways Usually when competing with thermal diffusion at the same temperature, irradiation enhanced diffusion tends to be much faster As a result, microstructure evolution such as phase decomposition under irradiation will become more rapid compared with the slow transformation under thermal aging condition s 1.4 .4 Synergistic Effect of Irradiation and Thermal Aging on Ferrite Decomposition The synergistic effect of thermal aging and neutron irradiation on ferrite phase deco mposition was fi rst introduced by Miller [36] in 1996 He concluded that neutron irradiation could facilitate the spinodal decomp osition at a very low dose of 0.03 dpa i n Fe 32% Cr alloy. For core internal component s exposed to neutron irradiation this irradiation enhanced thermal aging c an bring more inaccuracy to the evaluation of aging embrittlement Thus, it is critical to carry out more experiment s to study the effect of irradiation enhancement on the degradatio n of those components.

PAGE 43

43 I deally, to study the synergistic effect of thermal aging and neutron irradiation on the ferrite phase the material needs to be exposed to the actual working environment in a nuclear reactor Thus, the th ermal aging and irradiation can affect the microstructure simultaneously and interactively. Unfortunately, it is impractical to pursue a real time aging experiment for 60 years or longer In many cases, thermal aging has to be accelerated at a higher temperature to complete the exper iment in a reasonable period It has been reported that the activation energy of spinodal decomposition in a Fe Cr system has a wide range of 164 to 324 kJ/mole [44,54] By considering the low end of activation energy of 164 kJ/mole in a conservative manner, 10,000 hours aging at 400C will correspond to 79 years of service at 315C in a PWR To further study the synergistic effect of neutron irradiation and thermal aging, a neutron irradiat ion experiment can be performed on the aged materials. This method can provide a scientific insight for the understanding of the combined effect of thermal aging and neutron irradiation. 1.4 5 Micro S cratch T est for Fracture Toughness Extraction With the development of the atom probe tomography ( APT ) transmission electron microscopy ( TEM ) and focused ion beam ( FIB ) technique s the microstructur al characterization of the CASS/weld can be systematically performed However, it is very important to c orrelate the microstructure results with the corresponding changes in mechanical propert ies, such as hardness, tensile streng th and fracture toughness. Due to the dimension limit of the materials available in this research, it wa s quite impractical to carr y out mechanical testing such as im pact energy testing Such testing typically requires a large dimension specimen which is very material consuming. With the development of the micro scratch test for fracture toughness characterization t he

PAGE 44

44 sample prepar ation requires only a tiny dimension as big as the fingernail Such testing is a perfect match of the weldment sample s in this study 1. 5 S ummary CASS s and stainless steel welds have been extensively used in LWRs as primary pressure boundaries and core internal components They both have a duplex structure of the ferrite phase surrounded by the aus tenite matrix. Ferrite plays a major role in enhancing the strength, stress corrosion cracking and weldability. However during the designed 40 years of service of LWRs, ferrite is su sceptible to thermal aging indu ced phase decomposition. Recently bot h industry and regulators have strong incentive s to extend the ser vice lifetime of LWRs to 80 years As part of the material aging and degradation researc h and development program, this study is d edicated to perform a systematic study of the CASS/weld. Scientific data and analysis are established to help regulators perform better and more accurate assessment and evaluation on the licensing of LWRs. The scop e of this research is mainly focused on structural characterization of CASS/weld s low strain rate mechanical property testi ng and interpretation, along with subsequent structure property relationship investigation such as the st rengthening mechanism. The high strain rate fracture behavior of the materials is also included.

PAGE 45

45 CHAPTER 2 EXPERIMENT AL DETAIL S 2 .1 Material s and Treatment CF 3 steels were selected f r o m the ANL research program [17,55] with heat batch number 69. This batch of CF 3 wa s static cast slab with dimensions of 610 mm x 610 mm x 76 mm by Foundry of ESCO Corporation following ASTM A351 The slab dimension s were close to that of the pump impeller used in LWRs with Rockwell hardness of 83.7 The slab contain ed approximately 2 3 % volumetric fraction of delta ferrite measured by ferrite scope AUTO Test FE with probe type FSP 1 The documented averag e ferrite spacing wa s around 35 The chemical composition of CF 3 cast slab used in this study is listed in Table 2 1 All the CF 3 samples used in this study were obtained from the interior region of the same cast slab. Bulk material s were sectioned from cast slab and further sliced into small cuboid plate s with dimension s of Some of the bulk materials were further thermally aged. The thermal aging experiment was carried out at 400C for 10,000 hours in a box furnace. A Dualscope FMP100 was further used to measure the ferrite content of the CF 3 bulk samples via the magnetic induct ion method Probe FGAB1.3 Fe was used with a detection limit of 0.1%. Fifteen fields were analyzed for the measurement on unaged and aged CF 3 used in this study. The measured ferrite content is 2 3.6 rt [17] Table 2 1 Chemical composition (wt %) of the cast stainless steel CF3 (Ref. [17] ) Ni Si P S Mn C N Cr Mo Fe 8.59 1.13 0.015 0.005 0.63 0.023 0.028 20.18 0.34 Bal. Some of th e un aged and aged CF 3 plates were further gr ound and polished down to around 150 mm and punched in to 3 mm disc s using a disc punch from SPI

PAGE 46

46 Supplies with model type 17001 AB. B oth surface s of the disc were jet polished to remove the polishing strain layers. Th e discs were further seal ed into a helium fi lled capsule to be irradiated in the reactor T he neutron irradiation experiment was conducted in the Halden reactor in Norway. The irradiation temperature was controlled at 315 C with two sets of melting alloy temperature monitors installed in the irradiation assembly. Due to limited access to th e post irradiated samples, metallography of the irradiated sample s i s not available During the irradiation experiment, three fluence mon itor wires (Fe, Ni, and Al/Co alloy) were placed outside the irradiation capsule. After irradiation, dosimetry reading was performed by the Halden reactor researchers based on the se monitors Using the activation cross s ections determined previously the a ccumulated neutron fluence for the irradiation capsu le could be estimated. The obtained fas t neutron fluence (E > 1 MeV) wa s about to 5. 5 6x10 19 n/cm 2 with a dose rate of 2.8x10 9 dpa/s correspond ing to a displacement damage of 0.08 dpa for the sample [55] To simulate the irradiation under extreme condition, CF 3 discs in as cast condition were kept in 3 mm capsule s, which were further placed in a BOR 60 reactor for neutron irradiation The irradiation tempera ture was co ntrolled at 315 2 C. The irradiation damage calculation was performed by Research Institute for Atomic Reactors [56] Th e irradiation dose rate was estimated at 10 6 dpa/s An ASTM 304L stainless steel weldment fabri cated with ER308L filler metal wa s used in this study. The chemical composition of the weldment used is shown in Table 2 2. The 304L stainless steel plates were welded following a conventional submerged arc butt welding procedure with a double V joint design. Th e parameters for each weld pass

PAGE 47

47 were documented in the data package including current, voltage, travel speed, heat input and so on. On ly the pro perty of the 308L fusion zone wa s considered for the environmental degradation evaluation in this study. T he measured ferrite content of the fusion zone was 13.5 0.5 % using Fe ri scope FMP100 Some of the weldment joint s were thermally aged at 400 C for 2,226 h ours and 6000 h Table 2 2 Chemical composition (wt.%) of 308L weld fusion zone C Mn P S Si Ni Cr Mo Ti Fe N .023 2.06 .032 .006 .92 9.30 20.50 .10 <.01 Bal. .10 2 .2 OM & SEM Metallography A Tagremin 25 polisher with SiC sand pape r was used for the surface polishing finished with 1200 grit (P4000, 2.5 m ) for the non irradiated CF 3 steels and 308L weldment specimens. All grinding and polishing procedures follow ed the microprobe polishing standard by Fynn and Zipperian [57,58] After fine polishing with 1 diamond slurr y, the surfaces were etched with ferric chloride balanced with hydrochloric acid for a few seconds following the guide for special alloy etching [59 61] Due to the limited access to the neutron irradiated 3 mm CF 3 disc, the metallographic image of ir radiated specimens has not been prepared. Further SEM images should be prepared in the future to clarify the structural st ability of CF 3 under irradiation condition s T he CF 3 steel specimens used were in as cast condition with no prior heat treatment his tory. Typically, heat treatment such as solution annealing, is performed to remove the macro segregation and carbide precipitate and to control ferrite content The measured ferrite content stayed relatively the same throughout this research for CF 3. As concluded by Li [11] solution annealing o n cast stainless steels has little

PAGE 48

48 impact in the subsequent aging kineti cs at lower temperature. Thus, use of CF 3 casting with no heat treatment in this study is fully justified. A Bruker XFlash 6|30 EDS on a Tescan LERA3 Xe plasma source FIB/SEM was used for chemical composition analysis in this study on the unaged CF 3. The accelerating voltage used was 15 kV wit h a working distance at 30 mm. T he spot size used was 13 nm. At least five di fferent measurements were performed for the unaged CF 3. The EDS result is shown in Chapter 3. 2.3 TEM Crystallography and Characterization Around 2 mm thick slices of 308L weld were prepared from the fus ion zone. The sectioning plane wa s paral lel to the w elding surface. Th e slices were further gr oun d and polished to around 150 chips which were punched into 3 mm discs using a SPI Supplies disc punch. Those discs w ere further electropolished using a single jet polisher with model type M 550 Specimens we re finished with centra l perforation size of 100 30 The transparent regi on adjacent to the perforation wa s thin enough for TEM investigation. The electrolyte used was 10% perchloric acid balanced with methanol. A c old bath was used with dry ice mixed w ith methanol The bottom part of the plastic bucket of the polisher was submer ged in the cold bath. The temperature of the electrolyte was maintained at 18 2 C measured by a thermometer during the experiment. TEM specimens of the irradiated CF 3 steels and welds were prepared using FEI Quanta 3D FEG FIB on th e 3 mm irradiated discs A typical TEM specimen fabricated using FIB i s shown in Figure 2 1. A Tecnai G2 F30 FEG TEM in Idaho Falls, Idaho National Lab (INL) was used to characterize the microstructure of ferrite phase with an acceleration voltage of 3 00 kV.

PAGE 49

49 Figure 2 1 SEM images of a typical TEM sample on the grid fabricated using FIB. 2 .4 Atom P robe Microstructural Characterization & Analysis With the objective of d evelop ing a fundamental understanding o f ferrite microstructure evolution characterization was focused on the development of spinodal decomposition and G phase precipitation. Because of its atomic resolution [62,63] APT examination w as extensively used to quantify t he redistribution of Cr element i n the ferrite phase and the fine precipitates Chips with thickness around 2 mm were sectioned from the bulk CF 3 and weldment fusion zone using a high speed diamond saw A Techcut 5 high sp eed diamond saw from Allied high tech was selected for the experiment and a diamond T he rotation speed of the wheel wa s set at 3000 rpm, and t he feeding speed used wa C hips

PAGE 50

50 we re further ground to 150 200 using the Tegramin 25 polisher to remove the abrasive cutting damage layer The final step of sample surface grinding was finished with 800 grit sandpaper. A 3 mm disc punch from SPI Supplies with model type 17001 AB w a s used to fabricate 3 mm disc s from the steel chip s The sample surface s were further etched using a jet polishe r from Southbay Technology M550 to reveal the ferrite and austenite phase boundary. Each surface was polished with a 1.5 mm nozzle for 100 s at 182 C with a voltage and current at 75 V and 100 mA, respectively, in a mixture of 10% perchloric acid balanced with methanol. The discs prepared for irradiation were electropolished to reveal the ferrite and austenite phase boundary in ANL A jet polisher with model type of Struer s T enupol 5 was used Each surface was polished with a 1.5 mm nozzle for 100 s at 18 C with the voltage and current at 75 V and 100 mA respectively Electrolyte was a mixture of 5 % perchloric acid balanced with methanol After irradiation experiment, t h e 3 mm discs were shipped back to ANL from the reactor site s After a few more safety protocol communication s between ANL and I NL, irradiated samples were transported from ANL to INL user facility for microstructure characterization In INL, APT tips were prepared for both CF 3 steels and 308L weldment using FEI Quanta 3D FEG FIB focu sed on the ferrite phase of th e 3 mm etched disc S amples were finished using low energy ion beam cleaning with volt age of 2kV and current of 27pA to ensure a minimal Ga implantation. For e ach condition of CF 3 steels and 308L weldments a minimum of seven tips were fabricated The atom probe experiment was carried out in a local field atom probe ( LEAP ) 4000 XHR at 55 K with ion detection efficiency at 37%. The laser pulse rate was set at

PAGE 51

51 200 kHz with pulse energy at 60 pJ. The detection rate was kept between 0.1% 0.5% to allow for large dataset collection T he 3 D reconstruction and analysis were performed using integrated visualization analysis software ( IVAS ) 3.6. 8 following the standard procedure of the R econ Wizard in IVAS. SEM tip images were used for defining tip profiles. The evaporation field value was 33.0V/nm and a k factor of 3.30 was used. The FIB and APT instruments used for this study are both located in the Microscopy and Characterization Suite ( MaCS ) Center for Advanced Energy Studies ( CAES ) in Idaho Falls, Idaho More details of the APT sample fabri cation and LEAP experiment can be found in Append ix A. To quantify the nanoscale spinodal decomposition in ferrite phase, the frequency distribution radial distribution function ( RDF ) and Cr proxigram techniques were used to determine the degree of spinodal decomposition, the wavelength and the amplitu de. 2 4 1 Frequency Distribution The frequency distribution technique is often used to roughly analyze the extent of spinodal decomposition in ferrite phase. The pro cedure included index ing all the identified solute ions which were evenly divided into boxes with a certain number of ions (100/box in this study) that best reflects the distribution trend. For each 100 ion b ox, the concentration of Cr was calculated. The concentration s of all the boxes were histogramed from 0 100% t o generate the frequency distribution of Cr. Figure 2 2 shown a Cr frequency distribution of ferrite in DSS with aging time from 1 h to 1000 h at 450 C. The Cr map i s also illustrated in Figure 2 3 As can be observed, with the increase of aging time, the extent of Cr clustering behavior became greater. As a result the peak shift and peak broadening in Figure 2 2 are typically used to reflect this Cr distribution evolution trend A peak shift towards lower Cr concentration indicates the

PAGE 52

52 formation of a Cr depletion region. On the other hand, the peak broadening towards the higher Cr concentration s is re presentative of the Cr rich clusters Figure 2 2. Cr frequency distribution of ferrite in DSS with varied aging time. (Source: Ref. [64] ) Figure 2 3. Cr map in ferrite phase in DSS with varied aging time s (Source: Ref. [64] ) 2 .4 2 Radial Distribution Function The RDF method calculates the Cr concentration distribution as a function of radial distance from each referenced Cr ion. The clustering tendency c an be revealed based on the fluctuation of the RDF. The RDF method provides a unique quantitative evaluation on spinodal decomposition as compared with other methods, ( e.g. 1D

PAGE 53

53 concentration profiles and frequency distribution ), because it provides a 3 D anal ytical perspective by including all the identified Fe or Cr ions. Input parameters used to perform the RDF calculation in IVAS include the original reference ion type, bin width (nm) and maximum distance (nm). For Cr distribution fluctuation analys is, each and every detected Cr ions in the dataset is settled as the center reference ion [62] By default, the maximum analysis distance is 10 nm. The bin width as the increment ranges from 0.01 nm 1 nm with 0. 2 nm often used to best reveal the Cr cluste ring behavior with out introducing too much noise The algorithm calculate s the shell concentration and generate a histogram based on the bin width. Figure 2 4 Typical RDF plot of Cr Cr of ferrite in aged CF 3. Figure 2 4 shows a typical Cr Cr RDF curve for the ferrite phase in the aged CF 3. The inset in Figure 2 4 highlights the wavelength between two maxim a The

PAGE 54

54 maximum indicates a statistically closest distance between two Cr rich regions and corresponds the Cr concentration fluctuation wavelength. We can assume Cr concentration as a sinusoidal wave functio n [34] : (2 1) where, C o is the average Cr concentration in ferrite phase, A is the concentration fluctuation amplitude, is the wavelength, and is a position vector. 2 .4 .3 Proxigram of Cr Map Figure 2 5 Typical proxigram of Cr of spinodally decomposed ferrite (Source: Ref. [67]) The so called proxigram is a combination of the proximity and histogram technique s in performing concentration profile calculation s A T raditiona l 1 D concentration profile is typically calculated along a certain axis of the cuboid or cylinder. Proxigram, however, computes the concentration profile along varied distance s from a reference geometric surface. As long as there is a re ference surface, proxigram can be

PAGE 55

55 applied for the analysis It is capable of visualiz ing the gradual demixing phenome non of the Cr and Fe system Thus it is very useful for the amplitude and concentration profile study The first publication using this technique i s shown in Figure 2 5 [65] group from Oak Ridge National Lab ( ORNL ) A smooth Cr concentration profile s hifting from the left side of the the phase was demonstrated This can be interpreted as evidence of the diffusing interface forme d due to the spinodal decomposition in ferrite 2 .4 4 Cluster Analysis To quantify s precipitates in ferrite decomposition, a cluster analysis tool is used in IVAS software on the APT data. P d e d max N min L specified during the analysis Maximum separation method ( MSM ) is one of the commonly used approach es to dete rmine the optimal values of th e parameters. The actual application of this method in this study is focused on the G phase cluster detection and identification. G ph ase is typically enriched in Ni, Si and Mn. With the help of this method, the size, density, chemical composition and volume fraction of the G phase clusters c an be quantifie d. A typical procedure on how to carry out the analysis is describe d in the follow ing paragraph. Based on the preliminary analysis, Mn ions showed the best distribution contrast between matrix and clustered regions. Therefore, it was chosen as the ion for determining the solute cluster regions to select the desired orders (ions). The Mn Mn nearest neighbor distributions for orders of one to ten are plotted in Figure 2 6, which shows that 4 th order of nearest neighbor starts to shows a clear separation of ions between matrix and cluster regions. The left peak represents the ions in cluste r regions

PAGE 56

56 and the right peak represents the ions in a more diluted matrix regions. The 4 th order nearest neighbor distribution was then fitted using two Gaussian distributions, as shown in Figure 2 7. The value of d max is determined as the distance between zero and the crossover of two fitted Gaussian distributions. The value of d e is the distance between the peak of the left Gaussian distribution and the crossover of two Gaussian distributions. As suggested by Y. Chen et al. [66] envelope parameter L equals d max .To determine parameter N min the minimum of ions in a cluster was counted in the analysis. The curve of identified cluster number vs N min is plotted in Figure 2 8. As can be seen, a reasonable choice for N min is 20, which eliminates clusters containing only a few atoms while giving a stable number of clusters, as increasing N min does not significantly change the number of clusters. The precipitate mod el in this method uses the maximum distance of constituent ion positions along each of the precipitate axes instead of radius of gyration. Figure 2 9 reveals the results of the 2,226 h aged 308L ferrite G phase cluster size and density evaluation using the cluster analysis method. A mean size of 2.8 nm and number density at 6.9 x 10 23 /m 3 were identified based on the calculation. Fig ure 2 6 Mn Mn nearest neighbor distribut ions for orders from 1 to 10

PAGE 57

57 Fig ure 2 7 Gaussian peak deconvolution for the 4 th order nearest neighbor distribution and dete rmination of d max and d e Figure 2 8 Cluster analysis parameters selection of the N min in MSM.

PAGE 58

58 Figure 2 9 Cluster analysis results of ferrite in 308L weld aged for 2,226 h at 400 C 2 5 Mechanical Property Testing 2.5.1 Nano indentation Test Nano indentation has the advantage of using nano size d indent ation o ver a small area of interest to obtain hardness and Y Chips with thickness around 1 mm were sectioned from the bulk weldment fusion zone using the high speed diamond saw mentioned previously. The feeding speed used was to minimize cutting damage C hips were further ground to 150 m 200 m using th e Tegramin 25 polisher. The polishing paper used w as grinding was finished with 800 grit sandpaper (P2400, 6.5 m ). A 3 mm disc s punch from SPI Supplies w as used to fabricate 3 mm disc from the thin chip. To remove the surface strain induced by the abrasive polishing, the sample surface was further etched using a jet polisher from Southbay Technology M550. The electrolytic etching was

PAGE 59

59 performed with a 3 mm nozzle for 5 s at 20 C with voltage and current at a round 75 V and 100 mA respectively Electrolyte was a mixture of 10% perchloric acid balanced with methanol. A Nikon Eclipse LV100 optical microscop e was used to examine the sample surface and a digital camera from Digi tal Sight SD U3 was used for imaging Fig ure 2 1 0 Optical image of the sample surface of electropolished 308L weld at 500X As shown in Figure 2 1 0 the overall morphology of the ferrite phase is interconnected thin and elongated fish bone shape s surrounded by the austenite m atrix. A Hysitron triboindenter 95 0 was used for the load controlled nano indentation test in INL. Th e indenter used is a diamond Berkovich tip system by Bruker All indentat ions were performed after routine calibration of the tip area position and loading To avoi d interference between the austen ite and ferrite phase s a maximum load of 2 mN was set for the indentation with the average indentation depth at around 400 nm The s ample

PAGE 60

60 surface was scanned to obtain general information about the phase size and morpholog y The p osition of the indent was then set based on the scanned image. The indentation was divided into three stages : loading for 5 s, holding for 2 s, followed by another 5 s for unloading shown in Figure 2 11. A post indentation imag e generat ed by the in denter tip scanning over the sample surface is shown in Figure 2 1 2 The red circles indicate the indent identified within austenite phase and the green circles denot es the location of the intents i n the ferrite phase. An average of 45 indents were perform ed for each sample. Fig ure 2 11 Loading and unloading curve f or the nano indentation. To interpret the results of the indents, a direct comparison of the indent mapping and the scan image was performed to determine the number of each indent by visual recognition from t heir relative locations The raw data for each indent number was

PAGE 61

61 stored in a hys fil e and analyzed with Triboscan software to obtain parameters such as effective indent depth, hardness and reduced ela stic modulus Fig ure 2 1 2 Tip s can ned image of 308L weld with indent ations. For the Bercovich indenter, the hardness of the indent ed region can be calculated through equation Eq. 2 2 (2 2 ) w here P max is the maximum loading and A c is the actual contact area. For this experiment, the indenter tip contact area as a function of indentation depth was calibrated on a piece of quarts with known hardness and elastic modulus.

PAGE 62

62 2.5.2 Miniature Tensile Test The effect of thermal aging on the mechanical response of CASS was evaluated by miniature tensile tests Thermal aging was cond ucted at 400C for 10,000 hrs. This thermal aging condition is known to reduce the impact energies of CASS allo ys significantly, approaching their lower bounds rega rdle ss of their chemical compositions [17] Such micro scale uniaxial tensile test s have not been fully s tandardized [67,68] The samples were miniature f lat tensile specimens with a nominal ga uge section of 1.6 mm x 0.76 mm x 7.6 mm as shown in Figure 2 1 3 Two tests were performed at room temperature for unaged and aged CF 3 T he strain rate was 1x10 3 s 1 The test results were corrected with the elasti c modulus of stainless steels ( 210 GPa ) to compensate for the compliance of loading s train It must be pointed out that, according to the ASTM report, [68] the sample thickness should be around 4 5 times greater than the grain size. Thus, the test results in this study might not be representative of the actual polycrystalline tensile behavior. However, considering the average size of the ferrite phase at 10 the thickness is 5 times greater than ferrite phase size Thus, the miniature testing sample is sufficient to reveal the effect of ferrite hardening on the tensile behavi or of the duplex structure Subsequent fractography stud ies w ere performed using a Tescan LERA3 Xe plasma source FIB/SEM. Fig ure 2 13 Sample dimension for the sub size tensile test of CF 3.

PAGE 63

63 2.5.3 Micro Scratch Test The m icro scratch test was carried out to characterize the fracture toughness property of the material. Three groups of 308L stainless steel weldment samples were prepared with the following treatment conditions : as welded, aged for 2,226 h and aged for ( Table 3 1 ) Bulky specimens with dimensions of 10 mm x 10 mm x 2 mm were sectioned from bulk weldment fusion zone. Steel plates were further polished to make sure that the top and bottom surfaces were parallel. A 0.06 m silica collo idal suspension was used i n the final step to improve the smoothness of the testing surface. With an estimated activation energy of 164 kJ/mole for the weldment, these two aging experiments were equivalent to 17.6 year s service period at 31 5 C in a LWR according to Eq. 1 1. Thus, these two aging experiments were representative o f two distinctive degr ees of aging embrittlement. W ith the increase in aging time, the testing results should yield an obvious descending trend of fracture toughness Figure 2 14 Schematic of the micro scratch test The basic schematics of the testing are shown in Figure 2 14. A new indenter L 119 was used for the modified micro scratch test. It is diamond indenter with a conical spherical shape with a tip radius of 100 m, and half apex angle of 45. Experiments

PAGE 64

64 were conducted at room temperature with progressively increased load ing force using the Anton Parr M icro C ombi T ester which has a valid loading range at 0.1 mN t o 30 N with resolution at 0.1 mN. The maximum indent depth could reach 1 mm with depth measurement resolution at 0.3 nm. The loading force was applied from 0.03 N to 30 N with a loading increment rate of 1 N/s. The length of the scratch was controlled at e xact ly 3 mm with a scratch speed of 6 mm/min. The rate of data acquisition was set at 30 Hz. Three scratches were conducted for each condition to obtain an average penetration depth and sliding force. All samples were sectioned into dimension s slightly gre ater than 15 mm x 15 mm from the fusion zone of the weldment block using a Techcut 5 high speed diamond saw manufactured by Alli ed H igh T ech. A diamond wafer blade with diameter of thickness of was de rotation speed wa s 3000 rpm. The estimated abrasive cutting damage to the steel subsurface wa s very small [69,70] Further polishing were carried on a Struers Tegramin 25 to flatten both surface s while removing the sectioning damage layer. Samples were glued onto a steel plate holder using crystal bond. The rotation speed of the wheel wa s set at 150 rpm. The s ample surface was finished with 1200 grit SiC sandpaper. To eliminate the grinding damage layer, further polishing with a 3 m diamond slurry was used. All sample preparation s were performed at the University of Florida. Samples were then shipped to Anton Parr Inc. for the micro scratch test. To accurately simulate the scratching process on th e steel weld, the cal ibration material should have similar fracture mechanics. Therefore, ANSI 1045 steel was

PAGE 65

65 chosen with a referenced fracture toughness of 50 MPa(m) 1/ 2 for the indenter shape function calib ration in Eq. 2 3: ( 2 3 ) where f is the indenter shape function, d is the penetration depth, p(d) is the indenter perimeter as a function of penetration depth d and A(d) is horizontal projection area as a function of penetration depth d. B oth the spherical and conical part s of the tip w ere gradually in contact with the steel materials due to the intermediate indentation depth. Figure 2 15 Indenter shape function calibration with fitted curve. The ratio of penetration depth to indenter radius, d/R was used for the calculation The eq uation used for the calibration is shown in Eq. 2 4 ( 2 4 ) are fitting coefficients. The 0.0270 and = 4 x10 5 indicate that both the

PAGE 66

66 spherical and conical part s w ere involved in the scratching process shown in Figure 2 14. The coefficient of determination was 0.9969 representing a valid fit. To verify that this fitting is accurate in predicting the fracture toughness, a validation test was performed on paraffin wax. The measured fracture toughness was around 0.169 MPa(m) 1/ 2 with a literature value of 0.15 MPa(m) 1/ 2 extracted from three point bending test. With the error less than 13%, the calibration is quite acceptabl e. The assessment of the aging response of 308L weld was conducted by fractographic analysis using SEM. For the scratch surface analysis, the corresponding SEM images focusing on the rim of the scratches for each condition is shown in Figure 2 13. Overall, the micro scra tch test on fracture toughness wa s recently developed by Akono [71,72] The post scratch surface analysis for potential embrittlement interpretation was not yet fully established. In comparison, fractographi c analysis in other fracture testing method s, such as the tensile test has validated crack surface analysis with the failure mode justification. To symmetrically conduct the analysis without skewed conclusion, conservative features were used with only qualitative comparison s introduced. Th e cracks initiated at the bottom of the scratch groove were ignored for the aged sample due to the low density. The front and ending region of the scratch were not selected either. Cracks developed in these two areas c ould be either too localized or too trivia l to be detected by SEM. In fact for each condition, 3 scratches were made within nearby region distanced by at least 5 mm. For each scratch, exact ly the same testing parameter s, such as loading increase rate and horizontal sliding speed were applied. All three sc ratches were examined by SEM for each co ndition and showed good consistency

PAGE 67

67 among scratches. Here only one scratch was selected randomly with results at 0 tilting presented. Equi pment used for the exami nation wa s a Tescan LERA3 Xe plasma source FIB/SEM.

PAGE 68

68 CHAPTER 3 RESULTS 3.1 Metallography and Chemical Composition 3.1.1 CF 3 The unaged and aged CF 3 met allography i s shown in Figure 3 1. The darker region is the ferrite phase while the lighter area is the austenite phase As explained in paragraph 1.3.1, the scale ferrite phase resides at the core of the el ongated austenite dendrite arms [73,74] The ferrite morphology within the duplex st ructure can be observed as a discontinuous fis h bone shape. Co mbined with th e FEM modeling and experimental results by Schwarm and Wang [26,28] the duplex structure rather than the austenite dendrite structure should be the focus of the structure property relationship for such casting. Figure 3 1. Metallographic im age of CF 3 of a). unaged and b). aged for 10 kh (50X).

PAGE 69

69 Figure 3 1. Continued Figure 3 2. Meta llographic image of unaged CF 3 using SEM.

PAGE 70

70 Table 3 1 lists the APT measured chemical composition of the ferrite tips at four different conditions of CF 3. The standard deviations were calculated for all four conditions The concentrations of Fe, Cr, Ni, Mn and Si are reasonably consistent E ven for P, S and C with very low concentration s the measured values do not var y significantly considering a relative ly large measurement error for the se minor alloying element s. Table 3 1. APT measured chemical composition (wt. %) of the ferrite at different conditions. Condition Ni Si P S Mn C Cr Mo Fe As cast 5.61 1.36 0.028 0.0009 0.57 0.006 24.45 0.618 67.34 Aged 5.71 1.37 0.026 0.0015 0.56 0.011 23.61 0.658 68.05 Irradiated 5.30 1.39 0.042 0.0012 0.54 0.009 24.11 0.657 67.95 Aged &Irradiated 5.87 1.25 0.028 0.0015 0.55 0.013 23.06 0.664 68.19 Standard deviation 0.05 0.01 0.001 0.0003 0.002 0.0006 0.13 0.005 0.08 As cast (EDS) 6.2 1.9 1.33 24.06 0.31 66.19 In comparison, the EDS point scan was performed on the ferrite phase of different grains of the unaged CF 3 specimen shown in Figure 3 2 Due to the limitation of detection resolution, only major element s are included in the last row of Table 3 1. Overall, no significant difference in chemical c omposition is observed for the tested ferrite with respect to Fe, Cr and Ni compared with APT results. This indicates that the content of Fe and Cr in these specimens are relatively similar. A muc h greater content of Si and Mo i s observed in the EDS measurement. This c an be due to the large error in the measurem ent. 3.1.2 308L weld The metallography of the as welded and 2226 h a ged 308L weld ment is shown in Figure 3 3 by optical microscopy. The darker region is the ferrite phase while the lighter

PAGE 71

71 area is the austenite phase. Figure 3 4 shows the corresponding metallography using SEM Samples were obtained from the fusion zone near the welding center line The morphology of equiaxed dendritic structure can be observed for both as welded and aged samples as reported by Silva [75 77] and Davi d [78] The key objective of this study is to examine how the ferrite phase decomposition affect s the mechanical properties of the weldment for instance, the fracture toughness Thus, t he base metal, heat affected zone and columnar dendrite area are not the focus of this study. Manganese silicide inclusion i s observed in both ferrite and austenite phases similar to [79] Figure 3 3. Metallo graphic image of 30 8L weld a). as welded and b). aged for 2226 h (50X).

PAGE 72

72 Figure 3 3. Continued. Figure 3 4. Metallogra phic image of 308L weld a). as weld ed and b). aged using SEM.

PAGE 73

73 Figure 3 4. Continued. 3.2 Crystallography and Microstructure Characterization The environmental degradation s on ferrite in CF 3 steels and 308L weld s were revealed through spinodal decomposition and G phase precipitation. The crystallography of the ferrite and austenite phases is not the focus of this study. However, it is the basis of the nano feature characterization of spinodal decomposition and G phas e precipitation. Thus, only th e TEM diffraction pattern s to with precipita tion are shown in this section. The TEM results are further compared with the APT reconstruction of fe rrite phas e for better structural characterization. 3.2.1 Spinodal Decomposition Figure 3 5 shows the evolution of Cr distribution in the as cast end of life aged and end of life a ged with irradiation ferrite phase in CF 3 cast steel The atom ic maps were formulat ed from a slice of 10 nm i n thickness at the Y middle plane of each APT

PAGE 74

74 tip reconstruction The atom maps illustrate that Cr clustering was induced dur ing thermal aging at 400 C The extent of clustering was further enhanced by subsequent neutron irradiat ion. Figure 3 5 Cr distribution in ferrites upon different treatment conditions of CF 3. Figure 3 6 Fe Cr elemental frequency distribution s in ferrites for different conditions. To quantify spinodal decomposition, the Cr and Fe frequency distributions are plotted using the frequency distribution analysis illustrated in 2.3.1 as shown in Figure

PAGE 75

75 3 6 The atoms in solute clusters were excluded from the anal ysis, and the bin size of ions was 100. For comparison, theoretical normal distributi ons of Cr and Fe were also calculated based on the measured average concentrations in the examined APT specimens. Apparently, the as cast ferrite has nearly perfect nor mal distributions of Cr and Fe with profile peak s located at 2 5.24 and 6 5.88 at. %, respe ctively. This is consistent with the Cr atom map showing that there is no spinodal decomposition occurring in the as cast ferrite. For the aged and irradiated samples both of the Cr frequency profile peaks shift to left, while the profile is also broaden ing into the high concentration range. To quantify the extent of spinodal decomposition, the measured values of C sd as illustrated on Figure 3 6 are 0, 13.37, 13.37 and 14.97 at.% for conditions of as cast, aged, irradiated, and aged plus irradiated, respectively. For the Fe atoms, the concentration frequency profiles show a similar trend by deviating from a nominal distribution upon thermal aging, irradiation or a combination of those two. The low dose irradiation on thermal ly aged ferrite could promo te further spinodal decomposition a nd lead to a higher extent of de mixing of Cr and Fe atoms in the ferrite. To systematically analyze the spinodal decomposition in ferrite under high dose neutron irradiation the frequency distribution s of Fe and Cr were calculated and are shown in Figure 3 7 The bin size was set at 100 ions to best reveal the distribution trend. Peak broadening and shift of Cr and Fe distribution due to neutron irradiation was observed compared with the binomial distribution Such a phen omenon can be interpreted as a result of the demixing behavior of Fe and Cr element from the homogeneous state.

PAGE 76

76 Figure 3 7 Frequency distribution of Fe and Cr in ferrite of irradiated CF 3. RDF was applied to the Cr element distribution to extract the of the modulation The dependence of illustrated in Figure 3 8 The wavelength of as cast CF 3 at 0 dpa i s also added to the curve as a comparison. An overview of the Cr atom spatial distribution for all four irradiation conditions i s also illustrated. A chip of 10 nm was extracted from the cone shape tip for each condition to generate the Cr map The blue color represents the Cr atom s A sc aling morphology of the Cr atom distribution can be observed in all fou r irradiation conditions indicating spinodal decomposition. The darker blue region is the Cr

PAGE 77

77 the dark and white region s grows stronger with increasing irradiation dose. A size increase of the dark domain (Cr rich) and white domain (Cr diluted) was observed with increasing irradiation dose, while the white domain outgrew the dark domain at 40 dpa. Both the Cr map and the waveleng th calculated from RDF indicate an increasing trend of Cr clustering compared with as cast ferrite. A possible wavelength saturation trend was observed near 20 dp a with an estimated wavelength of 20.3 nm. Figure 3 8 Wavelength vs irradiation dose of neutron irradiated CF 3 To validate the capability and accuracy of APT in Cr clustering detection, TEM results of 308L aged at 400 C and irradiated at 0.08 dpa we re selected as shown in Figure 3 9 In comparison, the corresponding APT reconstru ction under same magnification of the Cr map in black dots are illus trated in the lower left corner in Figure

PAGE 78

78 3 9 The thickness of the Cr map envelop e was set to be 5 0 nm which is close to the actual thickness of the TEM lamella e The two bright field TEM images were taken at zone axis [001]. In both cases, the APT reconstruction maps show a good resem blance with the TEM mottle d phase s Both are direct indication s of the spinodal decomposition pattern of the ferrite phase. It can be pointed out that t he TEM contrast is not necessarily an exact match to APT reconstruction. The major reason for it is that the orienta tion of the Cr map in APT reconstruction is not necessarily aligned along the [001] direction No correlation has been successfully established for the interpretation of the APT and TEM results. Figure 3 9. TEM images of spinodal decomposit ion with mottled structure in ferrite of 308L weld, a) 400 C aged for 2226 h; b) irradiated at 0.08 dpa. The small figure in the lower left corner is the APT reconstruction image.

PAGE 79

79 Figure 3 9. Continued. 3. 2.2 G Phase Precipitate Overall, the precipitates are enriched in Ni, Mn an d Si for those ferrites in CF 3 and 308L weld under varied aging and low dose irradiation condition s Such enrichment is a representative elemental composition of a G phase precipitate. The size s were all less than 20 nm, with a spherical shape. Figure 3 10 shows a typical reconstruction of G phase clusters via an iso surface of the solute elements for the aged, irradiated and aged & irradiated ferrite in CF 3 E nhancement by neutron irradiation of G phase clustering in both size and density was revealed by cluster analysis quantification method

PAGE 80

80 Figure 3 10 G phase precipitates in aged, irradiated, and aged irradiated ferrites. The images are sized to have an identical scale, and the isovalue thresholds of Mn, Ni an d Si are 2, 10 and 6 %, respectively. Table 3 2 Average composition of G phase precipitate in CF 3 (at.%) Condition Mo Mn Ni Fe Cr Si Aged 0.67 3.09 13.72 51.30 24.50 6.22 Irradiated 1.55 5.60 26.62 31.76 19.11 11.35 Aged and Irradiated 1.33 7.98 30.10 29.65 15.75 13.42 T he G phase precipitation of high dose irradiated CF 3 steels is evident in Figure 3 11 and Figure 3 12 Neutron ir radiation has a strong impact on the G phase size increase On the other hand, t he G phase number density keeps decr easing. The compositions of major solute element s are listed in Figure 3 8 T he ratio of Ni : Si : Mn i s estimated to be 15:5:2 in all four conditions Pronounced Ni enrichment was found in the G phase clusters with an increase in concentration from 36.67% to 53.51% for irradiation dose s of 5 dpa and 40 dpa respectively. The average Ni comp osit ion in ferrite matrix i s less than 6 % as shown in Table 3 1

PAGE 81

81 Figure 3 11 Cluster size distribution of neutron irradiated CF 3. Figure 3 12 The ionic concentration of Ni, Si and Mn in G phase precipitate in ferrite of irradiated CF 3.

PAGE 82

82 A summary of the G phase clusters of ferrite in CF 3 steels and 308L welds i s presented in Table 3 3 T he G phase precipitation behavior in the ferrite phase of 308L welds is relatively more active than that in the CF 3 steel s ferrite with respect t o size and density for both aging condition and neutron irradiation condition s Table 3 3 Summary of G phase precipitation. Material Size /nm Density (10 12 /cc) As cast CF 3 0 0 10,000 h aged CF 3 2.7 7 0.08 dpa CF 3 4.0 10 A&I CF 3 4.6 11.5 As weld weld 0 0 2,226 h aged weld 2.8 69 0.08 dpa weld 3.1 45 With the help of TEM, the crystal lography of the G phase and its orientation relationship with ferrite was investigated The goal of this study wa s intended to cover all the samples examined in APT. Only the result for 20 dpa irradia ted CF 3 is presented. Data for t he remaining conditions will be obtained in the future As s hown in Figure 3 13 the diffraction pattern of ferrite of 20 dpa irradiated CF 3 exhibits a B.C.C. structure with strong reflection s The lattice parameter measured from Figure 3 13 is 2 .85 which is within the range of literature report s [1] Meanwhile a weaker F.C.C. diffraction reflecti on i s also observed in Figure 3 14 with a lattice parameter measured of 1 1 60 which matches previous results on G phase precipitation [39] T he lattice parameter of G phase is about 4 ti mes the size of the ferrite matrix in this study. Such a relationship corresponds quite well with literature report s [39] In addition, the cube on cube orientation of th e G phase and ferrite matrix was well confirmed in this study. As shown the zone axis direction for both G phase and ferrite matrix was at [001] indicating the direction parallelism. Further more, the lattice plane of (020) for G phase is

PAGE 83

83 aligns with that of ferrite phase (020). Such evidence strongly indicates that the lattice plane s (020) for G phase and ferrite are aligned. Figure 3 13 Diffraction pattern of ferrite matrix and G phase precipitate of CF 3 irradiated at 20 dpa. Due to the weak reflection of G phase spot, it is not easy to capture its diffraction near the direct beam. Thus the diffraction spot of G phase at ( ) G shown in Figure 3 14 was used around the ( ) f spot at g with zone axis at [001]. The dark field image captured the tiny particle and its homogeneous distribution within the ferrite matrix. The estimated size of the G pha se precipitate is about 3.5 nm.

PAGE 84

84 Figure 3 14 TEM image of G phase precipitate in ferrite of CF 3 irradiated at 20 dpa. 3. 3 Mechanical Testing Results 3.3.1 Nano indentation Results on Welds On average, the hardness of austenite for as welded and aged 308L is 3.49 GPa and 3.53 GPa, respectively, which represents an insignificant change in the hardness data shown in Figure 3 1 5 This result matche s well with the reported data [80] Austenite stays quite st able under the influence of thermal aging in an intermediate temperature range. As can be seen in Figure 2 1 2 the edge of the indent is very close

PAGE 85

85 to the ferrite/austenite phase boundary. Thus, it is impossible to rule out the influence of the nearby aust enite on the nano hardness measurement. In th is case, the results were inter pret ed as the nano hardness of the ferrite+austenite phase A significant increase in hardness was observed for the ferrite+austenite phase shown in Figure 3 15 ( 4.49 GPa for as w e lde d and 6.88 GPa for aged ) As can be seen, the ferrite hardening is the major reason for the increase of the deformation resistance of the ferrite+austenite phase Fig ure 3 15. Nano hardness of the austenite and ferrite + austenite phase of 308L weld aged at 0 h and 2226 h.

PAGE 86

86 3.3.2 Tensile Test on CF 3 The tensile test result s of as cast and aged CF 3 are presented in the inset table of Figure 3 1 6 The e ffect of thermal aging increased yield strength ( YS ) by 39 MPa and ultimate tensile strength ( U TS ) by 128 MPa approximately for CF 3 steel. Shown in Figure 3 1 6 a 20% increase in YS i s present in the strain range from 0 to 0.02. This phenomenon m ay be due to the age induced dislocation pinning in both ferrite and austenite [81] Right after the elastic deformation, strain hardening was initiated for both condition s with strain at 0.02 0.0 5. A reduction in ductility with 24% decrease in uniform elongation ( UE ) i s another indicator of the ferrite hardening effe ct. The a ustenite matrix is responsible for the major part of the plastic deformation. The SEM image of the tensile fracture surfac e i s illustrated in Figure 3 17 SEM images were taken at 5 kV with absorb ed current around 9 pA. The working distance was at 30 mm with only the secondary electron detector used for imaging The size of the probe was 15 nm. Ductile dimple fracture was observed in both unaged and aged CF 3 specimens. Figure 3 1 6 Stress strain curves of CF 3 with 24% ferrite tested at room te mperature.

PAGE 87

87 Figure 3 17. SEM fractography of CF 3 with 24% ferrite in a). unaged and b). aged condition

PAGE 88

88 3.3.3 Micro Scratch Test on Welds Fracture toughness results for the 308L welds under various aging conditions measured using the micro scratch test are shown in Table 3 4 The as weld ed condition has the highest value of 22.65 MPa(m) 1/ 2 while the 6,000 h aging condition has the lowest fracture toughness at 17.71 MPa(m) 1/ 2 In comparison the J ic values tested by 0.4T CT test on as w eld ed 308L cladding was estimated at 200 kJ/m 2 at room temperature [82] This value is quite typical for extreme ly tough material in steel welding. If the elastic modulus of 308L is assumed to be 150 GPa ( estimated based on previous nano indentation in this study ) the 308L weld plane strain fracture toughness could be as high as 165 MPa(m) 1 / 2 Obviously, the fracture toughness measurement via micro scratch test in this study is a bout an order of magnitude lower compared with literature report s Table 3 4 Fracture toughness of 308L weld measured from micro scratch test. Condition As weld Ag ed 2226 h Aged 6000 h K IC / MPa(m) 1/2 22.65 20.99 17.71 The micro scratch ed surface was further examined using the previously mentioned SEM data for tensile fractography. Enlarged image s were obtained to identify the surface relief pattern adjacent to the scratch. No crack was observed on either side of the rim for the as welde d condition shown in Figure 3 18 A) and B). In comparison, crack s started to emerge on the upper rim for the 2226 h aged weld shown in Figure 3 18 C) and D ) C rack development was observed all over the upper rim for the 6000 h aged sample shown in Figure 3 18 E) and F). Furthermore, the strain of the scratch propagated towards the nearby substrate region ( marked in the green cir cle ) causing a corrugated pattern in Figure 3 18 B) ri ght above the extruded rim. In comparison, the same region above the crack in Figure 3 18 D) induced only some quasi periodic

PAGE 89

89 surface relief patterns as a sign of local strain relief. No surface relief patterns were found where nearby cracking ha d initiate d for the 6000 h aged 308L weld in Figure 3 18 F). Figure 3 18. SEM images of the scratch of 308L weld for (A) and (B) as welded, (C) and (D) aged for 2226 h, (E) and (F) aged for 6000 h, images on the right are the red box area in images on the lef t.

PAGE 90

90 CHAPTER 4 DISCUSSION 4.1 Ma c rostructure and Metallurgi cal Analysis Due to the low grain boundary angle of the austenite, the current etching can clearly reveal only the austenite/ferrite phase boundary for CF 3 by optical microscopy The f errite phase was easily etched while austenite remained quite unaffected. The estimated grain size is on the scale of a few mm based on visual inspection Electron back scattering diffraction ( EBSD ) is known to recognize grains with different orientation s However, m ost commerci al EBSD equipment is incapable of work ing on grain size s this large. Samples would have to be tilted with the workin g distance as small as possible. Thus this method wa s not considered. 4.1.1 CF 3 The CF 3 in both unaged and aged samples exhib its a duplex structure of ferrite and austenite phases shown in the optical images in Figure 3 1 No obvious macro defects such as bubbles or large inclusions were observed. The average size of the ferrite phase is around a few Based on the chemical composition provided in Table 2 1, the equivalent Cr and Ni content for CF 3 was calculated to be 20.5 wt.% an d 9.4 wt.% following Eq. 4 1 and 4 2 [77,83] ( 4 1 ) ( 4 2 ) According to the pseudo binary phase diagram of Fe Cr Ni with Fe balanced at 70 wt.%, the phase transformation sequence of the CF 3 cast will be L >L + As can be seen, ferrite is the primary phase of the cast. Most of the ferrite would then transform into austenite similar to the diagram shown in Figure 1 4

PAGE 91

91 However, depending on the actual cooling rate of the casting, not all fer rite phase may have sufficient time to be transformed into austenite. In this case, for CF 3, around 24% ferrite is re ta ined in the dendrite core Figure 4 1. 70 wt.% Fe Cr Ni pseudo binary phase diagram. ( Source: Ref. [11] ) No obvious macro segregation was observed in the SEM images shown in Figure 3 3. The actual chemical composition variation for elements such as Fe, Cr and Ni in ferrite phases is quite sm all based on APT analysis. One reason for this can be that all those CF 3 samples were from the exact same interior region of cast slab Thus, it is quite possible for the sample to have very close microstructure and chemical composition. As explained by Louthan [84] the dendritic structure tends to have a certain degree of constitutional variation, especially between the dendritic arms and in terdendritic domains. This can be reflected in the chemical composi tions of austenite and ferrite hav ing distinctive Cr and Ni content Such a difference could also be traced

PAGE 92

92 back to the phase diagram and the solidification process As s hown in Figure 1 2, the binary phase diagram of Fe Cr has a very flat solidus line whe n the Cr content is less than 35%. This means that concentration variation will have a very small impact on the solidus temperature upon solidification. The addition of Ni element would expand the when Ni cont ent exceeds 11 wt. % as calculated by Li [11] and shown in Figu re 4 2. ha d very close chemical composition to that in this stud y for CF 3 shown in Table 2 1, is used to simulate the solidification pro cess for CF 3 in this study. The c ooling rate was estimated to be less than 10 C/s for static casting Non equilibrium solidification and unstable dendrite growth were assumed. As the initial ferrite solidifies from the liquid, the chemical composition of the ferrite would be C owing to solute partitioning based on the phase diagram in Figure 4 1 [85] The correspond ing Ni and Cr concentration s agree quite well with those measured in this study. Due to the enrichment in Cr with less Ni, the in itially formed fe rrite would be the last to transform into austenite phase as the transformation temperature is much lower compared with ferrite with composition of C Thus, th e retained ferrite s residing in the dendrite s tend to have chemical compositi on s similar to that of C In comparison, the austenite phase ten ds to have more Ni and less Cr. Fu [85] investigated the phase transformation by directional solidification on liquid stainless steel 304 He further concluded that depend ing on the cooling rate, the du plex structure could be formed via ite could be retained owing to the incomplete ferrite to austenite transformation. If the cooling rate is sufficient ly slow, ferrite will fully transform into austenite. Then a subsequent reverse

PAGE 93

93 transformation of austenite into ferrite will occur for samples with greater Cr content. Further more, EDS across the ferrite/austenite phase boundary clearly indicated that ferrite has greater Cr and less Ni content compared with the adjacent austenite. Mhuru [86] recently extracted APT results at the interface area of ferrite/austenite phases in unaged CF 3. A sharp transition of Cr and Ni across the phase boundary was observed with greater Cr and less Ni at the ferrite side. It has to be pointed out that the CF ution annealed at 1065 C for 2h followed by water quench. Annealing at such temperature will provide a good homogenization of the chemical composition of each phases with little impact on the phase transformation. Figure 4 1 only reveals the equilibrium c onstitution of the steel all oy above 800 C. Based on previous metallurgical analysis, the a ustenite phase with chemical composition around C or more to the right of Figure 4 1, is in its equilibrium state at around 400 C. Thus, long term thermal exposu re tends to have very limited effect on its phase stability at relatively low temperature. This conclusion has been frequently validated by various TEM images, diffraction patterns and APT results [87,88] Ai [89] performed XRD, EDS, SEM and TEM on thermally aged DSS 7MoPlus steel on both the ferrite and austenite phases. The only phase transformation observed wa s the mottled phase by spinodal decomposition in ferrite. 4.1.2 308L Weld The metallurgical phase solidification of 308L weld is very similar to that of CF 3. The weldment fusion zone has a much faster cooling rate compared with the casting. Because of the cooling rate difference, the varied extent of constitutional supercooling allowed for the trans formation planar >cellular >columnar dendrite >equiaxed dendrite illustrated in Figure 4 2 [76,77,90] Similar to that of CF 3, a duplex structure of ferrite

PAGE 94

94 and austenite was formed but with much smaller dimension [91] In this study, only the equiaxed region is shown in Figure 3 3 and Figure 3 4. Other regions of the as weld 308L are included in the Appendix D Figure 4 2 Variation in solidification mode in weld. ( Source: Ref. [76] ) 4.2 Microstructure Evolution in Ferrite Spinodal Decomposition 4.2 .1 Synergistic Effect of Spinodal Decomposition The long term thermal exposure of CF 3 at 400 C for 10 kh was studi ed to simulate the in reactor service degradation process The atom probe results of ferrite in Figure 3 1 successfully reveal the Cr clustering behavior during thermal exposure. Such a result is in cl ose correspondence with literature reports. The combined effect of thermal aging and neutron irradiation on the ferrite phase degradation is demonstrated on spinodal decomposition shown in Figure 3 1 As reviewed by Wilk es [92] it was experimentally and theoretically shown that irradiation induced disordering competes with the process of thermal ordering. The resulting phase structure is strongly

PAGE 95

95 dependent on the defect production rate and the irradiation temperature, which determine the concentration of defects and their mobility. Either the radiation accelerates the appro ach to thermal equilibrium, or radiation induced non equilibrium phase precipitation occur s. The mechanism of microstructural evolution can be further complicat ed by Mn, Mo, Si and other minor alloying elements One recent study shows that 6.4 MeV Fe 3+ ion irradiation at 300C can suppress spinodal decomposition by decreasing the fluctuation of Cr concentra tion in the Cr [93] However, Miller [36] studied the effect of neutron irradiation on spinodal decomposition of a Fe Cr model al loy, and it was found that the spinodal decomposition in a Fe 32% Cr alloy was significantly enhanced by neutron irradiation after 0.03 dpa at 290 C Th is controversy regarding ir radiation effect on the spinodal decomposition of aged ferrite can be attribu ted to the difference in incident particles, dose rate, temperature and the material itself. The dose rate in this study (10 9 dpa/s) is of the same order study. T herefore, it is not surprising that both studies show an enhancement of spinodal decomposition from neutron irradiation. 4.2 .2 Irradiation Enhancement E stimation After 10,000 hours aging at 400C a Cr Fe concentration profile nearly identical to that of 0.08 dpa neutron irradiation at 315C was indicated by the frequency d istribution in Figure 3 2 The Cr and Fe peaks both become broader after thermal aging or neutron irradiation. The Cr peak broadening to a lower concentration (<25.24 at.%) corresponds to the increased volume of the Cr depleted zone, and broadening to a higher concentration (>25.24%) indicates the increased volume of Cr enriched zones. This study demonstrates that low dose neutron irradiation c ould induce a spinodal decomposition of delta ferrite in a similar fashion to that induced by thermal aging. With

PAGE 96

96 the assumption that the individual aging and irradiation experiment s induced equivalent extent s of Fe Cr demixing, it is possible to estimate the irradiation enhancement rate of the system. By applying the Arrhenius equation in Eq. 1 1, with the activatio n energy of 164 kJ/mol aging at 400 C for 10,000 h is equivalent to aging at 315 C for 692,000 h During neutron irradiation at 0.08 dpa at 315 C with a dose rate of 2.8x10 9 dpa/s the total time consumed was 7,936 h. Thus for an estimated aging at 3 15 C for nearly 79 years of thermal diffusion, with the help of neutron irradiation at the same temperature less than a year is needed to reach the same extent of ferrite decomposition. This comparison suggests that neutron irradiation assisted diffusion kinetics in Fe Cr system is approximately 80 times faster on average. Such a simulated result is quite similar to what Miller [36] discovered in 1996. In fact based o n neutron irradiation physics, point defects introduced by irradiation normally ha ve a greater density compared with the thermally activated vacancy density. In addition the demixing b ehavior of Cr Fe is a diffusion dependent process especially via vacancy. It is obvious that neutron irradiation can have a direc t positive impact on the Cr Fe demixing process. Similar irradiation enhancement estimation s on the Fe Cr system were performed by Li [94] Kr + with 1 meV energy was used to irradiate the CF 8 alloy at 400 C with fluence of 1.88 x10 19 ions/s Subsequent AP T investigation was carried out on the ferrite phase. The steady state model was adopted to calculate the characteristic diffusion distance. A diffusion coefficient in ion irra diated ferrite was calculated to be 3 x10 16 cm 2 /s. With a referenced thermal diffusivity of 3 x10 19 cm 2 /s for the Fe Cr system [95] an enha ncement of 3 order s of magnitude by ion irradiation was concluded.

PAGE 97

97 However, the temperature s for both the calculat ed and referenced diffusivity were not clearly mentioned. The element for the diffusivity was not clearly stated either. In addition, the arti cle later concluded that ion irradiation would suppress the reaction of spinodal decomposition which contradicts the results of enhanced diffusion kinetics of the F e Cr system. Because of these reasons, the enhancement calculation from this article was not considered. 4.2 .3 S pinodal D ecomposition at H igh Dose Irradiation Neu tron irradiation was observ ed to synergistically enhance the ferrite decomposition process [96] The introduction of large quantities of point defects such as interstitials and vacancies could significantly affect the microchemistry of the ferrite phase. The correlation of the wavelength vs irradiation dose in Figure 3 4 show s a steady wavelength growth because of neutron irradiatio n. With the dramatic increase in vacancies, the diffusion kinetics of Cr atoms migrating through the ferrite matrix was significantly enhanced. Based on this plot, the wavelength growth rate gradually decreases. A possible wavelength saturation value at 20.03 nm is illustrated in Figure 3 8 In general, the wavelengths extracted for high dose irradiated CF 3 are much greater than those reported for CF 3 under elevated aging condition s One possible reason could be that the wavelength is dependent on the diffusivity of Cr atoms. Neutron ir radiation induces faster Cr diffusion and yields a greater wavelength. However, the low dose neutron irradiation induced an identical wavelength compared with aging at 400 C at around 6.8 nm This indicates that high dose irradiation modified more than ju st the diffusion kinetics of the Fe Cr system, and modified the microchemistry of the Cr migration as well

PAGE 98

98 Based on the classical spinodal decomposition theory established by Cahn [34] once the wavelength of the decomposition is observed, it should stay at a critical value from then on. However, in this study, an obvious wavelength evolution progression wa s observed. To p roperly explain the wavelength saturation behavior modified to include three stage spinodal decomposition from the homogeneous state toward thermal equilibrium as shown in Figure 4 3 The measured average Cr concentration is around 25% for ferrite and this value was used for the homogeneous state. For those Fe rich and Cr rich domains, the actual total concentration of Fe and Cr was measure d to be greater than 97%. To simplify the model the influence of other alloy element s, such as Ni and Si wa s not considered for equilibrium estimation. Figure 4 3. Schematics of (a) early stage, (b) intermediate stage and (c) late stage of spinodal decomposition.

PAGE 99

99 Figure 4 3. Continued To consolidate the thermal equilibrium chemical composition of each child phase long term aging at low temperature on the Fe Cr system is needed However, no related data h as been published yet. There have been several efforts for modeling and computation of this equilibriu m concentration of Fe Cr by Willia m, Chandra, Xiong and Zhou [29,35,97,98] at.%Cr was se lected. And as the data is much hard er to obtain and validate. It is commonly accepted that the chemical concentration of Cr should be greater than 95 at .%. Thus a value 95 at.% was selected for a conservative calculation. The equilibrium volume fraction of the Cr rich domain at 95% was calculated to roughly 17% based on the leve r law As can been seen in the three different stages for Fe Cr with 25 at. r eached its equ ilibrium concentration much earli er co mpared to while the wavelength increased rapidly during the early stage at its equilibrium status, the contribution to wavelength wa s limited to the shrinka ge of the interface. Both process es eventually slowed wav elength growth and arrived at late stage wavelength saturation.

PAGE 100

100 Krishan and Abromeit [99 101] proposed a model of spinodal decomposition under neutron irradiation of AB alloys. Because the neutron irradiation induced defect supersaturation, an amplification factor on both spinodal decomposition wavelength and amplitude was developed theory The critical wavele ngth was shifted to a larger value. An equation bas ed on the model of wavelength i s expressed in Eq. 4 3. ( 4 3 ) w here r is the recombination radius, n v is the averaged vacancy concentration, F is the parameter based on the defect formation rate. T he detailed interpr etation and clarification of the se parameters is quite complicated and not in the realm of material science. Figure 4 4 Cr proxigram s in ferrite of irradiated CF 3.

PAGE 101

101 To better visualize the shift of short/long range diffusion of spinodal decomposition from early to intermediate stage, the Cr pro xigram was plotted in Figure 4 4 The light blue vertical broken line indicates the medium positio n of the diffusing interface. The demixing of Fe and Cr through uphill diffusion corresponds with this phenomenon. No sign of nucleation and growth was present because a sharp interface was not observed A s mooth concentration transition from the dpa and 10 dpa dose s, where diffusion of Cr and Fe element be quite local In comparison, for 20 dpa and 40 dpa doses a lower plateau was formed with Cr concentration close to 10% On the one hand, th is plateau is actually the sign of could be depleted by e requires the diffusion of Cr via a much longer distance from nearby a Cr rich region such as Combined with the frequenc y distribution of Cr in Figure 3 3 a peak shifting saturation was observed near 10% for ferrite irradi ated at 40 dpa. Such a phenomenon suggests that a lower bound of Cr co ncentration was reached at 40 dpa. The above results and modeling agrees quite well with the data published by Hamaoka and Pareige [44,64] Hamaoka carried out aging experiment s on CF 3M with subsequent APT investigation s o f the ferrite phase. The Cr Fe demixing sequence of aging at 400 C for 100 h to 5000 h was compared via frequency distribution. Both p eak shift and peak broadening were captured similarly. This implies that the effect of neutron irradiation on Fe Cr demixing i s sim ilar to that of thermal aging. Due to the lack of wavel ength information, the trend of wavelength changes were not revealed in

PAGE 102

102 Thus, it was not taken into account. In comparison, Pareige experiment was performed by aging CF8M at 350 C and ATP was later carried out on the ferrite phase. A method of frequency distribution of distances developed by Novy [102] was used to extract the wavelength information of Cr modulation. Overall, there was a logarithmic relationship of increasing wavelength with increased aging time. An equivalent time exponent was estimated to be 0.16. To explain the w avelength increase, it was concluded the Fe Cr system was in its early stage even after 20 years of thermal aging at 350 C. In order to generate a more direct comparison of the wavelength of Cr modulation in the ferrite of th e neutron irradiated CF 3, the author attempted to apply the frequency distribution of distances but n o valid wavelength was successfully extracted One possible reason could be that the randomization of the region of interest was not well formulated. Another reason could be that the wavelength yielded in this study is too large making this extraction method invalid. Novy [103] carried out aging experiment s on Fe 20 at. % Cr model alloys at 500 C A non steady coarsening behavior of the A lower solubility limit at 500 C was found at 14 at.% which makes sense c ompare d with 10 at.% used in this study at 315 C The actual trend of the spinodal curve for Fe Cr phase diagram shown in Figure 1 2 implies that the lower the aging temperature the lower the content of Cr in the Cr at equilibrium. This eviden ce indicates that even under high dose neutron irradiation, the phase diagram of Cr Fe system was not significantly modified.

PAGE 103

103 4.2 .4 Spinodal Decomposition under TEM Currently it is commonly accepted that a mottled phase should be visualized more obviously with the bright field image captured along the [001] zone axis compared to other alignment directions [81,104] The details of the sp inodal decomposition of 308L weld are shown in Figure 3 3 The incident electron beam was aligned along [001] zone axis of the ferrite phase. The modulated contrast resembling the surface of an orange peel was visualized in TEM micrographs in both aged and irradiated 308L weld s Such an appearance corresponds to the finding by Nichol et al [105] However the fundamental mechanism of such a mottle d cont rast is still unknown. Recently, Westraadt [106] explained that this modulated contrast could be due to a small coherency strain which is more evident along the [001] direction. However, he also stated that other factors such as sample thickness fluctuation and ion milling defects could also i nduce a similar modulated contras t. The TEM image is a perfect mapping tool for APT data reconstruction. As shown in Figure 3 9 a) and b) the r econstructed small micrograph corresponds to the clustering of Cr element by APT Only Cr ion s were in clud ed as the dark region B oth the aged and i rradiated APT reconstruction highly mimics that of TEM results. Such a correspondence not only helps validate the APT reconstruction results, but also help s to poss ibly explain that the contrast in the phase chemical composition difference. The wavelength s of the Cr modulation calculated by RDF for aged and irradia ted 308L weld s show very close values at 6.4 nm and 6.8 nm for aged and irradiate d samples, respectively. Such a similarity in wavelength of Cr modulation was also indicated in the TEM micrograph s

PAGE 104

104 The imaging of spinodal decomposition via TE M has been dif ficult Weng [81] performed thermal aging experiment s on DSS 2205 at 475 C. Further TEM imaging with a FEG/TEM Tecnai F30 on the ferrite phase was performe d. The modulated contrast was obvious when the electron beam wa s in alignment with the <001> direct ion of ferrite. The reason for this was not clearly discussed in hi s report. He also mentioned that the mottle d phase. Th is might not necessarily be true in this study As c an be cle arly seen in the Cr only APT r econstruction, no oxidation occurred A similar mottled contrast by the APT reconstruction revealed that it could be due to the Z contras t or the composition s Recently, Westraadt [106] developed a STEM imaging method based on a similar mechanism. TEM experiment s w ere performed on aged Fe 36 wt.%Cr model alloys. The material was solution treated followed with aging at 500 C. A JEOL ARM 200F TEM was used for the imaging with operation voltage at 200 kV. The TEM foil was jet polished simil ar to th e aged specimen s in this research. The nano scale mottle contrast via STEM of the model alloy highly resembled the APT reconstruction in this study. Subseq uent compositional analysis proved that the mottled signal intensity corresponded to the STEM image which is known to be sensitive to the Z contrast. The modulated contrast caused by chemical composition difference clearly indicated the modulated morphology of spinodal decomposition in the ferrite phase. It is estimated that the TEM sa mple used in his study could be as thin as 10 nm. Otherwise, the 3D superimposition of the modulated structure would definitely dissipate the 2D projection. Up until now, there is no other TEM investigation on spinodal decomposition

PAGE 105

105 of ferrite under neutro n irradiation. Thus, little comparison coul d be made as to the irradiation induced spinodal decomposition. 4.3 Microstructure Evolution in Ferrite G Phase 4.3 .1 G Phase Formation under Irradiation and Aging in CF 3/308L weld As s hown in Figure 3 4 a direc t comparison between the aged CF 3 at 10,000 h with the ag ed and irradiation sample indicated that the combined effect could dramatically enhance the G phase in both size, density and volume fraction. T he spinodal deco mposition assisted flux could play a partial role in this result More importantly, the irradiation in duced defects provided a high energy status for G phase to nucleate inside the ferrite matrix where both interf acial and strain energy could counteract With the irradiation enhanced diffusi on on those solute elements such as Ni, Si and Mn, the G phase growth was also favored. Such an increase in both size and density directly l ed to the increase of its volume fraction. The 400 C thermal aging induce d more solute cluster nuclei while the lo w dose i rradiation promoted coarsening. T he synergistic effect of neutron irradiation and thermal aging on G phase precipitation has rarely been studied and reported. However, Li [94] performed similar experiment s with heavy ion irradiation on aged CF 8 with subsequent APT cluster analysis. The thermally aged CF 8 at 400 C for 10,000 h yielded G phase precipitates with size s of 4.8 nm and density of 9.7 x 10 22 /m 3 Such results are quite comparable to those for CF 3 under same aging condition s However, t he low dose heavy ion irradiation suppressed the size of the G phase, with insignificant impact on the number density. These contrary results could be due partially to the time spent to reach the accumulated irradiation dose level. Neutron irradi ation perf ormed in this study occurs at a much smaller dose rate than that observed for heavy ion irradiation Th e extended

PAGE 106

106 irradiation in this study would allow the irradiation induced defects to interact with the nearby atoms to enhance the diffusion kinetics. The densities of th e G phas e observed in the 308L weld summarized in Table 3 3 are about an order of magnitude greater than that of CF 3 ferrite. This is probably because a large number of favorable sites at ere not involved with G phas e nucleation in CF 3 ferrite. It was also found that irradiation weldment promotes a greater number of clusters as compared with irradiated CF 3 probably because the ferrite in 308L weld typically contains more Mn and less Si than that of CF 3. Concurrent evolution between solute clustering and spinodal decomposition was proposed by Mateo and studied by Danoix [39,107] It was claimed that the G phase phase interfaces and no particles were observed within the domains. Our study shows that the solute clusters are al l located in the Cr deple ted zone, as shown in Figure 4 5 As spinodal decomposition progresses, domains enriched in iron and domains enriched in chromium form over the parent solid solution. It seems that Ni and Mn are rejected from the Cr enriched domain interfaces and unsymmetrically grow into Fe rejected from either domain hase. However, due to its relatively low concentration in the solute clusters and the ferrite matrix, it is difficult to determine the domain from which Si was rejected Overall, a strong interaction between the spinodal decomposition and solute cluster nu cleation and growth was well confirmed in the irradiated and aged 308L weldment ferrite

PAGE 107

107 Figure 4 5 Iso surfaces of Mn (Gold) Ni (Green) Si (Gray) clusters and interfaces of Cr phases: Aged (33.58%Cr 7.47%Mn 13.39%Ni 5.6%Si) and Irradiated (33.57%Cr 6.72%Mn 14.31%Ni 5.74%Si) Takeuchi [43,54] performed two separate neutron irradiation experiment s on stainless steel electroslag weld overlay claddings with irradiation fluence at 7.2 x 10 19 n cm 2 .and 7.2 x 10 23 n cm 2 The corresponding irradiation temperature s were 280 C and 290 C respectively. His comparison of aging and neutron irradiation on ferrite decomposition in weld indicated that irradiation tend s to increase the activity of Si while

PAGE 108

108 aging yield s greater Ni clustering. Both are solute element s of the G phase cluster. Such a phenomenon was not identified in this study because the fact that G phase clustering in aged and irradiated weld s behave d quite similarly. In addition, the frequency dist ribution result s be over interpreted. An obvious issue is the binomial distribution which was not in agreement with the average element content. Thus, discovery was not considered in this study. 4.3 .2 G P hase Coarsening under High Dose Irradiation in CF 3 As s hown in Figure 3 5, the number density of the G pha se clusters decreased with high irradiation doses This could be largely due to the fact that the interface area shrinking, limiting the nucleation action of those Ni and Si clusters. The ch emical composition ratio of th e high dose irradiated ferrite has not been reported previously in C F 3. The average size of th e G phase clusters identified after 20 dpa was much greater than that observed in CF 3 under accelerated aging condition s using an atom probe by other researchers [107] Such evidence indicates that the G ph ase coarsening in ferrite was significantly enhanced under high dose neutron irradiation due Mn, no saturation or thermodynamic equilibrium trend was observed based on A PT data This phenomenon is quite different from a conve ntional precipitation scenario where an equilibrium chemical composition was observed. Similar high dose irradiation results on CASS have not been reported Intermediate dose ion irradiation was perfo rmed on aged CF 8 by Li [94] at 400 C with corresponding irradiation dose up to 3.82 dpa In the irradiation dose rang e of roughly 1 dpa 3.82 dpa, Li observed that the G phase precipitate exhibited a number density

PAGE 109

109 increase and mean size decrease. Such a trend is contrary to the result found in this study. Li did not explain the mechanism for such a phenomenon. The author believes that the impact of high energy with the precipitation ordering b ehavior due to irradiation damage. This could help explain the G phase mean size decrease. The number density increase is likely due to the competition of thermal nucleation and irradiation damage where thermal diffusion was more important than the irradi ation disordering. Thus, were not considered in this study. [44] long term (0 year 20 years) therma l aging experiment on CF8M can help compare the effect of thermal aging and high dose neutron irradiation on G phase precipitation A steady evolution of G phase in mean size was observed in CF8M aged at 350 C. Such a characteristic trend of precipitation behavior correspond s to the coarsening behavior of neutron irradiation observed in this study. As for the number den sity, a trend of initially increase followed by decrease was observed by Pareige with peak at 30,000 h aged condition. In comparison, the number density of G phase found in this study kept decreasing. However, the irradiation dose range did not strictly fo llow the low intermediate high dose trend similar to that of the thermal aging experiment in 40 dpa could be interpreted as corresponding to aging time greater than 30,000 h. such a correlation indicates that high dose neutron irradiation could easily accelerate the thermal aging phase decomposition process in respect of G phase precipitation. The number density of G phase found in CF 8M tends to be an order magnitude higher than that of CF 3. Part of the reason cou ld be due to the higher Mo content in CF 8M that could facilitate G phase clustering behavior as a

PAGE 110

110 solute element. The mean size increase rate in CF 8M wa s not as great as that of CF 3 under high dose irradiation ( 0.8 nm to 1.8 nm with corresponding aging time from 2000 h to 20,000 h ) [64] report. This could be explained by the enhanced n eutron irradiation diffusion of solute atoms which facilitates the coarsening behavior. The average solute concentration of Ni, Si an increasing trend with the increase of thermal aging time, which is similar to the trend observed in this study. In G phase stay relatively the same. Both Pareige and Hamaoka used a filter as a concentration threshold to categorize the G phase clusters. The f ilter concentration for ort was much smaller than that in Hamaoka 4.3 .3 TEM investigation and APT validation To compare the ATP reco nstruction results for the G phase precipitation with the TEM micrographs, a specific APT reconstruction of the 20 dpa irradiated CF 3 ferrite was performed. In Figure 4 6 the object in white denotes the G phase precipitate created from the Ni iso surface with ionic concentrat ion at 20 at.%. The background blue spheres indicate the Fe and Cr atoms as representative of the ferrite matrix. Overall, the number density of G phase reconstructed in IVAS corresponds to that of TEM results. However, the size of the G phase constructed from iso surface is much larger than that measured from TEM. This can be explained by the G phase in ferrite, in which only the core area contributes to the diffraction with the F.C.C. structure. Meanwhile, that enrichment in t he exterior region of G phase i s still coherent with the ferrite matrix. Thus, its existence i s not revealed un der TEM diffraction contrast. This means that the i so surface of Ni used to bind the G phase is relies highly on the

PAGE 111

111 clustering behavior of Ni. Thus it is impossible to tell where the phase boundary is based on its concentration in IVAS reconstruction In this case, the size s of Ni, Si and Mn clusters are much large r than the G phase precipitate size Figure 4 6 G phase precipitate in ferrite of CF 3 ir radiated at 20 dpa reconstructed in IVAS.

PAGE 112

112 To elucidate this interpretation discrepancy in IVAS reconstruction a more detailed 1 D concentration profile was represented as a cylinder with diameter of 2 nm passing through a quasi spherical Ni, Si and Mn clu ster center region shown in Figure 4 7 To best re veal the element concentration profile the length along the z axis of the cylinder was set at 20 nm. In addition, the step size for the profile was set at 0.6 nm. The calculated concentration profile along the z axis i s summarized in Figure 4 8 The line of is o surface concentration of Ni i s presented to mark the corresponding cluster size measurement and a reading of 7.1 nm can be obtained. Such a number is slightly smaller than the size identified by the MSM technique shown in Figure 4 8 I n MSM calculati on, the size of the clusters are based on the phase boundary which in Figure 5 12 for example, occurs at Ni w ith concentration around 5 at.%. In comparison, the estimated size and p osition of the actual G phase i s marked along a relative plateau of Si element. Inside this area, the Ni element does not seem to reach a steady level possibly because the Ni composition is more sensitive to the calculation step size at 0.6 nm. To summarize, the interface regio n detection is the major challenge causing the difference in TEM and APT measurement s of G phase size In this study cluster identif ication via the MSM technique i s clarified to be effectively extracting the cluster size and composition, rather than the G phase. Meanwhile the TEM technique has successfully identified the crystallography of the G phase and its orientation relationship with the ferr ite matrix. Both methods show good agreement for the number density of th e nano size d particles.

PAGE 113

113 Figure 4 7 Relative position of the G phase cluster from Ni iso surface and the cylinder. Figure 4 8 G phase precipitate in ferrite of CF 3 irradiated at 20 dpa reconstructed in IVAS. 4. 4 Effect of Thermal Aging and Neutron Irradiation Validation Because of the limited irradiation sample volume and access, no metallography of the irradiated CF 3 or 308L weld is available to reflect the impact of irradiation on

PAGE 114

114 macro structure Thus, it is impractical to describe th e macro scale impact of irradiation damage on the CASS/weld In the meantime, based on the metallography shown in Chapter 3 long term thermal exposure has shown negligible impact on the macrostructure of CF 3 and 308L weld, such as ferrite content and morphology. No dramatic phase transform ation of ferr ite to austenite was observed. As no met allography is available for th e neutron irradiated samples, it is hard to tell if any macro sc ale defects were induced. Future work is needed to clarify the existence of i rradiation induced segregation void and bubble Irradiation is known to significantly affect the microphysics and microchemi stry of the materials owing to irradiation induced defects The impact o n density related properties, such as free energy and diffusivity, in the long term, will be reflect ed in the microstructure evolution. Shown in section 4.2.2, neutron irradiation can dramatically enhance spinodal diffusion and increase the rate of the ongoing spinodal decomposition toward thermal equilibrium. Similar evidence has been reported by Miller [36] as neutron irradiation enhance d the reaction rate compared with that of thermal aging. Such an irradiation enhanced diffusion c an also be explained by Gladyshev and [99] theory. Kirihara performed experimental analysis on neutron irradiation enhanced self diffusion in gold. His theory indicated that n eutro n irradiation induced defects are not only associated with the free energy, but also with the changes in the entropy and enthalpy of the system. In the present study, it is quite obvious that the enthalpy of vacancy formation was dramatically decrease d as a result neutron irradiation.

PAGE 115

115 4. 5 M echanical Testing Analysis 4. 5 .1 Nano indentation The dislocation mobility in ferrite phase under low strain rate s can be inferred from the nano indentation results. Both spinodal decomposit ion and G phase precipitate can interact with the dislocation movement in the ferrite phase. Spinodal decomposition results in n anoscale Fe Because of the atomic size difference of Fe and Cr, a local strain fiel d can be induced owing to the lattice misfit Such fine scale coherent domains interconnect with each other within the ferrite phase and are known to increase the lattice friction stress for the dislocation movement [30] G phase precipitate can bring about precipitation hardening which plays a minor role in the ferrite phase strengthening Takeuchi [54] performed similar nano indentation test s on as received and a ged weld. An ENT 1100a nano indentation tester was used for the experiment with the maximum loading force set at 2 mN. The hardness measured by Takeuchi for as received and aged ferrite was 4.7 GPa and 5.2 GPa, respectively. The hardness of as received fer rite is in good agreement with the result of this study shown in Figure 3 12 at around 4.5 GPa. However, the aged ferrite in this study ha d a much greater hardness measurement. aged at 300 C for 2000 h. The 308L weld tested in this study was aged at 400 C for 2226 h. Thus it makes sense that the higher aging temperature enable d greater extent of phase decomposition in ferrite. This in turn can lead to greater hardening effect. Fujii [93] also performed n ano indentation experiment s on as received a nd thermally aged CF8M at 400 C for 10,000 h and 40,000 h. Both austenite and ferrite phase s were examined using an ENT 1100a The indentation with a Berk ovich diamond

PAGE 116

116 tip. T he maximum ind entation depth was kept at 300 nm The austenite hardness measured in CF 8M for as receiv ed and two aged condition s stay ed at about 3.0 GPa. Considering the testing error, these results are in good agreement with the finding in this study of 3.5 GPa. The ferrite hardness of as r eceived CF 8M was at around 4.8 GPa The hardness of aged 10,000 h and 40,000 h CF 8M ferrite were around 9.5 GPa and 10.5 GPa respectively. The hardness measurement of ferrite/austeni te in 2226 h aged 308L weld is reflective of a typical ferrite phase aged under th ese condition s Recently, Schwarm performed nano indentation on CF 3 in as received condition [27] The cast block was solution treated at 1065 C before testing The indentation was performed using a Hysitron 900 Triboindente r with a diamond Berkovich tip with the maximum loading at 4 mN. The measured hard ness for austenite and fer rite we re quite close at around 5 GPa. As his report did not include the post indentation SEM image of the tested sample, it is difficult to rationalize why the ferrite phase and the austenite phase have such close nano hardness values The present author suspects that, all of austenite phase as the Hysitron 90 0 tester has a low accuracy in adjusting the indentation position. Under such a ssumption the indentati on result not considered in this study. 4. 5 .2 Tensile Test Recently, Mburu and Schwarm [26,27,86,87] carried out similar miniature tensile test s on aged CF 3 steel to investigate the ferrite hardening phenomenon under thermal aging. The results indicate d that aging ha d litt le effect on the YS of CF 3 steel but increase d the UTS by close to 100 MPa Obviously, the impact of aging on UTS of CF 3 in their study agrees quite well with that in this study After performing APT analysis on both austenite and ferrite phase s Mburu a nd Schwarm found that austenite phase

PAGE 117

117 stayed quite hom ogeneous after long term aging, and they concluded that ferrite hardening is the primary reason for the increase in the UTS of the CF 3 miniature test In order to fully understand the tensile test beha vior of CF 3 steels schema tics of the composite for the dual phase of ferrite/austenite are illustrat ed in Figure 4 9 To simplify the tensile behavior of this duplex structure, a fiber reinforced composite schema tics were use d. The 24% ferrite in CF 3 can be treated as reinfo rcement for austenite with disc ntinuous (chopped) distribution in random short mode. Th is indicates that ferrite can reinforce the austenite matrix without severely affecting the matrix deformation. Thu s, the deformation process can be decomposed into the individual behavior of ferrite an d austenite. T he phase boundary was assumed to be negligible To achieve this, an exaggerated stress strain curve i s illustrated in Figure 4 9 to allow for a direct decoupling of the dual phase struc ture based on the composite model As can be seen, austenite is more strain sensitive than ferrite. Under tensile testing at l ow loading rate, austenite can deform more freely and favorably than ferrite s with the responsive strain distributed homogeneously within the austenite phases F or the as cast CF 3 steel, austenite shoulder s most of the deformation. In the meantime, ferrite can help increase the tensile strength with less plastic deformation. This assertion perfectly explains the dimple microvoid coa lescence failure of the fracture surface for as cast DSS [11,108] Crack formation and growth were restr ain ed within the austenite phase The ferrite phase is well isolated from plastic deformation indicated b y the blue dash ed line in Figure 4 9

PAGE 118

118 Figure 4 9 Decoupled stress strain curve of ferrite, austenite and CF 3 in exaggerated illustration. (Phase b oundary not considered) It is well known that aging at 400 C c an directly harden the ferrite phase with little impact on the austenite phase If the phase boundary still plays the same role in the strain transfer between the austenite and ferrite, the increase s in YS and UTS can be readily explain ed For the increased initial stage strain hardening due to thermal aging o ne possible explanation is that the phase boundary becomes l ess ductile after aging. This leads to more strain transferred from austen ite to ferrite. Such increased strain in ferrite causes it to deform plasti cally and the strain hardens resulting in rapid initial stage strain hardening. Wang [28] recently performed similar miniature tensile test s on a DSS with ferrite content of 15 wt.% under unaged and aged conditions. The effect s of thermal aging and strength enhancement trend me asured in his study were quite comparable to those in this study. His result indicate d that plastic deformation occurs initially in the austenite phase, regardless of the treatment condition. This conclusion is in good agreement with the illustration shown in Figure 3 1 2 where austenite exhibits plastic deformation earlier

PAGE 119

119 than ferrite. Fur ther deformation and fracture were correlated with the ferrite/a ustenite phase boundary in report. This include d the stress transfer and strain redistribution between the two phases Schwarm [26] performed FEM modeling on the elastic deformation stage of the austenit e and ferrite phases in as received CF 3 steel under applied stress. He further concluded that austenite is more prone to elastic deformation than ferrite. This makes sense because the ferrite phase tends to be stiffer than austenite. H is modeling also indicate d that the strain was unevenly distributed across the bulk material during elastic defor mation, especially at the austenite/ferrite phase boundaries. He further performed an EBSD examination on the surface of the tensile testing sample prior to fracture (strain = 45% ). The result suggest ed that the fer rite and austenite phases share similar e xtent of plastic deformation at low strain rate for tensile testing indicat ing that the phase boundary exhibits an excellent capability in coordinating the stress strain distribution between th e two phases in CF 3 during the late stage of the tensile test The fracture surface micrographs of the unaged and aged CF 3 in Figure 3 17 both show a predominant micro voi d coalescence dimple fracture, and a v ery s mall portion of river pattern can be observed in the aged tensile fracture surface. Fo rmation of the r iver pattern i s probably due to the involvement of the hardene d ferrite phase. Chen [55] performed compact tension test s on the unaged and aged CF 3 steels Similar dimple fracture was observed for both the unaged and aged CF 3 compact tension test fracture surface s with no clear evidence of embrittlement. 4.5 .3 Micro Scratch Test Compared with th e J IC values from literature measured via J integral and compact tension test on 308 weld by Mills and Alexander [109,110] the fracture

PAGE 120

120 toughness results measured in this study are several times smaller. Initially, this was considered to be because the fracture failur e mode is a combination of mode I crack opening a nd mode II crack sliding [71] As for the scratch test utilized in this study the fail ure mode resulted primarily from Type II sliding fracture toughness. However, a fter examin ing the morphology of all scratches, it was obser ved that cracks were not created in the sample surface during the tests As s hown in Figure 3 18 the bottom region of the scratch lan e has no sign of any cracks with only smooth surface This indicates that the measured tangential force is no longer a ref lection of the crack formation and stable propagation. It is rather a combin ation of the tip/surface friction force and the forward indentation resistance [71,72] a valid fracture evolution is critical to extract the fracture toughness. Only in this case, the tangential force is a reflection of the actual crack propagation. The present author further collaborated with An ton Parr Inc. to perform scratch test s on CF 3 aged for an extremely long time. Much greater indentation pressure s were appli ed. However, no obvious cracking at the scratch bottom was initiated. This indicates that under horizontal scratch loading the pre ferential response of 308L welds is plastic deformation rather than cracking. In this case, this test method seems to be ineffective for the high fracture toughness metal in this study. 4. 6 Structure Property Relationship Validation o n CASS/weld 4.6.1 Spin odal Decomposition Strengthening The results of nano indentation, tensile test and micro scratch test all indicate that long term aging exposure would harden the ferrite phase and decrease the overal l ductility of the material. T he austenite phase is relatively stable under low temperature thermal aging impact based on the phase diagram Thus, it ha s limited impact on the

PAGE 121

121 mechanical property degradation of CASS/weld. Such c orrelation has been conclusive ly validate d. The relationship between ferrite emb rittlement and its phase decomposition has been reported but quite inconclusive especially the underneath strengthening mechanism Weng [81] reporte d to have discovered twinning in the ferrite phase of DSS near the fracture surface due to t hermal aging under external impact loading Twinning in B.C.C. typically occurs via {112} planes and occur s at low temperature and high strain rate [111] No twinning was observed in the present study by TEM or stress strain curve However, Weng report agrees quite well with the observation of Cortie and [30,112] on hardness indentation impressions of Fe Cr binary alloy s, where the ferri te grain size wa s quite large at 50 100 Th e observed twin ning deformations all occurred under high strain. In comparison, the loading rate in this study wa s quite small Even though the underneath mechanism is still unknown, i t i s suspected that aging induced spinodal decomposition has a greater impact on lattice friction energy as a result of chemical strengt hening. Such strengthening has much high er impact on resolved shear stress for dislocation slip than twinning. Thus, for fully aged ferrite, a shift in plastic deformation mechanism from slip to twinning can be observed. Meyers [113] performed constitutive analysis over the onset of twinning in B.C.C. metal s, especially in Fe and found that the grain size play ed a major role in the twinning initiation in Fe grains Such a description corresponds perfectly with fi ndings in thi s study. The average phase sizes of ferrite in CF 3 steels and 308L weld s we re less than 10 and 1 respectively. It is quite impr actical for dislocation pile up induced

PAGE 122

122 stress concentration to accumulate sufficiently to exceed the critical twinning stress. Thus, it makes sense that no twinning wa s observed in this study. [114] group performed in situ TEM investigation on aged DSS ferrite. He further concluded that the dislocation mobility was hindered because of thermal aging related spinodal decompos ition. As Cahn and Kato [115,116] reported the amplitude and wavelength of Cr composition modulation can induce periodic variation of th e elastic modulus in Fe Cr binary alloy s Kato further calculated the amount of lattice misfit hardening and modulus hardening. The calculated result s agreed quite well with experimental measurement. Little is known concerning the scale of the spinodal cohe rency strain field even though it has been widely mentioned. Some researcher s [81] propose isotro pic distribution between the Cr rich zone and Cr depleted matrix. However, Nichol [105] concluded that the modulated structure yielded better contrast along the <001> direction. One major reason is the small atomic size difference betw een Fe and Cr [7] Thus, Weng [81] argued that the elastic coherence strain at the indiscernible interfaces i s negligible However the measurement of the spinodal strain by Li [11] show ed clear existence of an interconnected strain field in his g eometric phase analysis on aged ferrite Even though a strain field at be small, such strain field distribution i s not uniform within the ferrite phase. It is suspected that the accumulated effect on the strain field can be quite large, exceeding 30% in both tensile and compressive strain The distance of such high strain region s is about 20 nm. Although further validation of such strain field measurement is need help t o explain how the accumulated strain field play s an important role in the

PAGE 123

123 strengthen ing of the ferrite phase. The interaction between the dislocation strain field and the spinodal strain field can make it much more difficult for dislocation to mov e ,thus en hancing the hardness and strength of the ferrite phase. The reason for such non phase due to its low solubility [117,118] while Si is known to partition Such behavior can be a potential reason for the strain distribution non uniformity. 4.6.2 Particle Strengthening by G Phase G phase is also known to play a minor role in ferrite strengthening. Mburu [86] performed mechanical property testing on aged CF 3 with G phase present and on C F 8 with no G phase precipitation observ ed The changes in mechanical property were very similar. However, a number of other factors can affect both the microstructure and mechanical propert ies including chemical composition and the scale of spinodal decom position in ferrite. Thus, it is actually quite difficult to tell if G phase played a role or not as it normally emerges synergistically with spinodal decomposition in ferrite of duplex structured stainless steel alloys. Moreover, i t is quite difficult to decouple the impact of both phases. Yamada [119] performed aging experiment s at 400 C to introduce spinodal decomposition and G phase in ferrite. Aged sample s were then reheat ed at 550 C to eliminate the spinodal decomposition but not the G phase. The change in hardness show ed that G phase wa s responsible for the second stage ferrite hardening behavior result helps to c onfirm that G phase does contribute to the ferrite phase hardeni ng. T heoreti cally, it is possible to determine the impact of G phase based on its volume fraction average radius and strengthening mechanism For the aged CF 3 measured via APT G phase reconstruction, the volume fraction of G

PAGE 124

124 phase is less than 1%. In com parison, the spinod al decomposition morphology occurs throughout the ferrite phase. In this manner, it is safe to conclude that G phase played an insignificant role in the ferrite strengthening. 4.6.3 Other Strengthening Mechanism Th is author recently revi ewed the YS and hardness of Fe Cr binary tested at room temperature by Maykuth and Jaffee [120] shown in Figure 4 10. The plot indicates that a very small port ion of Fe in the Cr matrix can have a significant impact on its strength with peak Fe content at 25 at.%. If this s trengthening phenomeno n w ere true when the Cr rich zone size was a few nanometers, the spinodal decomposition could be explained accordingly. F or CF 3 with 25 at.% of Cr in the ferrite phase, as ferrite decompose s spinodally the summation of the nominal hardness of the Cr rich zone and matrix with less Cr will always have greater strength va lue than the homogeneous state, because the strength vs co mposition curve near 25 at.% face s downward. Unfortunately, the reason for this phenomenon of Fe strengthening on the C r matrix is still unknown. It i s suspected that a greater lattice friction force is needed for dis location to move through the Cr rich zone regardless of its size. To further quantify the possible strengthening magnitude, the following assumptions and discussions were made. Overall, previous efforts trying to correlate hardness or YS with wavelength or amplitude seem to be quite inconclusive and lacking a detailed strengthening mechanism. Thus, it is quite reasonable to assume that, after long term thermal exposure, fe rrite decomposes into the Cr depleted matrix and the Cr rich zone dispersions distanced by wavelength Such a simplification makes it possible to apply particle strengthening mechanisms to the system.

PAGE 125

125 Figure 4 10. Strength and hardness of unaged Fe Cr alloys. (Source: Ref. [112] ) To correlate the result of tensile test s on CF 3 steels and their microstructure evolution properly only the resolved shear stress quantification for pure slip should be considered and compar e it with the 0.2% YS of the ferrite phase based on the tensile test For the unaged condition, the value from Figur e 4 10 with Cr content of 25% are employed with value of 35,000 psi = 241 MPa S trengthening mechanism s such as solid solution, Hall Perch e ffect and G phase clustering were not considered for simplification. Under the composite model of CF 3, with ferrite content of 24%, the YS for austenite is 228 MP a. Considering that austenite hardness stays relatively constant, its YS can be set as constant too The corresponding strengthening in ferrite can be calculated accordingly to be 403 MPa = 58,000 psi based on the v alue measured in section 3.3.2. Before proceeding to calculate the resolved shear stress for pure slip mode in aged ferrite of CF 3, it is help ful to consider the Orowan bending theory. Although it is

PAGE 126

126 highly impossible for such occurrence to be true, it helps to clarify if the Cr rich zone is deformable or not The stress needed for dislocation to loop around t he 3 nm Cr rich zone to critical radius can be expressed in Eq. 4 4 ( 4 4 ) where is a structure coefficient at 0.5 for B.C.C. G is the shear modulus of the material at 77 GPa and b 0.248 nm. With the value of wavelength and zone size d in this study the estimated resolved shear stress = 4 GPa. Such a value is impractically larger than the observed value of 403 MPa. Thus, there is no doubt that dislocation can be better estimated to cut th rough the Cr rich zone. [112] evaluation of aged Fe Cr binary alloys the mechanism of friction stress augmentation due to the generation of the Cr rich zone was applied. He proposed that the higher Cr rich zone yields greater lattice friction force maximized at 75 at.% Cr thus inhibiting the dislocation slip movement The resolved shear stress for slip based on this mechanism accounted for around 5 0 % of the observed YS i n his report. Similarly, owing to thermal agi ng, the ferrite phase was simplified to decompos e into the 12 at.% Cr matrix a nd the Cr rich zone dispersion s of 45 at.% Cr based on Figure 3 6 The volume fraction of the Cr rich zone wa s estimated from the lev er law to be 39 .4 % The total friction al force F T acting on a unit length of a perfectly rigid dislocation is given by Eq. 4 5. ( 4 5 ) w here Cr and M refer to the corresponding shear stresses acting on the dislocation of the Cr rich zone and matrix, respectively and f Cr denotes the volume fraction of the Cr

PAGE 127

127 rich zone. Assuming that b Cr = b M = b, the total shearing s tress of slip can be expressed by Eq 4 6. ( 4 6 ) Estimation based on Figure 4 10 give s M at 20,000 psi and Cr at 62,000 psi The calculated total shearing stress is T = 36,500 psi = 252 MPa, which is approximately 63% of the measured value. Compared with the original unaged ferrite YS value, this is only a slight increase. Other strengthening mechanisms such as modulus strengthening, chemical strengthening and William [121] mo dified Mott and Nabarro theory on coherency strain hardening can a ll contribute to the fer rite hardening. Under the Cr rich zone dispersion assumption, the estimated modulus strengthenin g and chemical strengthening have quite limited impact on strengthening magnitude Williams [121] considered the interaction between the hydrostatic stress field of edge dislocation and pr ecipitate. The estimated critical shear stress was found to be the same order magnitude as T H owever, theory is more applicable to precipitate s with very small volume fraction. In the meantime, Kato [118] extended the study of lattice misfit hardening and modulus strengthening based on the modulated structure of Cr concentration fluctuation rather than Cr rich zone. y due to the spinodal decomposition in B.C.C. Fe Cr alloys i s ex pressed as Eq. 4 7. ( 4 7 ) (4 8)

PAGE 128

128 w here the former term denotes the magnitude of misf it hardening and the latter term indicates the modulus strengthening extent s, A is an amplitude factor, dimensionless variation of the lattice constant a with respect to the concentration of Cr and Y is expressed by the elastic moduli. For i sotropic elasticity, Y is given by Eq. 4 8 where E is the elastic modulus ( 0.3 for steel ), the shear modul us difference of the Fe rich and Cr rich zone s, b is th of 0.248 nm, a the value obtained from RDF i s used. Substituting all values leads to an incremental y of B.C.C. alloys [114] As can be seen, the at least 120% overestimated The main reason for it that, modulus st in theory which is way too linear thus causing unexpected overestimation when More future work is needed in this area to possibly combine the theories of Marcinkowski and Kato to obta in a more accurate strengthening mechanism. 4.6.4 Compact Tension Test and Impact Test of CF 3 Because of the limitation of the micro scratch test on the 308L weldment, no successful extract ion of the fracture toughness was achieved in this study However, the T compact tension test was performed on the CF 3 by Chen at AN L using specimens from the same cast slab as those used in this study. Th e details of these test s can be found in the referenced conference paper [88] Compared with unaged CF 3, aging at 400 C for 10,000 h dramatically reduced the J Q value from 320 kJ/m 2 to 170 kJ/m 2 Subsequent fractography investigation was also performed [55] As the microstructure of austenite was reported to show little change under thermal aging, the fracture

PAGE 129

129 toughness decrease due to thermal aging c an be purely attributed to the spinodal decomposition of ferrite phase. The synergi sti c effect of thermal aging and neutron irradiation on t he fracture toughness decrease wa s also confirmed. The extra amount of G phas e precipitation in ferrite can be one reason for th e decrease. Another reason can be the irradiation hardened austenite reported by Takeu chi [43,54] The amount of decrease in fracture toughness due to the low dose neutron irradiation is much larger compared to that measured on stainless steel weld by Tobita [82] As no detailed micro structure graphs were presented in his report, t he reason for this discrepency is still unknown. The full size Charpy V notch impact energy changes of C F 3 due to thermal aging was studie d by Chopra [17] Specimens tested were the same as the CF 3 steel s used in this study with heat number 69. The i mpact test used a much greater strain rate compared with that in the compact tension and tensile test. The effect of ferrite embrittlement on the fractography of CF 3 aged at 400 C for 10,000 h clearly reveal ed the trend of brittle cleavage failure throughout the fracture surface. C rack initiation in ferrite was assumed to be caused by either dislocation pile ups or twinning [81] report. Meanwhile, the dislocation pile up owing to spinodal decomposition constraints can be reinforced b y Li and Hsieh [11,114] findings This author tends to agree with the dislocation pile up mechanism.

PAGE 130

130 CHAPTER 5 CONCLUSIONS In this study, CASS/welds were investigated under varied agi ng and irradiation treatment Systematic metallurgical, metallogr aphic, crystallographic and microstructural characterization and analysis were c arried out by optical microscopy SEM, TEM and APT examination T he chemical composition was examin ed by APT. The corresponding mechanical properties were evaluate d by tensile test, nano hardness test and micro scratch test. The macrostructural transformation texture of dendritic austenit e and duplexed with ferrite sub grains was successfully identified. It was assumed that the austenite/ferrite duplex played the major role in de t ermining the mechanical properties of the steel rather than the dendrite grains. A structure propert y relationship on CASS/welds was explored. The ferrite phase microstructure evolution under aging and irradiation was characterized using APT. Part of the APT reconstruction results were further validated using TEM c haracterization with respect of spinodal decomposition and G phase precipitation. The tensile strength of CF 3 and hardness of weldment under varied aging condition s were successfully cha racterized by sub sized tensile testing and nano indentation Based on th e characterization and subsequent analysis, the following contributions have been achieved: A synergistic effect of neutron irradiation with thermal aging on the ferrite in CF 3 has been confirmed. The enhancement i n both spinodal decomposition and G phase precipitation was quantified for the firs t time. The neutron irradiation induced G phase precipitation is more active than spinodal decomposition.

PAGE 131

131 Response of ferrite to high d ose neutron irradiation was investigated for the first time. Irradiation dose range s, including threshold and the saturation dose levels for spinodal decomposition were established. Wavelength growth of spinodal decomposition vs neutron irradiation in f er rite of CF 3 was observed. The difference from the conventional wavelength migration under thermal aging was explained by three stage s evolution of Cr clustering Th e results obtained from this research can further help push forward the CASS/weld degradati on un der LWR working environment. Th e high dose neutron irradiation results helped provide valuable data point s for the establishment of full range neutron irradiation response of ferrite decomposition. Th e micro scratch test itself set a good example of fracture toughness extraction and interpretation. The volume fraction variation of G phase under neutron ir radiation was proposed and can act as one of the key parameter s for aging degradation assessment. The new methods and theories dev eloped this thesis include : First, the methodology of proxigram for Cr concentration profile extraction was developed and interpreted by the author The method was capable of determining the wavelength and amplitude of spinodal decomposi tion, in good agree ment with frequency distribution and RDF results. Second improvement o n the RDF calculat ion was successfully achieved by extend ing its range from 10 nm to 25 nm with additional Matlab code. This modification perfectly solve d the limitation of the IVAS built in algorithm. Third the extent of spinodal decomposition in ferrite of CASS/weld was quantified based on frequency distribution. By judging from the peak shift saturation, the equilibrium concentration of the Cr phase was accurately predi cted. Fourth, specific milling step s were develop ed f or the atom probe tip preparation on th e irradiated cast and weld samples Th e irradiated sample s tend to be very fragile The designed milling steps can help mak e high quality APT tips with large data s et extraction For features such as G phase clustering, a large r data set can allow for more accurate size and density quantification.

PAGE 132

132 Fifth a simple method f or estimating irradiation enhanced diffusion was achieved based on the aged and irradiated CF 3 i ndividually This estimation is within the acceptance of the enhancement rate. Th e result c an act as a benchmark for future estimation and validation Sixth, nano scale features such as Cr spinodal decomposition and G phase clusters identified from APT reconstruction were successfully correlated with the results from TEM. The good agreement of APT and TEM data indicates that data extracted from APT are quite accurate in spinodal decomposition wavelength and G phase precipitation number density prediction Takeuchi [54,122] performed APT examination on the 308L weldment under neutron irradiation condition s and observed Si rich cluster s in austenite phase. Similar characterizations are needed as future work on the irradiated austenite phase in both CF 3 and 308L to clarify this phenomenon

PAGE 133

133 APPENDIX A ATOM PROBE BASICS AND TIP FABRICATION There are two data acquisition modes in LEAP, the voltage mode and laser mode. Conductive materials such as stainless steels can be analyzed well in the voltage mode. The alignment of needle specimen to the electrode is achieved visually after sample loading. Laser mode is applicable to any materials, such as semiconductors and ceramics. The alignment of specimen apex to the la ser takes more steps including pre alignment, manual alignment and fine alignment of laser position and focus. For CASS/weld tip, both modes work fine. To maximize the data collection on those irradiated specimen s that are prone to fracture, laser mode i s preferred with detection rate kept between 0.1% 0.5 %. Sometimes the voltage mode i s used as well. The samples are placed in a high vacuum chamber with a cooling system in order to meet the strict working environment in LEAP. The pressure is usually kept extremely low at 10 11 torr. The temperature is dependent on the analyzed materials I n the case of stainless steel alloys, the temperature is set between 20 K and 80 K to achieve similar evaporation rates for both Fe and Cr atoms. In the present work, the temperature was se t to be 55 K. Data collection w ill stop once the sample tip fractures or the voltage reached the limit of 9 kV. Most of the time the tip fractures owing to the stress on the tip caused by the local electric field. An y small defects in th e tip can cause a fracture event. In case of a good run when tip did not fracture, the applied voltage will keep ramping up When the voltage becomes too high the local electrode starts to evaporate contaminates. lead to cata strophic tip fractures, which could damage the local electrode. Thus data collection automatically stop ped when voltage reached a set value

PAGE 134

134 A good quality tip is key to the success of the atom probe ex periment. Typically large data is not required for sp inodal decomposition analysis. On average, 10 millions of ions is sufficient to reveal the features such as the frequency distribution and RDF computation. However, for cluster analysis, the accuracy in size and density calculation decreases significantly when sample data size is too small. Thus for those atom probe experiments on CF 3 and 308 weld, large data set with approxima tely 20 million ions is highly nee ded. Tips that run good in LEAP with more than 20 million ions collection rely on several different factors including the tip geometry, material, treatment condition and the LEAP running environment. For stainless steel, the average data set obtained is around 10 million. Depend ing on the treatment condition as cast homogeneous CF 3 t ips can reach 20 million before fracture. In comparison, the neutron irradiated CF 3 tips tend to fracture more easily under high voltage due to the irradiation induced defects In general, a clean tip pillar with sharp tip top and smooth tapering transiti on, followed with near straight body is the ideal tip geometry. The detailed steps of the tip annular milling are summarized in Table A 1. The low kV step at 5 kV is the critical step. The key this step is to watch the tip shrinking and sharpening. The maj or purpose is to fully sputtering out the Pt layer without causing too much tip thinning. A sequence of FIB SEM images demonstrating the basic procedures are illustrated in Figure A 1 Starting from Figure A 1 a), ferrite with length greate r than 15 is ideal for Pt depo sition. For a maximum of 6 APT tips a cantilever with 15 is typical ly needed. After two steps of trench milling and a cantilever lift out using the FIB micromanipulator t he triangle coupon of ferrite wa s further transferred onto the S i tip holder. Shown in

PAGE 135

135 Figure A 1 b) and c), the part of the ferrite coupon was welded onto the Si tip holder imaged using I beam (Figure A 1 b) ) and E beam (Figure A 1 c) ) mode with stage tilt at 0 degree The dark box indicate s the ion milling region to separate the tip with the ferrite coupon. Figure A 1 d) illustrates the annular mi lling using high energy Ga+ ion to obtain the ideal cylinder shape of the tip finished with outer diameter of 350 nm. Figure A 1 e) shows a good tip finished with outer diame ter at 150 nm by the 5 kV Ga+ ion To reduce the Ga+ ion sputtering damage of the tip surface, a cleaning step at 2 kV is needed as the final step. Table A 1. LEAP Tip Annular Milling Steps Current (30 kV) OD(um) ID(um) Z 3 nA 6.0 4.00 0.3 1 nA 4.5 2.00 0.3 0.3 nA 4.0 1.5 0.2 0.3 nA 3.5 1.0 0.2 0.1 nA 3.0 0.75 0.2 ( measure ) 0.1 nA 2.5 0.50 0.2 (measure) 30 pA 2.0 0.30 0.2 (measure) 30 pA 1.5 0.25 0.2 (measure) 5 kV 48 pA 0 Watch the tips change shape 2 kV 27 pA 0 1 min would work the best Figure A 1 LEAP Tip fabrication steps in FIB SEM procedure.

PAGE 136

136 APPENDIX B MICRO SCRATCH FRACTURE TOUGHNESS The stress intensity factor defines the status of fracture when a crack is initiated. With the help of stress concentration f actor the stress intensity factor i s defined as follows by Irwin Griffith equation : ( B 1 ) w here Y is the geometry factor, denotes the applied stress. a is the crack length. K increases with the increase of crack length grows. When the crack gets great enough after initiation with the increase of external load, its rapid propagation becomes inevitable. This value of stress intensity factor where crack starts to be u nstable and propagate rapidly i s typically termed as fracture toughne ss, Kc Figure B 1. Relationship of fracture toughness and sample thickness.

PAGE 137

137 Kc is dependent on the material geometry. Shown in Figure B 1, with the increase of the sample thickness, the Kc, or the critical stress intensity factor will initially increas e linearly with increase of sample thickness then decrease. Further increase of the thickness will lead to the Kc reaching a lower satur ation level. This value of Kc i s determined as the property of the material, regardless of the testing sample dimension, geometry, testing method, as mode I plane strain fracture toughness K Ic Recently, researchers from Massach usetts Institute of Technology developed a new testing method of measur ing the fracture toughness via micro scratch tes t. During the test, a scratch i s produced in the sample using a typical indenter tip with vertical load and a constant speed. The tangential force, penetration depth, and vertical load are measured. With the vertical load increases, both tangential force and penetration depth increase gradually. (B 2) For linear elastic fracture mechanics model, a relationship between the tangential fo rce and the penetration depth i s analyzed shown in Eq. B 2 where both the perimeter, p and contact area A are a function of penetration depth d. To further quantify the relationship between the tangential force F T and penetration depth d, the value of F T vs d 3 /2 is plotte d [72] It turned out that a linear relationship between the two can be revealed when the penetration distance i s greater than half of its max. After further mathematical formulation, the slope of the F T vs d 3 / 2 is correlated w ith the fracture to ughness of the material Both shared the same unit of Pa(m) 1/ 2 Furthe r research conducted by the same group proposed the perimeter p, and the indenter projection area alongside the sliding direction A, to describe the impact of the ind enter tip

PAGE 138

138 geometry The product of the two parameter s with a factor of 2, denoted as 2pA is the probe shape function f. Further derivation proved that the probe shape function is proportional to the value of d 3 By plotting the tangential force, F T vs the square root of the probe shape function f a linear relationship can be observed. In addition, the slope of the plot was further proved to be the fracture toughness K c of the material.

PAGE 139

139 APPENDIX C THE PHYSICS OF NEUTRON IRRADIATION DAMAGE The fundamental physics of a single neutron irradiation with the cascading defects genera tion i s illustrated in Figure C 1. The incident neutron collides with the lattice atom, as the primary knock on atoms ( PKAs ) followed with kineti c energy transfer. Those PKAs then col lide with more nucleus that are in the passage of the PKA This process generates more point defects. The whole process happened within extremely short amount of time in about 10 11 s [52] As the energy of the incident neutron varies, the density of the induced point defects also changes. Some of irradiation induced Frenkel pairs will recombine through defects reaction due to thermal healing. The process of radiation damage event c an be summarized as the formation point defects and defect clusters. The i rradiation damage rate i s typically estimated from SRIM calculation based on the K P model. For steel material, pure iron with density of 7.8 g/cm 3 i s used for the SR I M calculation alongs ide with the neutron irradiation energy profile. Theoretically, the neutron irradiation damage i s given by the displacement rate shown in Eq. C 1 [52] ( C 1 ) where t he displacement rate R has the unit of number of displacements per unit volume per unit time N is the number density of the material, denotes the neutron flux and is the displacement cross section. By utilizing the K P model and several other neutron irradiation physics assumptions, R c an be estimated through Eq. C 2. [52] ( C 2 )

PAGE 140

140 w here is the scattering cross section, is the avera ge neutron irradiation energy and denotes the total neutron flux above Figure C 1. Schema tic of neutron irradiation damage in solid. ( Source: Ref. [123] ) T he schema tic shown in Figure C 1 is a B.C.C. ferrite crystal with {111} orientation The high energy incident neutron strikes the Fe atom via an i nelastic scattering process transferring a portion of its energy to the Fe atom or primary knockon atom (PKA) The As the PKA moves through the latt ice Fe atoms in its path are displaced to nearby interstitial sites. Such an effect leaves a long vacancy pathway surrounded b y the displaced interstitials, known as a cascading effect. Researcher [52] has estimated the amount of Fe atoms being displaced in a single PKA cascading event For a neutron energy of 0.5 MeV, approximately 350 Fe atoms were displaced per neutron. For a neutron irradiation

PAGE 141

141 damage rate of 10 6 dpa/s, its equiva lent effect is that each Fe atom will be displaced from its lattice position once every 12 days (10 6 seconds).

PAGE 142

142 APPENDIX D SUPPLEMENT RESULTS Figure D 1 Maximum size, mean size and volumetric number density of G phase precipitate s of CF 3 The data of G phase identified in CF 3 in APT reconstruction has already been summarized in the table 3 3. The mean size of the G phase clusters with error bar is shown in Figure D 1. Figure D 2 A s weld 308L OM image of the base metal and fusion zone

PAGE 143

143 As revealed in Figure 4 2, the columnar dendrite, heat affected zone and base metal are illustrated in Figure D 2. Some of the welding defects are labelled as those black irregular void area s

PAGE 144

144 LIST OF REFERENCES [1] P. Marshall, Austenitic stainless steels: microstructure and mechanical properties, Springer Science & Business Media, 1984. [2] J.L. Dossett, H.E. Boyer, Practical heat tre ating, Asm International, 2006. [3] W.D. Callister, D.G. Rethwisch, Materials scien ce and engineering, John Wiley & Sons NY, 2011. [4] B.L. Bramfitt, Structure Property relationships in irons and steels., Mater. Park OH ASM Int. 1998. (1998) 153 173. [5] H. Okamoto, L. Kacprzak, P.R. Subramanian, Binary alloy phase diagrams, ASM internat ional, 1996. [6] M. Maalekian, The effects of alloying elements on steels (I), Christ. Doppler Lab. Early Stages Precip. 23 (2007) 221 230. [7] F. Danoix, P. Auger, D. Blavette, Hardening of aged duplex stainless steels by spinodal decomposition, Microsc. Microanal. 10 (2004) 349 354. [8] I.V. Vernyhora, V.A. Tatarenko, S.M. Bokoch, Thermodynamics of fcc Ni Fe Alloys in a Static Applied Magnetic Fiel d, ISRN Thermodyn. 2012 (2012). [9] P.J. Grobner, W.C. Hagel, The effect of molybdenum on high temperature pr operties of 9 pct Cr steels, Metall. Trans. A. 11 (1980) 633 642. [10] F.C. Campbell, Structural composite mate rials, ASM international, 2010. [11] S. Li, Y. Wang, X. Wang, Effects of ferrite content on the mechanical properties of thermal aged duplex stai nless steels, Mater. Sci. Eng. A. 625 (2015) 186 193. [12] J.M. Vitek, S.A. David, Solidification and aging behavior of Types 308 and 308CRE stainless steel welds, Oak Ridge National Lab., TN, 1984. [13] S. Bonnet, J. Bourgoin, J. Champredonde, D. Guttmann M. Guttmann, Relationship between evolution of mechanical properties of various cast duplex stainless steels and metallurgical and aging parameters: outline of current EDF programmes, Mater. Sci. Technol. 6 (1990) 221 229. [14] O. Kamiya, K. Kumagai, T. Futami, Effect of ferrite morphology on fracture mode of stainless steel weld metal, Weld. Int. 8 (1994) 798 803. [15] L. Karlsson, H. akan Arcini, E. L. Bergquist, J. Weidow, J. Brjesson, Effects of alloying concepts on ferrite morphology and toughness o f lean duplex stainless steel weld metals, Weld. World. 54 (2010) R350 R359.

PAGE 145

145 [16] O.K. Chopra, H.M. Chung, Aging of cast duplex stainless steels in LWR systems, Nucl. Eng. Des. 89 (1985) 305 318. [17] O.K. Chopra, A. Sather, Initial assessment of the mecha nisms and significance of low temperature embrittlement of cast stainless steels in LWR systems, Nuclear Regulatory Commission, Washington, DC (USA). Div. of Engineering; Argonne National Lab., IL (USA), 1990. [18] O.K. Chopra, W.J. Shack, others, Mechanic al Properties of thermally aged cast stainless steels from shippingport reac tor components, Citeseer, 1995. [19] H.M. Chung, T.R. Leax, Embrittlement of laboratory and reactor aged CF3, CF8, and CF8M duplex stainless steels, Mater. Sci. Technol. 6 (1990) 2 49 262. [20] S.J. Zinkle, J.T. Busby, Structural materials for fission & fusion energy, Mater. Today. 12 (2009) 12 19. [21] S.J. Zinkle, G.S. Was, Materials challenges in nuclear energy, Acta Mater. 61 (2013) 735 758. [22] D. Testa, A. Kunkle, The Westinghouse pressurized water reactor nuclear power plant, Westinghouse Electr. Corp. Water React. Div. (1984). [23] O.K. Chopra, G. Ayrault, Aging degradation of cast stainless steel: status and program, Nucl. Eng. Des. 86 (1985) 69 77. [24] M.K. Miller, J. Bentley, S.S. Brenner, J.A. Spitznagel, Long term thermal aging of type CF 8 stainless steel, J. Phys. Colloq. 45 (1984) C9 385. [25] P.H. Pumphrey, K.N. Akhurst, Aging kinetics of CF3 cast stainless steel in temperature range 300 400 C, Mater. Sci. Te chnol. 6 (1990) 211 220. [26] S.C. Schwarm, S. Mburu, R.P. Kolli, D.E. Perea, J. Liu, S. Ankem, Mechanical and Microstructural Effects of Thermal Aging on Cast Duplex Stainless Steels by Experiment and Finite Element Method, in: Mech. Creep Behav. Adv. Mat er ., Springer, 2017: pp. 253 262. [27] S.C. Schwarm, R.P. Kolli, E. Aydogan, S. Mburu, S. Ankem, Characterization of phase properties and deformation in ferritic austenitic duplex stainless steels by nanoindentation and finite element method, Mater. Sci. E ng. A. 680 (2017) 359 367. [28] Y. Wang, Y.H. Yao, Z.P. Wang, Y.H. Jin, X.L. Zhang, J.N. Liu, Thermal ageing on the deformation and fracture mechanisms of a duplex stainless steel by quasi in situ tensile test under OM and SEM, Mater. Sci. Eng. A. 666 (201 6) 184 190. [29] D. Chandra, L.H. Schwartz, M ssbauer effect study of the 475 C decomposition of Fe Cr, Metall. Trans. 2 (1971) 511 519.

PAGE 146

146 [30] M.B. Cortie, H. Pollak, Embrittlement and aging at 475 C in an experimental ferritic stainless steel containing 38 wt.% chromium, Mater. Sci. Eng. A. 199 (1995) 153 163. [31] B. Trindade, R. Vilar, Influence of nickel on 475 C embrittlement of Fe Cr Ni alloys: Mssbauer effect study, Hyperfine Interact. 66 (1991) 351 358. [32] H. Kuwano, Y. Ishikawa, T. Yoshimura, Y. Hamaguchi, Characterization of the spinodal decomposition of Fe Cr alloys by Mssbauer spectroscopy, Hyperfine Interact. 69 (1992) 501 504. [33] H. Yamamoto, A study on the nature of aging of Fe Cr alloys by means of the Mssbauer effect, Jpn. J. Appl. Phys. 3 (1964) 745. [34] J.W. Cahn, On spinodal decomposition, Acta Metall. 9 (1961) 795 801. [35] J. Zhou, An Atom Probe Tomography Study of Phase Separation in Fe Cr Based Steels, KTH Royal Institute of Technology, 2014. [36] M.K. Miller, R.E. Stoller, K.F. Russell, Effect of neutron irradiation on the spinodal decomposition of Fe 32% Cr model alloy, J. Nucl. Mater. 230 (1996) 219 225. [37] R.A. Van Santen, The Ostwald step rule, J. Phys. Chem. 88 (1984) 5768 5769. [38] C.M. Wayman, Solid state phase tra nsformations, Annu. Rev. Mater. Sci. 1 (1971) 185 218. [39] A. Mateo, L. Llanes, M. Anglada, A. Redjaimia, G. Metauer, Characterization of the intermetallic G phase in an AISI 329 duplex stainless steel, J. Mater. Sci. 32 (1997) 4533 4540. [40] H.J. Beatti e, F.L. VerSnyder, A new complex phase in a high temperature alloy, Nature. 178 (1956) 208 209. [41] J.M. Vitek, S.A. David, D.J. Alexander, J.R. Keiser, R.K. Nanstad, Low temperature aging behavior of type 308 stainless steel weld metal, Acta Metall. Mate r. 39 (1991) 503 516. [42] Y. Matsukawa, T. Takeuchi, Y. Kakubo, T. Suzudo, H. Watanabe, H. Abe, T. Toyama, Y. Nagai, The two step nucleation of G phase in ferrite, Acta Mater. 116 (2016) 104 113. [43] T. Takeuchi, Y. Kakubo, Y. Matsukawa, Y. Nozawa, T. To yama, Y. Nagai, Y. Nishiyama, J. Katsuyama, Y. Yamaguchi, K. Onizawa, Effects of neutron irradiation on microstructures and hardness of stainless steel weld overlay cladding of nuclear reactor pressure vessels, J. Nucl. Mater. 449 (2014) 273 276.

PAGE 147

147 [44] C. P areige, S. Novy, S. Saillet, P. Pareige, Study of phase transformation and mechanical properties evolution of duplex stainless steels after long term thermal ageing (> 20years), J. Nucl. Mater. 411 (2011) 90 96. [45] P.J. Grobner, The 885 F (475 C) embritt lement of ferritic stainless steels, Metall. Trans. 4 (1973) 251 260. [46] A. Trautwein, W. Gysel, Influence of long time aging of CF8 and CF8M cast steel at temperatures between 300 and 500 C on impact toughness and structural properties, in: Stainl. Stee l C ast., ASTM International, 1982. [47] A.A. Tavassoli, A. Bisson, P. Soulat, Ferrite decomposition in austenitic stainless steel weld metals, Met. Sci. 18 (1984) 345 350. [48] O.K. Chopra, Degradation of LWR core internal materials due to neutron irradiat ion, US Nuclear Regulatory Commission, Office of Nuclear Regulatory Research, 2010. [49] F.M. Haggag, W.R. Corwin, R.K. Nanstad, Effects of irradiation on the fracture properties of stainless steel weld overlay cladding, Nucl. Eng. Des. 124 (1990) 129 141. [50] F. Haggag, W. Corwin, D. Alexander, R. Nanstad, Tensile and Charpy Impact Behavior of an Irradiated Three Wire Series Arc Stainless Steel Cladding, in: Eff. Radiat. Mater. 14th Int. Symp. Vol II, ASTM International, 1990. [51] J.S. Lee, I.S. Kim, R. Kasada, A. Kimura, Microstructural characteristics and embrittlement phenomena in neutron irradiated 309L stainless steel RPV clad, J. Nucl. Mater. 326 (2004) 38 46. [52] G.S. Was, Fundamentals of radiation materials science: met als and alloys, Springer, 2016. [53] S.J. Rothman, Effects of irradiation on diffusion in metals and alloys, in: Phase Transform. Irradiat., 1983. [54] T. Takeuchi, J. Kameda, Y. Nagai, T. Toyama, Y. Matsukawa, Y. Nishiyama, K. Onizawa, Microstructural changes of a thermally aged stainless steel submerged arc weld overlay cladding of nuclear reactor pressure vessels, J. Nucl. Mater. 425 (2012) 60 64. [55] Y. Chen, B. Alexandreanu, K. Natesan, Crack Growth Rate and Fracture Toughness Tests on Irradiated Cast Stainless Steels, Argonn e N ational Laboratory (ANL), 2013. [56] J.P. Massoud, P. Dubuisson, P. Scott, V.K. Chamardine, CIR II program: description of the Boris 6 and 7 experiments in the BOR 60 fast breeder reactor, EPRI Rep. 1011787 (2005).

PAGE 148

148 [57] G.W. Fynn, W.J.A. Powell, The cutting and polishing of e lectro optic materials, (1979). [58] D.C. Zipperian, Metallographic specimen pr eparation basics, Pace Technol. [59] D.C. Zipperian, Metallographic handbook, Chief Tech. Off. PACE Technol. Tucson AR. (2011). [60] B. Katharine, Smal l, Guide to Etching Speciality Alloys, Adv. Mater. Process. (2008). [61] G.F. Vander Voort, E.P. Manilova, Imaging phases in steels, Adv. Mater. Process. 163 (2005) 32 37. [62] M.K. Miller, E.A. Kenik, Atom probe tomography: A technique for nanoscale chara cterization, Microsc. Microanal. 10 (2004) 336 341. [63] M.K. Miller, R.G. Forbes, Tutorial review: atom probe tomography, Mater Charact. 60 (2009) 461 469. [64] T. Hamaoka, A. Nomoto, K. Nishida, K. Dohi, N. Soneda, Effects of aging temperature on G phase precipitation and ferrite phase decomposition in duplex stainless steel, Philos. Mag. 92 (2012) 4354 4375. [65] W. Guo, D.A. Garfinkel, J.D. Tucker, D. Haley, G.A. Young, J.D. Poplawsky, An atom probe perspective on phase separation and precipitation in d uplex stainless steels [66] Y. Chen, P.H. Chou, E.A. Marquis, Quantitative atom probe tomography characterization of microstructures in a proton irradiated 304 stainless steel, J. Nucl. Mater. 451 (2014) 130 136. [67] K. Kumar, A. Pooleery, K. Madhusoodan an, R.N. Singh, J.K. Chakravartty, B.K. Dutta, R.K. Sinha, Use of miniature tensile specimen for measurement of mechanical properties, Procedia Eng. 86 (2014) 899 909. [68] W.R. Corwin, S.T. Rosinski, E. Walle, others, Small specimen t est techniques, in: A STM, 1998. [69] Y. Chen, C. Bunget, L. Mears, T.R. Kurfess, Investigations in subsurface damage strengthened nickel based superalloy, Proc. Inst. Mech. Eng. Part B J. Eng. Manuf. 230 (2016) 1221 1233. [70] A. Johnson, D. Fisher, G. Vander Voort, Precision Cutting The Science of the Cut, Tech Notes Using Microstruct. Anal Solve Pract. Probl. 3 (2003). [71] A. T. Akono, F. J. Ulm, An improved technique for characterizing the fracture toughness via scratch test experiments, Wear. 313 (2014) 117 124.

PAGE 149

149 [72] A.T. Akono, P.M. Reis, F.J. Ulm, Scratching as a fracture process: From butter to steel, Phys. Rev. Lett. 106 (2011) 204302. [73] C.O. Ruud, A.A. Diaz, M.T. Anderson, Grain Structure Identification and Casting Parameters of Austenitic Stainle ss Steel (CASS) Piping, PNNL 19002 Richland Wash. Pac. Northwest Natl. Lab. (2009). [74] P. Ramuhalli, R.M. Meyer, A.D. Cinson, T.L. Moran, M. Prowant, B. Watson, M.T. Anderson, In situ Characterization of Cast Stainless Steel Microstructures, in: Proc. Ei ghth Int. Conf. NDE Relat. Struct. Integr. Nucl. Press. Co mponentsBerlin Ger. 2012, 2012. [75] C.C. Silva, J.T. de Assis, S. Philippov, J.P. Farias, Residual Stress, Microstructure and Hardness of Thin Walled Low Carbon Steel Pipes Welded Manually, Mater. Res. 19 (2016) 1215 1225. [76] S. Kou, Welding metallurgy, John Wiley & Son s, 2003. [77] K. Li, D. Li, D. Liu, G. Pei, L. Sun, Microstructure evolution and mechanical properties of multiple layer laser cladding coating of 308L stainless steel, Appl. Surf. Sci. 340 (2015) 143 150. [78] S.A. David, Ferrite morphology and variations in ferrite content in austenitic stainless steel welds, Oak Ridge National Lab., TN, 1981. [79] K.B. Alexander, M.K. Miller, D.J. Alexander, R.K. Nanstad, Microscopical evaluation of low temperature aging of type 308 stainless steel weldments, Mater. Sci. Technol. 6 (1990) 314 320. [80] L.J. Ayers, The hardening of Type 316L stainless steel welds with thermal aging, Massachusetts Institute of Technology, 2012. [81] K.L. Weng, H.R. C hen, J.R. Yang, The low temperature aging embrittlement in a 2205 duplex stainless steel, Mater. Sci. Eng. A. 379 (2004) 119 132. [82] T. Tobita, M. Udagawa, Y. Chimi, Y. Nishiyama, K. Onizawa, Effect of neutron irradiation on the mechanical properties of weld overlay cladding for reactor pressure vessel, J. Nucl. Mater. 452 (2014) 61 68. [83] R.L. Plaut, C. Herrera, D.M. Escriba, P.R. Rios, A.F. Padilha, A short review on wrought austenitic stainless steels at high temperatures: processing, microstructure, properties and performance, Mater. Res. 10 (2007) 453 460. [84] M.R. Louthan Jr, Optical metallography, ASM Handb. 10 (1986) 299 308. [85] J.W. Fu, Y.S. Yang, J.J. Guo, Formation of a blocky ferrite in Fe Cr Ni alloy during directional solidification, J. Cryst. Growth. 311 (2009) 3661 3666.

PAGE 150

150 [86] S. Mburu, R.P. Kolli, D.E. Perea, S.C. Schwarm, A. Eaton, J. Liu, S. Patel, J. Bartrand, S. Ankem, Effect of aging temperature on phase decomposition and mechanical properties in cast duplex stainless steels, Mater Sci. Eng. A. 690 (2017) 365 377. [87] S. Mburu, R.P. Kolli, D.E. Perea, S.C. Schwarm, A. Eaton, J. Liu, S. Patel, J. Bartrand, S. Ankem, Atom Probe Tomography Analysis of the Local Chemical Environment at the Austenite/Ferrite Inter faces of Cast Duplex Stainless Steels, in: Proc. 14th LACCEI Int. Multi Conf. Eng. Educ. [88] Y. Chen, W.Y. Chen, A.S. Rao, Z. Li, Y. Yang, B. Alexandreanu, K. Natesan, Fracture Resistance of Cast Austenitic Stainless Steels, in: 2016 24th Int. Conf. Nucl. Eng., American Socie ty of Mechanical Engineers, 20 16. [89] W.J. Ai, H.C. Kuan, J.J. Dong, C.T. Kwok, K.H. Lo, Short term spinodal decomposition its effects on corrosion behaviour of a duplex stainless steel and feasibility as a strengthening treatment, Mater. Corros. 68 (2017 ) 395 404. [90] S. Lampman, Weld integrity and performance, ASM Int. (1997). [91] S.A. David, J.M. Vitek, D.J. Alexander, Embrittlement of austenitic stainless steel welds, J. Nondestruct. Eval. 15 (1996) 129 136. [92] P. Wilkes, Phase stability under irra diation a review of theory and experiment, J. Nucl. Mater. 83 (1979) 166 175. [93] K. Fujii, K. Fukuya, Effects of radiation on spinodal decomposition of ferrite in duplex stainless steel, J. Nucl. Mater. 440 (2013) 612 616. [94] M. Li, M.K. Miller, W. Y. Chen, Phase stability in thermally aged CASS CF8 under heavy ion irradiation, J. Nucl. Mater. 462 (2015) 214 220. [95] B. Jnsson, Assessment of the mobilities of Cr, Fe and Ni in bcc Cr Fe Ni alloys, ISIJ Int. 35 (1995) 1415 1421. [96] Z. Li, W. Y. Lo, Y. Chen, J. Pakarinen, Y. Wu, T. Allen, Y. Yang, Irradiation response of delta ferrite in as cast and thermally aged cast stainless steel, J. Nucl. Mater. 466 (2015) 201 207. [97] R.O. Williams, H.W. Paxton, The nature of aging of binary iron chromium alloys around 500 C, J Iron Steel Inst. 185 (1957) 358 374. [98] W. Xiong, P. Hedstrm, M. Selleby, J. Odqvist, M. Thuvander, Q. Chen, An improved thermodynamic modeling of the Fe Cr system down to zero Kelvin coupled with key experiments, Calphad. 35 (2011) 355 366. [99] G. Gladyshev, Radiation Processes in Crystal Solid Solutions, Be ntham Science Publishers, 2012.

PAGE 151

151 [100] K. Krishan, C. Abromeit, Calculation of radiation induced instability in concentrated alloys, J. Phys. F Met. Phys. 14 (1984) 1103. [101] C. Ab romeit, K. Krishan, Influence of irradiation on the spinodal decomposition of a concentrated alloy, Acta Metall. 34 (1986) 1515 1524. [102] S. Novy, Mcanisme de vieillissement trs longue chance des aciers inoxydables austnoferritiq ues, Universit de Rouen, 2009. [103] S. Novy, P. Pareige, C. Pareige, Atomic scale analysis and phase separation understanding in a thermally aged Fe 20at.% Cr alloy, J. Nucl. Mater. 384 (2009) 96 102. [104] J.K. Sahu, U. Krupp, R.N. Ghosh, H. J. Christ, Effect of 475 C em brittlement on the mechanical properties of duplex stainless steel, Mater. Sci. Eng. A. 508 (2009) 1 14. [105] T.J. Nichol, A. Datta, G. Aggen, Embrittlement of ferritic stainless steels, Metall. Mater. Trans. A. 11 (1980) 573 585. [106] J.E. Westraadt, E. J. Olivier, J.H. Neethling, P. Hedstrm, J. Odqvist, X. Xu, A. Steuwer, A high resolution analytical scanning transmission electron microscopy study of the early stages of spinodal decomposition in binary Fe Cr, Mater. Charact. 109 (2015) 216 221. [107] F. Danoix, P. Auger, S. Chambreland, D. Blavette, A 3D study of G phase ferrite by tomographic atom probe analysis, Microsc. Microanal. Microstruct. 5 (1994) 121 132. [108] J. Y. Jeon, Y. J. Kim, J. W. Kim, S. Y. Lee, Effect of thermal ageing of CF8M on multi axial ductility and application to fracture toughness prediction, Fatigue Fract. Eng. Mater. Struct. 38 (2015) 1466 1477. [109] D.J. Alexander, J.M. Vitek, S.A. David, Long term aging of type 308 stainless steel w elds: Effects on properties and microstructure, Oak Ridge National Lab., TN (United States), 1994. [110] W.J. Mills, Fracture toughness of stainless steel welds, in: Fract. Mech. Ninet. S ymp., ASTM International, 1988. [111] C.R. Weinberger, B.L. Boyce, C. C. Battaile, Slip planes in bcc transition metals, Int. Mater. Rev. 58 (2013) 296 314. [112] M.J. Marcinkowski, A. Szirmae, R. Fisher, Effect of 500degreesc aging on deformation behavior of iron chromium alloy Trans. Metall. Soc. AIME. 230 (1964) 676.

PAGE 152

152 [11 3] M.A. Meyers, O. Vhringer, V.A. Lubarda, The onset of twinning in metals: a constitutive description, Acta Mater. 49 (2001) 4025 4039. [114] Y. C. Hsieh, L. Zhang, T. F. Chung, Y. T. Tsai, J. R. Yang, T. Ohmura, T. Suzuki, In situ transmission electron microscopy investigation of the deformation ferrite in a duplex stainless steel, Scr. Mater. 125 (2016) 44 48. [115] J.W. Cahn, HARDENING BY SPINODAL DECONPOSITION, in: Symp. ROLE Substruct. M ech. Behav. Met., 1963: p. 375. [116] M. Kato, Hardening by spinodally modulated structure in bcc alloys, Acta Metall. 29 (1981) 79 87. [117] 397. [118] Y Y. Chuang, Y.A. Chang, A thermodynamic analysis and calculation of the Fe Ni Cr phase diagram, Metall. Trans. A. 18 (1987) 733 745. [119] T. Yamada, S. Okano, H. Kuwano, Mechanical property and microstructural change by thermal aging of SCS14A cast duple x stainless steel, J. Nucl. Mater. 350 (2006) 47 55. [120] D.J. Maykuth, R.I. Jaffee, Influence of Chromium Metal Purity on the Properties of Chromium Alloys, Ductile Chromium Its Alloys ASM. (1957) 229 246. [121] R.O. Williams, Theory of precipitation har dening: Isotropically strained system, Acta Metall. 5 (1957) 385 392. [122] T. Takeuchi, Y. Kakubo, Y. Matsukawa, Y. Nozawa, Y. Nagai, Y. Nishiyama, J. Katsuyama, K. Onizawa, M. Suzuki, Effect of neutron irradiation on the microstructure of the stainless s teel electroslag weld overlay cladding of nuclear reactor pressure vessels, J. Nucl. Mater. 443 (2013) 266 273. [123] J.A. Brinkman, Production of atomic displacements by high energy particles, Am. J. Phys. 24 (1956) 246 267.

PAGE 153

153 BIOGRAPHICAL SKETCH Zhang bo Li was bo rn in Guangshui, Hubei p He received a B.S. degree in M aterial s Science and E ngineering from Huazhong University of Science and Technology in 2012. In August 2012, Zhangbo started a Master of Engineering prog ram in Materials Science and Engineering at University of Florid a. One year later, he joined Professor sferred to the Ph.D program at Universit y of Florida. After completing his Ph. D, he plans to work in academia or indus try