Citation
Densification Evolution and Properties Evaluation of UO2-Based Composites Prepared by Spark Plasma Sintering (SPS)

Material Information

Title:
Densification Evolution and Properties Evaluation of UO2-Based Composites Prepared by Spark Plasma Sintering (SPS)
Creator:
Chen, Zhichao
Place of Publication:
[Gainesville, Fla.]
Florida
Publisher:
University of Florida
Publication Date:
Language:
english
Physical Description:
1 online resource (144 p.)

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Mechanical Engineering
Mechanical and Aerospace Engineering
Committee Chair:
SUBHASH,GHATU
Committee Co-Chair:
CHEN,YOUPING
Committee Members:
TULENKO,JAMES S
YANG,YONG
Graduation Date:
12/18/2015

Subjects

Subjects / Keywords:
Composite particles ( jstor )
Densification ( jstor )
Graphite ( jstor )
Graphitization ( jstor )
Heating ( jstor )
High temperature ( jstor )
Plasmas ( jstor )
Residual stress ( jstor )
Sintering ( jstor )
Thermal conductivity ( jstor )
Mechanical and Aerospace Engineering -- Dissertations, Academic -- UF
diamond -- sintering -- uo2
Genre:
bibliography ( marcgt )
theses ( marcgt )
government publication (state, provincial, terriorial, dependent) ( marcgt )
born-digital ( sobekcm )
Electronic Thesis or Dissertation
Mechanical Engineering thesis, Ph.D.

Notes

Abstract:
In current nuclear industry, uranium dioxide (UO2) is the most widely used fuel material. However, the extremely low thermal conductivity of UO2 limits both efficiency and safety of the nuclear reactor. In order to increase the thermal conductivity of UO2, the idea of incorporating high thermal conductivity material into a UO2 matrix has been proposed. A new sintering technique called spark plasma sintering (SPS) has then been reported to fabricate the UO2 composite pellets successfully. The main objective of this research is to investigate the sintering evolution in SPS of fuel pellets and to study the feasibility of novel UO2-diamond composite fuel pellets. Firstly, master sintering curve (MSC) theory was utilized in order to study the densification evolution of UO2 and UO2 composites in SPS. For UO2-diamond composites, the theory was proven to be not suitable. For UO2 and UO2-SiC composite, MSC was successfully applied. The apparent activation energies for sintering is determined to be 140 KJ/mol for UO2 and 420 KJ/mol for UO2-SiC composite. The ability of the derived MSCs to control and predict final density in the sintered compact was demonstrated by additional experimental runs using the isothermal heating method. Second, the microstructure and properties of UO2-diamond composites was investigated. High density UO2-5 vol% diamond composite pellets were fabricated using the spark plasma sintering (SPS) technique at 1300oC-1600oC with a hold time of 5 minutes. The resultant density, chemical reaction, microstructure, thermal conductivity and elastic modulus of the sintered pellets were investigated. Third, micro-Raman spectroscopy (MRS) was utilized to study the phase transformation and residual stress in diamond particles within a UO2-diamond composite sintered by SPS. Graphitization of diamond was observed and the degree of graphitization was quantified. The relationship between Raman peak shift and stress intensity was derived. Last, high temperature aging test was performed for UO2-diamond composite pellet. The changes of microstructure and thermal property were investigated. ( 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, 2015.
Local:
Adviser: SUBHASH,GHATU.
Local:
Co-adviser: CHEN,YOUPING.
Statement of Responsibility:
by Zhichao Chen.

Record Information

Source Institution:
UFRGP
Rights Management:
Copyright Chen, Zhichao. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Classification:
LD1780 2015 ( lcc )

Downloads

This item has the following downloads:


Full Text

PAGE 1

1 DENSIFICATION EVOLUTION AND PROPERTIES EVALUATION OF UO 2 BASED COMPOSITES PREPARED BY SPARK PLASMA SINTERING (SPS) By ZHICHAO CHEN A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FU LFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2015

PAGE 2

2 © 2015 Zhichao Chen

PAGE 3

3 To my parents, Guangyue Ding and Jian Chen, for their love and support

PAGE 4

4 ACKNOWLEDGMENTS I would like to thank Dr. Ghatu Subhash , my advisor and the supervisory committee chair for his support and guidance . H is knowledge and enthusiasm in scientific exploration influenced me a lot d uring my three years of work . I would like to thank Professor James Tulenko , my supervisory committee member and research mentor , for making significant contributions on my work and knowledge in the nuclear area . I would also like to thank Dr . Yong Yang and Dr. Youping Chen , for be ing my supervisory committee member , reviewing my work and giving me significant advice along the way . T hanks to Dr. Paul Carpinone and Dr. Luisa Dempere in Major Analytical Instrumentation Center for their help on instrument u sage and advice on my experiments. Thanks to Dr. Kenneth McClee lan at Los Alamos National Laboratory for his professional review and feedback on my work. Also thanks to all my colleagues in Laboratory for the Development of Advanced Nuclear Fuels and Materials (LDANFM) in department of n uclear Engineering and Laborato ry for Dynamic Response of Advanced Materials (LDRAM) in department of mechanical and a erospace Engineering. My research would have been much more difficult with their kindly help. The funding for this research is supported by the Department of Energy ( DOE ) Office of Nuclear Energy University Programs. I would like to thank them for the financial support for this research. Last but not least, I would like to express my appreciation to my father Jian Chen, my mother Guangyue Ding, my grandfather Ronglin Ding and my grandmother Lingyan

PAGE 5

5 Su . Also t hanks to my girlfriend Huihui Wang, who has ex hibited more patience than I deserve. Without their love and support, this work could not be accomplished .

PAGE 6

6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF T ABLES ................................ ................................ ................................ ............ 8 LIST OF FIGURES ................................ ................................ ................................ .......... 9 LIST OF ABBREVIATIONS ................................ ................................ ........................... 12 ABSTRACT ................................ ................................ ................................ ................... 14 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 16 2 LITERATURE REVIEW ................................ ................................ .......................... 26 Structure and Properti es of UO 2 ................................ ................................ ............. 26 Structure and Properties of Diamond ................................ ................................ ...... 28 Effective Thermal Conductivity of Particulate Composites ................................ ...... 29 Kapitza Resistance ................................ ................................ ........................... 29 Effective Thermal Conductivity Models ................................ ............................ 30 Spark Plasma Sintering ................................ ................................ .......................... 31 3 DENSIFICATION EVOLUTION STUDY OF UO 2 AND UO 2 BASED COMPOSITES DURING SPS ................................ ................................ ................. 38 Background ................................ ................................ ................................ ............. 38 Master Sintering Curve (MSC) Theory ................................ ................................ .... 40 Experimental ................................ ................................ ................................ ........... 41 Starting Powders ................................ ................................ .............................. 41 Spark Plasma Sintering ................................ ................................ .................... 42 S hrinkage Data Process ................................ ................................ ................... 43 Results and Discussion ................................ ................................ ........................... 45 C onstruction of MSC ................................ ................................ ........................ 45 Validation and Application of MSC ................................ ................................ ... 48 Di scussion ................................ ................................ ................................ .............. 49 Conclusion ................................ ................................ ................................ .............. 51 4 MICROSTRUCTURE AND PROPERTY OF DIAMOND REINFORCED URANIUM DIOXIDE COMPOSITE FUEL PELLETS ................................ .............. 63 Background ................................ ................................ ................................ ............. 63 Experimental ................................ ................................ ................................ ........... 63 Powder Preparation ................................ ................................ .......................... 63

PAGE 7

7 Spark Plasma Sintering ................................ ................................ .................... 64 Characterization Methods ................................ ................................ ................. 64 Results and Discussion ................................ ................................ ........................... 66 Density ................................ ................................ ................................ ............. 66 Chemical Reaction ................................ ................................ ........................... 67 Microstructure ................................ ................................ ................................ ... 69 Dispersion of diamond particles ................................ ................................ . 69 UO2 diamond interface ................................ ................................ .............. 69 Grain size ................................ ................................ ................................ ... 7 0 Thermal Conductivity ................................ ................................ ........................ 71 ................................ ................................ .............................. 73 Conclusion ................................ ................................ ................................ .............. 75 5 MICRO RAMAN STUDY OF PHASE TRANSFORMATION AND RESIDUAL STRESS IN UO 2 DIAMOND COMPOSITE FUEL ................................ .................. 89 Background ................................ ................................ ................................ ............. 89 Derivation of Relationship between Raman Peak Shift and Residual Stress in Diamond Particles ................................ ................................ ............................... 92 Experimental ................................ ................................ ................................ ........... 94 Pellet Fabrication ................................ ................................ .............................. 94 Raman Spectroscopy ................................ ................................ ....................... 95 Results and Discussion ................................ ................................ ........................... 95 Gra phitization of Diamond within the Pellet ................................ ...................... 96 Residual Stress of Diamond within the Pellet ................................ ................... 99 Conclusion ................................ ................................ ................................ ............ 101 6 PERFORMANCE OF UO 2 DIAMOND COMPOSITE PELLET UNDER HIGH TEMPERATURE AGING TEST ................................ ................................ ............ 111 Background ................................ ................................ ................................ ........... 111 Experimental ................................ ................................ ................................ ......... 111 Results and Discussion ................................ ................................ ......................... 112 Conclusion ................................ ................................ ................................ ............ 113 7 CONCLUSION AND FUTURE WORK ................................ ................................ .. 120 Conclusion ................................ ................................ ................................ ............ 120 Future Work ................................ ................................ ................................ .......... 121 Wa shout Behavior Study ................................ ................................ ................ 121 Compression Test ................................ ................................ .......................... 125 LIST OF REFERENCE S ................................ ................................ ............................. 134 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 144

PAGE 8

8 LIST OF TABLES Table page 3 1 Parameters of the heating profiles to sinter a UO 2 pellet with 90% of the TD in SPS ................................ ................................ ................................ .................... 53 3 2 Density comparison of UO 2 and UO 2 SiC pellets under the same heating profile in SPS ................................ ................................ ................................ ...... 53 4 1 Properties of UO 2 and diamond. ................................ ................................ ......... 76 5 1 Physical properties of UO 2 and diamond ................................ .......................... 102 6 1 Weight and density change of UO 2 diamond pellet after aging test .................. 114 7 1 Change of pellet weight before and after washout test ................................ ..... 126

PAGE 9

9 LIST OF FIGURES Figure page 1 1 US elec tricity production costs in 1995 2008 . ................................ ..................... 23 1 2 Nuclear e lectrici ty production in recent years . ................................ .................... 24 1 3 The percentage of nuclear ele ctricity generation in different countries in 2012 . ................................ ................................ ................................ .................. 24 1 4 Thermal conductivity of 15 vol% particulate diamond reinforced cordierite matrix composites for a range of diamond particle size. ................................ ..... 25 2 1 Schem a tic of UO 2 structure. ................................ ................................ ................ 34 2 2 Thermal conductivity of a 95% relative density UO 2 pellet . ................................ . 34 2 3 Electrical conductivity of single crystalline UO 2 . ................................ ................. 35 2 4 Thermal conductivity of natural abundance (1.1% 13 C) diamond and isotopically enriched (0.1% 13 C and 0. 001% 13 C) diamond a s a function of temperature . ................................ ................................ ................................ ....... 36 2 5 Dr. Sinter® SPS 1030 system and schematic drawing of the die assemble. ..... 37 3 1 SEM images of starting powders. ................................ ................................ ....... 54 3 2 Heating profiles for UO 2 pellets under the constant heating rate method and isothermal method ................................ ................................ .............................. 55 3 3 Shrinkage curves of UO 2 and the graphite during SPS.. ................................ .... 56 3 4 Evolution of relative density as a function of temperature for different heating rates.. ................................ ................................ ................................ ................. 57 3 5 SEM images of fracture surfaces. ................................ ................................ ....... 58 3 6 curves for selected value of activation energy of UO 2 .. ............................ 59 3 7 Mean residual squares for various estimated activation energy values .............. 59 3 8 Master sintering curve construction. ................................ ................................ ... 60 3 9 Comparison of MSC predicted relative density and experimental value of UO 2 and UO 2 SiC pellets using isothermal heating method ............................... 61 3 10 SEM images of pellets sintered by isothermal heating method. ......................... 62

PAGE 10

10 4 1 SEM image of as received UO 2 powder ................................ ............................. 76 4 2 SEM images showing morphologies of diamond powder. ................................ .. 77 4 3 Influence of sintering temperature and diamond particle size on the density of UO 2 +5vol % diamond composite fuel ................................ ................................ .. 78 4 4 XRD spectra of starting powder and UO 2 70 vol% diamond pellets sintered at various temperatures.. ................................ ................................ ........................ 79 4 5 Fracture surfaces of UO 2 diamond pellets sintered at 1400 o C with different mean diamond particle sizes. ................................ ................................ ............. 80 4 6 UO 2 diamond Interface in composite pellets sintered at 1400 o C with differen t mean diamond particle sizes.. ................................ ................................ ............ 82 4 7 Polished and thermally etched surfaces of UO 2 diamond composites with different mean diamond particle sizes.. ................................ .............................. 84 4 8 Thermal conductivity UO 2 diamond pellets with various sizes of diamond particles.. ................................ ................................ ................................ ............ 86 4 9 Comparison of thermal conductivity of UO 2 and various UO 2 based composites ................................ ................................ ................................ ......... 87 4 10 2 diamond composites as a function of mean particle size of diamond.. ................................ ................................ ................................ 88 5 1 SEM image of as received p owders. ................................ ................................ 103 5 2 Image of SPSed pellets.. ................................ ................................ .................. 104 5 3 Raman spectrum of as received diamond powder. ................................ .......... 105 5 4 Image of location ................................ ................................ .............................. 106 5 5 Typical Raman spectra of on diamond in UO 2 diamond composite.. ................ 107 5 6 Graphitization degree in UO 2 diamond composite pellet ................................ .. 108 5 7 Comparison of diamond Raman peak position between as received diamond powder and sintered pellet ................................ ................................ ............... 109 5 8 Raman line scan. . ................................ ................................ ............................. 110 6 1 Appearance of UO 2 diamond composite pellet.. ................................ ............... 115 6 2 SEM images of surface of UO 2 diamond composite pellet. .............................. 116

PAGE 11

11 6 3 SEM images of interior of UO 2 diamond composite pellet after aging test. ...... 117 6 4 XRD spectra of U O 2 diamond composite pellet before and after thermal aging test. ................................ ................................ ................................ ......... 118 6 5 Change of thermal diffusivity of UO 2 diamond composite after aging test. ....... 119 7 1 Images of pellets (From left to right: UO 2 , UO 2 SiC and UO 2 diamond composite) under 5h washout test. ................................ ................................ ... 127 7 2 Images of pellets (From left to right: U O 2 , UO 2 SiC and UO 2 diamond composite) under 15h washout test. ................................ ................................ . 128 7 3 XRD spectra of pellets before and washout test.. ................................ ............. 129 7 4 XR D spectra of UO 2 and UO 2 composite pellets after 15 hours holding at 380 o C in 0.1% O 2 . ................................ ................................ ............................. 130 7 5 Crumbled UO 2 and UO 2 composites pellets after oxidation test. ...................... 131 7 6 XRD spectrum of crumbled UO 2 and UO 2 composite pellets after 20h washout test in 1% O 2 . ................................ ................................ ..................... 132 7 7 Stress Strain curve of UO 2 5 vol% diamond composite pellet und er compression test. ................................ ................................ ............................. 133

PAGE 12

12 LIST OF ABBREVIATIONS ASTM A merican S ociety for T esting and M aterials ATF CaF 2 CO 2 Accident tolerant fuel C alcium fluori t e Carbon dioxide CNT CTE Carbon nanotube Coefficient of thermal expansion DC Direct current EDS Energy dispersive spectroscopy FAST Field as sisted s intering t echnique FCC HP LWR LOCA Face centered cubic Hot press Light water reactor Loss of coolant accident MPCD MRS Mean Perpendicular Curve Distance Micro Raman spectroscopy MSC Master sinterin g curve NEI O/U Nuclear energy institute Oxygen/Uranium OM Optical microscope PDF PWR SEM SiC SPS Powder diffraction file Pressurized water reactor Scanning electron microscopy Silicon carbide Spark plasma sintering

PAGE 13

13 TD Theoretical d ensity XRD X ray diffraction UO 2 Uranium d ioxide

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 DENSIFICATION EVOLUTION AND PROPERTIES EVALUATION OF UO 2 BASED COMPOSITES PREPARED BY SPARK PLASMA SINTERING (SPS) By Zhichao Chen December 2015 Chair: Ghatu Subhash Major: Mechanical Engineering In current nuclear industry, u ranium dioxide (UO 2 ) is the most widely u sed fuel material. However, the extremely low thermal conductivity of UO 2 limits both effici ency and safety of the nuclear reactor . In order to increase the thermal conductivity of UO 2 , t he idea of incorporating high therm al conductivity material into a UO 2 matrix has been proposed. A new sinteri ng technique called spark plasma sintering (SPS) has the n been reported to fabricate the UO 2 composite pellets successfully . The main objective of this research is to investigate the sintering evolution in SPS of fu el pellets and to study the feasibility of novel UO 2 diamond composite fuel pellets . Firstly, m aster sintering c urve (MSC) theory was utilized in order to study the densification evolution of UO 2 and UO 2 composites in SPS. For UO 2 diamond composites, the t heory wa s proven to be not suitable. For UO 2 and UO 2 SiC composite , MSC wa s successfully applied . T he apparent act ivation energies for sintering is determined to be 140 KJ/mol for UO 2 and 420 KJ/mol for UO 2 SiC composite . The ability of the derived MSC s to control and predict final den sity in the sintered compact wa s demonstrated by additional experimental runs using the isothermal heating method.

PAGE 15

15 Second, t he microstructure and properties of UO 2 di amond composites was investigated . H igh density UO 2 5 vol % diamond composite pellets were fabricated using the spark plasma sintering (SPS) technique at 1300 o C 1600 o C with a hold time of 5 minutes. T he resultant density, chemical reaction, microstructure, thermal conductivity and elastic modulus of the sintered pe llets were investigated. Third, mi cro Raman spectroscopy (MRS) wa s utilized to study t he phase transformation and residual stress in diamond particles within a UO 2 diamond composite sintered by SPS . Graphitization of d iamond was observed and the degree o f graphitization wa s quantified. T he relationship between Raman peak shift and stress intensity was derived. Last , high temperature aging test was performed for UO 2 diamond composite pellet . The change s of microst ructure and thermal property were investig ated.

PAGE 16

16 CHAPTER 1 INTRODUCTION As one of the most fundamental requirements for all human activities, energy is the key to sustainable development. With global development of economy and society, the demand for energy is increasing dramatically year by year. Traditional energy sources such as coal and oil promote economic growth but lead to numerous negative effects on the environment and climate . In addition, we have to face the exhaustion of the coal and oil as they are nonrenewable resources. For t hese reas ons, the energy crisis has become one of the biggest challenges for humanity. It is of great importance to consider alternatives. Nuclear energy is one of the best energy sources discovered in the last century which may be able to address the present ener gy crisis. Compared with other energy sources, nuclear energy has lots of advantages. First of all, nuclear power is very cost effective. Figure 1 1 shows the cost of US electri city production in recent years reported by the US Nuclear Energy Institute (NE I) [ 1 ] . According to the data, the cost of electricity produced by nuclear power plant is only 1.9c/KWh in 2008, while electricity generated by coal, gas or oil cost 3.2 c/KWh, 8. 1c/KWh or 17.8c/KWh, respectively. Secondly, nuclear energy has low environmental impacts compared to the other aforementioned energy sources. Due to excessive use of fossil fuels, th e amount of greenhouse gases has significantly increased in the last cent ury . Due to this, t he danger of global warming has been empha sized nowadays. Nuclear power has been proven to be the most environmentally benign way of producing electricity on a large scale. With nuclear energy, the use of fossil fuels can be diminished a nd the emission of greenhouse gas

PAGE 17

17 can be limited. Only a small amount of land is needed to isolate the waste generated by nuclear energy stations. Commercial nuclear power started in the United States. Yankee Rowe, designed by Westinghouse, was the first f ully commercial pressurized water reactor (PWR) and started operation in 1960. It had a capacity of 250 MWe [ 2 ] . During the past half century, significant progress has been made in the development of the nuclear industry. Figure 1 2 shows rapid growth of nuclear electricity production f rom 1971 to 2013. A t present , over 435 commercial nuclear power reactor s ar e operating in 31 countries and more are under construction. Figure 1 3 shows t he percentage of nuclear electricity ge neration in different countries. The total production provides over 1 electricity [ 3 ] . Nuclear power is created by splitting of uranium atoms ( 235 U) in a process which is called fission. Uranium was discove red in 1 789 by Martin Klaproth and named after the planet Uranus. The energy generated by f ission releases can be used to generate electricity. In nuclear reactors, the compounds of uranium are often used rather than the metal uranium itself because the melting po int of metal uranium is relatively low (1132 o C). Among all uranium compounds such as uranium nitride, uranium phosphide and uranium carbi de, uranium dioxide (UO 2 ) is the most widely used fuel material . Compared to other materials, UO 2 has the advantage of high melting point (2865 o C) [ 4 ] , good high temperature stability, good chemical compatibility with both cladding and coolant and resistance to radiation [ 5 ] . However, the main drawback of a UO 2 fuel pellet is it s ext remely low thermal conductivity. A review of thermal conductivity of uranium dioxide will be given in

PAGE 18

18 Chapter 2.1. Due to its poor thermal conductivi t y , UO 2 fuel pellets will lead to a fairly high centerline temperature during operation in a reactor and a steep radial temperature gradient from centerline to the edge , typically around a 500 o K drop within a pellet radial distance of 4 5 mm. Many harmful phenomena can occur under such condition. First, a large t hermal stress can be induced by the temperature gradient. The stress will lead to plastic deformation of the fuel pellets. If the stress intensity is higher than the toughness of UO 2 , micro cr acks will also be formed within the pellets . Second, the high temperature encourages fission gases such as such as krypton and xenon to be released, which would swell the fuel pellets because of insolubility. A ddition ally , in a loss of coolant accident (LOCA), wherein the zi rconium cladding rapidly reacts with steam, t he reaction is accelerated due to the high temperature of the fuel pellets and insufficient coolant. Hydrogen gas can be released through this reaction: Zr + 2H 2 O ZrO 2 + 2 H 2 A build up of hydrogen gas can lead to an explosion. On March 11, 2011, the reaction above occurred in boiling water reactors of the Fukushi ma Daiichi Nuclear Power Plant in Japan after the cooling system was interrupted by an earthquake, leading to the well known Fukushima Daiichi nuclear disaster. After that, the concept of accident tolerant fuel (ATF) was proposed and a variety of approaches a re being examined. The attempted approaches can be divided into two categor ies, s ome of which focus on the improvement of Zr cladding, such as use of coatings or replacement of Zr alloy with other materials, while others focus on the improvement of t he fuel material [ 6 ] . Therefore, UO 2 fuel with enhanced thermal conductivity is of great interest. The importance of increasing the thermal conductivity of UO 2 fuel cannot be over

PAGE 19

19 emphasized. If high thermal conductivity UO 2 fuel can be achieved, all the problems mentioned ab ove can be significantly mitigated . Moreover, the total power and efficiency of a reactor can be increased and the safety of a reactor can be enhanced as a result. In order to increase the thermal conductivity of UO 2 , t he idea of incorporating high thermal conductivity material into a UO 2 matrix and making a composite has gained popularity in recent years. With the addition of a second phase material (particle, fiber, etc), high heat conducting percolation pathways can be formed, which make the heat conduction more efficient [ 7 ] . The idea of a high thermal conductivity composite has been successfully applied to many materials. Hasselman et al., [ 8 ] fabricated silicon carbide particulate reinforced alumina composite with good density and interfacial thermal conductance. Zhou et al., [ 9 ] added 1 20vo l% Si 3 N 4 particles into polyethylene matrix and increased the thermal conductivity from 0.2 to 1.0 W m 1 K 1 . For UO 2 matrix, numerous other second phase materials have been proposed in the literature. Ishimoto et al. , [ 10 ] considered BeO as a second phase and sintered UO 2 BeO composite in a reducing atmosphere at 2073K succe ssfully. Yeo et al. , [ 11 , 12 ] and Cartas et al. , [ 13 ] investigated the feasibility of silicon carbide (SiC) and carbon nanotube (CNT) , respectively. Lee et al. , [ 14 ] proposed the idea of UO 2 graphene composite and studied the thermal propertie s through numerical simulation. In order to successfully fabricate UO 2 based compos ite fuel with enhanced thermal properties, the sintering method is of great importance. The conventional sintering method for fabrication of pure UO 2 fuel pellets is pressureless sintering. In a reducing atmosphere (argon hydrogen), a maximum sintering tem perature of 1700 o C is

PAGE 20

20 required. While in an oxidizing atmosphere, pellets can be fully sintered at 300 400 o C lower. However, UO 2 based composite is very difficult to fabricate using conventional pressureless sintering method. I nitial attempts to sinter UO 2 powder with second phase high conductivity materials such as SiC have been met with limited success, especially in terms of reaching the desired 96.5% of the theoretical density (TD) in the final pellet [ 15 ] . The thermal conductivity of low density composite pellet prepared by conventional sintering method is even lower than pure UO 2 pellet [ 11 ] . In order to solve this problem, a non traditional sintering method, spark plasma sinte ring (SPS), is used. SPS is a novel sintering technique which utilizes an electrical field and uniaxial pressure to facilitate sintering. A review of this technique can be seen in Chapter 2.4. In 2013, Ge et al., [ 16 ] first shows the feasibility of UO 2 fabrication using SPS. Compared to conventional sintering method s , S PS has many advantages . In conventional sintering of UO 2, it is require d that the pe l let be held at maximum sintering temperature (1300 o C 1700 o C) for 2 4 hours. The enti re temperature cycle (heating time, hold time, and cooling time) may take 10 15 hours due to a slow heating rate of only around 5 o C/min [ 17 19 ] . However, SPS of UO 2 only requires a maximum sintering temperature of 1050 o C and a hold time as short as 30 seconds. The heating rate can be up to 200 o C/min. thus significantly reducing t h e t o tal processing time . A detailed comparison between SPS and traditional sinterin g of UO 2 can be found in the work of Ge et al [ 20 ] . Using SPS, high density UO 2 SiC composite pellets [ 11 ] and UO 2 CNT composite pellets [ 13 ] with minimum chemical reaction and excellent interfacial contact have also been fabricated. The maximum si ntering temperature for fabricating UO 2 -

PAGE 21

21 based composite s is found to be as low as 1350 o C and the hold time is found to be as low as 5 minutes. With the previous success of UO 2 based composites and the improvement of its thermal conductivity, many other second phase materials are being considered. Among all potential candidates , diamond has an extremely high thermal conductivity (about 22 W/cm·K at room temperature), which is the highest of any solid material [ 21 ] . Because of its sp 3 hybridization bond structure, diamond is found to be resistant to the radiation environment [ 22 ] . Due to the superior properties, diamond has been widely used as a reinforci n g phase in either metal matrix or cer amic matrix composites for elevated thermal conductivity. Hasselman et al., [ 23 ] added 15 vol% particulate diamond in a cordierite matrix with different diamond particle sizes. The variation of thermal conductivity from room temperature to 700 o C is shown in Figure 1 4. A max imum increase of 75% in thermal conductivity at 700 o C was obtained. This new composite significantly improve d the thermal performance of the substrate material for electronic packaging. In addition, diamond reinforced composites have a l so been used for heat sink applications. Schubert et al., [ 24 ] and Chu et al., [ 25 ] fabricated Cu/diamond composite using powder metallurgy and spark plasma sintering (SPS), respectively. They both emphasized the effect of Cr on the improvement of Cu diam ond interfacial bonding and thermal conductivity. The performance of diamond in multiphase material has also been evaluated. Tavangar et al., [ 26 ] studied a diamond reinforced Al Si composite system and d iscussed the effect of heat treatment condition on its thermal nuclear fuel and UO 2 diamond composite has never been considered. From the

PAGE 22

22 economic perspective, the price of s ynthetically produced diamond powder is much lower (~$2.5/g) than that of CNT (~$270/g) or graphene (~$125/g), which is beneficial for commercialization in the future . Therefore, diamond makes an ideal second phase for incorporation into UO 2 pellets to ach ieve a higher thermal conductivity in the composite. The main purpose of this work is to study the nature of t h e SPS process for nuclear fuel pellets and to evaluate the feasibility of novel UO 2 diamond composite fuel. First, the densification evolutions of UO 2 , UO 2 SiC and UO 2 diamond composites are explored. Although UO 2 and UO 2 based composite s have been successfully fabricated, densification evolutions of them during SPS have never been investigated. The m aster sintering curve theory is attempted to describe the densification evolution on SPS of nuclear materials. Second, an investigation into the microstructure and properties of UO 2 diamond composite pellets is presented. T h e a ppropriate diamond particle size and sintering parameters are determined. Third, the change of diamond particles in a UO 2 diamond composite during SPS is studied. Graphitization of diamond and residual stress in diamond particles are observed. Micro Raman spectroscopy (MRS) is utilized for quantitative analysis of graphitization degree and resid ual stress intensity. The structure property variation within a pellet is al so discussed. L astly , the performance of UO 2 diamond composite after a high temperature aging test is evaluated.

PAGE 23

23 Figure 1 1. US electricity production costs in 199 5 2008 [ 1 ] .

PAGE 24

24 Figure 1 2. Nuclear e lectricity production in recent years [ 3 ] . Figure 1 3 . The percentage of nuclear electricity generation in different countries in 2012 [ 27 ] .

PAGE 25

25 Figure 1 4. Thermal conductivity of 15 vol% particulate diamond reinforced cordierite matrix composites for a range of diamond particle size [ 23 ] .

PAGE 26

26 CHAPTER 2 LIT ERATURE REVIEW Structure and Properties of UO 2 The structure and properties of uranium dioxide ( UO 2 ) have been intensively studied. Natural ur anium dioxide is a crystalline powder which has a calcium fluori t e (CaF 2 ) crystal structure . The structure belongs to the Fm3m space group where U atoms r eside in face centered cubic ( FCC ) s ite s and O atoms occupy tetrahedral interstitial sites. A schematic of UO 2 crystal structure is shown in Figure 2 1 . The theoretical density of UO 2 is 10.96 g/cm 3 [ 28 ] . The ionic bonding between U and O atoms gives UO 2 a high melting point of 2850 °C [ 4 ] and good chemical stability , which makes this material a good candidate for nuclear fuel . Due to the crystal structure of UO 2 , excess o xygen atoms are easil y dissolve into or escape from the crystal. Therefore , either hypo stoichiometric UO 2 ( UO 1.65 UO 2. 00 ) or hyper stoichiometric UO 2 ( UO 2. 00 UO 2.25 ) can be achieved under different oxygen potentials [ 29 ] . Boyarchenkov et al., [ 30 ] investigated the cation self diff usion mechanisms in UO 2±x using molecular dynamics. He found that the cation would diffuse via different mechanisms under different conditions, either by formation of Frenkel defects of Schottky defects. The diffusion of oxygen atoms can lead to the accumu lation of defects. Therefore, the stoichiometry has a large influence on the thermal and mechanical properties of uranium dioxide. In nuclear reactors, the therm al conductivity of fuel pellet is one of the most important prope rties , as mentioned in Chapter 1 . Therefore, the thermal conductivity of solid UO 2 has been studied since 1960s . Brandt et al. , [ 31 ] first report ed a temperature dependence of the thermal conductivi ty of UO 2 in 1967 . Un like typical materials, in

PAGE 27

27 which thermal conductivity increases or decreases monotonically wit h the increase of temperature, thermal conductivity of UO 2 decreases with temperature and reach es its minimum value at about 2000K , then increa ses until the melting point 3140K. Hyland [ 32 ] a lso showed similar results . Fink [ 33 ] summarized a variety of reporte d experimental results and concluded that t he temperature dependent thermal conductivity can be calculated using the following equation (2 1) where is the thermal conductivity and t=T(K)/1000 , where T is the absolute temperature. The results are plotted in Figure 2 2. In order to diminish the effect of density, a ll exp erimental values were already converted to 95% of the theoretical density of UO 2 using the following equation ( 2 2) w here p is the porosity fraction, is the thermal conductivity of UO 2 with porosity p , is the thermal conductivity of 100% dense UO 2 and t=T(K)/1000 . As mentioned earlier, t he thermal conductivity of UO 2 is also influenced b y the O/U ratio. Both hyper stoichiometric and hypo stoichiometric phase have significantly lower thermal conductivi ty than stoi chio metric UO 2 [ 34 ] . Therefore, stoichiometric UO 2 pellets are required for use in order to get relatively high thermal conductivity. In this study, UO 2 and UO 2 based composites were fabricated using SPS, an electric field assisted sintering technique. For this reason, the electrical properties of UO 2 should also be discussed. UO 2 is an excellent semiconductor. The bandgap of UO 2

PAGE 28

28 is comparable to conve ntional silicon and GaAs [ 35 ] . The dielectric constant is almost double that of Si and GaAs. Although UO 2 has never been considered for semiconductor devices, UO 2 based semiconductor s are possible and could offer better performance. Figure 2 3 [ 35 ] shows the electrical con ductivity of single crystalline UO 2 as a function of temperature. With increases in temperature, the electrical conductivity decreases significantly. Therefore, during the spark plasma sintering process of UO 2 , it is expected that the current going through the UO 2 po wder will gradua l ly decrease and the current going throu gh the graphite die will gradually in c rease . The current distribution during SPS leads to a temperature gradient within the powder and affects the microstructure of sintered pellet [ 36 ] . Structure and Properties of Diamond Diamond is a metastable allotrope of carbon. The crystal structure is also FCC and it belongs to Fd m space group. Unlike graphite, which is formed by sp 2 hybr idization between two carbon atoms, diamond has sp 3 hybridization bond structure, which leads diamond to be more resistant to radiation rich environment s [ 22 ] . Diamond has an extremely high thermal conductivity ( about 22 W/cm·K at room temperature) , which is th e highest of any solid material [ 21 ] . Wei et al., [ 37 ] summarized the experimental data for natural and isotopically modified single crystal diamond and studied the temperature effect on the thermal conductivity, see Figure 2 4. It is found that the thermal conductivity increases wit h the increase of temperature in the low temperature range (below 100K), while i t starts to decrease wit h the increase of temperature in the high temperature range (above 100K). Under the same temperature conditions , a higher concentration of 12 C in diamond leads to a higher thermal

PAGE 29

29 conductivity. Similar results were also obtained from the work of Che et al., [ 38 ] who calculated thermal conductivity of diamond from molecular dynamic s simulations. Besides ultrahigh thermal conductivity, diamond also has supe rior mechanical properties. It is well modulus of diamond is around 1050 1200 GPa [ 39 ] , and t he toughness of diamo nd has been measured as 7.5 10 MPa·m 1/2 [ 40 ] . Effective Thermal Conductivity of Particulate Composites Thermal conductivity is a parameter which describes the heat conduction capability of a material. In a p article reinforced composite, the effective thermal conductivity of the composite has bee n systematically studied and several models have been constructed to predict its thermal conductivity value. Kapitza Resistance In a composite containing two constituents, the interfacial characteristics play a significant role in the thermal conductivity. Kapitza [ 41 ] firstly studied interfacial thermal behavior in liquid helium and found that the temperature difference at the interface was proportional to the normal part of heat flux. The proportionality coefficient was then defined as interfacial thermal resistance. The resistance is also called Kapitza resistance, which is a property of an interf ace to resist heat flow. In solid materials, heat is conducted by energy carriers, either phonon s or free electrons. D ue to the difference in vibrational properties in different materials, the energy carrier will scatter at the interface when it attempt s to traverse from one material to the other. Therefore, the Kapitza resistance exists even at an atomically perfect interface. In ceramic matrix composites, phonons are the dominant carrier due to a lack of free electrons. Phonon -

PAGE 30

30 phonon scattering occur s du e to crystalline defects and lattice imperfections , which determines the thermal conductivity of ceramic composites. Effective Thermal Conductivity Models Theoretical and empirical models have been constructed for various kinds of composite systems. In thi s study, only models suitable for ceramic matrix composites are discussed. A well known classical theoretical model was derived by Maxwell [ 42 ] : (2 3) where is the thermal conductivity, is volume fraction of second phase, the subscripts c , m and p represent composite, matrix and second phase particle s , asis of many other models. However, the applicability of this model is limited by several assumptions made during its derivation. First, the model is valid only for spherical shaped particles. Second ly , second phase particles should be randomly distributed in a homogeneous continuous medium with no interaction. Last, the effect of Kapitza resistance is ignored. Fricke [ 43 ] extended the theory to homogeneous ellipsoidal pa rticles and modified the equation as: (2 4) where (2 5) The in Eq (2 5) re presents the semi principal axis of the ellipsoid.

PAGE 31

31 Hasselman and Johnson [ 44 ] of particle size and Kapitza resistance into account. In his model, the effective thermal conductivity can be calculated in the following equation: (2 6) w here is the radius of particle and is the K apitza resistance. Ac cording to Eq (2 6), the thermal conductivity of a composite wou ld decrease with the decrease in size of second phase particles. Benvensite [ 45 ] also obtained the same result from on a micromechanical model. These model s have been widely used and validated in literature by comparing experimental value s with predicted results [ 23 ] . Spark Plasma Sintering The development of the s park plasma sintering (SPS) technique can be divided into three stages [ 46 ] : 1) 1930s 1970s: A sintering process using electrical energizing was developed in the United States in the 1930s. Little research was done on this process due to the lack of applica tion tech nology. 2) 1980s 1990s: Small e xperimental systems were developed with a maximum pressure of 5 tons and a maximum current of 800 A. 3) 2000s p resent : Large experimental systems were developed with a maximum pressure of 10 100 tons and a maximum current of 2000 20,000A. A variety of powde rs, such as metals [ 47 ] , ceramics [ 48 ] , polymers [ 49 ] and composite powders [ 50 ] have been compacted by the SPS technique. Even materials which are considered to be difficult to sinter by conventional methods can be sintered in short times and at relatively low temperatures to full density.

PAGE 32

32 The c onfiguration of SPS is similar to conventional hot press (HP) sintering. In both cases, powder is loaded into a graphite die with upper and lower punches. A uniaxial pressure is applied during sintering . The main difference is the heating method. In HP, th e die is heated by heat ing elements in the chamber, while in SPS, it is heated by a pulsed direct current ( DC ) which goes through the die and punch es . Figure 2 5 shows a picture of Dr. Sinter ® SPS 1030 system and a schem a tic of the die assembly. Munir et al., [ 51 ] concluded the significant features of SPS as a) high heating rate, b) the application of pressure and c) the effect of the current. Due to these features, the importance and advantages of the SPS has been proven by a l arge number of publications. Compared with conventional sintering methods which several hours of sintering time are needed, t ypical sintering t imes in SPS are on the order of 1 to 10 minutes at significantl y lower processing temperatures . Other advantages such as small grain size and homogeneity of final product are reported as well. The mechanisms of SPS have been studied in recent years. I t is pr oposed that during SPS, pulsed current causes joule heating at the inter particle contact areas and upon application of pressure, these particles fuse to form the final compact. Tokita [ 46 ] later concluded that the impact s of pulsed direct current (DC) energizing a s: a) spark p lasma, b) spark impact pressure, c) Joule heating and d) an electrical field diffusion effect. While b) d) have been accepted by most researchers now, i t is still in question as to whether a plasma is created in the SPS process [ 52 ] [ 53 ] on SPS of Fe Al alloy, the formation of plasma is proved by electrical discharge effect. However, according to t he study of Hulbert et al [ 54 ] , no pla sma can be detected using in situ atomic emission spectroscopy or in situ voltage measurements.

PAGE 33

33 Based on the understandi ng of SPS, several modeling studies have also been performed . Graeve et al., [ 55 ] constructed a model to describe the effect of a current on the reaction between powders in a compact. Anselmi Tamburini et al., [ 56 ] modeled the temperature and current distribution in the SPS for conducting and non conducting samples separately. Giuntini et al., [ 57 ] investigated the effect of graphite die/punch geometry and the amount of graphite spacers on the temperature distribution. It should be no ted that up to now, the ro le of current is limited to Jou le heat in all current models. Therefore, improvement on these models is still in progress.

PAGE 34

34 Figure 2 1 . Schematic of UO 2 structure . White circles are uranium atoms and bl ack circles are oxygen atoms. Figure 2 2 . Thermal conductivity of a 95% relative density UO 2 pellet [ 33 ] .

PAGE 35

35 Figure 2 3 . Electrical conductivity of single crystalline UO 2 [ 35 ] .

PAGE 36

36 Figure 2 4. Thermal conductivity of natural abundance (1.1% 13 C) diamond and isotopically enriched (0.1% 13 C and 0.001% 13 C) diamond as a function of temperature [ 37 ] .

PAGE 37

37 Figure 2 5. Dr. Sinter ® SPS 1030 system and sch ematic drawing of the die assembl e.

PAGE 38

38 C HAPTER 3 DENSIFICATION EVOLUTION STUDY OF UO 2 AND UO 2 BASED COMPOSITE S DURING SPS Background As discussed in Chapter 1 and C hapter 2, spark plasma sintering (SPS) has become a fairly promising approach for processing difficult to sinter powders. Recently, the application of SPS has been expanded to the area of nuclear fuel. Ge et al., [ 16 , 20 ] successfully fabricated urani um dioxide (UO 2 ) pellets with 96% of the theoretical density (TD) by SPS at a sintering temperature of 1050°C and a hold time as low as 0.5 minutes. These sintering conditions are significantly lower than the conventional sintering conditions of 1600°C for 4 hours [ 58 ] . Yeo et al. , [ 11 , 12 ] later utilized SPS to fabricate UO 2 SiC composite fuels with high density, which could not be sintered by conventional methods. While these nuclear fuel pellets have been successfully fabricated, the sintering kinetics during SPS have not been investigated. The control of the sintering process and the reproducibility of sintered compacts are crucial for quality assurance and, therefore, it is essential to understand the densification kinetics during the SPS process. Based on the type of diffusion mechanism, many sintering models have bee n proposed to describe the sintering process. Coble [ 59 ] developed an intermediate late stage sintering model for lattice diffusion. Johnson [ 60 ] derived the governing equations for calculating volume, grain boundary, and surface diffusion coefficients. Kang and Jung [ 61 ] grain boundary diffusion. These theories were verified by experimental examinations; however, they are either ju st limited to a single sintering stage, i.e., either initial, intermediate or final stage, or only applicable to single phase systems. The master

PAGE 39

39 sintering curve (MSC) concept, which was first proposed by Su and Johnson [ 62 ] , describes the sintering process in a new perspective. It is an experimentally derived model that regards sintering as a continuous process. Unlike the models mentioned above, MSC does not aim to describe sintering behavior in general, but describes the shrinkage behavior of a specific powder system. This makes MSC a useful engineering tool. In addition, MSC is applica ble to multi phase systems, both metals and ceramics. Su and Johnson [ 63 ] constructed the MSC for Al 2 O 3 TiO 2 and Al 2 O 3 ZrO 2 systems, and Park et al. [ 64 ] successfully applied this theory to tungsten heavy alloys. In addition, MSC has been introduced in industry to guide the manufacturing process. For example, Deborah et al. , [ 65 ] demonstrated the applicability of master sinteri ng curve theory in an industrial furnace. While MSC theory has been widely used to control and understand knowledge, the theory is only recently beginning to be applied for SP S process [ 66 , 67 ] and has never been applie d to nuclear materials. Field assisted sintering methods such as SPS and microwave sintering, have been known to utilize significantly lower processing times and temperatures compared to conventional methods. Therefore, the activation energy and the kineti cs of densification are expected to be dramatically different. To further understand these issues, the current manuscript deals with the development of MSCs for nuclear materials fabricated by SPS. The primary goal of the present work is to investigate the applicability of master sintering curve theory to spark plasma sintering of UO 2 , UO 2 SiC and UO 2 diamond powders. The apparent activation energies for UO 2 and UO 2 SiC composite are determined and the MSCs are

PAGE 40

40 constructed. The approach has been validated b y using isothermal heating methods. However, the MSC of UO 2 diamond is proven not to be constructed. Finally, the activation energy of sintering UO 2 derived for SPS process is compared to that for the conventional sintering method and the difference in the se values are rationalized. Master Sintering Curve ( MSC ) Theory Master sintering curve theory is derived from a combined stage sintering model originally proposed by Hansen et al [ 68 ] . In the development of MSC, two major assumptions were made [ 62 ] . First, the microstructural evolution depends on density only; and second, one of the diffusion mechanisms, i.e, either volume diffusion or grain boundary diffusion, dominates the sintering process. It shoul d be noted that another assumption, i.e., the sintering body undergoes isotropic linear shrinkage, was made in the original work of Su and Johnson [ 62 ] , but this assumption was found to be unnecessary by Diantonio and Ewsuk [ 69 ] . As mentioned earlier, the MSC was derived from the densification rate equation of the combined stage sintering model [ 68 ] . Based on the assumptions above, the sintering rate equation can be expressed in the following form [ 62 ] : ( 3 1 ) Where is the relative density, t is the sintering time, is the surface energy, is the atomic volume, is the diffusion coefficient, is lumped scaling parameter that relates geometric features, G is the grain diameter, n is a constant which depends on the sintering mechanism, Q is the apparent activation energy, R is the gas constant, and T is the absolute temperature.

PAGE 41

41 By separating all the micro structural and materi al parameters from thermal parameters, Eq. ( 3 1) can be rearranged and integrated as ( 3 2) The term on the left hand side represents the evolution of microstructure, which is independent of the thermal history of the powder compact, and can be ex pressed as a function of density, . The right hand side represents the thermal history, which . Therefore, Eq. ( 3 2) can be reduced to ( 3 3) The above empirical relationship betw een and is then defined as the master sintering curve (MSC), which is distinct for a given powder. For the construction of a MSC of a given powder, the experimental density evolution data, i.e., the integral on the left hand side of Eq. ( 3 2), sho uld be known. This requires careful experimental measurement of the volume change of powder during the sintering process. The goal of this st udy is to develop MSC for different powder systems mentioned above based on such experimental data. The experimenta l procedure and results will be discussed in detail in the following. Experimenta l Starting P owders UO 2 powder, obtained from AREVA, Hanford, WA, was reported to have a bulk density of 2.3g/cm 3 , tap density of 2.65g/cm 3 , mean diameter of 2.4µm, and a BET s urface area of 3.11m 2 /g. The oxygen to metal ( O/M ) ratio was determined to be UO 2.16 using ASTM equilibration method (C1430 07). In the next step, SiC and diamond powder particles were used as second phase to fabricate UO 2 SiC and UO 2 diamond

PAGE 42

42 composite r espectively . The SiC powder was obtained from Alfa Aesar Inc, Ward Hill, MA, and had a mean particle diameter of 1µm. The diamond p owder was obtained from Advanced Abrasives, Pennsauken, NJ and had a mean diameter of Figure 3 1 (a) (c) reveals SEM imag es of as received UO 2 , SiC and diamond powders. A 5 vol% SiC powder or 5 vol% diamond powder was blended with UO 2 powder in a SPEX 8000 shaker for 1 hour , respectively . The SPEX jar is of 38 mm (1.5 inch) internal diameter and 45.7 mm (1.8 inch) length, a nd is made of zirconium oxide. To disperse the second phase particles, a blending aid, 2,3 Dihydroperfluoropentane, which acted as a lubricant for powder particles [ 70 ] , was used. The blending aid evaporated in air after the mixing process, and no residual contamination was left in the final mixture [ 11 ] . Spark Plasma S intering A Dr. Sinter® SPS 1030 system was used to perform the spark plasma sintering process. For SPS, 4g of the mixture powder was loaded into a graphite die with 12.7 mm i nner diameter, and then graphite punches were inserted into both ends of the die. The assembly was placed into the SPS chamber, and a vacuum of 10 Pa was maintained in the chamber during the sintering process. An axial pressure of 36 MPa was applied on th e punches and sintering was carried out at prescribed conditions. T he temperature was measured by a pyrometer focused on the surface of the graphite die. The programmed heating schedule initiated at 600 o C, as that is the lower detection limit of the pyrome ter. More details of these method can be found in the work of Ge et al [ 16 ] . Two differ ent heating schedules were used: ( i ) the constant heating rate method and ( ii ) the isothermal heating method. The constant heating rate method was used to const ruct the master sintering curve. T hree heating rates of 50 o C /min, 100 o C /min and

PAGE 43

43 150 o C /min were used to reach the maximum sintering temperature of 1200 o C for pure UO 2 and 1400 o C for UO 2 based composite. The current was then turned off , and the pellets were naturally cooled down in the vacuum chamber. If the MSC can be constructed, the isothermal heating method was employed to validate the constructed master sintering curve. The p owders were sintered by heatin g at a rate of 75 o C /min up to a maximum temperature of 750 o C to 1050 o C and then hold for 5 minutes at the prescribed maximum sintering temperature . Details on selection of maximum sintering temperature will be discussed later. The heating profiles for UO 2 powder under the two methods are plotted in Figure 3 2, which clearly illustrate the difference between these two sintering method. S hrinkage Data P rocess For the development of time dependent relative density curves, actual shrinkage data of the given pow der compact was required. In SPS, the axial displacement data of the bottom punch during sintering of the powder can be recorded. However, in order to take into account the thermal expansion of punches and the die blocks [ 71 73 ] , the following procedures were adopted for each sample. Initially, the powder was sintered in the graphite die and the axial displacement of the bottom punch was recorded. This displacement data was a combination of powder densification and the total expansion of graphite punches and die at high temperatures. To separate the graphite expansion, the entire die assembly was cooled down to room temperature and reheated to the same temperature wi th the sintered compact inside the die. As the length of sintered pellet is only about 3 mm, 30 times less than the length of punches, the thermal expansion of sintered pellets can be assumed negligible. Therefore, the displacement recorded in this second cycle corresponds only to the

PAGE 44

44 thermal expansion of the graphite punches and die. The length of the sintered sample was measured after the first and second run and found to be the same, indicating that no further densification of the sintered sample occurre d during the second run. Finally, the powder shrinkage data were obtained by subtracting the thermal expansion of the graphite during the second run from the original displacement data in the first run. Typical raw data for UO 2 and the graphite are shown i n Figure 3 2. In terms of the die displacement, positive values indicate shrinkage while negative values indicate expansion. The powder shrinkage curve (solid line) was obtained by subtracting the graphite expansion curve (dash line) from the original tota l specimen plus graphite shrinkage curve (dash dot line). With regard to the convention of shrinkage evolution being considered positive, it is extremely common in SPS literature to depict powder shrinkage as a positive quantity [ 16 , 48 ] . However, most of these papers do not report the die expansion (which is opposite in sign for shrinkage) and hence no confusion was ge nerated. However, the recent results by other authors [ 71 , 73 ] which considered die expansion have followed the norm of powder shrinkage as positive and reported the die expansion as a negative quantity. Thus, the adopted sign convention confirms to the established norm. The time dependent relative density data was then calculated based on the following equatio ns: ( 3 4) ( 3 5) Where is the instantaneous relative density, is the final density which is measured by the Archimedes method, is the final length of the sintered pellet, is the

PAGE 45

45 instantaneous length, and is the displacement difference between final state and instantaneous state, i.e., actual UO 2 shrinkage curve which can be read from Figure 3 2. Results and D iscussion Construction of M SC The time dependent relative densities of the sintered powder versus temperature for different heating rates for UO 2 , UO 2 SiC composite and UO 2 diamond composite are plotted in Figure 3 4 ( a) to 3 4 ( c) respectively. The temperature started from 600 o C, at whi ch point no densification has taken place. It can be seen that for all powder systems, the relative density at any temperature showed a modest dependence on the heating rate for a given powder. However, the shapes of densification curves are different. For pure UO 2 and UO 2 SiC composite, all densification curves have regular sigmoidal shape, while for UO 2 diamond composite, a plateau can be seen when the density is about 85% TD. The reason why this plateau appeared will be discussed later. At the same heati ng rate, all powder systems started to densify at about the same temperature, but the densification rate s of UO 2 SiC and UO 2 diamond composites were much lower than that of UO 2 . For a constant heating rate of 50°C/min, 95% of the TD could be achieved at 11 00°C for sintering a pure UO 2 pellet, whereas a 1350°C temperature was required for sintering a UO 2 SiC pellet and 1400 o C was required for sintering a UO 2 diamond composite . This difference can be quantified by the MSC , if applicable , and will be discussed later. Ge et al. , [ 16 ] reported that the densification range for UO 2 (where rapid increase of densification occurred) was around 720 °C to 1000°C , whereas from Figure 3 3, obvious further densification could still be seen until 1 100°C. It is more appropriate to modify this range to 1100°C, as the shrinkage curve

PAGE 46

46 in the present study also accounts for punch and die expansion at high temperature, which was not considered in the previous study [ 16 ] . Thus, t he current study provided more accurate limits for the densification process. A typical microstructure for sintered UO 2 , UO 2 SiC and UO 2 diamond pellet at a rate of 150°C/min is shown in Figure. 3 5 ( a) to 3 5 ( c) respectively. L ow level s of porosity (95.5% of TD for UO 2 , 96.0% of TD for UO 2 SiC and 95.7% for UO 2 diamond) are seen in all images, showing that high density pellets can be achieved under current sintering conditions. It is instructive to note here that no more than 96.5% of TD is needed in the UO 2 pellet used in nuclear industry [ 74 ] . Therefore, the curves in Figure 3 4 can represent the entire densification ev olutions of the given powder and can be used to construct the MSC. Another important parameter for constructing the master sintering curve is the apparent activation energy Q in Eq. ( 3 2). As discussed earlier, for a given powder compact, with the appropr iate value of Q , the densification behavior becomes independent of thermal history, which means that the data with different heating rates can converge to a single curve. To accomplish this, various values of activation energ ies were chosen, and values were calculated from Eq. ( 3 2). The curve s of UO 2 with selected activation energy values for all heating profile s are shown in Figure 3 5. It is seen that the three curves do not converge when Q equals 50 KJ/mol and 250 KJ/mol, whereas they tends to almost collapse into a single curve for all three heating rates when Q is around 150 KJ/mol. To obtain more precise values of activation energy, detailed analys e s w ere performed on UO 2 , UO 2 SiC and UO 2 diamond powder system s for a range of Q valu es.

PAGE 47

47 To quantify the convergence, various statistical analysis methods and tools have been used by many researchers. Teng et al. , [ 75 ] performed a sigmoidal curve fit for the data and developed a Visual Basic macro in Microsoft Excel and automated the computation. Pouchly and Maca [ 76 ] stated an alternative criterion, the so called Mean Perpendicular Curve Distance (MPCD), and developed specific software for calculation. In our study, a polynomial curve fit was made to all the data point s by Origin , and the mean of residual squares of all points with respect to that line were calcu lated. Figure 3 6 reveals the mean residual squares with different assumed values of Q, from which the minimum is identified when Q =140 KJ/mol for pure UO 2 , Q =42 0 KJ/mol for UO 2 SiC composite and Q =480 KJ/mol for UO 2 diamond composite . Note that the mean residual squares of UO 2 diamond composite remains high compared to UO 2 and UO 2 SiC composite, and the results directly influence the construction of MSC. The curve s with calculated Q were plotted in Figure 3 8 . It is seen that for UO 2 and UO 2 SiC composite, despite a 3 fold difference in heating rates, individual curves have merged reasonably close to a single curve. While for UO 2 diamond composite, curve s with different heating rates cannot merge. Recall that in Figure 3 3(c), a plateau showed in the densification curves of UO 2 diamond composite. Therefore, it can be assumed that potential chemical reaction occurred during the sintering of this composite. As the MSC theory is based on the assumption that only one of the diffusion mechanisms dominates the sintering process, the violation of this assumption may lead to the failure of construction of MSC. The potential chemical reaction during spark plasma si ntering of UO 2 diamond composite will be studied in next chapter.

PAGE 48

48 Validation and A pplication of MSC As discussed before, the MSC cannot be applied to UO 2 diamond composite successfully. Therefore, in this and next section, the discussion towards the valid ation and application of MSC is only within UO 2 and UO 2 SiC composite. Before applying MSC to guide a manufacturing process , t he validity of the constructed curves should be verifie d with various experimental runs . To accomplish this, a new set of experime nts in which pellets were sintered by isothermal hold at maximum sintering temperature was conducted. W ith a heating rate of 75°C/min, which was chosen randomly, pellets were heated up to a selected temperature and h e ld for 5 minutes. Compared with the con stant heating rate method used previously, lower maximum sintering temperatures were used here to guarantee that the powders were in the sintering range during the 5 minute hold time. According to the densification curves in Figure 3 3, the densification r ange (where rapid increase of densification occurred) was around 720 °C to 1100°C for UO 2 , and 750°C to 1350°C for UO 2 SiC composite. Therefore, in the isothermal sintering method, 750°C , 800 °C and 850 °C were selected as the maximum sintering temperatures for UO 2 , and 900 °C, 1000°C and 1050°C were selected for UO 2 SiC composite. After sintering, the final densities were measured by the Archimedes method. The results are shown in Figure 3 8 , where the relative densities of pellets made in this method are no w plotted on the MSC obtained from the previous method. It can be seen that the relative densities of the pellets made using these new conditions fall right on the constructed MSC s , confirming the validity of the MSCs derived from the earlier experiments . Typical microstructures of the selected pellets produced under isothermal conditions are shown in Fig ure s. 3 9(a) and 3 9(b). Note that significantly lower

PAGE 49

49 temperatures were used in this method, and hence the densities obtained in these pellets were lower than what were obtaine d previously. Compared with Figure 3 4, it is clearly seen that the levels of porosities, especially the inter granular porosities, are much higher. Discussion The constructed MSC can be a useful tool to model and control the spark p lasma sintering process. First, as any given heating profile corresponds to a specific value on the MSC, the final density is predicted. Second, if a desired density is given, the way to design various heating profiles for this objective is achievable. An example of designing the heating profiles which can be utilized to achieve a UO 2 pellet with 90% of the TD is discussed here. With the goal of 90% of the TD, the corresponding value of work of sintering, , can be easil y found to be 16.6 through Figure 3 7. Then, a combination of heating rate, maximum sintering temperature and hold ti me can be determined by Eq ( 3 2), to achieve the required . The data corresponding to these heating profiles are shown in Table 3 1. In addition to the ability to predict processing conditions for a given theoretical density, the presence of the rapid grain growth regime can also be detect ed by the MSC. Note that in Figure 3 7, the curves of sintering UO 2 with different heating rates do not converge well in the high density range (>95% of the TD). Recall that MSC is valid only when one sintering mechanism dominates the process. It is reasonable to conclude that when the density exceeds 95%, rapid grain growth occurs during sintering, and the dominant diffusion mechanism is d ifferent. Ge et al. , [ 20 ] studied the microstructure evolution of UO 2 during SPS process, and concluded t hat the grain size remains nearly the same (0.4 µm to 1 µm) in the medium density range (70% TD 95% TD), while it

PAGE 50

50 increased rapidly (3 µm to 9 µm) in the high density range (96% TD 97% TD), which is consistent with the statement here. On the contrary, the curves for UO 2 SiC composites fit well in the entire density range, which shows that no exaggerated grain growth occurred in UO 2 matrix due to the pinning effect of SiC particles. This result was also consistent with the work of Yeo et al. , [ 11 ] , where very small grain sizes were observed even in a pellet with 97.8% of the TD of UO 2 SiC pellet. The MSC is unique for a given powder, as clearly seen in Figure 3 7. With the addition of 5 vol% SiC particles, the apparent activation energy increases dramatically from 140 KJ/mol to 420 KJ/mol, and the MSC is drastical ly different. The MSCs for both UO 2 and UO 2 SiC give a quantitative description on the sinterability and densification of these two materials. Table 2 shows an example of the density results of UO 2 and UO 2 SiC pellets sintered at 850°C and held for 2 minut es. Under the same heating profile, the extent of densification of different materials can be easily compared. Due the presence of the SiC particles, the grain boundary diffusion path of UO 2 grains was diminished and thus led to the poor densification of t he composite. Similar behavior was also seen in Si 3 N 4 based ceramics as well [ 77 ] . In other words, compared to UO 2 , more ene rgy, either higher sintering temperature or longer hold time, was required for sintering the UO 2 SiC composite to the desired high density. Thus, MSC provides a quantitative analysis to characterize the effect of second phase particles on the densification during the sintering process. Finally, a brief discussion on the apparent activation energy Q of UO 2 by SPS is warranted. As master sintering curve is a model that describes the whole densification process, the Q value calculated in this study can be seen as an average value for the

PAGE 51

51 entire temperature range. It is noted that compared with the Q value calculated for conventional sintering methods (299 KJ/mol 429 KJ/mol) [ 78 ] , the a ctivation energy during SPS (140 KJ/mol) is significantly lower, which could possibly be attributed to the presence of coupled direct current field and pressure field in SPS. The effect of the current field has been noted and discussed most recently by man y researchers. Cologna et al. , [ 79 ] proposed that an avalanche of Frenkel point defects, both anions and cations, can be nucleated under an electric field, and this phenomenon has been observ ed in ceramics, such as yttria [ 80 ] and titania [ 81 ] . The aggregation of these point defects accelerates the diffusional transport rate and lowers the activation energy. Second, the pressure is also a significant factor that distinguishes SPS from conventional sintering methods. The presence of pressure could increase the number of grain grain contacts and change the grain shape [ 82 ] . This effect has also been o bserved in SPS by Ge et al. , [ 20 ] during their studies on densification and grain growth. All of these ef fects could lead to a lower activation energy. However, it is still difficult to quantify the effect of current field and pressure on activation energies. Raj et al. [ 83 ] stated that the current field, as compared with the applied pressure, cannot significantly influence activation energy for diffusion during ceramic sintering in SPS. The above studies clearly indicate t hat further study is required on the role of pressure and electric field in SPS. Conclusion MSC theory has been applied to UO 2 and UO 2 based composites s intered by SPS. Due to potential chemical reactions, the master sintering curve of UO 2 diamond composit e cannot be constructed, while the MSCs of UO 2 and UO 2 SiC composite were constructed sufccessfully. Utilizing just one temperature and time dependent parameter

PAGE 52

52 ( ), the densification process wa s quantified . The activation energy for SPS has been determined to be 140 KJ/mol for UO 2 and 420 KJ/mol for UO 2 SiC composite. The master sintering curve is proven to be an effective tool to describe the densification beha vi or of UO 2 and UO 2 SiC powder compact during SPS . The ability of MSC to control and predict density has been demonstrated by a few experimental runs using the isothermal method. In addition, it is shown that the MSC can be used to compare the sinterability of different powder compacts in SPS .

PAGE 53

53 Table 3 1. Parameters of the heating profiles to sinter a UO 2 pellet with 90% of the TD in SPS Goal: 90% of the TD Sintering profile Heating rate (°C/min) Maximum sintering temperature (°C) Hold time (min) 1 75 8 00 6.0 2 75 850 2.6 3 125 850 3.0 Table 3 2. D ensity comparison of UO 2 and UO 2 SiC pellets under the same heating profile in SPS Heating rate (°C/min) Maximum sintering temperature (°C) Hold time (min) Predicted density (% of TD) Experimental density (% of TD) UO 2 100 850 2 88.8 89.8 +/ 0.4 UO 2 SiC 100 850 2 69.0 70.8 +/ 1.0

PAGE 54

54 A ) B ) C ) Figure 3 1. SEM images of starting powders. A ) UO 2 powder . B ) SiC powder. C ) D iamond powder . Photo courtesy of author .

PAGE 55

55 Figure 3 2 . Heating profiles for UO 2 pellets under the constant heating rate method and isothermal method

PAGE 56

56 F igure 3 3 . Shrinkage curves of UO 2 and the graphite during SPS. The actual shrinkage curve for UO 2 is obtained by subtracting gra phite expansion from total shrinkage curve.

PAGE 57

57 A ) B ) C ) F igure 3 4 . Evolution of relative density as a function of temperature for different heating rates. A ) UO 2 , B ) UO 2 SiC co mposite . C ) UO 2 diamond composite. Note that a plateau appeared in densification curves of UO 2 diamond composite.

PAGE 58

58 A ) B ) C ) F igure 3 5 . SEM images of fracture surfaces. A ) UO 2 si ntered to 1200°C at 150°C/min, B ) UO 2 SiC sinte red to 1400°C at 150°C/min C ) UO 2 diamond sintered to 1400°C at 150°C/min . Photo courtesy of author .

PAGE 59

59 Figure 3 curve s for selected value of activation energy of UO 2 . Note that for Q=150 KJ/mol, the relative density values tend to collapse into a single curve for all three heating rates. F igure 3 7. Mean residual squa res for various estimated activation energy values

PAGE 60

60 A) B) F igure 3 8. Master sintering curve construction. A ) P ure UO 2 and UO 2 SiC composite. B ) UO 2 diamond composite

PAGE 61

61 F igure 3 9. Comparison of MSC predicted relative density and experimental value of UO 2 and UO 2 SiC pellets using isothermal heating method

PAGE 62

62 A ) B ) F igure 3 10. SEM images of pellets sintered by isothermal heating method. A ) UO 2 isother mal at 800°C for 5 minutes . B ) UO 2 SiC isothermal at 1000°C for 5 minutes . Photo courtesy of author .

PAGE 63

63 CHAPTER 4 MICROSTRUCTURE AND PROPERTY OF DIAMOND REINFORCED URANIUM DIOXIDE COMPOSITE FUEL PELLETS Background In C hapter 3 , UO 2 diamond composite was fabricated using SPS and the densification evo lution was studied. However, the microstructure and properties of UO 2 diamond composite pellet has never been investigated. Compared to other second phase materials such as silicon carbide or carbon nanotube, whether diamond would be a be tter candidate is remain unknown. Therefore, t he objective of this work is to fabricate UO 2 diamond fuel pellets using SPS using different diamond particles and investigate the microstructure and properties of this novel composite fuel. Experimental Powde r P reparation Figures 4 1 and 4 2 show SEM images of the as received UO 2 and diamond powders, respectively. The uranium dioxide powder was obtained from AREVA Federal Services, Hanford, Washington, USA. The powder was reported to have a bulk density of 2.3 g/cm 3 , mean particle diameter of 2.4µm and a BET surface area of 3.11m 2 /g. The grain size was dete rmined to be 200 400 nm, see Figures 4 1. The O/U ratio was determined to be 2.11 using ASTM equilibration method (C1430 07). The diamond powder was obtained f rom Advanced Abrasives, Pennsauken, NJ. Four sizes of diamond particles with a mean particle size of 0.25 µm, 3 µm, 12 µm and 25 µm, were used in this study. A SPEX 8000 shaker was used to blend UO 2 and a 5vol% diamond powder for 1 hour with a blending ai d 2,3 Dihydroperfluoropentane. The use of a blending aid was proven to be effective and non contaminative to the final powder , see Chapter 3 .

PAGE 64

64 S park Plasma S intering A graphite die with inner diameter of 12.7 mm and outer diameter of 31.5 mm was used to si nter the mixed powder. A thin graphite foil was inserted into a graphite die and then 4g of UO 2 +diamond mixture was poured into the die. Graphite punches were used on either side of the die to hold the powder. The graphite foil prevents reaction between th e powders and the die. A Dr. Sinter ® SPS 1030 spark plasma system (SPS) machine was used to sinter the blended powder. The entire die assembly with powder and punches was placed in the SPS chamber. During sintering, a vacuum of less than 10 Pa was maintain ed in the chamber and the temperature was measured using a pyrometer that focused on the surface of the graphite die. A temperature ramp rate of 100 o C/min was employed and the maximum sintering temperature was varied from 1300 o C to 1600 o C with a hold time of 5 minutes for various powder mixtures. An axial pressure of 36 MPa was applied when the maximum sintering temperature was reached, and the pressure was released when the cooling process started. More detailed description of this method can be found in r ecent publications [ 11 , 12 , 16 , 84 ] . It is important to note that although the stoichiometry of the starting uranium dioxide powder was 2.11, the O/U ratio has changed to be 2.00 due to the reducing environment in SPS. Similar result was first observed in fabricating pure UO 2 p ellets using SPS [ 20 ] where a detailed analysis regarding the change in stoichiometry in UO 2 during SPS can be found. Thus, SPS eliminates additional post sintering steps commonly used in conventional sintering method of UO 2 pellets. Characterization M ethods Upon cooling to room temperature, the pellets were removed from the graphite dies. The density of eac h sintered pellet was measured by the Archimedes method and

PAGE 65

65 converted to relative density. The theoretical density of the composite can be calculated from the rule of mixture, i.e., ( 4 1) where is the theoretical density of the composite, , and are density of diamond, volume fraction of diamond, and density of UO 2 , respectively. The density and relevant values of diamond and UO 2 are listed in Table 4 1. X ray diffractio n (XRD) measurements were conducted to investigate the chemical reactions which may have occurred between UO 2 and diamond during the sintering process. As the detection limit of XRD is around 5wt% of chemical compounds and the short processing time of 5 mi nutes may not be enough to form reaction compounds at sufficient quantity, the volume fraction of diamond was increased to 70% in the composite for this specific study only. The mixing process and sintering procedures were the same as the other pellets. X Ray diffraction patterns were ® angle was scanned from 20 o to 80 o with a step size of 0.005 o . The software X'Pert HighScore Plus® was used for peak fitting. To characterize the microstructure of the sintered pellets, scanning electron microscopy (JEOF 6335L and FEI XL 40 FEG SEM) was used. The secondary electron mode was chosen for the images and the accelerating voltage was varied between 10 to 15 KV. The thermal conductivi ty of the sintered composite pellets we re calculated using the following equation ( 4 2)

PAGE 66

66 where is the thermal diffus ivity, is the specific heat and is the density of the composite. was measured at three different temperatures, i.e., 100 o C , 500 o C and 900 o C, using a laser flash method (Ant er Flashline ® 3000). Specific heat was calculated using the Neumann Kopp rule [ 85 ] : ( 4 3) where and are the specific heat of diamond and UO 2 , respectively, and is the weight fraction of diamond. determined using the ultrasonic measu rement method, an accurate and non destructive technique. Using a pulser/receiver system (Model 5072PR, Olympus, Waltham, MA), both longitudinal velocity and shear velocity of the pellets were measured. d by the following equation [ 86 ] : ( 4 4) ( 4 5) where is the density of the pellet. Four measurements were performed on each composite and the average value was calculated. Results and D iscussion Density D ensity has a huge influence on the thermal and mechanical properties of the composites. Therefore it is important to obtain high density composite pellets. Figure 4 3 shows the variation in the relative density of the UO 2 diamond composites as a

PAGE 67

67 function o f sintering temperature and diamond particle size. A horizontal line at 96.5% of the theoretical density (TD), which is the requirement for a fuel pellet to be used in a nuclear reactor [ 87 ] , is plotted as the reference density. For all diamond particle sizes, the densi ty of the sintered pellets increased with increasing sintering temperature up to 1500 o C . While when the maximum sintering temperature achieve 1600 o C, a slightly decrease in density can be observed. The decrease may be due to the formation of micro cracks a nd the chemical reaction at high temperature, which will be discussed in the following sections. For nano size diamond particles (0.25 µm), the pellet density is consistently lower than other composites sintered under the same sintering conditions. A pelle t with a 96.5% of the TD cannot be achieved even at a temperature of 1600 o C. While for micro size diamond particles (3 can be achieved when the maximum sintering temperature is 1400 o C or above. The poor densification behavior of UO 2 0.25 µm diamond composite could be explained by the microstructure of the sintered pellet, which w ill be discussed later. Chemical Reaction Understanding the potential chemical reactions between diamond and UO 2 during SPS is critical to an evaluation of the performance potential of the composite pellet in the reactor environment. The chemical reaction products, (if any) may badly influence the thermal and mechanical performance of the sintered pellets. Figure 4 4 shows the XRD spectra of UO 2 and diamond powder as well as the UO 2 Diamond composite at four different temperatures. For the spectra of starti ng powders, all characteristic peaks of UO 2 (Powder Diffraction File (PDF) 00 005 0550) and diamond (PDF 01 075 0223) were clearly detected, revealing the cubic structure of UO 2 and diamond. The crystal planes were identified and marked in the plot. For th e UO 2 70

PAGE 68

68 vol% diamond composite sintered at 1300 o C, no other peaks, except UO 2 and diamond peaks, were observed. However, when the maximum sintering temperature was increased to 1400 o C, a clear graphite peak was detected. The formation of graphite may be a n indication of the graphitization of diamond. Although it is known that the graphitization rate of diamond is extremely slow below 1600 o C in an inert gas environment [ 88 ] , this temperature could decrease significantly if high pressure is applied. During SPS, a pressure of 36 MPa was applied at the maximum sintering temperature, which could facilitate the graphitization process. When the maximum sintering temperature reached 1500 o C or higher, not only graphite, but also uranium carbide (UC) peaks were observed. The following reactions between UO 2 and carbon are expected [ 89 ] : ( 4 6) Note that there is CO gas formed in these reactions. Recall that slightly decrease in density was observed in pellets sintered at 1600 o C, which may be due to the formation of gas. During the hold time at such high temperature, the sintering process was almost done, but some micro holes were left because of the formation of CO. The increase of porosity thus led to a decrease in final density. Clearly, higher sintering temperatures are not preferable due to the formation of graphi te and other reaction products. Sintering temperatures well below 1500 o C are recommended for fabrication of UO 2 diamond composite pellets.

PAGE 69

69 Microstructure Dispersion of diamond particles The dispersion of diamond particles in the composite pellets sintered at 1400 o C can be seen in Figure 4 5. The diamond particles with mean size of 3 were seen to be relatively uniformly distributed in the UO 2 matrix from Fig ure s. 4 5(b) to 4 5(d), while the agglomeration of 0.25 µm nano size diamond parti cles was observ ed in Figure 4 5(a). The agglomeration could be due to the poor mixing. Longer mixing time (6 hour) was also attempted but no obvious improvement was observed. Therefore, the current milling process is proven not to be an appropriate approac h for mixing nano sized diamond in UO 2 powder. The agglomeration of diamond particles may be a reason for the poo r densification observed in Figure 4 3 in UO 2 diamond composite with a diamond particle size of 0.25 µm. In the region of diamond agglomeratio n, UO 2 diamond interfaces were replaced instead by diamond diamond interfaces. Because the diamond is the most covalently bonded material, the sintering of diamond particles is much more difficult. As a result, the pellets did not sinter fully and loose di amond powder regions were seen in the microstructure of UO 2 0.25 µm composite pellet, see Fig ure 4 5(a). This led to a poor measured density in the sintered nano diamond composite pellets as well. UO2 diamond interface The UO 2 diamond interfaces in various UO 2 diamond composites are shown in Figure 4 6. The diamond particles had good interfacial contact with the UO 2 matrix in pellets containing 0.25 µm , 3 cracks ond particles. The formation of micro cracks were due to a mismatch in coefficients of thermal expansion (CTE) between the

PAGE 70

70 UO 2 and the diamond particles. It is noted that the CTE of UO 2 and diamond is 9.93 × 10 6 K 1 and 1 × 10 6 K 1 at room temperature, respe ctively. Although the CTE would slightly change with temperature, a large difference in CTE between UO 2 and diamond existed at all temperatures [ 90 , 91 ] . Due to this mismatch, diamond particles were subjected to a large compressive stress during the cooling process. The larger the particle size was, the greater the c ompressive stress would be. Micro cracks are initiated from the interface when the tensile stress in UO 2 exceeds a critical value in the UO 2 matrix. Similar phenomena were also seen in UO 2 SiC composites where such CTEs mismatch existed between matrix and second phase particles [ 12 ] . The induced micro cracks adversely infl uence the thermal and mechanical properties of the composite pellets, which is discussed next. Grain size Figure 4 7 reveals the polished and thermally etched surfaces of UO 2 diamond composite pellets sintered at 1400 o C with a hold time of 5 minutes. Unifo rm grain sizes were revealed in pellets w ith 3 µm diamond (2.6 µm, Figure 4 7(d)), 12 µm diamond (2.9 µm, Figure 4 7(e )) and 25 µm diamond (3.1 µm, Figure 4 7(f)), while an obvious non uniform grain size distribution was observed in the UO 2 0.25 µm diamond pellet (Figure 4 7(a) 4 7(c)). Typical UO 2 grain sizes were no larger than 3.5 µm in all these microstructures. However, in the work of Ge et al [ 20 ] , it was shown that the grain size of UO 2 pellets sintered by SPS can be as large as 8.9 µm, much larger than the current UO 2 grain sizes in UO 2 diamond composites. The results are reasonable when the pinning e ffect of second phase particles are considered. During sintering, the UO 2 grains start to grow and once grain boundary reaches a diamond particle, the grain boundary migration was pinned by that particle. Therefore, smaller grains were formed

PAGE 71

71 in the UO 2 matrix. Since the volume fraction of diamond is the same in all composites, a higher surface to volume ratio of diamond particles exists in composite containing smaller diamond particles. Therefore, the intensity of pinning effect would be more in comp osite with 0.25 µm diamond. It can be seen that within these four UO 2 diamond composites, the pellet with 25 µm diamond particles has the largest grain size, which is in agreement with the above argument. The formation of th e non uniform grain size in Fig ure 4 7(a) can be owing to the uneven distribution of nano diamond particles. Recall that in Fig ure 4 5(a), the agglomeration of diamond particles was observed. Therefore, the region A which lacks diamond particles has much larger grain size (2.5 µm) than the region B (0.7 µm) which has well distributed diamond particles. Thermal C onductivity As discussed in the previous section (Eq. ( 4 2)), the thermal conductivity is a product of thermal diffusivity, specific heat and pellet density. The specific heat of UO 2 5 vol% diamond composite was calculated, using Eq . ( 4 3) and the values listed in Table 4 1, to be 266.3, 327.5 and 339.7 J/kg·K at 100 o C, 500 o C and 900 o C, respectively, and then the thermal conductivity was calculated. Figure 4 8 shows the thermal c onductivity for UO 2 diamond composites sintered at 1400 o C with various sizes of diamond particles. The dashed lines represent the thermal conductivity of pure UO 2 in the literature at the specified temperatures. Compared to pure UO 2 , an increase in thermal conductivity was thermal conductivity than pure UO 2 . The maximum increase in thermal conduc tivity was

PAGE 72

72 38.3% and 34.2% at 100 o C, 500 o C and 900 o C, respectively. For particle reinforced composites, several theoretical models were utilized for estimating the effective thermal conductivity [ 44 , 92 ] [ 44 ] , which has been proven to be effective by many researchers, is given by the following equation: ( 4 7) where is the thermal conductivity, subscripts c , d and m are composite, particle and matrix, respectively, V d is the volume fraction of particles, a is the radius of particle, and h is the interfacial thermal conductance. If the volume fraction of particles is fixed, it is easy to prove that the thermal conductivity of the composite is a monotonically increasing functio n of . Theoretically, the interfacial thermal conductance between matrix and second phase particles is related to the density and phonon velocity of the two materials [ 93 ] . Therefore, the interfacial thermal conductance between UO 2 and diamond is assumed to be a constant. The thermal conductivity, then, would increase with the increase i n diamond particle size based on this model. However, the experimental value s in Figure 4 2 pellet. This result is because of t he imperfections in the microstructure. Recall from Figure 4 6(c) and 4 6(d) that at the interfaces between the UO 2 and diamond of sintered pellets, micro cracks were observed in UO 2 2 composite. These micro cracks have an adverse influence on the thermal conductivity.

PAGE 73

73 result in even lower thermal conductivity values. A comparison of the thermal conductivity of the UO 2 diamond composite and oth er UO 2 based composites publishe d in literature is shown in Figure 4 9. The density of these pellets were not normalized to the same value because a density of 96% 98% was achieved in all the pellets discussed. Among these composites, UO 2 diamond was foun d to have the highest thermal conductivity at the same measured temperature. odulus Although an increase in the thermal conductivity is the primary motive for fabricating UO 2 diamond composite pellets, the mechanical property of the pellets is al 2 diamod composites, therefore, is measured and compared with pure UO 2 pellets. Figure 4 2 5 vol% diamond composite pellets with different diamond particle sizes. All pellets were sintered at 1400 o C with a hold time of 5 minutes. The values of pure UO 2 pellets from different publications are also plotted for comparison. Note that Gatt et al [ 94 ] , and Padel and Noyion [ 95 ] fabricated the pellets using conventional sintering method, while Ge et al [ 16 ] used spark plasma sintering (SPS) technique to sinter UO 2 pellets. Due to the good agreement between their results, it has been concluded that the sintering technique has 2 pelle ts. The UO 2 diamond composites are 2 due to the extremely high theoretically have the same value if the diamond conce ntration is the same. However, it is seen in the figure that only pellets with the 3 µm diamond particles revealed a higher

PAGE 74

74 2 pellets, while the pellets with 0.25 µm, 12 µm and 25 µm diamond showed a much lower value. It is highly sensitive to the micro cracks, therefore, the results are consistent with the observation of micro crack presence in those pellets. For UO 2 0.25 µm diamond pellet, no c racks were revealed by SEM (Figure 4 6(a)). The composite may be due to its low density compared to other pellets. The variation in nductivity results shown in Figure 4 8. Clearly, both thermal and mechanical properties are s trongly influenced by the formation of micro cracks, and therefore the significance of avoiding them is self evident. From the above experimental and characterization results, it is seen that 3 µm diamond particles provide optimal density, microstructure, thermal conductivity, and mechanical property. Such composites are candidate materials for accident tolerant fuel because those pellets can operate at lower temperatures, reduced fission gas release and can dissipate residual heat more quickly in a loss of coolant accident (LOCA) in nuclear reactor. For the design of UO 2 diamond composite pellet, the content of diamond is to be optimized in the future. It is a trade off because more diamond may theoretically give a better thermal conductivity but lower nucl ear reactivity due to the reduction in UO 2 content in the pellet. Therefore, more studies and calculations are still needed. In addition, our future research will include investigation of mechanical integrity and chemical stability of these composite pelle ts at long duration high temperature exposure and eventually testing their effectiveness in advanced test reactor at Idaho National Lab.

PAGE 75

75 Conclusion The concept of high thermal conductivity UO 2 diamond composite fuel pellets has been demonstrated. Using the spark plasma sintering technique, UO 2 5 vol% diamond composite fuel pellets were successfully sintered. A maximum sintering temperature of 1400 o C and a hold time of 5 minutes are recommended in order to achieve a high density pellet with minimal UO 2 diamo nd chemical reaction. Four diamond particle sizes ranging from 0.25 µm to 25 µm were investigated. However, pellets with 0.25 µm diamond particles were not acceptable due to the poor mixing and the resulting low densification. On the other hand, micro crac ks were observed when the diamond particle size was larger t thermal and mechanical properties of the composites. Therefore, a 3 µm diamond is recommended for the UO 2 diamond composite pellet in order to obtain the best thermal and mechanical performance. For the UO 2 d iamond composite pellets, the maximum increase in thermal conductivity was found to be 41.6%, 38.3% and 34.2% at 100 o C, 500 o C and 900 o C, respectively. These results are higher than those reported in literature for other UO 2 based composites. Therefore, di amond could probably be a more effective candidate for UO 2 composite fuel pellets.

PAGE 76

76 Table 4 1. Properties of UO 2 and diamond. Materials Density (g/cm 3 ) Specific heat, (J/Kg K) [ 96 , 97 ] 100 o C 500 o C 900 o C UO 2 10.96 258 305 314 Diamond 3.52 757 1687 1849 Fig ure 4 1. SEM image of as received UO 2 powder . Photo courtesy of author .

PAGE 77

77 A ) B ) C ) Fig ure 4 2. SEM images showing morphologies of diamond powde r . A) 0.25 µm, B) 3 µm, C) 12µm, and D ) 25 µm. Photo courtesy of author .

PAGE 78

78 D ) Figure 4 2. Continued. Figure 4 3. I nfluence of sintering temperature and diamond particle size on the density of UO 2 +5vol% diamond composite fuel

PAGE 79

79 Figure 4 4. XRD spectra of starting powder and UO 2 70 vol% diamond pellets sintered at various temperatures. Note that formation of UC and graphite phases is revealed at higher temperature.

PAGE 80

80 A ) B ) Figure 4 5. Fracture surfaces of UO 2 diamond pellets sintered at 1400 o C w ith different mean diamond particle sizes. A ) 0.25 µm , B ) 3 Note that aggl omeration of diamond particles is seen in A ). Photo courtesy of author .

PAGE 81

81 C ) D ) Figure 4 5. Continued.

PAGE 82

82 A ) B ) Figure 4 6. UO 2 diamond Interface in composite pellets sintered at 1400 o C with different mean diamond particle size s. A ) 0.25 µm , B ) 3 Photo courtesy of author .

PAGE 83

83 C ) D ) Figure 4 6. Continued.

PAGE 84

84 A ) B ) C ) Figure 4 7 . Polished and thermally etched surfaces of UO 2 diamond composites wi th different mean diamond particle size s. A ) 0.25 µm , D ) 3 B) and C ) are enlarged pictures of areas in A ).The grain sizes are indicated on the top left of each image. Photo courtesy of author .

PAGE 85

85 D ) E ) F ) Figure 4 7 . Continued.

PAGE 86

86 Figure 4 8 . Thermal conductivity UO 2 diamond pellets with various sizes of diamond particles. The dotted lines refer to UO 2 literature values [ 96 ] at each temperature.

PAGE 87

87 Figure 4 9 . Comparison of thermal conductivity of UO 2 and various UO 2 based composites

PAGE 88

88 Figure 4 2 diamond composites as a function of mean pure UO 2 pellets reported from literature.

PAGE 89

89 CHAPTER 5 MICRO RAMAN STUDY OF PHASE TRANSFORMATIO N AND RESIDUAL STRESS IN UO 2 DIAMOND COMPOSITE FUEL Background The development of enhanced thermal conductivity and accident tolerant uranium oxide (UO 2 ) fuel is a major priority in the nuclear industry. In an effo r t to increase the thermal conductivity o f UO 2 , the concept of fabricating UO 2 based composites by adding a high thermal conductivity second phase, has been investigated in recent years. Many materials such as silicon carbide (SiC) [ 11 ] , carbon nanotubes (CNT) [ 13 ] and diamond have been investigated and high density composites pellets have been successfully fabricated utilizing a new sintering technique called spark plasma sintering (SPS). Among these materials, UO 2 diamond composite pellets are of particular interest best due to the extremely high thermal conductivity of diamond. Nevertheless, changes in the state of diamond particles and their properties during the SPS processing of UO 2 d iamond composites remains a concern. It is known that diamond is a metastable allotrope of carbon and it has a propensity to convert to graphite even under normal atmospheric conditions [ 98 ] . Although the conversion rate is negligible under room temperature, the rate is expected to increase by orders of magnitude under high temperatures. In order to fulfill the density requirement of a nuclear fuel pellet (at least 96.5% of the theoretical density), a temperature of 1400 o C or above is required for processing of UO 2 diamond composite pellets using SPS [ 99 ] . Under such high sintering temperatures, graphitization of diamond may be possible. The fo rmation of graphite has a huge influence on the thermal and mechanical properties of the

PAGE 90

90 composite fuel and therefore, the understanding of graphitization of diamond in UO 2 diamond composites is of considerable interest . Due to superior thermal properties, diamond and diamond related materials have been used in many high temperature devices such as light emitters, high temperature electronics and micro electromechinal systems (MEMS) [ 100 ] . In operation of a nuclear reactor, fuel pellets are subjected to a high temperature up to 1600 o C [ 101 ] . The high temperature stability of diamond is an important consideration in all these app lications . Another factor that influence properties and performance of any composite material is residual stress in its constituent phases. As a composite, the thermal expansion mismatch between the matrix and the second phase particles can always lead to the formation of residual stress. In UO 2 diamond composite fuel pellets, due to the large difference in the coefficient of thermal expansion (CTE) between UO 2 ( 9.9 .1 × 10 6 K 1 ) and diamond ( 1.1 × 10 6 K 1 ) [ 90 , 91 ] , the diamond particles are subjected to a large compressive stress upon cooling from high processing temperature during the SPS . The residual stress may cause interfacial debonding and micro cracks which negatively impact the properties of the composit e during the reactor operation and throughout spent fuel waste disposal process [ 102 ] . Therefore, determination of the processing induced residual stress in diamond particles is a necessary first step in assessing the implementation of this new composite fuel. Micro Raman spectroscopy (MRS) is a powerful technique for chemical and mechanical analysis for inorganic materials. Utilizing MRS, phase transformation [ 103 105 ] and residual stresses [ 104 , 106 108 ] have been effectively studied in various ceramics. Gogotsi et al. , [ 104 ] studied phase transformation and residual stress an d

PAGE 91

91 silicon thin films. Jannotti et al. , [ 105 , 107 ] characterized the residual stress and compositional chan ge in B 4 C ceramic s . Ghosh et al. , [ 103 ] observed the amorphization of B 4 C during dynamic indentation by the appearance of amorphou s carbon peaks in the Raman spectrum . DiGregorio et al. , [ 106 ] measured the residual stress in Si C whiskers within an alu mina matrix. Shafiq and Subhash [ 108 ] studied surface biaxial stress in ZrB 2 SiC composites using MRS and verified the accuracy of MRS results using digital image correlation (DIC) technique. Diamond, which is also a Raman active material, can be characterized through MRS as well. The residual stress in diamond particles can be characterized based on the shift of diamond peak in a Raman spectrum [ 109 ] , and the formation of graphite can be characterized by the appearance of characteristic graphite peaks [ 110 ] . Unlike other characterization technique s such as X ray diffraction (XRD), MRS can be used to analyze micro scale features due to its high spatial resolution. Therefore, the purpose of this study is to utilize MRS and analyze the phase transformation and residual stress of diamond particles within a UO 2 diamond composite pellet. As a first step we will derive a relationship between Raman peak shift and the induced residual stress in the particle. We will utilize this r elationship to measure the level of residual stress at several locations in a UO 2 diamond pellet. In the next step, we will investigate if diamond has undergone graphitization during high temperature processing of the composite by noting the appearance of graphitic peaks in the Raman spectrum.

PAGE 92

92 Derivation of Relationship b etween Raman P eak Shift and Residual Stress i n Diamond Particles It is known that the residual stress can be characterized using MRS by noting the shift in Raman peaks from the virgin state (stress free state) to its current state. If the peak shifts to the left of the virgin state in the Raman spectrum, a tensile stress is indicated; while if the peak shifts to the right, a compressive stress is indicated. The relationship between stress an d Raman peak shift can be expressed: ( 5 1) where is the residual stress, is a coefficient and is the difference in Raman peak position between virgin state and stressed state. The main objective here is to derive the value of for diamond particles in SPS. Recall that Raman scattering is a result of interaction between the incident light and the phonons in the sample. The frequencies of three optical phonon modes are the Eigen values of the following secular equation [ 111 ] . ( 5 2) where p , q and r are material constants, and are residual normal and shear strain ( 5 3) where S ij are the compliance constants.

PAGE 93

93 For diamond, which has cubic structure ( 5 4) For simplicity, one neglects all shear component of stress and strain, substituting ( 5 3) ( 5 4) into ( 5 2) and solving for the eigenvalue, ( 5 5) It is known that d iamond has a triple degenerate optical phonon with F 2g symmetry [ 112 ] , therefore the three eigenvalues should be the same. The values of p and q are not experimentally available in the literature, but they can be calculated by physical properties of diamond below. The mode Grüneisen parameter 0 and anharmonicity parameter s was defined and measured by the following equations [ 113 ] , ( 5 6) ( 5 7) where W 0 is the Raman peak position of stress free diamond. Solving ( 5 6) and ( 5 7) gives, ( 5 8) ( 5 9) The diamond compliance constants needed in Eq ( 5 5) are reported as [ 109 ] ( 5 10) ( 5 11)

PAGE 94

94 For small deformation, the relationship between the frequencies of deformed crystal and undeformed crystal can be expressed by the peak shift ( ), peak position of the stress free diamond ( ), and its Eigenvalue ( ) ( 5 12) Substituting ( 5 5),( 5 8) ( 5 11) to ( 5 12), ( 5 13) Assuming each diamond particle to be a sphere and subjected to a bi axial stress condition during SPS, , (5 14 ) Substituting (5 14) into (5 13 ), the relationship between residual stress and Raman peak shift can be expressed as ( 5 15 ) The negative represents that a peak shift to a higher wavenumber indicates compressive stress. Experimental Pellet F abrication The uranium dioxide powder obtained from AREVA Federal Services (Hanford, Washington, USA) had a grain size 100 400nm . The diamond powder was obtained from Advanced Abrasives (Pennsauken, NJ, USA) and had a mean particle size of 3 µm. Figures 5 1(a ) and 5 1(b) show SEM images of the as received UO 2 and diamond powders, respectively. A SPEX 8000 shaker was used to blend U O 2 and a 5 vol% diamond powder for 1 hour with a blending aid 2,3 Dihydroperfluoropentane. The use of a blending aid was

PAGE 95

95 proven to be effective and non contaminative to the final powder after dried in an oven at around 40 o C. The powder was placed in a grap hite die and Dr. Sinter ® SPS 1030 spark plasma sintering (SPS) system was used to sinter the blended powder at a maximum sintering temperature of 1400 o C for 5 minutes at a pressure of 36MPa. These sintering parameters were chosen because they were found to provide the best thermal and mechanical properties for UO 2 diamond composite pellets as discussed previously [ 114 ] . The resulting pellet has a diameter of 12.5 mm and a thickness of 3.5 mm, see Figure 5 2(a). A more detailed description of this method can be found elsewhere [ 11 , 12 , 16 , 84 ] . Raman S pectroscopy A Raman spectrometer (Renishaw Invia® model. Hoffman Estates, IL) with a 532 nm laser and spectral resolution as low as 0.38 cm 1 was used. In order to avoid the influence of laser heating and assure sample integrity, only 10% of the total laser power was used. A Si standard was used for calibration of peak position before testing the UO2 diamond sample. The pellet was cut vertically and the cross section of the pellet was poli shed by successive smaller grinding medium. Raman scanning was performed at five locations (A E) along the centerline of the pellet shown in Figure 5 2(b). Results and D iscussion The spectrum of as received pure diamond powder is shown in Figure 5 3. Only a sharp peak centered at around 1331.6+/ 0.3 cm 1 was observed between 900 cm 1 to 1700 cm 1 range. It is well known that diamond has a triple degenerate peak at around 1332 cm 1 and the result here is in good agreement with other literature s [ 115 , 116 ] .

PAGE 96

96 G raphitizat ion of Diamond within the P ellet An image of location A is shown in Figure 5 4(a). The white region is UO 2 matrix while dark regions are diamond particles. A Raman map scan was performed at this location and a schematic is shown in Figure 5 4(b). Raman spe ctra of all grids were obtained. Obviously some grids are on the UO 2 matrix and some grids are on the diamond particles. The spectra showing information of diamond particles were then collected in order to study the phase change of diamond. Similar experim ents were performed on other locations and around two hundred spectra which measured the surfaces of diamond particles were collected for each location. Three types of spectra are observed and the typical Raman spectra are shown in Figure 5 5(a) 5 5(c). On ly one peak is observed in Figure 5 5(a). The sharp peak at around 1332cm 1 is the characteristic peak of diamond as shown before. In Figure 5 5(b), however, not only diamond peak, but a peak at around 1580cm 1 is shown. This peak is a characteristic peak of graphite and is called G (graphite) peak [ 117 ] . In Figure 5 5(c), another peak at around 1350cm 1 was observed. It is also a characteristic peak of graphite which is called D (disordered graphite) peak [ 117 ] . The appearance of D peak and G peak clearly indicated that some diamond has been changed to graphite. In order to study the degree of graphitization, a quantitative analy sis was performed. Wang et al . , [ 109 ] studied t he relative intensities of diamond and graphite peaks in Raman spectroscopy and derived the following relationship ) ( 5 16 ) where C d is the diamond purity, A d is the area of the curve beneath the 1332cm 1 diamond peak and A i are the areas of peaks corresponding to the graphitic bands.

PAGE 97

97 Ferreira, et al ., [ 118 ] later used Equation (5 16 ) and studied the graphitization in diamond film grown on a silicon substrate. Utilizing the method above, th e degree of graphitization in UO 2 diamond composite can be calculated. All peaks were fitted by mixed Gaussian and Lorentzian functions and the peak areas were calculated using the software WiRE 3.4 supplied by the manufacturer of Renishaw spectrometer. T he degree of graphitization for each location along the centerline of the cross section was then determined by averaging C d of all spectra and the result is shown in Figure 5 6. At all locations within the pellet, more than 10% of the diamond has been tran sformed to graphite. However, a large difference in the degree of graphitization was observed between the surface and the interior of the pellet. On both surfaces, around 22% of the diamond was transformed to graphite, while in the interior of the pellet, only 12% 14% of diamond was transformed. The graphitization of diamond has been a subject of interest since 1960s. Davies and Evans [ 119 ] showed that under zer o pressure and below 1500 o C, the transformation of diamond to graphite was not detectable experimentally and the graphitization rate is relatively slow between 1500 o C 1960 o C. Below 1900 o C, only 5% of diamond was transformed after 40 minutes of exposure. Ho wes [ 120 ] also stated the rate of graphitization would not start to increase rapidly until 2100 o C. In our study on SPS processed UO 2 diamond composite, however, the graphitization of diamond did occur at a maximum sintering temperature of 1400 o C and a significant amount of diamond has been transformed for a hold time as short as 5 minutes. The reason for this difference could be explained by the nature of spark plasma sintering process. First of all, the programmed temperature of 1400 o C is controlled by a

PAGE 98

98 pyrometer, which measures the temperature on the surface of the die during SPS but not directly on the powder being sintered. However, there is a temperature gradient between the surface of the die and the surface of the UO 2 diamond pel let and the temperature difference can be several hundred degrees [ 121 ] . Secondly, as UO 2 is a semiconductor, current still passes through the pellet during SPS and signific ant Joule heat can be generated along the particle interfaces within the pellet. Therefore, the actual temperature in pellet can be significantly higher than 1400 o C. Last but not the least, as a field assisted sintering method, the existence of electric fi eld in SPS may also facilitate the phase transformation of diamond to graphite. But this assumption has not been verified yet, since no literature has reported the effect of electrical field on ntioned that unlike the zero pressure environment in literature, a uniaxial pressure of 36 MPa was applied during SPS. High pressure may retard the graphitization process based on the pressure temperature phase diagram of carbon [ 122 ] . Davies and Evans [ 119 ] studied the transformation of diamond under 4.8 GPa. Compared to zero pressure environment, an increase of 90 o C 130 o C of temperature was required in order to achie ve the same graphitization rate. However, the pressure in SPS is relatively low and t he effect of pressure should be limited. Because of the above complexities in SPS, it is reasonable to assume that graphitization indeed occurs under the current sintering conditions. The large variation in the degree of graphitization of diamond within th e pellet is also expected to be a result of SPS. Due to the low thermal conductivity of the UO 2 , the temperature on the surfaces of the pellet, which are in contact with the graphite punches, can be much higher than the temperature in the interior of the p ellet. Similar

PAGE 99

99 result was observed by Ge [ 36 ] in sintering pure UO 2 using SPS. Exaggerated grain growth occurred only on the very top surface of the pellet , due to higher temperature than in the interior of the pellet. Residual Stress of Diamond within the P e llet In section 5. 2, it was shown that the residual stress in diamond particle is related to the shift of diamond peak, see Eq uation ( 5 15 ). The peak range ( 1315 cm 1 1345 cm 1 ) was then checked carefully for diamond particles in all locations in Figure 5 2(b) and a typical spectrum is shown in Figure 5 7. Compared to the stress free diamond powder, the peak of diamond particles in the composite clearly shifts to higher frequencies, indicating a compressive stress in the diamond particle. In order to study the variation of stress within diamond particles, line scans across a diamond particle at each location were performed. A typical scan position is shown in F igure 5 8(a) and the variation of the diamond peak position is shown in Figure 5 8(b). The shifts have been related to stress values using E quation ( 5 15 ). The negative sign indicates that the stress is compressive. For diamond particles at all locations, it is observed that the distribution of stress is not uniform in a given diamond particle and varies within the region. The edges of the diamond particles are subjected to a higher compressive stress than the interior. The stress intensity also shows a di fference between the surface and the interior of the pellet. On both surfaces (Location A and E), the maximum stress value (950 MPa and 1.16 GPa) are higher than the maximum stresses in the interior of the pellet (Location B: 314 MPa, C: 625 MPa & D : 732 M Pa ) . Recall that in S ection 5. 4.1, it was discussed that during SPS, the temperatures on the pellet surfaces were much higher. The diamond

PAGE 100

100 particles were cooled down from a higher temperature and therefore led to a larger residual stress after SPS. Therma l stress in a particulate composite can also be predicted by the equation below using the classical theory of elasticity [ 123 ] , ( 5 17 ) where , and E modulus, respectively. Subscript m and p represe nt matrix and particulate materials, respectively. is the particle volume fraction. is the change of temperature. The corresponding values for UO 2 matrix and diamond particles are listed in the Table below. The processing induced stress is then calculated to be 2.87 GPa in UO 2 diamond composite. The experimental result does not agree with the value predicted by classical elasticity theory. On one hand, a uniform stress distribution was expected within a diamond particle based on elasticity t heory. However, an obvious variation of stress intensity in a particle was observed. This may be due to the irregular shape of the study on silicon [ 107 ] . Note that although few points on some diamond particles indicate tensile stress, it is conceivable that such small tensile stress indeed develops in diamond. On the other hand, the experimental stress is smaller than the predict value. The difference could probably be explained by the assumptions made in the derivation of Eq (5 15 ). In theoretical calculation, the matrix was assumed to be infinite and uniform everywhere, however, the conditions c ould not be satisfied in the UO 2 diamond

PAGE 101

101 composite pellet. In addition, the matrix was assumed to be pure UO 2 and only one diamond particle was analyzed, while in the UO 2 diamond composite pellet, diamond particles were evenly distributed everywhere (see F igure 5 4(a)). Last but not the least, graphitization of diamo nd which has been discussed in S ection 5. 4.1 could lead to a nd. According to the experimental results above, both severe graphitization and large residual stress are observed on the surfaces of UO 2 diamond pellet, which will have a negative influence on the thermal and mechanical properties of the pellet. Therefor e, post processing of the pellet to remove the surface may be required, if necessary. Conclusion Micro Raman spectroscopy was utilized to study the phase transformation and residual stress in diamond particles within a UO 2 diamond composite fabricated by s park plasma sintering. An equation relating Raman peak shift and induced stress intensity was derived. A non uniform distribution of compressive residual stress was observed in diamond particles at all locations within the pellet. For the study of phase tr ansformation, the graphitization of diamond occurred at all locations within the pellet. Quantitative analysis was performed to calculate the degree of graphitization. Severe graphitization and large residual stress were found in the diamond particles on b oth surfaces of the pellet.

PAGE 102

102 Table 5 1. Physical properties of UO 2 and diamond UO 2 [ 94 , 96 ] Diamond [ 97 , 124 ] Thermal expansion coefficient, 9.9×10 6 / o C 1.1×10 6 / o C 0.316 0.2 200 GPa 1200 GPa Temperature change, 1375 o C

PAGE 103

103 A ) B ) Fig ure 5 1. SEM image of as received powders. A ) UO 2 and B ) diamond . Photo courtesy of author .

PAGE 104

104 A ) B ) Figure 5 2 . Image of SPSed pellets. A ) UO 2 diamond composi te pellet sintered by SPS. Photo courtesy of author . B ) S chematic diagram showing the cross section of the pellet and the locations being measured.

PAGE 105

105 Figure 5 3 . Raman spectrum of as received diamond powder.

PAGE 106

106 A ) B ) Figure 5 4 . Image of location A. A ) Image of location A in UO 2 diamond composite showing UO2 matrix and diamond par ticles; B ) Scheme of map scan performed on location A . Photo courtesy of author.

PAGE 107

107 A) B) Figure 5 5. Typical Raman spectra of on diamond in UO 2 diamond composite . A) UO 2 peak and Diamond peak only. B) UO 2 , D ia mond and G peak. C) UO 2, D iamond, G peak and D peak .

PAGE 108

108 C) Figure 5 5 . Continued Figure 5 6 . Graphitization degree in UO 2 diamond composite pellet

PAGE 109

109 Figure 5 7 . Comparison of diamond Raman peak position between as received diamond powder and sintered pellet

PAGE 110

110 A ) B) Figure 5 8. Raman line scan. A) Schematic of Line scan across a diamond particle . Photo courtesy of author . B ) Line scan results showing str ess distribution at location A E .

PAGE 111

111 CHAPTER 6 PERFORMANCE OF UO 2 DIAMOND COMPOSITE PELLET UNDER HIGH TEMPERATURE AGING TEST Background UO 2 diamond composite fuel pellets has been successfully fabricated using SPS and the microstructure and properties of th e composi te pellets has been studied in C hapter 4. However, the environment in a nuclear reactor is relatively different. First, the temperature of a nuclear fuel is high. While the edge of fuel is just about 300 o C, there is a steep temperature gradient in the fuel and centerline temperature can be up to 1600 o C 1700 o C. Second, a high radiation level is maintained in the reactor. In addition, the fuel pellets would be subject to external pressure during operation. Under such complicated environment, the mic rostructure and properties of the composite pellets could be changed easily. Therefore, it is important to evaluate the performance of this novel composite under such environment. Due to the limitation of laboratory conditions , it is impossible to fully si mulate the environment in a nuclear fuel. In this study, a thermal aging test was performed and the influence of high temperature on the performance of UO 2 di amond composite was investigated. Experimental The uranium dioxide powder was obtained from AREVA Federal Services (Hanford, Washington, USA) with a grain size 100 400 nm. The diamond powder was obtained from Advanced Abrasives (Pennsauken, NJ, USA) with a mean particle size of 3 µm. The mixed powder was fabricated with a maximum sintering temperature of 1400 o C at a constant heating and cooling rate of 100 o C/min using SPS. At the

PAGE 112

112 maximum temperature, a pressure of 36MPa was applied and held for 5 minutes. Detailed powder mixing and SPS procedures can be found in section 3.3 or section 4.2. A Lindberg ® b lue M tube furnace was used for the thermal aging test. The furnace was heated up to 1400°C for held for 10 hour s. 1400 °C was chosen because it was the predicted fuel centerline temperature of UO 2 diamond composite in operational nuclear reactor environmen t. The atmosphere is ultrahigh purity argon gas (no more than 1 ppm O 2 ). The heating/cooling rate was set to be 2.6 °C/min . Therefore, the total furnace utilization time was about 27 hours. The pellet was polished by successive smaller grinding medium befo re aging test. The microstructure of the pellet was characterized by scanning electron microscopy (SEM) (JEOF 6335L and JEOF 6400). The potential chemical reaction was characterized using X Ray diffraction ( ® di ffractometer) . The th ermal diffusivity was measured by a laser flash method (Anter Flashline ® 3000). Results and D iscussion Figure 6 1 (a) and 6 1 (b) shows the picture of the pellet before and after aging test, respectively. After The high temperature aging, the pellet is still intact and no obvious cracks or chips can be seen. The change of weight and density is shown in Table 6 1. A slight decrease of weight and a slight increase of relative density is observed. In order to investigate the reason of those changes, microstruct ure of the pellet was characterized. Figure 6 2(a) and 6 2 (b) shows the SEM images of the polished surface of the pellet before and after aging test, respectively. It can be seen from Figure 6 2(b) that many diamond particles on the surface are disappeared and some dents are left. Note that the diamond particles were on the surface with good interfacial contact

PAGE 113

113 before aging test, see Figure 6 1(a). Therefore, it is proven that the diamond particles were not removed during polish process. The only reason cou ld be the existence of minor oxygen. Despite less than 1 ppm in the atmosphere, the diamond particles were oxidized and became carbon oxide/dioxide at such high temperature. The oxidation of diamond can be the explanation of the decrease of weight. The ima ge of the interior of the pellet is shown in Figure 6 3. Diamond is still in good contact UO 2 matrix, showing that the composite can survive under current environment. Figure 6 4 shows the XRD spectra of the composite pellet before and after thermal aging test . Note that only characteristic peaks of UO 2 ( PDF 00 005 0550) were detected, while the diamond peaks were not detected. It may be due to detection limit of XRD. The diamond content is just 5vol% so that the signal of diamond is relatively low. Compare d to the spectrum before aging test, no new peaks was observed, which can prove that there was no detectable chemical reaction during the thermal aging test. The change of thermal diffusivity was also measured and the result is shown in Figure 6 5. After h igh temperature aging test, the thermal diffusivity was reduced by 2.3%, 1.9% and 0.9% at 100 o C, 500 o C and 900 o C, respectively. A possible reason for the decrease is the loss of surface diamond. However, the thermal diffusivity of the composite pellet is s till higher than pure UO 2 due to the existence of diamond in the interior. Conclusion UO 2 3 um diamond composite pellet can survive after 10 hours duration at 1400 o C. While the pellet remained intact in appearance, oxidation of surface diamond was observed , which led to a slight decrease of weight and thermal conductivity. No potential chemical reaction was occurred during the thermal aging process. Despite the

PAGE 114

114 changes on the surface, the interior of the pellet remained unchanged and the thermal diffus i vi ty of the composite after thermal aging was still higher than pure UO 2 pellet . Table 6 1 . Weight and density change of UO 2 diamond pellet after aging test Weight (g) Relative Density (%) Before test 3.6637 96.3 After test 3.6604 96.6

PAGE 115

115 A ) B ) Figur e 6 1 . Appearance of UO 2 diamond composite p ellet . A) B efore thermal aging test. B ) A fter thermal aging test. Photo courtesy of author .

PAGE 116

116 A ) B ) Figure 6 2 . SEM images of surface of UO 2 diamond composite pellet . A) Before aging test. B ) A fter aging test . Photo courtesy of author .

PAGE 117

11 7 Figure 6 3. SEM images of interior of UO 2 diamond composite pellet after aging test . Photo courtesy of author .

PAGE 118

118 Figure 6 4. XRD spectra of U O 2 diamond composite pellet before and after thermal aging test .

PAGE 119

119 Figure 6 5. Change of thermal diffusivity of UO 2 diamond composite after aging test .

PAGE 120

120 CHAPTER 7 CONCLUSION AND FUTURE WORK Conclusion T he densification evolution in SPS of fuel pellets and the feasibility of SPS of nove l UO 2 diamond composite fuel pellets has been investigated. M aster sintering c urve (MSC) theory was utilized in order to study the densification evolution of UO 2 and UO 2 composites in SPS. For UO 2 diamond composites, the theory was proven to be not suitabl e. For UO 2 and UO 2 SiC composite , MSC was successfully applied. By applying the constant heating rate method, t he apparent act ivation energies for sintering was determined to be 140 KJ/mol for UO 2 and 420 KJ/mol for UO 2 SiC composite . The ability of the de rived MSC s to control and predict final density in the sintered compact was demonstrated by additional experimental runs using the isothermal heating method. The reason for significantly lower activation energy in SPS processed UO 2 pellets compared to conv entional sintering has been rationalized on the basis of field activation in SPS process. T he microstructure and properties of UO 2 diamond composites were investigated. H igh density UO 2 5 vol% diamond composite pellets were fabricated using the spark plas ma sintering (SPS) technique. Diamond particles with nano size ( 0.25 µm ) and several micro sizes (3 and were mixed with UO 2 powder and sintered using SPS at 1300 o C 1600 o C with a hold time of 5 minutes. T he resultant density, chemical reac tion, microstructure, thermal conductivity of the sintered pellets were investigated. The pellets with 3 diamond particles had a uniform distribution of particles as well as better thermal and mechanical properties compared to other s. A n increase in thermal conductivity of up to 41.6%, 38.3% and

PAGE 121

121 34.2% at 100 o C, 500 o C and 900 o C, respectively, w as measured in the UO 2 diamond composite pellets c ompared to the pure UO 2 fuel pellets . M icro Raman spectroscopy (MRS) was utilized to study t h e phase transformation and residual stress in diamond particles within a UO 2 diamond composite sintered by spark plasma sintering (SPS). Graphitization of diamond was observed and the degree of graphitization was quantified. T he relationship between Raman peak shift and stress intensity was derived. A non uniform distribution of compressive residual stress was shown in diam ond particles. Severe graphitization and large residual stress were found in the diamond particles on both surfaces of the pellet. T he p erformance of UO 2 diamond pellets after high temperature aging test was investigated. The composite pellet was proven to survive after 10 hours duration at 1400 o C. While the pellet remained intact in appearance, oxidation of surface diamond was observed, w hich led to a slight decrease of weight and thermal conductivity. No potential chemical reaction was occurred during the thermal aging process. Despite the changes on the surface, the interior of the pellet remained unchanged and the thermal diffusibity of the composite after thermal aging was still higher than pure UO 2 pellet . Future W ork Washout B ehavior Study In operation of a nuclear reactor, t he reliability of the fuel pellets pl ays a significant role on the availability factor of nuclear plants. In re cent years, fuel washout was observed due to the direct contact of th e fuel pellets with the coolant and became a severe issue. The washout of fuel pellets can cause the increase of background activity, which may take up to ten years to be eliminated.

PAGE 122

122 The fuel washout is mainly due to the oxidation of the fuel pellets. Carter and Lay [ 125 ] investigated the oxidation reduction quality of UO 2 and found that oxidation rate of UO 2 was very fast under CO 2 CO mi xtures. Verrall et al. [ 126 ] did further study and showed that the oxidation of UO 2 started and propagated along grain boundaries. T he grains finally stripped fr om the pellet s and led to dissolution when they were in contact with the cooling water . In order to study washout behavior of fuel pellets, Delafoy and Zemek [ 127 ] from AREVA performed two experiments. In the first part, the oxidation behavior of the pellets under different oxygen partial pressures (0.01% to 1%) was studied. Thermogravimetric analyzer (TGA) was used and the oxidation rate was determined. In the second part, the corrosion behavior of the pellets was studied by autoclave leaching test. An autoclave was used t o simulate the in reactor environment (pressure, te mperature and oxygen potential) in the case of a cladding fail ure . After the test, the mass change of the pellets was measured and phase change was characterized. In the preliminary experiment, full size pure UO 2 , UO 2 10vol% SiC com posite and UO 2 10vol % diamond composite were fabricated by SPS . The detailed fabricati on procedures can be found in previous chapters. The SiC whiske rs were obtained from Advanced Composite Materials, Greer, SC and the 3µm diamond was obtained from Advanced Abrasives, Pennsauken, NJ . The washout test was conducted in a Lindberg blue M tu be furnace in 0.1% O 2 99.9% N 2 atmosphere at 380°C. For characterization requirement, pellets with flat surfaces (no dish and chamfer) were fabricated. One surface of each pellet was polished.

PAGE 123

123 First, three pellets (1 UO 2 , 1 UO 2 SiC and 1 UO 2 diamond compos ite) were tested with a hold time of 5 hours. All pellets were intact and there was on change in appearance after test, see Figure 7 1. Second, another three pellets (1 UO 2 , 1 UO 2 SiC and 1 UO 2 diamond composite) were tested with a hold time of 15 hours. A fter test, UO 2 SiC and UO 2 diamond composite pellets were intact but the pure UO 2 pellet was broken into two pieces, see Figure 7 2. T he change of weight for each pel let is listed in the Table 7 1 . The decrease of weight is probably due to the oxidation of residual graphite. CO or CO 2 was formed and escaped as gas. Diamond in UO 2 diamond composite could also have been oxidized to CO or CO 2 , which led to a larger decrease of weight in UO 2 diamond pellets, see Table 7 1. X Ray diffraction technique was used t o characterize the phase and crystal structure of all pellets. The results are shown in Figure 7 3 (a) 7 3 (c ). All UO 2 characteristic peaks are observed while SiC or diamond peaks are not observed due to the detection limi t of XRD. It is seen in Figure 7 3 ( a) that a small peak appeared next to each characteristic peak of UO 2 after 5 hours. It is known that when UO 2 is exposed in the oxidizing atmosphere, oxygen is easily dissolved into the fluorite structure and transforms to the hyper stoichiometric phase U O 2+x , Teske et al . , [ 128 ] found that the lattice parameter a o f UO 2+x subjected to the equation a= 0.132x + 5.4705 and decreases with the increase of x . In a cubic structure, the interplanar spacing d decreases with the decrease of lattice parameter a . Then, according to Bragg diffraction Equation , the diffract i o n angle is expected to increase. Therefore, the new peaks which appeared on the right of each original UO 2 peak clearly indicate the

PAGE 124

124 formation of UO 2+x . When the hold time increased to 15 hours, the intensity of all new peaks were much higher an d even larger than original UO 2 peaks, indicating a more severe oxidation of UO 2 . Similar results are seen in UO 2 Si C and UO 2 diamond composites, see Figure 7 3(b) and 7 3(c). Original UO 2 peaks also shifted to the right with the increase of hold time, whi ch indicated that the amount of UO 2.00 was reduced gradually. The XRD spectra of UO 2 , UO 2 Si C and UO 2 diamond after 15 hours washout test is summarized in Figure 7 4 . It should be noted that a peak at around 22 degree appeared in all pellets after washout test, which is not a characteristic peak of UO 2 . The reason is still unknown and more study on it is required . Another test was performed in 1% oxygen 99% helium gas atmosphere at 380 °C for 20 hours. After experiment, it is found that all pellets were crum bled into powders after the oxidation test, see Figure 7 5 . The total weight of all pellets before test is 13.6587g, while the weight after test is 14.0480g. The increase of weight indicated the oxidation of UO 2 . The resulting products was analyzed by XRD and t he spectrum is shown in Figure 7 6 . U 3 O 8 and SiO 2 peaks are clearly observed. While UO 2 , SiC and diamond peaks are disappeared. It is shown that the UO 2 has been completely oxidized to U 3 O 8 and the pellets could not survive under current condition. F uture work will be focused on the improvement of oxidation behavior of the fuel pellets. It has been proved that for fuel pellets fabricated by conventional sintering method, the matrix grain size has a large decisive influence on the oxidation behavior [ 127 ] . However, whether the conclusion can be applied to SPSed fuel pellets and the

PAGE 125

125 way to improve the oxidation behavior are still i n question and should be studied in the future. Compression T est Compressive strength is the capacity of a material to withstand loads tending to reduce size, which is another important property for nuclear fuel. Also in industry, it is normal required tha t the pellets are capable of surging 100lbs force without chipping or cracking. Therefore, to get accurate compressive strength for the new UO 2 composite pellet is important. In the initial study, a static compression test for UO 2 diamond composite pellet was performed using a servo hydraulic universal testing machine (MTS) and followed by ASTM standard 1424 10. After polishing both top and bottom surfaces, UO 2 diamond pellets were sandwiched between two tungsten carbide discs and subject to uniaxial loadi ng with a constant displacement rate. The compressive strength was then determined by the maximum compressive stress which a pellet was capable of sustaining. Figure 7 7 shows a stress strain curve of UO 2 5 vol% diamond composite pellet under compression test and a maximum stress of 433 MPa is obtained. It should parallelism and non perpendicularity of the sample. It is di fficult to guarantee the parallel ism perpendicularity using the current hand polish method. Therefore, sample prepa ration process is still need to be improved and the second phase effect on the compressive strength of UO 2 composite fuel pellets is to be sy stematical studied.

PAGE 126

126 Table 7 1. Change of pellet weight before and after washout test Materials Hold Time (h) W eight before test (g) Weight after test (g) Percentage c hange of weight UO 2 5 4.4822 4.4813 0.020% UO 2 10vol% SiC 5 4.7120 4.7094 0.055% U O 2 10vol% diamond 5 4.6069 4.6 034 0.076% UO 2 15 4.7047 4.7023 0.051% UO 2 10vol% SiC 15 4.7327 4.7016 0.023% UO 2 10vol% diamond 15 4.6841 4.6800 0.086%

PAGE 127

127 A) B) Figure 7 1 . Images of p ellets (From left to right: UO 2 , UO 2 SiC and UO 2 diamond compos ite) under 5h washout test. A) B efore test. B ) A fter test . Photo courtesy of author .

PAGE 128

128 A ) B ) Figure 7 2 . Images of pellets (From left to right: UO 2 , UO 2 SiC and UO 2 diamond composite) under 15h washout test. A) Before test. B) After test . Photo courtesy of author .

PAGE 129

129 A) B) Figure 7 3. XRD spect ra of pellets before and washout test . A) UO 2 pellet. B) UO 2 SiC composite C) UO 2 diamond composite.

PAGE 130

130 C) Figure 7 3. Continued . Figure 7 4 . XRD spect ra of UO 2 and UO 2 composite pellets after 15 hours holding at 380 o C in 0.1% O 2 .

PAGE 131

131 Figure 7 5 . Crumbled UO 2 and UO 2 composites pellets after oxidation test . Photo courtesy of author .

PAGE 132

132 Figure 7 6 . XRD spectrum of c rumbled UO 2 and UO 2 composite pellets after 20h washout test in 1% O 2 .

PAGE 133

133 Figure 7 7 . Stress Strain curve of UO 2 5 vol% diamond composite pellet under compression test .

PAGE 134

134 LIST OF REFERENCE S 1. "The economics of nuclear power", World Nucle ar Association, December 201 1 [ http://www.energybc.ca/cache/nuclear/nuclear2/www.world nuclear.org/info/inf02.html ] . 2. "Outl ine History of Nuclear Energy", World Nuclear Association, March 2014 [ http://www.world nuclear.org/info/Current and Future Generation/Outline His tory of Nuclear Energy/ ] . 3. "Nuclear Power in the World Today", World Nuclear Association, Febrary 2015 [ http://www.world nuclear.org/inf o/Current and Future Generation/Nuclear Power in the World Today/ ] . 4. Carbajo JJ, Yoder GL, Popov SG, Ivanov VK: A review of the thermophysical properties of MOX and UO2 fuels. Journal of Nuclear Materials 2001, 299(3):181 198. 5. Glasstone S, Sesonske A : Nuclear reactor engineering. Journal Name: Nucl Sci Eng; (United States); Journal Volume: 81:3 1982, Pages: 484 485. 6. Zinkle SJ, Terrani KA, Gehin JC, Ott LJ, Snead LL: Accident tolerant fuels for LWRs: A perspective. Journal of Nuclear Materials 2014, 448(1 3):374 379. 7. Yeo S: UO2 SIC composite reactor fuels with enhanced thermal and mechanical poperties prepared by spark plasma sintering. [Gainesville, Fla: University of Florida; 2013. 8. Hasselman D, Donaldson KY, Geiger AL: Effect of reinforcement particle size on the thermal cond uctivity of a particulate silicon carbide reinforced aluminum matrix composite. Journal of the American Ceramic Society 1992, 75(11):3137 3140. 9. Zhou W, Wang C, Ai T, Wu K, Zhao F, Gu H: A novel fiber rei nforced polyethylene composite with added silicon nitride particles for enhanced thermal conductivity. Composites Part A: Applied Science and Manufacturing 2009, 40(6 7):830 836. 10. Ishimoto S, Hirai M, Ito K, Korei Y: Thermal Conductivity of UO2 BeO Pell et. Journal of Nuclear Science and Technology 1996, 33(2):134 140. 11. Yeo S, McKenna E, Baney R, Subhash G, Tulenko J: Enhanced thermal conductivity of uranium dioxide silicon carbide composite fuel pellets prepared by Spark Plasma Sintering (SPS). Journa l of Nuclear Materials 2013, 433(1 3):66 73.

PAGE 135

135 12. Yeo S, Baney R, Subhash G, Tulenko J: The influence of SiC particle size and volume fraction on the thermal conductivity of spark plasma sintered UO2 SiC composites. Journal of Nuclear Materials 2013, 442(1 3):245 252. 13. Cartas A, Wang H, Subhash G, Baney R, Tulenko J: Influence of Carbon Nanotube Dispersion in UO 2 Carbon Nanotube Ceramic Matrix Composites Utilizing Spark Plasma Sintering. Nuclear Technology 2015, 189(3). 14. Lee SW, Kim HT, Bang IC: Perfo rmance evaluation of UO2/graphene composite fuel and SiC cladding during LBLOCA using MARS KS. Nuclear Engineering and Design 2013, 257(0):139 145. 15. Wang J: Developing a high thermal conductivity nuclear fuel with silicon carbide additives. University o f Florida; 2008. 16. Ge L, Subhash G, Baney RH, Tulenko JS, McKenna E: Densification of uranium dioxide fuel pellets prepared by spark plasma sintering (SPS). Journal of Nuclear Materials 2013, 435:1 9. 17. Carrea AJ: Sintering of uranium dioxide in an atm osphere of controlled hydrogen content. Journal of Nuclear Materials 1963, 8(2):275 277. 18. Williams J, Barnes E, Scott R, Hall A: Sintering of uranium oxides of composition UO2 to U3O8 in various atmospheres. Journal of Nuclear Materials 1959, 1(1):28 38 . 19. Stuart WI, Adams RB: Effect of sintering atmosphere on the density of uranium dioxide pellets. Journal of Nuclear Materials 1975, 58(2):201 204. 20. Ge L, Subhash G, Baney RH, Tulenko JS: Influence of processing parameters on thermal conductivity of uranium dioxide pellets prepared by spark plasma sintering. Journal of the European Ceramic Society 2014, 34(7):1791 1801. 21. Slack GA: Thermal Conductivity of Pure and Impure Silicon, Silicon Carbide, and Diamond. Journal of Applied Physics 1964, 35(12): 3460 3466. 22. Dilawar N, Kapil R, Vankar VD, Avasthi DK, Kabiraj D, Mehta GK: Radiation hardness of polycrystalline diamond thin films irradiated with 100 MeV I7+ ions. Thin Solid Films 1997, 305(1 2):88 94. 23. Hasselman D, Donaldson KY, Liu J, Gauckler LJ, Ownby PD: Thermal Conductivity of a Particulate Diamond Reinforced Cordierite Matrix Composite. Journal of the American Ceramic Society 1994, 77(7):1757 1760. 24. Interfacial c haracterization of Cu/diamond composites prepared by powder metallurgy for heat sink applications. Scripta Materialia 2008, 58(4):263 266.

PAGE 136

136 25. Chu K, Liu Z, Jia C, Chen H, Liang X, Gao W, Tian W, Guo H: Thermal conductivity of SPS consolidated Cu/diamond c omposites with Cr coated diamond particles. Journal of Alloys and Compounds 2010, 490(1 2):453 458. 26. Tavangar R, Molina JM, Weber L: Assessing predictive schemes for thermal conductivity against diamond reinforced silver matrix composites at intermediat e phase contrast. Scripta Materialia 2007, 56(5):357 360. 27. "World Nuclear Power Reactors & Uranium Requirements", World Nuclear Association, January 2014 [ http://www.world nuclear.org/info/Facts and Figures/World Nuclear Power Reactors and Uranium Requirements/ ] 28. Kapoor K, Ahmad A, Laksminarayana A, Hemanth Rao GVS: Fracture properties of sintered UO2 ceramic pellets with duplex m icrostructure. Journal of Nuclear Materials 2007, 366(1 2):87 98. 29. Moore E, Guéneau C, Crocombette J P: Diffusion model of the non stoichiometric uranium dioxide. Journal of Solid State Chemistry 2013, 203(0):145 153. 30. Boyarchenkov A, Potashnikov S, Nekrasov K, Kupryazhkin AY: Investigation of cation self diffusion mechanisms in UO 2±x using molecular dynamics. Journal of Nuclear Materials 2013, 442(1):148 161. 31. Brandt R, Neuer G: Thermal conductivity and thermal radiation properties of UO 2. Journ al of Non Equilibrium Thermodynamics 1976, 1(1):3 24. 32. Hyland G: Thermal conductivity of solid UO< sub> 2: Critique and recommendation. Journal of Nuclear Materials 1983, 113(2):125 132. 33. Fink J: Thermophysical properties of uranium dioxide. Jo urnal of Nuclear Materials 2000, 279(1):1 18. 34. Nichenko S, Staicu D: Molecular Dynamics study of the effects of non stoichiometry and oxygen Frenkel pairs on the thermal conductivity of uranium dioxide. Journal of Nuclear Materials 2013, 433(1 3):297 30 4. 35. Meek T, Hu M, Haire MJ: Semiconductive properties of uranium oxides. In: Waste Management Symposium: 2001 . 10 18. 36. Ge L: Processing of Uranium Dioxide Nuclear Fuel Pellets Using Spark Plasma Sintering. [Gainesville, Fla: University of Florida; 20 14. 37. Wei L, Kuo PK, Thomas RL, Anthony TR, Banholzer WF: Thermal conductivity of isotopically modified single crystal diamond. Physical Review Letters 1993, 70(24):3764 3767.

PAGE 137

137 38. re lated materials from molecular dynamics simulations. The Journal of Chemical Physics 2000, 113(16):6888 6900. 39. Klein CA, Cardinale GF: Young's modulus and Poisson's ratio of CVD diamond. Diamond and Related Materials 1993, 2(5):918 923. 40. Lee J, Novik ov N: Innovative superhard materials and sustainable coatings for advanced manufacturing, vol. 200: Springer Science & Business Media; 2005. 41. Kapitza PL: Heat Transfer and Superfluidity of Helium II. Physical Review 1941, 60(4):354 355. 42. Maxwell JC: Electricity and magnetism. 1954. 43. Fricke H: A mathematical treatment of the electric conductivity and capacity of disperse systems I. The electric conductivity of a suspension of homogeneous spheroids. Physical Review 1924, 24(5):575. 44. Hasselman DPH, Johnson LF: Effective Thermal Conductivity of Composites with Interfacial Thermal Barrier Resistance. Journal of Composite Materials 1987, 21(6):508 515. 45. Benveniste Y: Effective thermal conductivity of composites with a thermal contact resistance betw een the constituents: Nondilute case. Journal of Applied Physics 1987, 61(8):2840 2843. 46. Tokita M: Mechanism of spark plasma sintering. In: Proceeding of NEDO International Symposium on Functionally Graded Materials: 1999 . Japan: 22. 47. Lagos MA, Agote I: SPS synthesis and consolidation of TiAl alloys from elemental powders: Microstructure evolution. Intermetallics 2013, 36(0):51 56. 48. Hayun S, Paris V, Mitrani R, Kalabukhov S, Dariel MP, Zaretsky E, Frage N: Microstructure and mechanical properties o f silicon carbide processed by Spark Plasma Sintering (SPS). Ceramics International 2012, 38(8):6335 6340. 49. Gu YW, Khor KA, Cheang P: Bone like apatite layer formation on hydroxyapatite prepared by spark plasma sintering (SPS). Biomaterials 2004, 25(18) :4127 4134. 50. Li W, Gao L: Fabrication of HAp ZrO2 (3Y) nano composite by SPS. Biomaterials 2003, 24(6):937 940. 51. Munir ZA, Anselmi Tamburini U, Ohyanagi M: The effect of electric field and pressure on the synthesis and consolidation of materials: A r eview of the spark plasma sintering method. Journal of Materials Science 2006, 41(3):763 777.

PAGE 138

138 52. Hulbert DM, Anders A, Dudina DV, Andersson J, Jiang D, Unuvar C, Anselmi plasma sinter Journal of Applied Physics 2008, 104(3):033305. 53. Groza JR, Zavaliangos A: Sintering activation by external electrical field. Materials Science and Engineering: A 2000, 287(2):171 177. 54. Hulbert DM, Anders A, Andersson J, Lavernia EJ, Mukherjee A K: A discussion on the absence of plasma in spark plasma sintering. Scripta Materialia 2009, 60(10):835 838. 55. Graeve OA, Carrillo Heian EM, Feng A, Munir ZA: Modeling of wave configuration during electrically ignited combustion synthesis. Journal of Mat erials Research 2001, 16(01):93 100. 56. Anselmi Tamburini U, Gennari S, Garay JE, Munir ZA: Fundamental investigations on the spark plasma sintering/synthesis process: II. Modeling of current and temperature distributions. Materials Science and Engineerin g: A 2005, 394(1 2):139 148. 57. Giuntini D, Olevsky EA, Garcia Cardona C, Maximenko AL, Yurlova MS, Haines CD, Martin DG, Kapoor D: Localized overheating phenomena and optimization of spark plasma sintering tooling design. Materials 2013, 6(7):2612 2632. 58. Kutty TRG, Hegde PV, Khan KB, Basak U, Pillai SN, Sengupta AK, Jain GC, Majumdar S, Kamath HS, Purushotham DSC: Densification behaviour of UO2 in six different atmospheres. Journal of Nuclear Materials 2002, 305(2 3):159 168. 59. Coble RL: Sintering Cr ystalline Solids. I. Intermediate and Final State Diffusion Models. Journal of Applied Physics 1961, 32:787 792. 60. Johnson DL: New Meth od of Obtaining Volume, Grain Boundary, and Surface Diffusion Coefficients from Sintering Data. Journal of Applied Physics 1969, 40(1):192 200. 61. Kang S JL, Jung Y I: Sintering kinetics at final stage sintering: model calculation and map construction. Ac ta Materialia 2004, 52(15):4573 4578. 62. Su H, Johnson DL: Master Sintering Curve: A Practical Approach to Sintering. Journal of the American Ceramic Society , 79:3211 3217. 63. Su H, Johnson DL: A practical approach to sintering. American Ceramic Society Bulletin 1997, 76. 64. Park SJ, Johnson JL, Wu Y, Kwon Y S, Lee S, German RM: Analysis of the effect of solubility on the densification behavior of tungsten heavy alloys using the master sintering curve approach. International Journal of Refractory Metals and Hard Materials 2013, 37(0):52 59.

PAGE 139

139 65. Blaine DC, Park S, German RM, LaSalle J, Nandi H: Verifying the Master Sintering Curve on an Industrial Furnace. ADVANCES IN POWDER METALLURGY AND PARTICULATE MATERIALS 2005, 1:1. 66. Guillon O, Langer J: Master si ntering curve applied to the Field Assisted Sintering Technique. Journal of materials science 2010, 45(19):5191 5195. 67. Enneti RK, Bothara MG, Park S J, Atre SV: Development of master sintering curve for field assisted sintering of HfB< sub> 2 20Si C. Ceramics International 2012, 38(5):4369 4372. 68. Hansen JD, Rusin RP, Teng MH, Johnson DL: Combined Stage Sintering Model. Journal of the American Ceramic Society , 75:1129 1135. 69. DiAntonio CB, Ewsuk KG: Controlled and Predicted Ceramic Sintering Thr ough Master Sintering Curve Theory.15 23. 70. Anzelmo J, Seyfarth A, Arias L: Approaching a universal sample preparation method for XRF analysis of powder materials. Adv X ray Anal 2001, 44:368 373. 71. Xiong Y, Fu ZY, Wang H, Wang YC, Zhang QJ: Microstruc ture and IR transmittance of spark plasma sintering translucent AlN ceramics with CaF2 additive. Materials Science and Engineering: B 2005, 123(1):57 62. 72. Zhu LH, Huang QW, Zhao HF: Preparation of nanocrystalline WC 10Co 0.8VC by spark plasma sintering. Journal of Materials Science Letters 2003, 22(22):1631 1633. 73. Borodianska H, Demirskyi D, Sakka Y, Badica P, Vasylkiv O: Grain boundary diffusion driven spark plasma sintering of nanocrystalline zirconia. Ceramics International 2012, 38(5):4385 4389. 7 4. Assmann H, Bairiot H: Process and product control of oxide powder and pellets for reactor fuel application. Guidebook on Quality Control of Water Reactor Fuel, Tech Report Series 1983(221):149. 75. Teng MH, Lai YC, Chen YT: A computer program of master sintering curve model to accurately predict sintering results. 2012. 76. Pouchly V, Maca K: Master sintering curve: A practical approach to its construction. Science of Sintering 2010, 42:25 32. 77. Kang S JL: Sintering: densification, grain growth and mic rostructure: Butterworth Heinemann; 2004. 78. Dehaudt P, Bourgeois L, Chevrel H: Activation energy of UO2 and UO2+x sintering. Journal of Nuclear Materials 2001, 299(3):250 259.

PAGE 140

140 79. Cologna M, Francis JSC, Raj R: Field assisted and flash sintering of alumi na and its relationship to conductivity and MgO doping. Journal of the European Ceramic Society 2011, 31(15):2827 2837. 80. Yoshida H, Sakka Y, Yamamoto T, Lebrun J M, Raj R: Densification behaviour and microstructural development in undoped yttria prepare d by flash sintering. Journal of the European Ceramic Society 2014, 34(4):991 1000. 81. Jha SK, Raj R: The Effect of Electric Field on Sintering and Electrical Conductivity of Titania. Journal of the American Ceramic Society 2014, 97(2):527 534. 82. Bothar a MG, Park S J, Sudarshan TS, Radhakrishnan R: Sintering Behavior of Nanocrystalline Silicon Carbide Using a Plasma Pressure Compaction System: Master Sintering Curve Analysis. Metallurgical and Materials Transactions A 2014, 41:3252 3261. 83. Raj R, Colog na M, Francis JSC: Influence of Externally Imposed and Internally Generated Electrical Fields on Grain Growth, Diffusional Creep, Sintering and Related Phenomena in Ceramics. Journal of the American Ceramic Society 2011, 94(7):1941 1965. 84. Chen Z, Subhas h G, Tulenko JS: Master sintering curves for UO2 and UO2 SiC composite processed by spark plasma sintering. Journal of Nuclear Materials 2014, 454(1 3):427 433. 85. Rudajevová A, von Buch F, Mordike BL: Thermal diffusivity and thermal conductivity of MgSc alloys. Journal of Alloys and Compounds 1999, 292(1 2):27 30. 86. Phani KK, Sanyal D, Sengupta AK: Estimation of elastic properties of nuclear fuel material using longitudinal ultrasonic velocity A new approach. Journal of Nuclear Materials 2007, 366(1 2 ):129 136. 87. Assmann H, Bairiot H: Guidebook on Quality Control of Water Reactor Fuel. IAEA technical series 1983(221). 88. Davies G, Evans T: Graphitization of diamond at zero pressure and at a high pressure. Proceedings of the Royal Society of London S eries A, Mathematical and Physical Sciences 1972:413 427. 89. Gossé S, Guéneau C, Alpettaz T, Chatain S, Chatillon C, Le Guyadec F: Kinetic study of the UO2/C interaction by high temperature mass spectrometry. Nuclear Engineering and Design 2008, 238(11):2 866 2876. 90. Kempter CP, Elliott RO: Thermal Expansion of UN , UO2 , UO2 ThO2 , and ThO2 . The Journal of Chemical Physics 1959, 30(6):1524 1526.

PAGE 141

141 91. Xu CH, Wang CZ, Chan CT, Ho KM: Theory of the thermal expansion of Si and diamond. Physical Review B 1991, 43(6):5024 5027. 92. Nan C W, Birringer R: Determining the Kapitza resistance and the thermal conductivity of polycrystals: A simple model. Physical Review B 1998, 57(14):8264 8268. 93. Liu Y, Hu C, Feng W, Men J, Cheng L, Zhang L: Microstructure and properties of diamond/SiC composites prepared by tape casting and chemical vapor infiltration process. Journal of the European Ceramic Society 2014, 34(15):3489 3498. 94. Gatt JM, Monerie Y, Laux D, Baron D: Elastic behavior of porous ceramics: applicatio n to nuclear fuel materials. Journal of Nuclear Materials 2005, 336(2 3):145 155. 95. Padel A, De Novion C: Constantes elastiques des carbures, nitrures et oxydes d'uranium et de plutonium. Journal of Nuclear Materials 1969, 33(1):40 51. 96. Fink JK: Therm ophysical properties of uranium dioxide. Journal of Nuclear Materials 2000, 279(1):1 18. 97. Reeber R, Wang K: Thermal expansion, molar volume and specific heat of diamond from 0 to 3000k. JEM 1996, 25(1):63 67. 98. Fischbach D: The kinetics and mechanism of graphitization. Chemistry and physics of carbon 1971, 7:1 105. 99. Chen Z, Subhash G, Tulenko JS: Spark plasma sintering of diamond reinforced uranium dioxide composite fuel pellets. Nuclear Engineering and Design 2015, 294:52 59. 100. Willander M, Frie sel M, Wahab Q u, Straumal B: Silicon carbide and diamond for high temperature device applications. J Mater Sci: Mater Electron 2006, 17(1):1 25. 101. Bagger C, Mogensen M, Walker CT: Temperature measurements in high burnup UO2 nuclear fuel: Implications f or thermal conductivity, grain growth and gas release. Journal of Nuclear Materials 1994, 211(1):11 29. 102. Lindgren LE, Häggblad HÅ, Josefson BL, Karlsson L: Thermo mechanical FE analysis of residual stresses and stress redistribution in butt welding of a copper canister for spent nuclear fuel. Nuclear Engineering and Design 2002, 212(1 3):401 408. 103. Ghosh D, Subhash G, Lee CH, Yap YK: Strain induced formation of carbon and boron clusters in boron carbide during dynamic indentation. Applied Physics Let ters 2007, 91(6):061910.

PAGE 142

142 104. Gogotsi Y, Baek C, Kirscht F: Raman microspectroscopy study of processing induced phase transformations and residual stress in silicon. Semiconductor Science and Technology 1999, 14(10):936. 105. Jannotti P, Subhash G, Zheng J Q, Halls V, Karandikar PG, Salamone S, Aghajanian MK: Raman spectroscopic characterization of the core rim structure in reaction bonded boron carbide ceramics. Applied Physics Letters 2015, 106(4):041903. 106. DiGregorio J, Furtak T, Petrovic J: A techniqu e for measuring residual stress in SiC whiskers within an alumina matrix through Raman spectroscopy. Journal of applied physics 1992, 71(7):3524 3531. 107. Jannotti P, Subhash G: Micro Raman Spectroscopic Evaluation of Residual Microstresses in Reaction Bo nded Boron Carbide Ceramics. In: Dynamic Behavior of Materials, Volume 1. Springer International Publishing; 2015: 39 43. 108. Shafiq M, Subhash G: A Novel Technique for the Determination of Surface Biaxial Stress under External Confinement Using Raman Spe ctroscopy. Exp Mech 2014, 54(5):763 774. 109. Wang W, Polo M, Sanchez G, Cifre J, Esteve J: Internal stress and strain in heavily boron doped diamond films grown by microwave plasma and hot filament chemical vapor depositi on. Journal of applied physics 199 6, 80(3):1846 1850. 110. Ferrari AC: Raman spectroscopy of graphene and graphite: disorder, electron phonon coupling, doping and nonadiabatic effects. Solid state communications 2007, 143(1):47 57. 111. Ganesan S, Maradudin A, Oitmaa J: A lattice theory of morphic effects in crystals of the diamond structure. Annals of Physics 1970, 56(2):556 594. 112. Prawer S, Nemanich RJ: Raman spectroscopy of diamond and doped diamond. Philosophical Transactions of the Royal Society of London A: Mathematical, Physical a nd Engineering Sciences 2004, 362(1824):2537 2565. 113. Grimsditch M, Anastassakis E, Cardona M: Effect of uniaxial stress on the zone center optical phonon of diamond. Physical Review B 1978, 18(2):901. 114. Chen Z, Subhash G, Tulenko j: Spark Plasma Sint ering of Diamond Reinforced Uranium Dioxide Composite Fuel Pellets. Nuclear Engineering and Design 2015. 115. Krishnamurti D: The Raman spectrum of diamond. In: Proceedings of the Indian Academy of Sciences, Section A: 1954 . Indian Academy of Sciences: 211 216. 116. Krishnan R: The raman spectrum of diamond. In: Proceedings of the Indian Academy of Sciences, Section A: 1944 . Indian Academy of Sciences: 216 223.

PAGE 143

143 117. Goresy Ahmed E, Gillet P, Chen M, Künstler F, Graup G, Stähle V: In situ discovery of shock induced graphite diamond phase transition in gneisses from the Ries Crater, Germany. In: American Mineralogist. vol. 86; 2001: 611. 118. Ferreira NG, Abramof E, Leite NF, Corat EJ, Trava Airoldi VJ: Analysis of residual stress in diamond films by x ray dif fraction and micro Raman spectroscopy. Journal of Applied Physics 2002, 91(4):2466 2472. 119. Evans T, James P: A study of the transformation of diamond to graphite. In: Proceedings of the Royal Society of London A: Mathematical, Physical and Engineering S ciences: 1964 . The Royal Society: 260 269. 120. Howes V: The graphitization of diamond. Proceedings of the Physical Society 1962, 80(3):648. 121. Wang C, Cheng L, Zhao Z: FEM analysis of the temperature and stress distribution in spark plasma sintering: Modelling and experimental validation. Computational Materials Science 2010, 49(2):351 362. 122. Bundy F, Bassett W, Weathers M, Hemley R, Mao H, Goncharov A: The pressure temperature phase and transformation diagram for carbon; updated through 1994. Carbon 1996, 34(2):141 153. 123. Brooksbank D, Andrews K: Stresses associated with duplex Oxide Sulphide Inclusions in Steel. J Iron Steel Inst 1970 , 208(6):582 586. 124. Klein CA, Cardinale GF: Diamond 1992 Young's modulus and Poisson's ratio of CVD diamond. Diamond and Related Materials 1993, 2(5):918 923. 125. Carter R, Lay K: Surface controlled oxidation reduction of UO 2. Journal of Nuclear Mater ials 1970, 36(1):77 86. 126. Verrall R, He Z, Mouris J: Characterization of fuel oxidation in rods with clad holes. Journal of nuclear materials 2005, 344(1):240 245. 127. Delafoy C, Zemek M: Washout behaviour of chromia doped UO 2 and gadolinia fuels in L WR environments. Advanced Fuel Pellet Materials and Fuel Rod Design for Water Cooled Reactors 2010:127. 128. Teske K, Ullmann H, Rettig D: Investigation of the oxygen activity of oxide fuels and fuel fission product systems by solid electrolyte techniques. Part I: Qualification and limitations of the method. Journal of Nuclear Materials 1983, 116(2 3):260 266.

PAGE 144

144 BIOGRAPHICAL SKETCH He received his bachelor degree in materials scie nce from Shandong University, China degree in materials engineering in May 2012. He then enrolled in the PhD program in mechanical engineering and received his Doctor of P hilosophy degree from the University of Florida in December 2015.