Microstructure Evaluation and Mechanical Behavior of High Niobium Containing Titanium Aluminides

MISSING IMAGE

Material Information

Title:
Microstructure Evaluation and Mechanical Behavior of High Niobium Containing Titanium Aluminides
Physical Description:
1 online resource (139 p.)
Language:
english
Creator:
Bean, Glenn E
Publisher:
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

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

Subjects

Subjects / Keywords:
alloy -- aluminide -- aluminum -- compression -- deformation -- gamma -- mechanical -- niobium -- sigma -- testing -- titanium
Materials Science and Engineering -- Dissertations, Academic -- UF
Genre:
Materials Science and Engineering thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract:
Ti-Al-Nb-based alloys with gamma(TiAl) + sigma (Nb2Al) microstructure have shown promise for potential high temperature applications due to their high specific strength. Recent research has been aimed towards increasing strength and operating temperatures through microstructural refinement and control. Alloys with 10 - 30% sigma-phase have been investigated, exploring relationships between chemistry, microstructure development, and flow behavior. Alloys with composition Ti-45Al-xNb-5Cr-1Mo (where x = 15, 20, 25 at%) have been produced, characterized, and tested at high temperature under compression. Processing, microstructure and mechanical property relationships are thoroughly investigated to reveal a significant connection between phase stability, morphology and their resultant effects on mechanical properties. Phase transformation temperatures and stability ranges were predicted using the ThermoCalc software program and a titanium aluminide database, investigated through thermal analysis, and alloys were heat treated to develop an ultrafine gamma+sigma microstructure. It has been demonstrated that microstructural development in these alloys is sensitive to composition and processing parameters, and heating and cooling rates are vital to the modification of gamma+sigma microstructure in these alloys. Towards the goal of designing a high-Nb titanium aluminide with ultrafine, disconnected gamma+sigma morphology, it has been established that microstructural control can be accomplished in alloys containing 15-25at% Nb through targeted chemistry and processing controls. The strength and flow softening characteristics show strain rate sensitivity that is also affected by temperature. From the standpoint of microstructure development and mechanical behavior at elevated temperature, the most favorable results are obtained with the 20 at% Nb alloy, which produces a combination of high strength and fine disconnected gamma+sigma microstructure. Microstructural analysis reveals the gamma-phase is primarily responsible for the alloy's accommodation of deformation to large strains under high temperature compression in gamma+sigma alloys with gamma-TiAl as the primary phase, and the scale and morphology of the gamma-phase correlates strongly with deformation and failure mechanisms in these alloys.
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.
Statement of Responsibility:
by Glenn E Bean.
Thesis:
Thesis (Ph.D.)--University of Florida, 2014.
Local:
Adviser: MYERS,MICHELE V.
Local:
Co-adviser: PATTERSON,BURTON ROE.

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
lcc - LD1780 2014
System ID:
UFE0046558:00001


This item is only available as the following downloads:


Full Text

PAGE 1

MICROSTRUCTURE EVALUATION AND MECHANICAL BEHAVIOR OF HIGH NIOBIUM CONTAINING TITANIUM ALUMINIDES By GLENN ESTEP BEAN JR. A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2014

PAGE 2

2014 Glenn Estep Bean Jr.

PAGE 3

To my mother, my father, Samantha and Fereshteh

PAGE 4

4 ACKNOWLEDGMENTS To my first advisor, the late Dr. Fereshteh Ebrahimi, I owe a great deal of my success and who I am today as a researcher, as a scientist, and as a person. She always had a way of bringing out the best in a person by expecting mastery of whichever topic was at hand, and leading by example. Never before have I met someone so intelligent and persistent, while also being so caring and living l ife so energetically. Fereshteh pushed me to my limits and expected me to succeed, and she is the sole reason that I was able to make the quick and difficult transition into graduate school and scientific research. She made me prove to myself that if som ething can be achieved, I can accomplish it. For this, and for her belief in me, I cannot fully express how grateful I will always be. To my advisor, Dr. Michele Manuel, I must express my thanks for easing my transition into her research group, and allowi ng me to continue to pursue my research and explore other opportunities as part of the materials design group. Because of her, I was able to not only finish my dissertation research, but also learn what it is like to create and pursue different re search p rojects and directions. I would also like to thank my fellow researchers in the MDPL. Their advice, assistance, and distractions helped me make it through this process, and keep an even keel. Particular thanks to Dr. sure of working with from the beginning, and to my advisory committee for their invaluable guidance and discussions. To my parents I would like to express my gratitude for all of their support, and even more so for all the decisions they made over the yea rs, both big and small, that allowed me to make it here. They never questioned whether I would succeed; they knew I would, so they never accepted anything but my best. Whether it was helping me

PAGE 5

5 understand homework in grade school, moving so we could atte nd better schools, or just not letting me realize how big of a nerd I really was during my childhood, I can never say thank you enough. To Samantha, I need to offer a very special acknowledgement. She believed in me and supported me throughout the stress, the long hours, the coffee, the rewrites, and my scientific ramblings even though she usually prefers to do science with living things. Now I can finally make up for lost time, and I am looking forward to a lifetime of doing just that.

PAGE 6

6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 8 LIST O F FIGURES ................................ ................................ ................................ .......... 9 ABSTRACT ................................ ................................ ................................ ................... 13 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 15 2 BACK GROUND ................................ ................................ ................................ ...... 18 2.1 Phases of Interest in Titanium Aluminides ................................ ........................ 18 2.2 Current Microstructures of Interest ................................ ................................ .... 21 2 ) Alloys ................................ ........................ 21 ................................ ............................. 24 phase ................................ ............................ 24 2.2.2.2 Past research into the Ti Al Nb system ................................ .......... 27 ................................ ................................ 29 2.3 High Temperature Deformation Mechanisms ................................ ................... 32 2.3.1 Strain Hardening ................................ ................................ ..................... 32 2.3.2 Dynamic Recovery and Recrystallization ................................ ................ 33 2.3.3 Interfacial Sliding ................................ ................................ ..................... 34 2.3.4 Dislocation Climb ................................ ................................ ..................... 35 2.4 Strengthening Mechanisms ................................ ................................ .............. 35 2.4.1 Solid Solution Strengthening ................................ ................................ ... 35 2.3.2 Grain Boundary Strengthening ................................ ................................ 37 2.3.3 Precipitation Strengthening ................................ ................................ ..... 38 3 EXPERIMENTAL METHODS ................................ ................................ ................. 41 3.1 Ra w Materials and Alloy Fabrication ................................ ................................ 41 3.1.1 Arc Melting ................................ ................................ .............................. 41 3.1.2 Externally prepared alloys ................................ ................................ ....... 43 3.2 Alloy Characterization ................................ ................................ ....................... 43 3. 2.1 Electron Probe Microanalysis (EPMA) ................................ ..................... 43 3.2.2 Differential Scanning Calorimetry (DSC) ................................ ................. 43 3.2.3 X Ray Diffraction (XRD) ................................ ................................ .......... 44 3.2.4 Optical Microscopy ................................ ................................ .................. 45 3.2. 5 Scanning Electron Microscopy (SEM) ................................ ..................... 45 3.3 Alloy Processing ................................ ................................ ............................... 45 3.3.1 Thermal Processing ................................ ................................ ................. 45

PAGE 7

7 3.13.1 Sample Machining ................................ ................................ ................. 46 3.4 Mechanical Testing ................................ ................................ ........................... 48 3.4.1 Compression and Tension Testing ................................ .......................... 48 3.4.2 Loading and Operation for High Temperature Testing ............................ 51 4 ALLOY SELECTION AND HEAT TREATMENT DESIGN IN TI AL NB (CR MO) ALLOYS ................................ ................................ ................................ .................. 53 4.1 Computational Analysis on the Effect of Nb on Phase Transformation Behavior ................................ ................................ ................................ .............. 53 4.2 Phase Fraction and Driving Force Predictions for Aging Treatments ................ 57 5 MICROSTRUCTURE DEVELOPMENT AND STABILITY IN Ti 45Al xNb 5Cr 1Mo ALLOYS ................................ ................................ ................................ 69 5.1 Thermal Processing ................................ ................................ .......................... 69 ................................ .................... 76 5.2.2.1 Aged Microstructures ................................ ................................ ..... 77 5.2.2.2 Effect of Precipitation Sequence ................................ .................... 80 6 MECHANICAL PROPERTIES OF TI AL NB BASED ALLOYS .............................. 89 6.1 Microstructure and Mechanical Properties ................................ ........................ 89 6.1.1 Flow Behavior ................................ ................................ .......................... 90 6.1.2 Deformed Microstructure ................................ ................................ ......... 94 6.2 Deformation mechanisms ................................ ................................ ................. 95 6.3 Failure Mechanisms ................................ ................................ .......................... 98 6.3.1 Effect of Microstructure Scale ................................ ................................ 102 6.4 Tensile Testing ................................ ................................ ............................... 104 7 SUMMAR Y AND CONCLUSIONS ................................ ................................ ........ 111 8 FUTURE WORK ................................ ................................ ................................ ... 115 APPENDIX A CHEMICAL ANALYSIS OF ALLOYS ................................ ................................ .... 120 B THERMAL ANALYSIS (DSC) ................................ ................................ ............... 121 C MECHANICAL TESTING DATA ................................ ................................ ........... 126 LIST OF REFERENCES ................................ ................................ ............................. 132 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 139

PAGE 8

8 LIST OF TABLES Table page 4 1 Phase transformation temperatures upon heating (C), calculated from ThermoCalc Ti Al Nb database [38] ................................ ................................ ... 56 4 2 Transformation temperatures (C) measured via DSC for quin ary alloys (addition of 5 at% Cr, 1 at% Mo), compared with transformation temperatures for ternary alloys calculated via ThermoCalc Ti Al Nb database (in parenthesis) ................................ ................................ ................................ ........ 63 6 1 Test temperature and strain rate along with the corresponding 0.2% offset yield stress and maximum stress (MPa) for 15 at% Nb samples tested in compression at 700 and 800C ................................ ................................ .......... 92 6 2 Test temperature and strain rate dependence of 0.2% offset yield stress, maximum stress (MPa), and true compressive strain to failure f ) for 20 at% Nb samples tested in compression at 700, 800, and 900C ............................... 93 6 3 Test temperature and strain rate dependence of 0.2% of fset yield stress, maximum stress (MPa), and true compressive strain to failure f ) for 25 at% Nb samples tested in compression at 700, 800, and 900C ............................... 93 A 1 Summary of EMPA results of 15Nb alloy from Certificate of analysis; Sophisticated Alloys, Inc., nationally certified external laboratory. .................... 120 A 2 Summary of EMPA results of 20Nb alloy from Certificate of analysis; Sophisticated Alloys, Inc., nationally certified external laboratory. .................... 120 A 3 Summary of EMPA results of 25Nb alloy from Certificate of Analysis; Sophisticated Alloys, Inc., nationally certified external laboratory. .................... 120 C 1 Calculated values for n, Q for compression testing of alloys at high temperature. Note change in stress exponent for 25Nb alloy at 900C; apparent change i n mechanism leads to differences in n, Q ............................ 131

PAGE 9

9 LIST OF FIGURES Figure page 2 1 Phases of interest in the Ti Al Nb system. ................................ .......................... 20 2 2 2 microstructures.. ................................ ........ 21 2 3 SEM micrographs illustrating difference in microstructure scale and ................................ ......................... 26 2 4 High temperature compression testing of 27Nb 33Ti 40Al (alloy #2) in the single aged condition. ................................ ................................ ......................... 28 2 5 Creep testing of 27Nb 33Ti 40Al samples in SA and DA condition at 1000C. .. 29 2 6 Dislocation pile up at a grain boundary.. ................................ ............................ 37 2 7 Orowan Looping. ................................ ................................ ................................ 39 2 8 Schematic representation of precipitate strengthening mechanism vs precipitate size ................................ ................................ ................................ .... 40 3 1 Images illustrating fabrication of compression samples. ................................ ..... 47 3 2 Schematic illustrating sample dimensions for sub size tensile specimen, based on ASTM E8 standard ................................ ................................ ............. 47 3 3 Sub size tensile sample, prepared for mechanical testing with 0.3 m polish. .... 48 3 4 Photographs of tensile specimen loaded in grips. ................................ .............. 50 4 1 Calculated Ti Al line highlighted ................................ ................................ ................................ .... 55 4 2 Isopleth sections of Ti Al Nb ternary along 45 at% Al indicating compositions which meet microstructure development requir ements. ................................ ..... 56 4 3 Phase fraction of stable phases with respect to temperature in Ti 45Al 15Nb alloy ................................ ................................ ................................ .................... 58 4 4 Phase fraction of stable phases with respect to temperature in Ti 45Al 25Nb alloy ................................ ................................ ................................ .................... 59 4 5 respect to temperature for 15, 20 and 25 at% Nb alloys ................................ ..... 60

PAGE 10

10 4 6 DSC data upon initial heating of alloys with 15, 20, and 25 at% Nb from phase nucleation and dissolution phase stability temperature regimes ................... 65 4 7 Comparison of DSC curves of Ti 45Al 20Nb 5Cr 1Mo alloy to calculated isopleth section for Ti 45Al 20Nb alloy, illustrating how calculated transformation temperatures were correlated with experimental results. ............ 67 4 8 Phase transforma tion temperature measured in DSC upon heating overlaid on Ti Al Nb isopleth section at 0.45 Al illustration suppression of transformation temperatures with substitution of 5 at% Cr and 1 at% Mo .......... 67 5 1 Heat treatment profile for experimental alloys. Temperatures for solution treatment (T soln ) and aging (T age ) determined by DSC analysis .......................... 70 5 2 Etching reveals solution treated and quenched microstructure of alloy with 15 phase grain size and boundaries. ................................ ........ 72 5 3 XRD of alloys with 15, 20, and 25% Nb in solution treated and quenched (S&Q) state, showing ................................ ........................... 73 5 4 As quenched microstructures of Ti Al Nb Cr grain boundaries with increasing Nb. A) 15 at% Nb, B) 20 at% Nb, and C) 25 at% Nb ................................ ..................... 74 5 5 SEM images of solution treated and quenched 25% Nb alloy. ........................... 75 5 6 grains indicative of solution treatment in the single phase region. ...................... 76 5 7 SEM micrographs of 15 at% Nb alloy aged at 1000C for 1 hour. ...................... 77 5 8 XRD of alloys with 15, 20, and 25% Nb in aged state, microstructure ................................ ................................ ................................ ..... 78 5 9 respect to Nb content. ................................ ................................ ........................ 79 5 10 SEM micrographs of 15 at% Nb alloy. ................................ ................................ 82 5 11 Microstruct phase, indicating homogeneous nucleation in the majority of the alloy with some heterogeneous nucleation between phase laths ................................ .... 83 5 12 Nb alloy . ................................ ................................ ................................ ............ 84

PAGE 11

11 5 13 Effect of aging on microhardness in 20 and 25 at% Nb alloys. 20 at% Nb alloy aged at 950C, 25 at% Nb alloy aged at 1125C ................................ ....... 85 5 14 Effect of aging on microstructure scale for 20 and 25 at% Nb alloys. ................. 87 5 15 Effect of aging on microstructure scale for 20 at% Nb alloy aged at 950C. ....... 88 6 1 TiAl matrix with disconnected Nb 2 Al particles ................................ ................................ ................................ 90 6 2 Compression testing results of 15 at% Nb alloy at strain rates of 3x10 3 to 3x10 5 s 1 ................................ ................................ ................................ ............. 91 6 3 Strain rate dependence of strength, comparing results of current testing of 15Nb alloy at 700 and 800C to TiAlNb alloy with 0.6V f phase and typical 2 microstructure ................................ ........................... 91 6 4 SEM micrographs of 15 at% Nb alloy before and after compression testing. ..... 94 6 5 Detail of sample deformed at 800C, 3x10 3 s 1 interfacial microcracking. ................................ ................................ .................... 98 6 6 Electron channeling contrast revealing with phase morphology. ................................ ................................ ........... 99 6 7 grain morphology. ................................ ................................ ............................... 99 6 8 Comparison of flow curves of 15 at% Nb alloy with varying microstructure. ..... 103 6 9 Tensile sample of 15 at% Nb alloy tested at 800C, 10 4 s 1 showing high amount of local plastic deformation; brittle tensile fracture. .............................. 105 6 10 SEM of fracture surface of tensile dogbone tested at 800C, 10 4 s 1 showing fracture initiation in high atomic weight inclusion. ................................ ............. 106 6 11 SEM of fracture surface of tensile dogbone tested at room temperature and 10 4 s 1 showing facing sides of fracture surface. Fracture initiates on sample face, and follows faceted path. ................................ ................................ ......... 107 6 12 SEM micrograph illustrating topography of phase fracture, indicative of mixed transgranular and intergranular modes. ................................ ................. 107 6 13 SEM micrograph of tensile sample face, showing crack branching near fracture surface, following path through phase ................................ 108 6 14 Detail of tensile sample face, illustrating microcracking solely through coarse phase fracture .......................... 109

PAGE 12

12 6 15 Comparison of microstructure scale and morphology .............................. 109 B 1 DSC of as cast 15Nb alloy, cycled three times. ................................ ................ 122 B 2 DSC of as cast 20Nb alloy, cycled three times. ................................ ................ 122 B 3 DSC of as cast 25Nb alloy, cycled three times. ................................ ................ 123 B 4 DSC curve upon heating of solution treated and quenched 15Nb alloy ............ 123 B 5 DSC curve upon heating of solution treated and quenched 20Nb alloy ............ 124 B 6 DSC curve upon heating of solution treated and quenched 25Nb alloy ............ 124 B 7 Comparison of DSC curves produced upon initial heating of solution treated and quenched 15, 20, and 25 at% Nb alloys ................................ .................... 125 C 1 Compression testing of Ti 45Al 15Nb 5Cr 1Mo alloy at 700C ............. 126 C 2 Compression testing of Ti 45Al 15Nb 5Cr 1Mo alloy at 800C ............. 126 C 3 Compression testing summary of Ti 45Al 20Nb 5Cr 1Mo alloy at 700C at strain rates of 10 2 10 3 and 10 4 s 1 ................................ .................. 127 C 4 Compression testing summary of Ti 45Al 20Nb 5Cr 1Mo alloy at 800C at strain rates of 10 2 10 3 and 10 4 s 1 ................................ .................. 127 C 5 Compression testing summary of Ti 45Al 20Nb 5Cr 1Mo alloy at 900C at strain rates of 10 2 10 3 and 10 4 s 1 ................................ .................. 128 C 6 Compression testing summary of Ti 45Al 25Nb 5Cr 1M o alloy at 700C at strain rates of 10 2 10 3 and 10 4 s 1 ................................ .................. 128 C 7 Compression testing summary of Ti 45Al 25Nb 5Cr 1Mo alloy at 800C at strain rates of 10 2 10 3 and 10 4 s 1 ................................ .................. 129 C 8 Compression testing summary of Ti 45Al 25Nb 5Cr 1Mo alloy at 900C at strain rates of 10 3 and 10 4 s 1 ................................ .......................... 129 C 9 Strain rate dependence of strength for 15Nb alloy, showing stress ex ponent determination ................................ ................................ ................................ .... 130 C 10 Strain rate dependence of strength for 20Nb alloy, showing stress exponent determination ................................ ................................ ................................ .... 130 C 11 Strain rate dependence of strength for 25Nb alloy, showing stress exponent determination ................................ ................................ ................................ ... 131

PAGE 13

13 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 MICROSTRUCTURE EVALUATION AND MECHANICAL BEHAVIOR OF HIGH NIOBIUM CONTAINING TITANIUM ALUMINIDES By Glenn Estep Bean Jr May 2014 Chair: Michele V. Myers Major: Materials Science and Engineering Ti Al Nb based alloys with 2 Al) microstructure have shown promise for potential high temperature applications due to their high specific strength. Recent research has been aimed towards increasing strength and operating temperatures through microstructural refinement and control. Alloys with 10 phase have been investigated, exploring relationships between chemistry, microstructure development, and flow behavior. Alloys with composition Ti 45Al xNb 5Cr 1Mo (where x = 15, 20, 25 at%) have been produced, characterized, and te sted at high temperature under compression. Processing, microstructure and mechanical property relationships are thoroughly investigated to reveal a significant connection between phase stability, morphology and their resultant effects on mechanical prope rties. Phase transformation temperatures and stability ranges were predicted using the ThermoCalc software program and a titanium aluminide database, investigated through thermal analysis, and microstru cture. It has been demonstrated that microstructural development in these alloys is sensitive to composition and processing parameters, and heating and cooling rates are vital to the designing a

PAGE 14

14 high established that microstructural control can be accomplished in alloys containing 15 25at% Nb through targeted chemistry and processing controls. The strength and flow softening characteristics show strain rate sensitivity that is also affected by temperature. From the standpoint of microstructure development and mechanical behavior at elevated temperature, the most favorable results are obtained with the 20 a t% Nb alloy, which produces a combination of high strength and fine disconnected microstructure. phase is primarily responsible for high temperature TiAl as the primary phase and the scale and morphology of the phase correlates strongly with deformation and failure mechanisms in these alloys

PAGE 15

15 CHAPTER 1 INTRODUCTION Advancements in the research and development of lightweight intermetallics have led to increases in the overall efficiency and performance of jet turbine and power production engines [ 1 5 ] TiAl + 2 Ti 3 Al titanium aluminide s have increased high temperature strength while demons trating excellent properties below 750C, owing primarily to alloying and microstructural control [ 2 4 6 9 ] Alloying with elements such as Cr, Mo, Nb, and Ru have been shown to improve the mechanical properties by modifying phase transformations and their resultant microstructures [ 7 10 ] In particular, additions of Nb in concentrations up to 10 at.% have resulted in 2 alloys, as well as improvement in strength, ductility [ 6 ] and oxidation resistance [ 11 ] Further increasing the Nb content leads to the development Nb 2 Al phase, which can be utilized to improve high temperature properties of titanium aluminides [ 1 12 14 ] It has been shown that two phase exhibit high temperature strength superior to that of 2 alloys [ 1 12 ] but have limited room temperature ductility and fracture toughness phase [ 14 ] Toughness was enhanced by incorporatin g ductile phase particles into the matrix [ 14 ] or by producing [ 13 ] D eformation of alloys with 6 0% phase [ 1 ] phase) to

PAGE 16

16 mechanisms operant at high temperature, and deformation should be more dominated phase. phase volume fraction Ti Al properties at room temperature [ 10 15 16 ] microstru phase precipitates improves strength and ductility at room temperature through grain refinement and precipitation strengthening [ 16 ] Additionally, these precipitates may help to pin grain boundaries at elevated temperatures, reducing the extent of grain boundary sliding, which has been found to dominate high temperature mechanical behavior [ 12 17 18 ] High temperature strength is typically diminished due to the increased influence of grain boundary sliding in finer microstructures [ 17 19 21 ] howev er by controlling the volume phase while maintaining a disconnected morphology, high temperature strength may be increased without significantly diminishing the improvements gained at room temperature. Understanding the effects of alloying on microstructure and high temperature deformation of these materials is of critical importance in order to enable the production of alloys with tunable microstructures, which can lead to desirable high temperature properties. In order to produce an ul trafine microstructure, alloying additions of Cr and Mo are necessary to modify phase stability and transformation kinetics, which can in turn be utilized to control microstructural scale [ 10 15 22 23 ] Cr is a strong phase stabilizer and is therefore favorable for microstructural development [ 10 ] but the retention of phase in the final microstructure can be detrimental for high temperature properties [ 7

PAGE 17

17 11 ] Mo is also a phase stabilizer and should partition between phases due to its large size, slowing down kinetics during microstructure development [ 8 ] While preliminary testing of Ti Al Nb Cr temperature has shown promising strength and fracture toughness [ 16 ] their high temperature properties are largely unknown.

PAGE 18

18 CHAPTER 2 BACKGROUND 2 titanium aluminides are used in the low temperature regions of jet turbines, such as the last stages of the low temperature turbine in the General Electric GeNx engine [ 5 ] In this application, weight reduction translates directly into improved efficiency and cost savings. For these reasons, the development of titanium aluminide alloys is of interest, and the Ti Al Nb system shows promise for improved high temperature strength [ 1 2 6 24 ] A great amount of research has gone into the development of titanium aluminides over the past few decades, focused mainly on alloys with TiAl + 2 Ti 3 Al microstructures, including research into chemistry processing properties relationships [ 25 27 ] The effects of alloying additions such as Nb, V, Ta, and Cr on mi crostructure development and mechanical behavior of 2 alloys was investigated [ 28 32 ] as well as the introduction of Nb or Ta rich phase and its contribution to hig h temperature mechanical beha vior [ 11 12 33 34 ] Crystal structures 2 ha 2 interface [ 25 29 35 37 ] Throughout the development of these alloys, modification of composition, processing, and microstructure have led to improvements in alloy performance [ 2 4 7 16 26 27 33 ] 2.1 Phases of Interest in Titanium Aluminides Important phases in the Ti Al Nb system that are commonly seen in alloy 2 [ 1 2 4 10 11 38 ] Ti phase is a BCC solid solution that is stable in the binary TiAl system at elevated temperatures for Al content below approximately 45 at.% and is important for forming operations at

PAGE 19

19 elevated temperature and machining if retained to r oom temperature. It is also important in the development of some two phase microstructures in titanium aluminides [ 16 23 39 ] which will be discussed further in the following sections. However, it has been observed that the r phase in the final microstructure is detrimental to toughness at ambient temperatures and elevated temperature deformation resistance [ 4 7 11 ] In order to improve the high temperature performance of TiAl based alloys, research has been done to optimize alloying chemistry and microstructure, with two phas e microstructures being commonly employed in order to improve both ductility and strength. Titanium aluminides should exhibit a balance of room temperature ductility and fracture toughness, as well as high temperature deformation resistance [ 4 7 ] which has been the objective of recent research in this sy stem [ 1 10 14 16 22 38 40 ] Th TiAl phase is an intermetallic phase with face centered tetragonal (FCT) L1 0 structure [ 2 41 ] in which the Ti and Al atoms occupy alternating layers. Stoichiometric TiAl has a c/a ratio of 1.015 ( with a=4. 00 and c=4.06 ) which varies from 1.01 1.03 with Al content fro m ~35 57at% [ 27 ] Single TiAl at room temperature has poor ductility and fracture toughness [ 2 ] Because of its poor 2 Ti 3 Al /TiAl 3 phase, an ord ered intermetallic with a hexagonal D0 19 structure [ 2 7 41 ] Thes e two 2 improves ductility [ 2 ] will be discussed in the following section. Nb 2 Al phase is very complex, with a unit cell containing 30 atoms [ 42 43 ] This phase is hard and brittle, exhibiting no ductility at temperatures up to 700C even when toughened by a more ductile second phase [ 1 12 14 ] However, the effects of a

PAGE 20

20 phase precipitate on strength and ductility are unknown, and wi ll be the focus of this study. In addition to in the Ti Al Nb system, the presence of phases has been documented in Fe Cr, Mo; Zr Al; Mg Cu, Zn, Ni; and other transition metal alloy systems [ 35 43 48 ] where the complex tetragonal crystal structure and its effect on microstructure have been studied. Applications of the phase range from strengthening precipitates [ 1 16 36 43 ] to thin films for diffusion barriers [ 49 ] All of these phases are portrayed schematically in Figure 2 1. Figure 2 1 : Phases of interest in the T i Al Nb system. A) BCC, B) TiAl C) 2 Ti 3 Al /TiAl 3 and D) Nb 2 Al phases. Adapted from [ 2 4 43 ] It is worth noting at this point that all compositions discussed in this document will be expressed in atomic percent (at%).

PAGE 21

21 2.2 Current Microstructures of Interest 2.2.1 Gamma P l 2 ) A lloys 2 titanium aluminides, research has been done to optimize the alloying chemistry and microstructure, with two phase microstructures being commonly employed in order to i mprove their ductility and strength [ 2 4 6 9 ] 2 alloys include near gamma (NG), fully lamellar (FL), nearly lamellar (NL), duplex (DP), pseudo duplex (PS DP), and equiaxed [ 2 4 6 8 ] Examples of some of these microstructures can be seen in Figure 2 2. Figure 2 2 : SEM micrographs illustrating + 2 microstructures A) nearly lamellar, B) fully lamellar, C) duplex, and D) pseudo duplex Adapted from [ 41 50 ]

PAGE 22

22 By modifying the morphology of these alloys, its mechanical properties at room 2 microstructure 2 platelets, and is generally characterized by h igh fracture toughness, and high resistance to crack propagation when compared with DP microstructures [ 2 42 ] The DP microstructure is a combination of single 2 colonies, and is characterized by a higher tensile strength, higher ductility, and longer fatigue life. This microstructure is developed upon cooling phase field at moderate cooling rates with the transformation of prior + 2 [ 51 ] An equiaxed microstructure can also be formed, with grains and large 2 particles at the grain boundaries, and exhibits higher strength and ductility than DP microstructure, ho wever the FL microstructure has been shown to significantly outperform the equiaxed microstructure [ 8 11 51 ] Effect of Alloying Additions In t he d evelopment of TiAl based alloys, there are many competing parameters must be taken into account when choosing alloying additions. Some of these issues include changes in phase stability, precipitates, microstructure morphology, transformation kinetics and mechanical properties Niobium additions strengthen TiAl based alloys, improve high temperature def ormation properties, and improve oxidation resistance [ 1 11 24 ] The addition of Nb increases 2 content and refines th e microstructure increasing strength due to the Hall Petch mechanism [ 11 19 ] Nb is assisted dislocation climb and reducing steady state creep rate and increasing high temperature strength [ 11 52 ] Alloys with additions of 4 12 at% Nb and Ta produce + 2

PAGE 23

23 microstructure, and show that for all amounts of Ta and Nb additions, the alloys with lower Al content (45 vs. 47 a t%) exhibit higher strength, but lower plasticity, with a quaternary alloys of Ti 45Al 4Nb 4Ta showing the best combination of strength and plasticity [ 53 ] The addition of stabilizing elements such as Cr Mo, Si, Ta, and W [ 37 54 57 ] can be essential in retaining this phase upon quenching, which is an integral part of microstructural control, as will be discussed in the following section [ 10 23 39 ] In this work, Cr will be used, as it is a strong stabilizer, h owever Cr can also have a detrimental effect on high temperature properties above 700C due to the retention of phase, which has a very open structure at hig h temperature [ 7 11 58 ] Conversely, at temperatures of 700C and below the presence of phase improves these properties [ 2 7 ] Cr is also known to increase ductility in based alloys [ 2 3 7 ] Molybdenum is alloyed into titanium aluminides due to its low diffusivity, which limits the kinetics of phase transformations by partitioning between phases [ 16 ] The addition of 0.66 at.% Mo to a Ti 24Al 17Nb + 2 alloy improves high temperature deformation resistance, and increasing the Mo content to 2.3 at.% results in an additional order of magnitude enhancement in steady state creep rate but embrittles the alloy [ 8 ] This improvement in high temperature deformation is also attributable to the reduction in 2 2 interfacial sliding due to the lower volume fraction of the phase. It was found that the alloying with 1 at% Mo yields comparable improvements at elevated temperatures without the accompanying embrittlement suffered with 2.3 at% Mo [ 8 ] For stabilizing

PAGE 24

24 properties of Mo are of the most interest [ 8 16 39 ] form the targeted microstructures. 2.2.2 Gamma Plus 2.2.2.1 phase By increasing the Nb content in Ti Al Nb based alloys, high temperature strength is improved through both solid solution strengthening effects of Nb as well as precipitation strengthening through the introduct phase in lieu of the 2 phase [ 4 6 24 ] Due to the potential for significant improvements in high temperature In the Ti Al Ta system it was also found that a (Ti,Ta) 2 Al phase can be produc ed in Ta rich alloys, and research by Weaver and Kaufman shows that the formation of and composition [ 59 ] Kim et al. showed that Ti Al Ta alloys containing 32 40 at% Al and approximately 25 at% Ta produced a fine two phase + microstructure when splat quenched alloys were re annealed below 1200C [ 60 ] Additionally, it was found that alloys with composition Ti 50Al ch exists as a single phase [ 61 ] Investigation into the nature of phase transformations and relationships in Ti Al based systems [ 15 38 59 61 66 ] is instrumental in the design of alloy composition and microstructure. Alloys in the Ti Al Nb system phase prove to significantly improve 2 a lloys, however room temperature ductility and fracture toughness are somewhat limited [ 12 14 40 ] When tested at

PAGE 25

25 700C the alloy (27Nb 33Ti 40Al) suffered brittle failure and at 900C there was approximately 10% total strain before failu re [ 1 ] At 1000C, a yield strength of approximately 450 MPa was achieved at 3.5 x 10 4 s 1 which is a large improvement over 2 alloys which generally have yield strength of ranging from 275 400 MPa at similar strain rates at 700C [ 1 2 4 ] Analysis of samples strained to 55% showed phase, which was more prominent at lower strain rates [ 1 ] based alloys is limited, due to the fact that at 60% phase is the connected phase, acting as the load bearing matrix phase and determining the mechanical properties of the alloy, and it has been shown that by including ductile second phase particles of phase, fracture toughness can be enhanced [ 1 13 14 ] It has also been proven that the scale and morphology of the microstructure has a significant effect on the strength and ductility of the alloy, and can be changed by modifying the heat treatment schedule [ 1 12 ] An alloy with composition Ti 40Al 27Nb was solution treated at 1450C for 2 hours then oil quenched. If subsequent single aging (SA) was carri morphology. However if a two step double aging (DA) treatment was carried out, first at 1300C for 12 hours and then 1200C for 14 hours, the result is a much coarser pha se morphology. Comparison of these microstructures can be seen in Figure 2 3. microstructure, then during aging at 1200C, the [ 1 ] This coarse disconnected microstructure is due to the improved diffusion at a higher aging temperature and additional time allowed for the coarsening and coalescence of

PAGE 26

26 phase but d espite the large difference in morphology, there is nearly equivalent V f phase in both microstructures [ 1 12 ] Figure 2 3: SEM micrographs illustrating difference in microstructure scale and morphology between microstructures. A) single aged; B) dou ble aged [ 1 ] (Courtesy of B.J.G. deAragao) The single aged alloy can be approximated as a ductile phase toughened composite, and it has b een shown that the high temperature deformation behavior is insensitive to the properties of the weak, disconnected phase when there is a strong connected matrix [ 1 ] From analysis of the microstructure and flow properties, it was phase in the single aged microstructure [ 1 12 ] This grain boundary sliding was seen to a greater extent in the finer SA microstructure due to the greater amount of / interfaces. In contrast, the active mechanisms in th e double aged alloy with a disconnected DP phase, as well as grain boundary sliding [ 1 12 ] Additionally, from compression testing at 1000C, it was phase occurred which is indicative of the characteristic serrated displayed in the flow

PAGE 27

27 curves [ 1 12 67 68 ] Tensile behavior of the SA alloy is presented in Figure 2 4, which shows the transition from brittle to ductile behavior. 2.2.2.2 Past research into the Ti Al Nb s ystem In addition to the previously discussed research, there were many investigations into the thermal, microstructura l, and mechanical characteristics of alloys in the Ti Al Nb system under the direction of Dr. Fereshteh Ebrahimi [ 1 10 12 16 23 33 38 40 52 65 66 69 70 ] Early research into the development of high temperature Nb based alloys was conducted Gomez [ 33 ] His study was into alloys with compositions of 27Nb 33Ti 40Al (alloy #2) and 42Nb 28Ti 30Al (alloy #4) (all composition s in at%) which were ar c melted, solution treated in the single in alloys #2 and #4, respectively. In compression testing, alloy #2 yielded preceding fracture at 700C, showing higher fracture strengths of 1200 1900 MPa, compared with alloy #4, which fractured without plastic deformation with strengths ranging from 1100 1500 MPa at 900C. Fracture strength was found to be insensitive to microstructure in bo th alloys, though yield strength of alloy #2 decreased with increasing phase content [ 13 14 33 40 ] Researc h conducted by u ndergraduate researcher B. phase, while fracture toughness is inversely phase volume fraction [ 36 ] Toug h ness was also evaluated through testing of 4 point bending specimen with chevron notch, and values calculated using the curve fitting model and indentation data were found to be within 20% of the measured values [ 14 71 ] Th ere was increased

PAGE 28

28 K1C values specimen from SA to DA samples, from 4.4 to 6.3 due to the phase in the DA alloys which does not continuously propagate a crack. Fractography shows that cracks propagate intergranularly a phase concentration were encountered indicating both the potential of high internal stresses at the boundary as well as the brittle nature of the large phase precipitates [ 1 33 ] Fig ure 2 4: High temperature compression testing of 27Nb 33Ti 40Al (alloy #2) in the single aged condition Adapted from [ 12 ] Following t his, research was conducted by m a de Aragao into the high temperature deformation of 27Nb 33Ti 40Al alloys, as detailed in section 2.2.2. The major findings of this research show that compression testing at

PAGE 29

29 phase also underwent fragmentation during deformation, and the main mechanism responsible for this behavior was grain boundary sliding. Creep testing was also carried out in uniaxial compressive stress for both single aged and double aged microstructures yielding very similar steady state creep rates, as seen in F igure 2 5 [ 1 12 13 ] Figure 2 5: Creep testing of 27Nb 33Ti 40Al samples in SA and DA condition at 1000C Adapted from [ 1 ] development Since the high V f of conn phase led to fragmentation and limited ductility, alloys with lower volume fractions have been explored [ 10 15 16 23 39 70 ] By reducing grain size, strength and ductility can both be improved [ 19 72 ] and by m phase, the brittle nature of the alloy c an be mitigated.

PAGE 30

30 Design of a duplex phase precipitates as well as phase grain boundaries may be effective in achie ving these improvements [ 13 73 ] Already, several alloys have been explored with approximately 0.15 V f phase which have shown an increase in both strength and fracture toughness, and has shown that by microstructural refinement, Ti Al Nb alloys [ 10 15 16 ] through suitable composition and heat treatment control. In order to create the desired ultrafine micr ostructure, its evolution must be carefully controlled. For this reason, a two step heat treatment was developed, which will be discussed in further detail in following chapters. Unlike the previously developed alloys, these alloys do not cool to room te mperature as single phase upon casting. Therefore, the alloy is first solution treated in the single phase solid solution region, then quenched to room temperature to retain the phase. This step is to ensure homogeneity of the alloy and to enable t he phase phase, producing an ultrafine microstruct ure. Two preliminary alloys have been investigated, with compositions Ti 45Al 18Nb and Ti 45Al 27Nb. In these alloys, there has been difficulty in retaining pure phase phase precipitated. In order to retain phase upon quenching, alloying with Cr, a known stabilizer was explored [ 4 10 ] Replacing 5 at% Nb with Cr has been shown to effectively stabilize the

PAGE 31

31 phase in the Ti 45Al 22Nb 5Cr alloy, but was ineffective in alloy Ti 45Al 13Nb 5Cr, phase laths at the phase grain boundar ies [ 10 23 39 ] In Ti 45Al 22Nb 5Cr, the phase is fully retained upon quenching to room temperature but phase laths still grew f rom the grain boundaries immediately upon aging. Also, there is a very small range of aluminum composition in which retaining single phase upon quenching is poss ible phase precipitation. Varying the aging temperature has been shown to change the microstructural scale as well. When Ti 45Al 22Nb phase was produced. When aged at 1050C, the resulting microstructure was coarser with phase, though the V f phase was the same, approximately 30% [ 16 ] Along with reducing the amount of Nb in the alloy when compared with past research, further alloying with Mo was conducted. The purpose of adding Mo to the alloy is to slow the kinetic s of phase transformations upon quenching and at elevated temperatures, allowing for stability in the microstructure and high temperature performance. This was studied with composition Ti 45Al 14Nb 5Cr 1Mo. Although the phase upon quen ching was not fully suppressed, a similar method to that used with Ti 45Al 22Nb 5Cr may be used to test the performance of the alloy without the effects of grain boundaries. That is, samples can be machined such a way as to avoid the presence of prior p hase grain boundaries In this way, Ti 45Al 14Nb 5Cr 1Mo produced an ultrafine microstructure, leading to compressive ductility of 15 30% and yield strength of about 1300 MPa at room temperature, and fracture toughness increased to about 15 [ 16 ]

PAGE 32

32 temperature has shown promising strength and fracture toughness [ 16 ] their high temperature propertie alloys with 0.6 V f phase [ 1 12 ] phase) mechanisms operant at high temperature. Deformation should be more heavily influenced phase, but the importance of competing deformation mechanisms is unknown. It is the goal of this research to investigate the microstructure microstruc ture through compression testing and the evaluation of microstructural evolution occurring during heat treatment and deformation. 2.3 High Temperature Deformation Mechanisms 2.3.1 Strain Hardening Strain hardening is the increase in dislocation density as strain is induced in a material. Strain hardening decreases with temperature, but increases with strain rate, as described by the Orowan equation ( 2 1) for a polycrystalline metal, where is the shear strain rate [ 20 21 ] As tensile strain rate ( ) increases, dislocation density ( ) or average dislocation velocity ( ) must also ( ). Since dislocation velocity is controlled by vacancy diffusion, at constant temperature the density of mobile dislocations must increase with increasing strain rate

PAGE 33

33 [ 20 74 75 ] At higher temperature, vacancy diffusion is easier so dislocation velocity can be higher resulting in a lower work hardening rate, as fewer dislocations are necessary to maintain the given strain rate [ 2 11 19 20 74 75 ] 2.3. 2 Dynamic Recovery and Recrystallization Dynamic recovery and recrystallization are important softening mecha nisms that can occur during deformation above 0.5 homologous temperature in metals. Dynamic recovery is defined by the movement of dislocations, during plastic deformation, to lower energy arrangements, often to grain boundaries, or to form sub grain stru ctures [ 21 ] Similar recovery can occur under static conditions during annealing after plastic deformation, but during deformation at high temperature the effects of dynamic recovery increase due to increased dislocation mobility, and reduces the effective strain hardening rate in the material [ 21 ] It has been found that dynamic recovery occurs more strongly in materials with high sta cking fault energy, suggesting that thermally activated cross slip is the controlling mechanism [ 21 ] In materials where dynamic recovery does not occur quickly, dynamic recrystallization can occur. Softening due to dynamic recrystallization results in a sharper drop in stress, and dependent on the amount of strain requir ed to fully recrystallize the metal, the flow curve can take on one of two characteristic shapes. These characteristics include a peak stress value followed by flow softening and oscillations in the flow stress [ 76 ] or a peak stress followed by continuous flow softening to a steady state value [ 21 ] During dynamic recrystallization, dislocation free grains are formed and as the material is straine d further, the dislocation density increases in these grains up to a critical value, at which point recrystallization is initiated once again [ 76 ] If one wave of

PAGE 34

3 4 recrystallization is completed before the critical strain to nucleate recrystallization is reached again, this leads to periodic oscillations in the flow stress. If ins tead the critical strain to complete recrystallization is greater than that to nucleate recrystallization, then a second wave of recrystallization will begin before the first is completed, leading to a flow curve with a peak stress and softening to a stead y state value. Another factor that can affect the dynamic recrystallization behavior of this material is alloying additions [ 67 76 ] If alloying additions are in solid solution the solute drag effect may hinder the gro wth of recrystallized grains in addition to potential for grain boundary pinning from the Nb 2 Al particles [ 1 19 21 74 ] Addi tionally, particle stimulated nucleation (PSN) of dynamic recrystallization is a phenomenon observed in precipitation and dispersion strengthened alloys with particles approximately 1 m or larger [ 74 ] 2.3. 3 Interfacial Sliding Strain accommodation can also occur by interfacial, or grain boundary sliding, especially at elevated temperature, and for very fine grained metals [ 76 77 ] Grain b oundary sliding is the shear induced deformation between two adjacent grains, resulting in deformation that is highly local to the interface between the grains. However, it is not able to operate as a singular deformation; interfacial sliding must be acco mpanied by either grain boundary diffusion or volume diffusion through the grains [ 77 ] If diffus ional flow operated independently, it would lead to grain elongation and separation under applied strain, but grain boundary sliding is able to accommodate the deformation and keep the grains together [ 77 ]

PAGE 35

35 2.3.4 Dislocation Climb At room temperature, deformation is typically controlled by dislocation glide, but at elevated temperature, edge dislocatio ns can move out of their slip planes through dislocation climb [ 19 21 78 ] This thermally activated mechanism is controlled by diffusion, and allows edge dislocations to move perpendicular to its slip plane in order to reach a parallel plane directly above or below. Positive climb occurs when a vacancy diffuses to the disloca tion and an atom moving to the vacant site, where negative climb is the movement of atoms to the extra half plane, creating vacancies that diffuse away from the dislocation [ 19 21 ] Positive climb is associated with a compressive stress in the slip direction, while negative climb is associa ted with a tensile stress [ 19 ] Both positive and negative climb result in the formation of jogs [ 19 21 78 ] which act as sources and sinks for vacancies [ 78 ] 2 .4 Strengthening Mechanisms 2.4 .1 Solid Solution Strengthening One method of strengthening observed in metals is solid solution strengthening, accomplished through the introduction of s olute atoms into solution. These solute atoms alter the crystal structure of the material and, when compared with a pure metal, increase strength. Solute atoms can either occupy the same lattice sites as the solvent metal, known as substitution al atoms in the lattice, known as interstitial solid solution. Interstitial solid solutions occur when the solute atom is substantially smaller than the metal it is being alloyed with, commonly occur ring elements such as carbon, boron, nitrogen, or oxygen. Substitutional solid solution occurs when solute and solvent atoms are close in size, such as additions of niobium or chromium to titanium alloys.

PAGE 36

36 The ability of elements to form substitutional sol id solutions is governed by the Hume Rothery rules, which state that the controlling factors are relative size, valence, electronegativity, and crystal structure of the two atoms [ 19 ] The work of Hume Rothery found that in solid solutions, a size difference of more than 15% will reduce solubility to less than 1%, but when valence is taken into account, it is found that metals of lower valence have less solubili ty in metals of higher valence when compared to the reverse situation. Elements must also have similar electronegativity, as those that do not are more likely to form intermetallics than solid solutions. Lastly, in order for there to be complete mutual s olubility, the elements must have the same crystal structure. Solute atoms can strengthen alloys by interacting with dislocations in a few ways. Substitutional solutes cause local changes in lattice parameter, and the resulting misfit stress causes an ela stic interaction with the dislocation line. This local change in crystal structure can also affect the local modulus of the material. Solute can also impede dislocation motion by their attraction to the strain fields surrounding the dislocations. Since there is a localized compression on the matrix above a dislocation line, and a tensile state below it, solute atoms are attracted to this site. Smaller atoms can reside above the dislocation line in order to reduce the amount of compressive strain experie nced in the matrix. Likewise, larger solutes are attracted to the area below dislocation lines. This attraction between solutes and dislocations causes the motion of dislocations through the metal to be slowed. One typical case of such attraction is wit h Cottrell atmospheres, where carbon interstitial solutes are attracted to the dislocation it must overcome this energetically favorable condition, or mo v e slowly en ough to allow

PAGE 37

37 for the diffusion of the Cottrell atmosphere along with the dislocation. Solutes can similarly interact with other dislocation line phenomena such as stacking faults and anti phase boundaries, impeding their motion through the alloy during d eformation. 2.3.2 Grain Boundary Strengthening Another method of strengthening commonly employed in alloy development is grain boundary strengthening. Grain boundaries act as obstacles to the movement of dislocations, so by decreasing grain size strength increases due to the increase in the number of these obstacles. The amount of strengthening can be predicted using the Hall Petch relationship which describes that yield strength is inversely proportional to grain size. In this equation, is the yield strength of the material, is the friction stress representing the overall resistance of the crystal lattice to dislocation movement, is a material constant, and is the average diameter of the g rains [ 19 ] ( 2 2 ) This relationship is based on the dislocation pile up theory, whereby dislocations pile up and must induce a critical stress in the neighboring grain in order to pro duce additional dislocations in that neighboring grain that will then propagate Figure 2 6: Dislocatio n pile up at a grain boundary. A dapted from [ 19 ]

PAGE 38

38 As the size decreases, the pile up stress induced by the dislocations decreases, so an increased amount applied stress is required to cause slip in the neighboring grain. Dislocation pileup is illustrated schematically in Figure 2 6 below, where L is the dislocation pile up length, and the location in the neighboring grain (point P ) is described by r 2.3.3 Precipitation Strengthening To increase the strength of metal alloys, it is common to introduce second phase precipitates into the microstructure. These particles increase the strength of the material by impeding the motion of dislocations through the matrix. When a dislocation encounters a precipitate, it must either cut or bypass the particle. For the dislocation line to shear a precipitate, the precipitate must be deformable and have a coherent or semi coherent interface with the matrix. Important factor s that determine the ability of p recipitates to be cut are its P e i rls stress, elastic modulus, and stacking fault energy. These factors affect the dislocation line energy as it passes through the precipitate and determine if it is energetically feasible to cut the particle rather than bypass it. For a dislocation line to bypass a precipitate it must bow around it and form a loop, and each loop m ake it more difficult for the next dislocation line to bypass the particle by effectively increasing the precipitate size, and is known as the Orowan strengthening mechanism for dislocation movement through a precipitate containing matrix illustrated in F igure 2 7 Other precipitate strengthening mechanisms include misfit strain, chemical hardening, and strengthening due to ordered structures. Misfit strains, due to the mismatch in size of the particle and matrix, can either attract or repel dislocations increasing the force required for them to approach or escape the precipitate as it bows.

PAGE 39

39 Strengthening due to ordered particles is also important in high temperature metals. As the dislocation line passes through an ordered precipitate, anti phase boun daries (APBs) will be formed as it is sheared, which can increase the energy of the dislocation line and make shearing more difficult [ 19 ] Also, by shearing a particle, a step is created, which increases the surface energy of the particle. This may lead to shear localization, since the cross sectional area of the particle along that plane is smaller, so it is requires less force to propagate subsequent disloca tion lines. Figure 2 7: Orowan Looping. A dapted from [ 21 ] Small pa rticles are more effective at strengthening by dislocations cutting them, but as they become larger and more difficult to cut, dislocations must bypass them, and larger particles are effective for Orowan strengthening This tradeoff is illustrated in Figu re 2 8 While cutting of precipitates depends highly on precipitate properties, bowing of dislocations requires a non deformable particles and is dependent on precipitate spacing, as described in Equation 1 2, where is the amount of stress required t o bow the dislocation line between the precipitates, G is the shear modulus, b the burgers vector of the dislocation, and the inter particle spacing. (1 2)

PAGE 40

40 Controlling the size of precipitates through heat treatment, therefore, allows for maximization in the effectiveness of precipitation strengthening. This becomes an issue with over aging of precipitates, especially for high temperature materials that may experience operating temperatures near those used in heat treatment. Therefore, it important to ensure stable precipitates in order to retain the desired strengthening Figure 2 8: Schematic representation of p recipitate strengthening mechanism vs precipitate size One very important aspect of precipitation strengthening to note that the dislocation mechanisms discussed in this section are relevant for deformation at lower temperature. During high temperature deformation, dislocation climb is active, which changes the dislocation particle i nteractions, may limit the amount of dislocation cutting and bowing that occurs, dependent on particle size and spacing, and particles may act as a source and sink of dislocations [ 19 21 76 ] Larger particles are more difficult to overcome via climb, and have a greater effect on deformation and recrystallization behavior in high temperature alloys [ 21 76 ]

PAGE 41

41 CHAPTER 3 EXPERIME NTAL METHODS 3.1 Raw Materials and Alloy Fabrication Metals for materials for alloy fabrication were obtained from a combination of sources. Elemental forms of Ti, Al, Nb, Cr, and Mo were purchased through Alfa Aesar in the following forms. Ti granules, 9 9.99% purity (metals basis excluding Na and K). Al wire, 2.0mm (0.08in) diameter, annealed, Puratronic 99.999% purity. Nb wire, 1.0mm (0.04in) diameter, 99.8% purity (metals basis). Chromium lumps, 99% purity. Molybdenum wire, 1.0mm (0.04in) diameter, 99.95% purity. Raw material preparation and cleaning will be detailed in the following section. 3.1.1 Arc Melting Raw materials were arc melted to fabricate 5 25g buttons and bars using a MRF ABJ 900 large bell jar arc melter with water cooled copper hear th and Miller GoldStar 652 power supply with 800 ampere capacity. Non consumable tungsten electrode arc melting was conducted under approximately 5 psi positive pressure argon (Ar) environment in order to minimize the potential for oxidation. Pressure was set via pressure relief valve, with flow rate of Ar between 5 30 mL/min, as regulated by secondary operator of arc melter. In order to ensure sufficient gas flow, pressure was set to 25 30 ps i at the argon tank regulator. Arc melting was selected as the alloy fabrication method of choice primarily due to the very high melting points of the metals being alloyed, as well as their susceptibility to oxidation. Melting points of pure Ti, Nb, Cr, and Mo are all well in excess of 1200 C, which is the limit of the available box furnaces. Additionally, these elements have a

PAGE 42

42 high reactivity with oxygen at elevated temperature, and the formation of oxides during the alloy fabrication process would seve rely alter its material properties. Raw materials were prepared by first washing in a soap water solution and rinsing, then washing an ultrasonic bath in 3% Nital, and rinsed with 2 Propanol. This was done to remove any dirt, oils, or oxide that the mater ials may have encountered or developed during handling. The exception to this is Mo, for which the Nital wash is omitted since it would cause a surface oxide to evolve Metals were then dried completely before being placed into the arc melter. Nb, which was in either shaving or powder form, was placed at the bottom of the hearth with pure Ti layered above so the arc would not blow them out of the way upon initial melting. Due to the complete mutual solubility of these two elements [ 79 ] they will combine readily upon melting. Al was placed on the side so that the arc would not directly strike it, but would be indirectly heated when the Nb and Ti are melted and then be drawn into the solution. Cr and Mo w ere scattered on top, as there were relatively small amounts of each of these alloying elements and there is not a risk of these heavy metals being perturbed by the arc upon melting. Melting of the alloys was conducted at 450 amperes current and pure Ti sa mples were used for oxygen gettering. After starting the arc, the getter was first melted, and then each sample was melted for 15 20 seconds, with the getter being melted again between each one in order to reduce the risk of oxidation. Samples were check ed after each melting to ensure complete melting, no oxidation, and they were flipped over to prepare for subsequent melting. Each sample was melted a total of six times in order to

PAGE 43

43 ensure homogeneity. After melting was completed, composition was confirm ed by electron probe microanalysis (EPMA). 3.1.2 Externally prepared alloys Once alloy compositions were selected from evaluation of the small scale samples prepared via arc melting, larger 250g buttons were fabricated external ly. Alloys of composition Ti 45Al xNb 5Cr 1Mo (where x = 15, 20, 25 at%) were fabricated by Sophisticated Alloys Incorporated with alloy composition verified independently via inductively coupled plasma ( ICP ) mass spectroscopy Each alloy composition was within 0.4 at% of the nomina l composition. A summary of ICP results can be found in Appendix A. These large buttons were used for thermal and mechanical analysis. 3.2 Alloy Characterization 3.2.1 Electron Probe Microanalysis (EPMA) EPMA of the experimental alloys was conducted on a JEOL Superprobe 733 at the Major Analytical Instrumentation Center (MAIC) at the University of Florida. The system was calibrated with Ti, Al, Nb, Cr, and Mo standards, and was operated by Wayne Acree of the MAIC staff. Readings were taken using 1 m spo t size at sets of ten randomly chosen spots on each alloy to obtain an average bulk reading. Since microstructure scale is sufficiently smaller than spot size, an average of these readings results in reasonable bulk composition quantities. 3.2.2 Different ial Scanning Calorimetry (DSC) Samples of arc melted buttons were sectioned, heat treated, and tested using a Setaram Setsys EVO 1750 Differential Thermal Analysis (DTA)/Differential Scanning Calorimetry (DSC)/Thermal Gravimetric Analysis (TGA) system wit h DSC 1600 rod. Testing was conducted under flowing helium (He) environment with temperature

PAGE 44

44 ramped at 10 C/min between 600 1550 C. Calibration of the DSC system was completed through the Setsoft2000 software for varying temperature ranges using indium (In), Al, copper (Cu), and nickel (Ni) standards. Samples ranging from 100 200mg were prepared with a very flat bottom face in order to ensure good thermal contact with the alumina 100 L crucibles. Before each run, an empty crucible is cycled through th e program and this output is used as a baseline subtraction for the data collected in the subsequent run using the same crucible loaded with a sample. Prior to heating the sample chamber, it is pumped and purged with He gas four times to clear any atmosph ere or contaminants that may have entered the system upon sample loading. DSC was conducted on 15, 20, and 25 at% Nb samples in the as cast and solution treated conditions. Samples for DSC analysis were cut from the solution treated alloy using a low sp eed diamond saw, and DSC of these alloys was conducted under the same conditions as described above, but the first heating cycle, which shows the phase transformations from the metastable quenched in phase, was used to determine metastable phase transforma tion temperatures. Analysis of DSC results paired with XRD was used for determination of phase transformation temperatures and stability regions of these alloys in order to design heat treatments. 3.2.3 X Ray Diffraction (XRD) Phase identification was conducted on alloy samples before and after heat treatment using an XPert Powder X Ray diffractometer. Samples were powdered in preparation for XRD analysis using a mortar and pestle, and double sided tape was used to affix the resultant powder to a glass slide for mounting in the diffractometer. 120.

PAGE 45

45 3.2.4 Optical Microscopy Alloys were sectioned using an Allied High Tech low speed saw with low concentration diamond metal bonded wafering blades, mounted in Bakelite resin and polished to a 0.03 micron finish with alumina polishing media. Optical microscopy was conducted on solution treated and quenched samples using a Leica DM2500 These samples were etched using a modified Kro 3 8% HF, 82% H 2 0) to reveal the grain boundaries and precipitates. ImageJ analysis was conducted on optical and SEM images to characterize phase fractions. 3.2.5 Scanning Electron Microscopy (SEM) Alloy samples in the solution treate d state as well as the aged state were sectioned mounting in Buehler KonductoMet, and polished to a 0.03 m finish for microstructure analysis. For SEM and EPMA, a silica finishing polish was used in order to avoid any surface contamination by alumina slur ry that may interfere with composition readings. Scanning electron microscopy (SEM) was conducted using a FEI XL 40 FEG SEM, using 15 kV accelerating voltage, spot size of 4.0 and working distance of 10mm. Energy dispersive spectroscopy (EDS) was also co nducted on this system in spot, line, and mapping modes in order to visualize the distribution of alloying elements throughout the microstructure. 3.3 Alloy Processing 3.3.1 Thermal Processing Solution treatment of alloy samples was carried out using a cus tom built vertical tube furnace with drop quenching capability. The system was built around a Centorr CM 1700 Split VTF with an alumina tube, vacuum, and gas system for heat treatment of samples in an oxygen gettered flowing Ar environment. Samples were he ated at a rate

PAGE 46

46 of 9 C/min and quenched into agitated water directly from the hot zone of the furnace to retain the solution treated microstructure to room temperature. Aging of samples was accomplished through hydrogen torch encapsulation in quartz tubes under Ar environment, heating in an ATS 3150 box furnace, and furnace cooling. Samples were solution treated and aged for two hours, at temperatures detailed in the following chapters. 3.13.1 Sample Machining The 250g buttons were first sectioned into lar ge pieces using a Leco MSX205M cutoff saw, then sectioned into thin (2 4mm) plates using an Allied low speed saw using low concentration diamond metal bonded wafering blades (60 20095). The low speed saw enabled very precise cuts to be made, which is esse ntial in the fabrication of small compression samples. Grinding and polishing of samples was carried out using a Leco Varipol VP 50 grinding wheel, grit papers from 60 1200 grit, and polishing cloths with alumina and silica polishing to achieve a 0.03 m f inish for microstructure analysis samples. Mechanical testing samples were polished to a 5 m finish to minimize friction between the sample faces and the platens used in compression testing. Initial experimental compression samples were fabricated using these tools in addition to a planer to produce 2x2x3mm rectangular samples with parallel faces. Samples from the 250g buttons were machined using electric discharge machining (EDM) wire cutting by Triad EDM in Dunnellon FL. For compression samples, cyli nders with a diameter of 2.5mm were cut from the as cast alloys, heat treated, then cut, ground, and polished into the appropriate geometry, as depicted in Figure 3 1. EDM was conducted by Joseph Hy Hyatovik from specifications programmed via AutoCAD draw ings. A schematic of the dogbone dimensions is illustrated in Figure 3 2, with final

PAGE 47

47 sub size specimen pictured in Figure 3 3. In this way, reproducibility was maximized between different alloys and machining operations. Figure 3 1: Images illustrating fabrication of compression samples. A ) Cylinders cut from 250g button via EDM, B ) schematic of compression samples sectioned from these cylinders after heat treatment (Courtesy of author) Figure 3 2: Schematic illustrating sample dimensions for sub size tensile specimen, based on ASTM E 8 standard [ 80 ]

PAGE 48

48 Figure 3 3: Sub size tensile sample, prepared for mechanical testing with 0.3 m polish (Courtesy of author) Tensile samples were also machined via EDM from the 250g buttons by Triad ED M. Plates as well as dogbones were cut at varying stages during the thermal treatment of these alloys, as further discussed in the tensile testing section (6.4) of this document. 3.4 Mechanical Testing 3.4.1 Compression and Tension Testing Compression tes ting was conducted using an MTS 810 servo hydraulic testing system with 2kip and 20kip load cells and attached Oxy Gon FR200 universal materials testing furnace system. This system also has a Pfeiffer TCP300 turbo pump and Centorr Model 2A inert gas purif ier. This system is capable of reaching 1600 C and being run under variable pressure inert oxygen gettered argon environment. Graphite and silicon carbide compression fixtures were used for elevated temperature testing. Compression samples were tested i n two geometries rectangular samples of size 2x2x3mm and cylindrical samples with 2.5mm diameter and a height of 3mm. Care was taken to ensure that these sub size specimens had parallel faces, to minimize testing error. This was accomplished using a co mbination of planers and levelers to a 0.03 micron finish with

PAGE 49

49 same alumina polishing media used for metallographic preparation Boron nitride spray was used on the silicon carbide platens as high t emperature lubricant. Testing was run in displacement controlled mode at a set strain rate, calculated from individual sample geometry. The controller for the MTS 810 system uses ramp rates programmed in terms of percent full scale range (%FSR) per second as well as a has interchangeable range and load cartridges that define the scaling of the standard +/ 10 Volt readings from the load and stress transducers. For testing of compression samples, the +/ Scale displacement cartridge was used, with a 2kip (2,000 lb. ) cartridge and load cell. In order to program the appropriate ramp for compression testing, the sample height ( desired final strain ( ) and desired strain rate ( ) must be known in addition to the FSR of the current displacement cartridge in use. From this, a strain rate can be translated into %FSR/second, and final strain into %FSR. This is done using a form of the follow ing equations [3 1 and 3 2], where care must be taken to maintain consistent units, and the forced negative value for level indicates compression. in displacement transducer as well as an e xternal single arm extensometer, and data was compliance rates of interest for the high temperature compression testing. (3 1) (3 2) Data acquisition was handled using a National Instruments (NI) analog to digital (A2D) converter and data acquisition (DAQ) card installed in a Windows system running

PAGE 50

50 NI LabView software. The LabView virtual instrum ent was constructed to save the load, displacement, and strain reado uts from the MTS 810 system to a text file after converting displacement to metric units to allow for more precise measurement (there is a limitation on the number of significant digits s aved to the right of the decimal place) Post test data analysis and interpretation was conducted using a combination Microsoft Excel and WaveMetrics Igor Pro software packages, with custom macros for baseline subtraction and compliance correction of data when necessary. Figure 3 4: Photographs of tensile specimen loaded in grips. A ) Tensile specimen loaded in vice style grips for room temperature testing; B ) Specimen loaded in collar style grips for elevated temperature testing (Courtesy of author) T ensile testing at high temperature was conducted on the aforementioned MTS system, with collar style grips, and room temperature testing was conducted on an Instron electro mechanical load frame, using custom machined vice style grips, pictured

PAGE 51

51 in Figure 3 4. Displacement was measured using a laser extensometer, calibrated before every set of testing, and measurements of samples were taken before and after testing to verify machine readout. 3.4.2 Loading and Operation for High Temperature Testing High temp erature compression testing under inert environment requires a few addition considerations when compared with room temperature testing. In addition to high temperature fixtures and lubricants, the samples required specific preparation and loading. After cleaning samples, they were loaded between the platens with the MTS controller in displacement control. This allows for precise loading and alignment of the small samples and the load on the samples can be monitored as the sample is put under a small amou nt of compression (5 10 MPa). The furnace O ring is cleaned, greased, and the furnace is closed and sealed. The furnace chamber is then evacuated using the turbo pump then filled with the inert gettered argon gas to about 10 psi below ambient pressure. environment, then the furnace is pressurized to positive 2psi, regulated by relief valve. At this point, the error on the load control module is zeroed and the system can be switched to load control while still under hydraulic power. The furnace is then heated to the desired temperature at a rate of 20 30 C/min. Holding the sample in compression under load control during heating allow for the hydraulic actuators to compensate for therm al expansion of the sample and fixtures without applying additional pressure on the sample. Once the furnace has reached the desired temperature, it is allowed 20 minutes to equilibrate before switching back to the displacement controller and running the desired testing program.

PAGE 52

52 Tests were run at 700 and 800 C at nominal strain rates of 10 3 10 4 and 10 5 s 1 under displacement control. Testing profiles were programmed as ramp functions with displacement rate and final displacement level expressed in terms of percent full scale range (FSR). Compression testing was conducted using the +/ 0.5 inch FSR module, with strain rate calculations carried out using a combination of MathCAD and MS Excel. Furnace temperature is measured and regulated by two inde pendent thermocouples which keep the testing chamber within 1 C of the set point. System compliance was determined at these temperatures and strain rates and used to correct the measured displacement results of all compression testing. An external extens ometer with +/ 0.016 inch FSR was used for finer strain measurement. After testing is completed, samples were unloaded to approximately 100MPa and the system was returned to load control before the furnace is cooled. In this way, neither the sample nor pl atens will fall into the furnace heating elements upon cooling and contraction of the fixtures and sample. Once the furnace has cooled to room temperature, gas flow is stopped, the system is returned to displacement mode, and the furnace is opened. At th is point, the sample can be removed and sectioned for microstructure analysis.

PAGE 53

53 CHAPTER 4 ALLOY SELECTION AND HEAT TREATMENT DESIGN IN TI AL NB (CR MO) ALLOYS Traditional alloy development cycles last approximately 20 30 years from initial investigation to implementation in industrial settings One way to reduce the amount of time needed to develop, evaluate, and implement a new alloy system is through the use of computational tools for use in alloy design and selection. In this method of a lloy development, specific compositions and microstructures can be targeted in order to produce the desired properties, and the number of experiment s can be greatly reduced by first predicting microstructure development and alloy performance through the us e of thermodynamic and science based modeling. In this work, thermodynamic prediction capabilities are leveraged in order to select specific alloying ranges and inform thermal treatments required to produce the targeted microstructures. 4.1 Computational Analysis on the Effect of Nb on Phase Transformation Behavior The ThermoCalc software platform [ 81 ] pro vides a powerful tool for thermodynamic calculation for the understanding of material properties and processes, allowing the user to simultaneously explore process variables such as alloy chemistry and temperature. In collaboration with Professor Hans Sei an optimized Ti Al Nb ThermoCalc database, developed by Cupid et al [ 38 ] was used to calculate alloy phase equilibrium and relevant processing characteristics such as the r elative contributions of Ti and Nb to phase transformation temperatures as well as to help select heat treatment temperatures necessary to produce the desired microstructural morphology.

PAGE 54

54 In order to meet the objective of producing a very fine, disconnect microstructure through thermal processing, a high temperature single phase phase region is required, as well as a lower produced upon quenching and aging [ 10 15 ] The initial retention of phase upon quenching is important in preventing any phase transformations upon cooling, the scale and morphology of the final microstructure can be controlled via the subsequent aging step, allowing for production of finer microstructure scale than possible through equilibrium phase transformations [ 1 10 82 ] The aging treatment temperatures and times can also be selected in order to modify the microstructur al scale, with increased nucleation and slower growth kinetics at lower temperatures being favorable to the ology. In this way, strength should be increased by the Hall Petch mechanism [ 19 21 ] as well as the ductility by the microstructure refinement and separation of phase regions [ 14 16 19 40 73 ] It should be noted that all calculations in this chapter are performed in the absence of Cr and Mo Ternary Ti Al Cr and Ti Al Mo databases are in the process of roup and combined with the Ti Al Nb database to produce quaternary and quinary databases [ 38 65 66 ] These alloying additions will affect the thermodynamics and kinetics of the reactions taking place in this system, thus it is possible that actual transformation temperatures in these alloys will be offset from those calculated in this section. H owever since optimized databases of this higher order system are not currently available, these predictions based upon the optimized ternary database calculations are the most appropriate way to model potential alloy behavior.

PAGE 55

55 Initial selection of 45 at% Al wa s fixed in order to enable a wide range of composition s and phase fraction flexibility. Previous work has shown that alloys in this range have the ability to retain upon quenching [ 10 15 ] and the potential for improvement of environmental protection with increased Al content [ 83 ] By leveraging the optimized Ti Al Nb database for ThermoCalc [ 38 ] compositional ranges of interest within the Ti Al Nb system were identified. A calculated isotherm at 1000C, Figure 4 1, Furthermore, there exists a range of composition s phase content below approximately 30%, which should aid in retaining di phase morphology [ 1 14 40 ] Figure 4 1: Calculated Ti Al line highlighted [ 38 ] calculated lines of equal phase phase; targeted region of <30% indicated as shaded region.

PAGE 56

56 Taking these constraints into account, an isopleth at 45 at% Al was calculated to determine temperature regimes for each of the phases of interest for microstructure development in this alloy system. Plotting this isopleth yields a secti on of the ternary, illustrated in Figure 4 2, which shows a region from approximately 15 25 at% Nb in which both of these alloy requirements are met, and transformation temperatures were calculated for alloys with 15, 20, and 25 at% Nb, as summarized in Ta ble 4 1. Figure 4 2: Isopleth sections of Ti Al Nb ternary along 45 at% Al indicating compositions which meet microstructure development requirements A) 0 30 at% Nb; B) T argeted 15 25 at% Nb range showing high phase with lower temperature Calculated using ThermoCalc [ 38 ] Table 4 1 : Phase transformatio n temperatures upon heating (C), calculated from ThermoCalc Ti Al Nb database [ 38 ]

PAGE 57

57 In order to solution treat the alloys they should be heated to a temperature above the final solid state transformation temperature ( ), and below melting. These solution treatment temperatures are calculated to be in the range of 1460 1540C, 1505 1560C, and 1560 1580C for alloys with 15, 20, and 25 at% Nb, respectively. Agin g temperatures for these alloys should be below the lowest temperature solid state transformation reported in Table 4 1 ( 4.2 Phase Fraction and Driving Force Predictions for Aging Treatments Using the ThermoCalc software platform and phases can be calculated upon cooling for each of the specific ternary compositions. Looking at the phase fraction analysis for the alloys at either end of the range, with co mpositions of 40Ti 45Al 15Nb and 25Ti 45Al 25Nb (at%), it can be seen dramatically in the range of 900 1000 C and 700 1100C, respectively (Figures 4 3 and 4 4) More important ly there is a temperature regime in which these phases are mutually stable. One key difference in the phase fraction analyses is that for the 15 at% Nb alloy there is nucleation of the 2 phase at temperatures below about 950C, where this phase is not pr esent in the 25 at% Nb alloy. This is consistent with the calculated isopleth section at 45Al, and by referencing that calculation, it is seen that the nucleation of the 2 phase is only potentially relevant in the development of the 15 at% Nb alloy. Due to this factor, if a sample is to be aged into a two microstructure, it should be at temperatures above 950C.

PAGE 58

58 Another key difference between the 15 and 25 at% Nb alloys is the phase es. The 15 at% Nb alloy has a calculated phase fraction range of 0.22 0.23 at 900 1000C, while the 25 at% Nb alloy phase fraction range of 0.37 0.4 at 700 1100C. The higher phase is consistent with expectations based on the relative + ield on the calculated isopleth (Figure 4 phase fraction with respect to temperature paired with the absence of 2 phase nucleation at lower temperature may make increased Nb concentration more favorable for aging treatment and production of the two phase + microstructure in this ternary system Both of these alloys also have a high temperature single phase region, with a decrease in the single pha se temperature stability regime with increase in Nb content. Now that the presence of all required phases has been calculated predictions regarding aging behavior can be made. Figure 4 3 : Phase fraction of stable phases with respect to temperature in T i 45Al 15Nb alloy, calculated by ThermoCalc [ 38 ]

PAGE 59

59 Figure 4 4 : Phase fraction of stable phases with respect to temperature in Ti 45Al 25Nb alloy, calculated by ThermoCalc [ 38 ] As previously discussed, the aging of these alloys will occur from a solution treated and quenched in metastable solid solution phase In order to predict how nucleation and growth may occur during the aging of these alloys, the change in driving phases from a metastable phase was calculated. This information can be used to indicate what type of microstructure can be achieved, since high driving force for nucleation results in a denser, more uniform distribution of precipitates in the microstructure, while lower driving force for nucleation results in a less precipitate dense microstructure due to fewer nucleation sites [ 21 ] It was seen phases throughout the entire temperature range over which the calculation was conducted as presented in F igure 4 5

PAGE 60

60 Figure 4 5 respect to temperature for 15, 20 and 25 at% Nb al loys, calculated by ThermoCalc [ 38 ] phase for the majority of the temperature regime above 400C, and with increasing Nb, this driving force increases slightly. These calculations assume an artificially stabilized phase, so it

PAGE 61

61 phase stability regions. In an experimental setting, the phase is more stable at these temperatures, so these nucleation events will not actually occur, but the relative driving f orces can be used to give insight into nucleation behavior. Using this information to compare driving forces at various temperatures shows that at lower temperatures there is a higher driving force for nucleation, which would result in finer microstructur al scale. Also important is the fact that at all temperatures below about 400C, there is a phase. This could indicate that even with phase nucleation would be s ignificantly higher than at elevated temperatures, and could lead to its nucleation. The differences in driving forces can be used to understand what would happen without the effects of kinetics or crystal defects as a first step in understanding what may occur in the alloys experimentally. As previously mentioned, in order to accomplish microstructure control in this phases. From these ternary phase diagrams, and isop leth sections (Figures 4 1, 4 2), we are able to predict that in order to meet all of the required alloy characteristics ( at approximately 15 25 at %. In order to develop a fine two phase microstructure, it is to increase the driving force for nucleation which results in a oy performance [ 10 16 ] Wh ile these predictions allow for a satisfactory understanding of ternary alloy behavior, it has been

PAGE 62

62 stabilizing elements such as Cr can be essential in the retention of phase. It has been demonstrated tha t additions of about 5at% Cr to a high Nb Ti Al Nb alloys stabilizes the phase and enable it to be retained upon quenching after solution treatment [ 10 ] and that up to 1 at % Mo may improve high temperature microstructural stability due to its slow d iffusion rate in this system, stabilizer [ 8 65 84 ] T he effects of variations in Ti, Al, or Nb can be calculat ed using ThermoCalc [ 38 ] and though the addition of quaternary and quinary alloying elements cannot be predicted using the same database the calculations can be used as a guide for alloy design These additions, along with variation in Nb content, necessarily affect the driving force for phase transformations and requiring the measurement of transformatio n temperatures in the quinary alloys currently under study. As it is not the focus of this work to undertake the optimization of quaternary and quinary ThermoCalc databases, the ternary calculations prove a sufficient first approximation of alloy behavior for initial selection and design work. More information about the quinary alloys of interest can be obtained experimentally and compared with the ternary calculations in order to determine the effect of alloying additions of Cr and Mo on phase transforma tion temperatures, and whether they have any effect on the trends predicted with respect to varying Nb concentration. 4.3 Experimental Analysis on the Effect of Nb on Phase Transformation Behavior In the previous section, phase transformation temperature s were established for the Ti Al Nb ternary alloy system. These calculations will be verified experimentally

PAGE 63

63 using DSC Due to the low am ou n t of Mo and Cr additions, it is anticipated that they will minimally affect and alter the phase transformation seq uence from that predicted using computational thermodynamics. Previous work has shown that the addition of Cr to a ternary Ti Al Nb alloy phase transformation upon heating, as well as the temperature of the reverse reactions upon cooling [ 10 ] The addition of 5 at % Cr did not affect the phase nucleated was reduced by approximately 100C. In this work, it has been seen that the addition of Cr and Mo suppress the transformation temperatures by an average of 50 100C upon heating. Of particular interest for microstructure development is the temperature rang precipitate from the metastable phase, which is necessary to select an appropriate aging temperature. Table 4 2 : Transformation temperatures (C) measured via DSC for quinary alloys (addition of 5 at% Cr 1 at% Mo), compared with transformation temperatures for ternary alloys calculated via ThermoCalc Ti Al Nb database (in parenthesis) The initial heating cycle from the DSC data (Figure 4 6 ) was used to determine the temperatures at which the as cast microstructures, which solidify as a mixture of non equilibrium phases due to the fast cooling rates, transform into the high temperature phase and its subsequent melting temperature. These samples were cycled several times to determine the equilibrium phase transformat ions. Cycled curves

PAGE 64

64 can be found in Appendix B, and it was observed that transformation temperatures for equilibrium transformations remain stable, while the initial exothermic peaks are diminished or do not appear for the subsequent cycles. This indicat es that they are metastable phase transformations. Of primary interest is the final transformation before melting occurs, which is the transformation to single phase After determination of this temperature range for all alloys, each alloy was solution treated at their respective phase temperatures and water quenched to retain this phase to room temperature. Similarly, the first DSC heating cycle of the quenched alloy was used to study the ( transformation from the metastable phase, so that aging treatments may be designed. From this heating curve it is seen that the first exothermic peak upon phase. The multiple endothermic peaks at higher temperatures correspo nd to the dissolution of these two phases and the re phase upon heating, and the final plateau is the single previous research [ 10 23 39 ] and phase identification was conducted in the regions of interest using XRD analysis presented in Chapter 5 These transformation poin ts are illustrated in Figure 4 7 and the results of thermal analysis are summarized in Table 4 2, and compared with transformation temperatures calcu lated using the ternary ThermoCalc database From the DSC results it can be seen that all transformation temperatures in the ) increase with increasing Nb content, with the most significant increase in the range of 20 25 at % Nb. That is, the temperature of all equilibrium phase transformations increase with Nb

PAGE 65

65 are stable increases, while th at for phase decreases with increasing Nb content. Figure 4 6 : DSC data upon initial heating of alloys with 15, 20, and 25 at% Nb from metastable phase nucleation and dissolution phase stability temperature regimes These trends are consistent when comparing quinary Ti Al (Nb,Cr,Mo) alloys with the calculated ternary Ti Al Nb isopleth section, which show upward sloping phase

PAGE 66

66 ) transformation path in this Nb content range. The sole exception is a larger increase between the higher Nb content alloys than in the ternary system. This can be stabilization effects of Cr and Mo additions as, it has been seen that by adding 5 at% Cr to ternary Ti Al Nb alloys, transformation temperatures are depressed by 10 100C [ 10 ] By comparing the transformation temperatures measured in the DSC cycling of the quinary alloy with those predicted by the ternary calculations, as illustrated in Table 4 2, a similar depression in transformation temperatures is seen. phase nucleation temperatures with increase in Nb, with more pronounced decrease occurring phase. While this large change in alloy behavior was not predicted based upon ternary calculations it phase formation upon quenching or aging treatments, especially in the case of the 25 at% Nb alloy. Notable feature s of the DSC curves produced through heating of t he metastable 6 ) include the two exothermic peaks at temperatures below 1000 C in all three alloys tested, as well as multiple endothermic peaks which indicate solid state transformations o ccurring before melting. The irregular shape of the dissolution peak upon heating of the 15 at% Nb alloy indicates a double peak, seen in Figure 4 6 more stable to higher temperat ure s and these peaks separate as phase dissolves at higher temperature. Also as Nb content increases, the temperature at which each of the exothermic transformations occur decrease, with a larger decrease in the first reaction. The temperatures at which the endothermic dissolution reactions begin to take place increase with increasing Nb. The increase in dissolution temperature with Nb is

PAGE 67

67 not stable until higher temperat ure. Figure 4 7 : Comparison of DSC curves of Ti 45Al 20Nb 5Cr 1Mo alloy to calculated isopleth section for Ti 45Al 20Nb alloy, illustrating how calculated transformation temperatures were correlated with experimental results. Figure 4 8 : P hase transformation temperature measured in DSC upon heating overlaid on Ti Al Nb isopleth section at 0.45 Al (calculated via ThermoCalc [ 38 ] ) illustration suppression of transformation temperatures with substitution of 5 at% Cr and 1 at% Mo for Nb In the 25at% Nb alloy, evidence of a third peak is seen, manifesting as double exothermic peaks near 800 and 850C, as w ell as an additional shoulder within the

PAGE 68

68 dissolution peaks near 1350C. One explanation for this additional peak is the non phase is not stable tion sequence is highly dependent on factors such as quenching rate, local alloy composition, and local precipitation sequence. It is thought that the additional peaks are due to the nucleation phase directly from The increased temperature is consistent with the additional energy requirement for homogenous nucleation of phase, and the higher dissolution temperature may be due to slight composition variation pha se produced through homogenous versus heterogeneous nucleation. Upon further thermal cycling of the alloy ( Appendix B phase is of equilibrium composition. Thus, not only is there a lack of exothermic precipitation phase dissolution. The effect of alloying additions on phase transformation temperature depression is summarized schematically in Figure 4 8, in which the shifted phase field boundaries are overlaid on the ternary isopleth calculated via the ThermoCalc Ti Al Nb database.

PAGE 69

69 CHAPTER 5 MICROSTRUCTURE DEVELOPMENT AND STA BILITY IN Ti 45Al xNb 5Cr 1Mo ALLOYS In order to improve the high temperature performance of TiAl based alloys, research has been performed to optimize alloy chemistry and microstructure, with two phase microstructures being commonly employed in o rder to concurrently improve ductility and strength [ 1 4 6 10 58 82 ] In the previous chapter, alloy selection and heat treatment design was detailed in order to produce a method by which the microstructure can be controlled and tailored. It has been seen experimentally that slow cooling of alloys in the Ti Al Nb system with Nb in the range of 10 15 at% will phase precipitate without heat treating for prohibitively long times. Even if heat t reated for sufficiently long periods of time, the resultant microstructure would be coarse, yielding unfavorable mechanical properties, and not fully benefiting from increased strength and ductility that would be seen if precipitation strengthening with a fine dispersion of Nb rich phase [ 1 12 14 40 ] phase which is known to be brittle and when it is the connected phase, acts to control the mechanical properties [ 1 12 14 40 43 ] phase morphology, so that alloys may retain the benefits of precipitation strengthening phase alloys. 5.1 Thermal Processing In order to overcome the challenges in producing a fine, phase morphology, two stage heat treatments were designe d such that alloys are solution

PAGE 70

70 treated in the single phase field regime as illustrated schematically in Figure 5 1 A nalysis of DSC data led to the determination of 1400, 1450, and 1500C as solution treatment temperatures for alloys with 15, 20 and 25 at% Nb, respectively as these temperatures lie fully within the single phase region Likewise, aging temperatures of 1000, 950, and 1125C were chosen for these alloys as these temperatures lie fully within the two regime . In this way, it is ensured that the alloy will be within the single region and will have adequate thermal energy available to fully solution treat. Aging temperatures were chosen to be fully ctive alloys, as illustrated in Figure 4 6 in the preceding chapter. Figure 5 1: Heat treatment profile for experimental alloy s Temperatures for solution treatment (T soln ) and aging (T age ) determined by DSC analysis

PAGE 71

71 5.2 Microstructure Development 5.2.1 Solution Treating and Quenching of Single Phase For microstructure modification in this alloy system, it is essential to have control phases. In order to accomplish this, it is important to quench the sing achieved though sufficiently severe quench as well as alloying additions [ 10 39 82 ] In these ways, the thermodynamics and kinetics of the react ion are altered. The stabilizing elements Cr and Mo were added and have been proven effective in stabilizing this phase upon quenching [ 10 82 ] In previous research, solution treatment and quenching of ternary Ti 44.5Al 26.5Nb alloy produced a phase could not be suppressed upon quenching, but with the substitution of Cr for Nb, a Ti 43.5Al 22.5Nb 5Cr alloy was able to retain solely phase to room temperature upon quenching [ 10 ] While reduction in Nb content when compared with alloys in this system that have been studied in the past [ 1 14 40 ] lowers the amount of stabilizing elements present in the alloy, it also moves the composition towards a larger, lower temperature area of the single phase region, which improves the ability to s olution treat. Solution treatment of alloys in the range currently under study produces a very large phase scale, with grain size on the order of millimeters. Etching of the quenched microstructures with a 10% HNO 3 8%HF ) reve als the location of the phase, which resides mostly along grain boundaries in the case of lower Nb content alloys, as illustrated in Figure 5 2. This

PAGE 72

72 relatively large microstructure scale is always seen after successf ul solution treatment of alloys in this system in the single phase region and quenching. Figure 5 2: Etching reveals solution treated and quenched microstructure of alloy with phase grain size and boundaries (Courtesy of author) In the alloys currently being studied, it was found that with increasing Nb content phase formation at the grain boundaries. This is illustrated in Figure 5 4 with XRD analysis in Figure 5 3, and can be attributed to the fact that all phase transformation are shifted to higher temperatures as summarized by the DSC results in Figure 4 6 Due to the higher phase is stable relative to the phase, the alloy is in a temperature range favorable for the nuclea tion starting temperature is controlled, so the quenching of 25 at% Nb samples from higher temperature compared to lower Nb alloys (1500C rather than 1400 or 1450C) causes T is higher for the 25 at% Nb alloy, this reduces the thermal gradient and cooling rate at

PAGE 73

73 lower temperature when compared with quenching from lower solution treatment temp erature. The less severe quench, paired with the high driving force for phase nucleation leads to the formation of this phase throughout the microstructure. Figure 5 3 : XRD of alloys with 15, 20, and 25% Nb in solution treated and quenched (S&Q) state showing + microstructure

PAGE 74

74 Figure 5 4 : As que nched microstructures of Ti Al Nb Cr Mo alloys showing increase in grain boundaries with increasing Nb A) 15 at% Nb B) 20 at% Nb and C) 25 at% Nb (Courtesy of author) Successful retention of the phase to room temperature is dependent on several factors. First is the ability to rapidly cool the alloy samples to room temperature from the solution treatment temperature. The vertical tube furnace setup employed in solut ion treatment allows for quenching directly from the hot zone, minimizing the amount of cooling that can occur before the samples come into contact with the quenching media. Initial quenching was done into a bucket with room temperature water, which prove d to be a sufficiently severe quench for the 15 at% Nb alloy, retaining phase to room temperature for samples below 3mm in thickness. Another factor affecting cooling rate is the sample s size. Samples that are very small, such as the 3 mm thick sample s that were used in the initial rounds of heat treatment are suitably thin for quick conduction and removal of heat due to their large surface to volume ratio. Not only can heat be removed quickly from the larger faces, but thinner samples have lower amou nt of heat stored within the sample before it comes in contact with the quenching media, enabling the water to remain closer to its initial temperature throughout heat removal from the sample. Thicker samples (~5mm) were solution treated, and proved too t hick for quick removal of heat by water quench, and more extensive transformation phase was observed. Similar results occurred when several thin samples were

PAGE 75

75 solution treated and quenched at the same time and this behavior holds for higher Nb concen tration alloys, as illustrated in Figure 5 5. Figure 5 5 : SEM images of solution treated and quenched 25% Nb alloy A) P artial suppression of phase transformation in thin (2 mm) sample, B) G rowth of phase in thick (6 mm) sample (Courtesy of author) Later solution treatments employed an agitated water quenching media, in which a water pump was used to force additional water into the quench bucket from a secondary reservoir and provide some amount of stirring to the quench media. This improve d heat removal from thin samples, aiding the retention of phase to room temperature in 15 and 20 at phase is observed only where it heterogeneously nucleates on the grain boundaries, however in the 25 at % Nb alloy, the phase throughout the phase in addition to heterogeneous nucleation at the phase could be a result of either no t fully solution treating the alloy, or of cooling somewhat before quenching. Observation of the phase grain structure of the 25 at% Nb alloy reveals that it is equiaxed and in the 1 2 mm range, similar to alloys with lower Nb

PAGE 76

76 content. A microscopic vi ew of this equiaxed structure can be seen in Figure 5 4 and a macroscopic image of the 25 at% Nb samples can be seen in Figure 5 6 Figure 5 6 grains indicative of solution treatm ent in the single phase region (Courtesy of author) If the heat treatment had occurred in the two phase single phase region, then growth of phase into large equiaxed grains would not have occurred. Therefore it is probable that there was some cooling of the sample before quenching, or that cooling rate upon quenching was slow enough to allow for phase before reaching room temperature. As discussed previously, ph ase nucleates upon cooling from the phase in the 25 at% Nb alloy, there is less undercooling upon quench in samples with phase nucleation and growth during quenching. 5.2.2 Aging to Produce phase upon aging. This transformation is not well studied, but it is known phase precipitates quickly from the prior phase gr ain boundaries, growing in

PAGE 77

77 widmanstatten morphology, often upon quenching [ 23 ] as shown in Figure 5 5 The nucleation occurs has been determined through DSC, and is seen as the region between the two nucleation events and the first F igure 4 6 in the phase regime will result in finer microstructure due to the increased nucleation rate and slower grain growth, while higher temperature will result in decreased nucleation and faster growth, a trend that has been found to hold in similar alloy systems with 22.6 at% Nb [ 16 ] 5.2.2.1 Aged Microstructures Figure 5 7 : SEM micrographs of 15 at% Nb alloy aged at 1000C for 1 hour A) Ultrafine microstructure produced via simultaneous nucleation phase ; B) detail of phase size and morphology (Courtesy of author) Initial aging treatments of the 15 at% Nb alloy at 1000 C showed successful development of the ultrafine microstructure, as illustrated in Figure 5 7 The morphology is greatly refined when compared with alloys w + upon slow heating to the aging temperature. The modification of microstructure through precipitation sequence will be discussed in further depth in the following sections.

PAGE 78

78 Analogous heat treatments conducted on alloys with 20 and 25 at % Nb showed phase content, as well as increased microstructure scale as illustrated in Figure 5 9 With more availability of Nb, increased amount and g rowth of phase is seen, and the microstructure transitions to one with a more connected phase. Figure 5 8 : XRD of alloys with 15, 20, and 25% Nb in aged state, showing microstructure

PAGE 79

79 Figure 5 9 : Change in scale and phase of aged s with respect to Nb content A) 15 at% Nb with12% B) 20 at% Nb with 25% and C) 2 5 at% Nb with 33% (Courtesy of author) In order to verify the phases present in the aged state, XRD was conducted in the aged state, and peak id phases present in the microstructure, as seen in Figure 5 8 It was previously believed that there may be 2 phase as thermodynamic calculations sh ow should occur in the 15 at% Nb alloy upon cooling. 2 phase in the 15 at% Nb alloy indicates that the additions of Cr and Mo phases such that further transformation will not occur upon near equilibrium cooling. Al ternatively, it could indicate that the alloying additions modify the thermodynamics of the 2 phase boundary line is shifted to lower Nb concentration or lower temperature. The information gained from XRD and DSC da ta suggest that the latter outcome is more 2 cycling of the 15 at% Nb alloy, where this transformation would occur upon near equilibrium heating and cooling, and all other phase transformation temperatures are depressed by 50 100C.

PAGE 80

80 5.2.2.2 Effect of Precipitation Sequence phase by holding at lower temperature (400 500 C phase formation will be suppressed as phase despite [ 23 39 ] However, if aging is phase is found to occur from the ecipitation, presence of the phase phase during the aging of these microstructures. The effect of precipitation sequence is evident from the microstructures produced in 15 and 20 at% Nb alloys, as seen in Figure 5 9 phase was suppressed upon quenching, resulting in a very fine, equiaxed microstructure. It has been experimentally observed that upon quenching there can be formation of oundaries, as seen in Figure 5 4 Upon aging of these alloys, t grains phase that forming upon quenching e w ould coarsen [ 10 23 39 ] Thermodynamic calculations conducted using an optimized ternary Ti Al Nb database [ 38 ] phase when increasing Nb content from 15 to 25 at%, as illustrated in Figure 4 4. For phase for mos t temperatures above about 400C. In ternary alloys with 15 at% Nb content, it is

PAGE 81

81 phase will have higher driving force within the highest temperatures, phase first upon cooling. phase for most of their stability regime, including the aging temperature range, these calculations only reflect phase, and does not take into account heterogeneous nucleation. It has been observed that upon quenching there i s phase grain boundaries. When quickly heated phase has enough energy to begin growth, phase be seen in the aged microstructure of 15 and 20 at% Nb alloys by observing differences phase grain boundaries Near these boundaries, a coarser acicular morphology is seen, while away from them a phase at phase within the grains. This va riation in microstructure scale and morphology with respect to location relative to prior phase grain boundaries is illustrated in Figure 5 10 phase in the 15 at % alloy are indicative of a combination of homogenous a nd heterogeneous nucleation. The fine, evenly phase, seen especially in the 15 at% Nb alloy, indicates phase to a very fine scale. Additionally,

PAGE 82

82 phase preci pitates can be seen, which indicate the formation of phase laths. This is illustrated in Figure 5 1 1 in which contrast within the phase grains is due to electron channeling contrast, and some sub surfa ce phase is visible. Additionally, the presence of annealing twins has been documented in + microstructures [ 12 33 ] such as those potentially represented in Figures 5 1 9 5 11 and 6 6 in the following chapter, which alter the contrast of the phase grains in these micrographs. Figure 5 10 : SEM micrographs of 15 at% Nb alloy. A) F ine microstructure away from B) C oarse microstructure near prior duri ng quenching (Courtesy of author) phase will continue to grow into large laths if phase is responsible for phase can phase is nucleated fir st, as it is with slower cooling leading to a coarser lath

PAGE 83

83 is illustrated in Figure 5 1 2 which compares the aged microstructures of a sample of 15 phase (Figure 5 1 2 A) to that of one phase upon quenching (Figure 5 1 2 B) During aging, competing nucleation and growth of both phases is occ urring, so the order in which these reactions are allowed to occur plays a very strong role in the development of the alloy microstructure. Figure 5 1 1 phase, indicating homogeneous nucleation in the majority of the alloy with some heterogeneous nucleation between phase laths (highlighted) (Courtesy of author) phase retention results in finer, more uniform microstructure after nced alloy performance [ 10 16 ] This fine microstructure is expected to result in higher strength alloys due to reduction in strain localization which would otherwise occur in reg phase grain boundaries [ 16 ] Further discussion of the effects of microstructure morphology on mechanical properties will be presented in Chapter 6 In order to

PAGE 84

84 hase necessary, as can be clearly seen through observation of microstructure development, presented in Figures 5 4 and 5 9 where phase upon quenching and produce an ultrafine microstructure, while the 25 at% phase upon quenching and results in a much coarser microstructure, and a modified morphology. Figure 5 1 2 Nb alloy; A phase formed upon quenching coarsens into lath shaped morphology, B phase nucleation suppressed upon quenching results in fine, equiaxed morphology (Courtesy of author) 5.2.3 Long term aging and microstructural stability Aging treatment study was carried out by undergraduate University Scholar student Cameron Palmer on alloys with 20 and 25 at% Nb in order to determine the evolution of microstructure scale with aging time. Additionally, microhardness testing was conducted to correlate with effects on alloy strength and to determine peak aging times. Samples with 20% at Nb were aged at 950C, and samples of 25% Nb alloy were aged at 1125C, as they were for previous microstructure development, for times ranging from 10 to 10,000 minutes. Aged samp les were air cooled and polished to a

PAGE 85

85 0.3 the sample to changes in alloy strength with aging. Figure 5 1 3 : Effect of aging on microhardness in 20 and 25 at% Nb alloys. 20 at % Nb alloy aged at 950C, 25 at% Nb alloy aged at 1125C It is important to note here that the grain size of the 20 at% Nb alloy are much smaller than that of the 25 at% Nb alloy, which may correlate to higher hardness values. In each alloy, there is a re latively short peak aging time, around 10 minutes of aging which indicates a very short incubation time before nucleation begins. This is followed with a decrease in hardness values, indicating over aging and then a leveling off of after this time, as su mmarized in Figure 5 1 3 A comparison of each of the alloys after 10 minutes and 1000 minutes of aging in illustrated in Figure 5 1 4. The short peak aging from the metastable phase and fast kinetics of this transformation Since the phase is stable only at temperatures above 1330 and 1475C for alloys with 20 and 25 at% Nb, respectively, there is a large driving force when these alloys are heat treated approxima tely 350C below these temperatures, coupled with a large amount of energy available at these high temperatures. The decrease in hardness after 10 minutes is indicative of over aging, which can be due to the increase in precipitate size and change

PAGE 86

86 in inte rfacial coherency with the matrix [ 19 21 ] Additionally, the decrease in hardness of the 20 at% Nb alloy followed by an increase could be attributed to the dissolution of phase precipitates, similar to a reversion process seen in some aluminum alloys [ 21 ] phase precipitates which strengthen the alloy. De termination of peak aging time is important for microstructure development of an alloy, and can be used to produce the highest room temperature strength possible. However, for elevated temperature applications, the stability of the microstructure is of gr eater interest. It was observed that alloys were over aged after the one hour heat treatment that has been employed for microstructure development in these alloys, but more importantly the hardness values were relatively stable as aging time progresses. In the case of the 25 at% Nb alloy, this stability in hardness would not be expected due to the large change in microstructure scale, but could be attributed to the strengthening from the phase and its size relative to the microhardness indenter. For th e 20 at% Nb alloy, the stability in hardness is understandable due to the fact that microstructure scale remains relatively small. In order to determine longer term microstructure stability, 20 at% Nb samples were aged for 10,000 minutes and it was observ ed that while there was some coarsening of the microstructure, it is not to the extent that would be expected for continuous growth of microstructure scale over an additional order of magnitude of aging time. Figure 5 1 5 shows SEM micrographs of the 20 at % Nb alloy after aging for 10,000 minutes which support that the material maintained relative stability in the microstructure as well as hardness.

PAGE 87

87 Figure 5 1 4 : Effect of aging on microstructure scale for 20 and 25 at% Nb alloys. A) 20 at% Nb alloy aged at 950C for 10 minutes, ( B ) 20 at% Nb alloy aged at 950C for 1000 minutes, C) 25 a t% Nb alloy aged at 1125C for 10 minutes and ( D ) 25 at% Nb alloy aged at 1125C for 1000 minutes. (Courtesy of author) 10 and 1000 minutes, correlating with decreased hardness levels. However, even after 10,000 minutes of aging, grain size is limited to 1 at% Nb alloy for 1000 minutes. The goal of generating a relatively fine microstructure through solution treatment and aging was accomplished, and the aging study helped characterize that the strength may be retained at extended aging times, even well past peak aging time Other mechanical properties would be altered by changes in microstructural morphology and scale over time, such as toughness and ductility, and this would be an appropriate focus for future work in this alloy system.

PAGE 88

88 Figure 5 1 5 : Effec t of aging on microstructure scale for 20 a t% Nb alloy aged at 950C. Aging times of A) 10 minutes B) 1000 minutes, and C ) 10,000 minutes (Courtesy of author)

PAGE 89

89 CHAPTER 6 MECHANICAL PROPERTIES OF TI AL NB BASED ALLOYS 6.1 Microstructure and Mechanical Properties As discussed in the previous chapter, solution treatment in the single phase [ 15 16 23 38 ] and subsequent quenching retains the phase, TiAl and allowing for better control over microstructural scale [ 10 15 23 ] Aging in the two TiAl + Nb 2 Al region, can produce a phase wi th disconnected phase [ 10 16 23 38 ] with a representative microstructure and XRD evaluation of the 15 at% Nb alloy illustrated in Figure 6 1. In the Ti Al Nb system, it is known that Ti and Al form several intermetallic phases, as do Nb and Al [ 2 4 38 79 85 ] It is also known that Ti and Nb have complete mutual solubility [ 79 ] Additions of Cr and Mo are expected to have solubility with both Ti and Nb as well, due to their similar electronegativity and size [ 19 21 38 65 66 ] The small additions of these alloying elements should offer some solid solution strengthening, but it is not expected to be as significant as the increase due to strengthening from precipitation of the Nb rich phase. Microstructure morphology and scale are expected to have a significant effect on strength, as well as deformation and failure mechanisms in this alloy system, du e to Hall Petch, solid solution, and precipitation strengthening effects in addition to the phase [ 12 14 40 ] during low temperature deformation. phase volume fraction and micrographs were analyzed to determine relative phas e grains ranging from

PAGE 90

90 0.1 1 m in diameter and approximately 0.12 area fraction (A f phase particles ranging in size from approximately 50 500 nm, while increasing the Nb content to 25 at% increases the A f phase to approximately 0 .33, with much la rger, contiguous phase ranging in size from 500 nm 5m. Figure 6 1 : Representative TiAl matrix (darker gray) with Nb 2 Al particles ( white ) (Courtesy of author) 6.1.1 Flow Behavior Initial compression tes ting was conducted on the 15 at% Nb alloy, and a s shown in Figure 6 2, all flow curves show a linear elastic region followed by a transition to plastic deformation with no sharp yield point. All samples were tested to approximately 40% engineering strain, where tests were interrupted with no macroscopic failure of the alloy. At testing temperature of 800 C, changes in flow behavior as a function of strain rate are most discernible. At lower strain rates, a peak stress followed by flow softening is observ ed, as well as oscillations in the curve that increase in period as the strain rate decreases. However, at faster strain rates this oscillatory behavior is diminished. At 700 C there is a moderate increase in yield stress and a change from flow softening to work hardening behavior for compression at the highest strain rate. At both

PAGE 91

91 temperatures, as the strain rate increases, the flow stress and work hardening rate increase. C ompression of samples at 700 and 800 C yields stress exponents of 13.1 and 7.7 respectively, as illustrated in Figure 6 3 where they are compared with strain rate dependent strength of high phase content alloys [ 1 ] The 0.2% offset yield stress and maximum flow stresses of all tests can be found in Table 6 1. Figure 6 2: Compression testing results of 15 at% Nb alloy at strain rates of 3x10 3 to 3x10 5 s 1 A ) 700 C B ) 800 C Figure 6 3: Strain rate dependence of strength comparing results of current testing of 15Nb alloy at 700 and 800C to TiAlNb alloy with 0.6V f phase [ 1 ] and typical 2 microstructure at 700C [ 2 4 ] as well as a high Nb 2 alloy with improv ed high temperature strength at 815C [ 6 ]

PAGE 92

92 Table 6 1: Test temperature and strain rate along with the corresponding 0.2% offset yield stress and maximum stress (MPa) for 15 at% Nb sampl es tested in compression at 700 and 800C Temperature (C) Strain Rate (s 1 ) 0. 2% Offset y max 700 3x10 3 1010 1510 3x10 4 840 1240 3x10 5 690 880 800 3x10 3 780 1010 3x10 4 680 725 3x10 5 415 485 Compression testing of 20 and 25 at% Nb alloys at temperatures of 700, 800, and 900C show similar trends with respect to strain rate, temperature, and flow softening. With increased Nb content, strength increases at all strain rates and temperatures tested. This can be explained by the higher amount of precipitation phase concentration. Ho wever, at the highest strain rate and lowest temperature (700C, 10 2 s 1 ) the 20 at% Nb alloy shows limited compressive strain to failure when compared with other testing conditions, failing after only 5% true compressive strain. At the same strain rate at 800C, the alloy accommodated over 50% compressive strain before failure. Under all other temperatures and strain rates tested, the tests were interrupted at approximately 50% engineering strain, at which point the flow behavior of the alloy was determ ined. At testing temperatures of 700 and 800C, the 25 at% Nb alloy fractured before 50% engineering strain. At 700C, failure occurred in a nearly brittle manner at 10 2 s 1 strain rate, and total compressive strain at failure was between15 and 20% tru e strain at the slower strain rates. Increasing the testing temperature to 800C increases failure strain to 15% at 10 2 s 1 and to approximately 30% and 55% for strain rates of 10 3 and 10 4 s 1 re spectively. At this temperature, a peak stress followe d by flow softening is seen, similar to lower Nb content alloys at this temperature.

PAGE 93

93 Table 6 2: Test temperature and strain rate dependence of 0 .2% offset yield stress maximum stress (MPa) and true compressive strain to failure f ) for 20 at% Nb sampl es tested in compression at 700, 800 and 900 C Temperature (C) Strain Rate (s 1 ) y max f 700 10 2 1300 1630 0.05 10 3 1220 1485 10 4 1080 1275 800 10 2 1010 1210 0.4 10 3 820 900 10 4 590 635 900 10 2 560 655 10 3 420 490 10 4 285 310 Table 6 3 : Test temperature and strain rate dependence of 0 .2% offset yield stress maximum stress (MPa) and true compressive strain to failure f ) for 25 at% Nb sampl es tested in compression at 700, 800 and 900 C Temperature (C) Strain Rate (s 1 ) 0.2% Offset y max failure 700 10 2 1800 1960 0.01 10 3 1350 1770 0.13 10 4 1380 1780 0.17 800 10 2 1280 1775 0.19 10 3 1060 1400 0.3 10 4 750 980 0.55 900 10 3 625 750 10 4 265 280 At 900C, there is a significant decrease in strength in both the 20 and 25 at% Nb alloys. This is evident in the drop in yield strength by nearly half between 800 and 900C at all strain rates. The drop is even more exaggerated at the slowest strain ra te in the 25 at% Nb alloy, in which the yield strength drops by approximately 65%. This drop can be attributed partially to the increased Nb concentration in this alloy, but the most important factor affecting the decrease in strength is the connected nat ure of the phase The microstructure of the 25 at% Nb alloy has increased phase concentration, size, and contiguity when compared with lower Nb alloys, as illustrated in Figure 5 7 in the previous chapter. Due to its more connected nature, the alloy behavior

PAGE 94

94 is more controlled by the phase as seen in previous research in high ph ase alloys, where phase is found to undergo a softening transition near 900C [ 1 12 14 ] 6.1.2 Deformed Microstructure SEM observation of the deformed microstructure of the 15 at% Nb alloy shows that it does not undergo signific ant coarsening during compression at 700 or 800C when compared to the non deformed microstructure as displayed in Figure 6 4. There was no fragmentation, pancaking, or flattening of particles normal to the loading direction, which were both observed in pr evious study of microstructures with 0.6 V f phase [ 1 ] With the exception of microcracking, the deformed microstructure appears identical to the non phase and some cracking at t opening at various angles with respect to the loading direction due to local re orientation of strain due to the hard second phase particles [ 86 ] Figure 6 4: SEM micrographs of 15 at% Nb alloy before and after compression testing. A ) Non deformed sample; B ) S ample deformed at 800 C, 3x10 4 s 1 to approximately 40% engineering strain. Arrows indicate microcracking in the phase, which appears as the darker phase in these micrographs [ 82 ] (Courtesy of author)

PAGE 95

95 6.2 Deformation mechanisms Analysis of flow curves obtained at varying strain rates and temperatures can lead to insight into the deformation mechanisms that may be activ e in these alloys The effect of strain rate on work hardening is evident in these alloys, especially in testing of the 15 at% Nb alloy at 700 and 800C. Calculated s tress exponents (reported in Appendix C) fall into the range of dislocation recovery mechanisms, which include dynamic recovery, dynamic recrystallization, interfacial sliding, and diffusion assisted dislocation motion [6]. This is supported by the flow curves, which indicate that dynamic recovery and recrystallization play important role s in deformation These deformation mechanisms could be confirmed through targeted transmission electron microscopy (TEM) work, but the current testing setup does not allow for quenching of samples directly from the testing stage for analysis of interrupt ed tests. Since quenching is not possible, some of the dislocations that are present in a deformed sample will be annealed out upon furnace cooling to room temperature which would change the dislocation substructures. Evidence of dynamic recrystallization can be seen most prominently in the flow curves of this alloy tested at 800 C and lower strain rates. At 3x10 5 s 1 multiple oscillations in the flow stress can be seen as a function of increasing strain due to the nucleation and propagation of dynamic recrystallization waves through the material. Since at this strain rate one wave of dynamic recrystallization can be fully completed before next wave begins, oscillations in the flow stress are evident. At higher strain rates, the critical dislocation density required to nucleate dynamic recrystallization is reached before the material can completely recrystallize, so oscillations are not visible in the flow curves as the material is continuously being recrystallized.

PAGE 96

96 Previous work with alloys containin g 0.6 V f phase has shown that dynamic phase deformation by dislocation glide and twinning [ 1 ] Due to the phase in the 0.6 V f alloy, high temperature deformation was co grain boundary sliding [ 1 14 ] but by reducing the phase fraction and phase in the alloys currently under study, as evidenced by analysis of the deformed microstructure, which will be discussed in further det ail in the following section. At phase acts as a hard phase, as it has been found in literature [ 1 40 ] that it undergoes a transition to a deformable phase at temperatures above 900C, as previously mentioned. While deformation of t his alloy is controlled by the properties of the continuous phase, the disconnected phase particles contribute to strengthening and limit the size of recrystallized grains. Similar to microstructure development upon aging of the alloy, the dispersion phase grain boundaries and limit coarsening [ 76 ] interface is not likely to be fully coherent, as dislocations at / interface or propagate through the phase they may be particles that they can overcome through dislocation climb or Orowan looping. These processes are more difficult in this temperature regime when compared / interfacial sliding, resulting in increased work hardening and retention of strength [ 2 11 19 21 72 ]

PAGE 97

97 The alloying additions Cr and Mo in solid solution may slow the growth of recrystallized grains and Nb 2 Al particles may exert grain boundary pinning [ 74 ] As discussed in Chapter 2, particle stimulated nucleation (PSN) o f dynamic recrystallization is observed in two phase and dispersion strengthened alloys, though it is unlikely to contribute significantly to nucleation of recrystallization in these alloys. Since second phase particles must be approximately 1 m or large r to initiate recrystallization [ 76 ] and the largest phase particles in th e 15 at% Nb alloy microstructures are approximately 500 nm, with many smaller particles in the regime of 10 50 nm. However, the presence of the ultrafine phase particles may have the effect of reducing the amount of strain required to initiate recrystal lization by increasing strain in the matrix even though local dislocation densities are not large enough at the precipitate matrix interface for nucleation at the particles. Since particles smaller than 1 been found to retard recrystallization und er certain conditions, systematic study would be required to determine the regimes of particle size and spacing in which recrystallization would be retarded or accelerated. At this point the effect of the phase particles upon recrystallization is unclea r, but it has been seen that they have a strong influence in the phase grain size through their pinning effect. In higher Nb concentration alloys similar flow behavior is observed, with a peak stress at low strain followed by significant f low softening. The exceptions to this are samples tested at the fastest strain rate (10 2 s 1 ) at 700C. At this temperature, there is not sufficient time for the accommodation of deformation. That is, at this temperature and strain rate, dislocation ge neration and mobility is not sufficient for dislocation

PAGE 98

98 recovery mechanisms to be active, so the induced strain results in the opening of microcracks, and complete fracture of the samples. 6.3 Failure Mechanisms phase in relationship to phase is vital to understanding the deformation and failure behavior in these two phase alloys and a nalysis of deformed samples shows that microcracking occurs throughout the deformed microstructures. Cracking can be used as an indication of what types of / sliding. Additionally, grain boundary incompatibility and internal stress could cause grain itself. In Figures 6 4 and 6 5, the cracks appear to be phase, indicating that it is the strain accommodating phase. Figure 6 5: Detail of sample deformed at 800C, 3x10 3 s 1 / interfacial microcracking (Courtesy of author) Dislocation formation during plastic deformation can be attributed the anisotropic phase leading to the production of geometrically necessary dislocations at grain boundaries, in addition to dislocation pileup. Since compression testing is being carri ed out at elevated temperatures, competing recovery mechanisms would also be

PAGE 99

99 active. Recovery mechanisms reduce the number of dislocations and amount of residual stresses [ 19 74 76 ] but when comp ressed to 50% engineering strain, there / Figure 6 6: Electron channeling contrast revealing + with phase morphology (Courtesy of author) Figure 6 7: Schematic relationship between observed microcracking with respect to grain morphology A ) equiaxed B ) ultrafine equiaxed C grain morphology [ 82 ] (Courtesy of author)

PAGE 100

100 Under specific beam conditions, the phase morphology is visible through electron channeling contrast, as seen in Figure 6 6, and when imaged in SE or BSE modes. However, phase morphology is not readi ly apparent in all micrographs of these alloy s, so the particular cracking behavior in the microstructure can be used as evidence to discern the relationship of phase morphology to deformation and failure mechanisms. Figure 6 7 schematically details the types of microcracking that may be observed and how they correlate with phase morphology and damage. If the majority of cracks are large and span the length phase particles, it could indicate that there is a larger, equiaxed phase (a). If there are short, randomly oriented cracks as well as cracks that form an angle or triple point, this could indicate finer equiaxed phase grains (b). If there are a series of parallel cracks, then this could indicate acicular phase morphology (c) which could locally share a common orientation relationship, due to the nature of their growth from the metastable phase [ 10 23 ] In samples of 15 at% Nb tested at 800C, SEM evaluation shows microcracking phase comprised mostly of short cracks that do not span the entire phase particles, as well as cracks opening up at characteristic triple points, which suggests a fine equiaxed morphology of particles as presented schematically in Figure 6 7(b). This microcracking pattern also indicates / interfacial sliding may be the main deformation mechanism active at 800C. This is cons grain morphology visible in SEM via channeling contrast. In compressive fracture of a brittle material, microcracks will align with the direction of maximum compressive load, which for single phase samples is the loading

PAGE 101

101 axis [ 87 ] It can be seen in Figures 6 4 and 6 5 (in which the loading direct ion is vertical with respect to the micrographs) that this is not the case for deformation of this two phase material, but instead there is uniform microcracking throughout the microstructure. Additionally, flow curves indicate pseudo plastic behavior. In ceramics or intermetallics, such uniform cracking is usually only seen when there is sufficient confining stress to prevent axial splitting or shear failure [ 87 89 ] Since t his alloy was tested only in uniaxial compression with no external confining stress, this behavior may be attributed to the re distribution of stresses due to the reinforcement phase, as well as the small grain size [ 19 86 ] In this two phase microstructure, confinement stress is provided locally by the hard reinforcement phase, which imposes a triaxial stress state due to strain compatibility constraint of the interface [ 16 86 ] This also change s the stress state such that the loading direction is not necessarily that of maximum compressive stress, therefore modifying the orientation of microcracks relative to the sample geometry and loading. No fragmentation or cracking phase was present in this alloy, in with higher V f phase [ 1 ] As expected, a continuous ultrafine TiAl matrix with disconnected Nb 2 Al results in increased ductility [ 16 ] This is evidenced by the high amount of compressive strain, more than 40% strain at 800 C, accommodated by the material with no evidence of large scale cracking. In samples of 15 at% Nb alloy tested at 700C, there is not a clear indication of phase morphology based solely on microcracking, as they do not appear to

PAGE 102

102 phase and is more severe in coarser regions of the microstructure, consistent with localization of strain phase also seen at 800C, as illustrated in Figure 6 8. However, there is still significant strain accommodation in these alloys at 700C, as these compression tests were also interrupted at approximately 40% engineering strain. By contrast, al phase matrix show very poor ductility, even at elevated temperatures, suffering brittle failure in compression testing at 700C [ 1 ] 6.3.1 Effect of Microstructure Scale Since the refinement of microstructure scale is expected to affect mechanical behavior in these alloys, the effect of modifying microstructure scale on deformation and failure behavior was studied. In order to determine the effect of microstructure scale, samples were heat treated to produce a coarser as detailed in the section 5.2.2. Alloy samples were also cut from solution treated and aged slices of the alloy such that prior phase grain boundaries were not included in the sample s. In this way, samples were produced that have uniform microstructure throughout. These phase crystal, removing any effects the prior grain boundaries have on microstruct ure scale and strain localization. Compression testing of single crystal samples of fine and coarse samples reported in Section 6.1. As illustrated in Figure 6 8, both f ine microstructures have similar scale and morphology, while the coarse microstructure contains lath phase morphology phase. It can also be seen that the coarseness of the

PAGE 103

103 microstructure or presence of prior grain boundarie s does not have a strong effect on yield strength, but does affect flow softening behavior. Single crystal samples have less flow softening than the polycrystalline sample, which can be attributed to their lack of heavy localized deformation near prior grain boundaries. In polycrystalline samples, there is localization of strain near these boundaries due to inhomogeneous microstructure scale, which facilitates the nucleation of recovery and recrystallization. Figure 6 8: Comparison of f low curves of 15 at% Nb alloy with varying microstructure. Samples prepared to produce A ) ultrafine equiaxed polycrystalline microstructure, B ) ultrafine equiaxed single crystal microstructure or C ) coarse single crystal microstructure with enlarged lath shaped grai n morphology (Courtesy of author) Since deformation is more homogeneously distributed in the single crystal samples, higher levels of strain are required to initiate recovery mechanisms. It is also seen that the coarse microstructure has less flow soften ing than the fine single crystal sample, which can be phase increase s the distance that dislocations must travel before annihilating at a grain boundary, and increase s oundary sliding when compared with

PAGE 104

104 phase. With comparable strength and reduction in flow softening, a coarser microstructure may make it a better candidate for high temperature deformation resistance. 6.4 Tensile Testing Compression testing indic ates that an ultrafine microstructure with disconnected phase yields a balance of strength and strain accommodation at high temperature. However, the tensile behavior of these alloys has not been previously studied. One reason that tensile testing has been a challenge is alloy fabrication and heat treatment. Due to the severe nature of the thermal treatment these alloys must undergo in order to into final geometry un til heat treatment is completed. Initial attempts at tensile sample fabrication proved that heat treatment of dogbone specimen causes high thermal stresses during quenching, and high residual stresses afterwards. Due to this, if cracking does not occur u pon quenching, the dogbone is highly susceptible to cracking upon further handling, usually initiating at the fillet. Similarly, heat treatment of large plates is difficult due to the large thermal gradients experienced during quenching. Under both of th ese conditions, significant sample cracking occurred during quench, making them unsuitable for further processing. It was found that the most reliable method for sample heat treatment and machining is to section the alloy to rectangular samples near the f inal dimensions of the dogbone, solution treat, age, and then machine the final geometry. In this way, the alloy samples are sufficiently small so that even heat removal can occur during quenching, and the simple geometry removes any stress concentration that would otherwise occur at the fillet.

PAGE 105

105 Initial tensile testing of these samples at elevated temperature resulted in premature cracking and failure in the fillet region of the dogbone. In a sample of 15 at% Nb alloy tested at 800C, 10 4 s 1 it was fou nd that this failure was the result of highly localized strain due to the collar style grips used. This stress concentrator induced a high amount of deformation, initiating cracking through the sample as illustrated in Figure 6 9. The majority of deforma tion occurred within the grip region, plastically deforming the sample to the inner contour of the collar grips. Since deformation occurred outside of the gage section, meaningful flow data was not able to be obtained. Analysis of a second tensile test u nder the same conditions resulted in failure near the fillet region as well. SEM analysis of this sample, presented in Figure 6 10, found this failure to be due to a Nb rich inclusion near the fillet region. Due to the high amount of local stress placed on the dogbone sample by collar style grips, it was determined that additional testing should be conducted using more traditional grips that would distribute the load more evenly on the grip section of the samples. Figure 6 9: Tensile sample of 15 at% Nb alloy tested at 800C, 10 4 s 1 showing high amount of local plastic deformation; brittle tensile fracture. Note: additional cracking within grip region occurred during sample removal from grips (Courtesy of author)

PAGE 106

106 Figure 6 10: SEM of fracture surfa ce of tensile dogbone tested at 800C, 10 4 s 1 showing fracture initiation in high atomic weight inclusion A) Secondary electron mode; B) backscatter mode (Courtesy of author) Further tensile testing was conducted on dogbone samples at room temperature and these samples also fractured near the fillet region before any plastic deformation occurred. Observation of the fracture surface using optical and electron microscopy (Figure 6 11) indicate a mixed mode fracture that appears macroscopically brittle and microscopically ductile. Crack initiation begins on the sample face and appears to follow a few cracking paths, giving the fracture surface a faceted appearance. This could be explained by fracture along previous phase grain boundaries that would p rovide an easy cracking path. However, since the microstructure is fully transformed to very fine there is no immediately distinct pattern of the fracture surface on a macroscopic scale. C loser examination of the fracture surface (Figure 6 12) doe s not reflect classical intergranular fracture, microvoid coalescence, or cleavage failure modes. Instead, it has characteristics of a combination of a tearing topography surface fracture and quasi cleavage fracture [ 19 21 90 ] which is indicative of some plasticity phase during fracture despite the apparent brittle fracture

PAGE 107

107 Figure 6 11: SEM of fracture surface of tensile dogbone tested at room temperature and 10 4 s 1 showing facing sides of fracture surface. Fracture initiates on sample face, and follows faceted path. (Courtesy of author) Figure 6 12: SEM micrograph illustrating topography of phase fracture, indicative of mixed transgranular and intergranular modes (Courtesy of author)

PAGE 108

108 Figure 6 13: SEM micrograph of tensile sample face, showing crack branching near fracture surface, following path through phase (Courtesy of author) Figure 6 13 shows the macroscopic crack path through the tensile sample, as well as some crack branching that occurred during deformation. The crack appears to follow a pr phase grain boundary, and continues to follow this path while the fracture surface deviates. This major crack phase orientation and morphology, as do es

PAGE 109

109 microcracking near the fracture surface. Further analysis of cracking near the fracture surface and approximately 100 that cracking occurs phase. Figure 6 14: Detail of tensile sample face, illustrating microcracking solely through phase as well as lack phase fracture A) N ear fracture surface; B) away from fracture surface, (Courtesy of author) Figure 6 15: Comparison of + micros tructure scale and morphology A ) near failure site (100 200 m) and B ) far from failure site (1 2 mm) (Courtesy of author) phase while loc al finer regions do not experience the same level of cracking, as illustrated in Figure 6 14. Comparison of microstructure scale near to the fracture surface (100 200 2 mm) reveals a large difference in micros tructure scale (Figure 6 15). This confirms that fracture occurred

PAGE 110

110 phase grain boundary, since it has been demonstrated that coarser phase nucleation upon quenching.

PAGE 111

111 CHAPTER 7 SUMMARY AND CONCLUSIONS Through thermal analysis, heat treatment, and microstructur al evaluation was conducted on alloys in the Ti Al Nb Cr Mo system with high Nb concentration (15 25 at%) Alloys of composition Ti 45Al xNb 5Cr 1Mo (where x = 15, 20, 25 at%) were prepared via arc melting, solution treated, water quenched, and aged to produce + microstructure. Alloys with 15, 20, and 25 at% Nb were solution treated at 1400, 1450, and 1500C and aged at 1000, 950, and 1125C respectively. This evaluation has found that increasing Nb content has several effects on phase transformation and microstructure development. Primarily, increasing Nb increases phase transformation temperatures for equilibrium phase transformations, suppresses metast able ( + transformation temperatures, increases amount of TiAl form ed upon quench ing as Nb stabilizes all phases to higher temperature, and increases phase fraction of Nb phase. Ultrafine, relatively equiaxed microstructure was successf ully produced in 15 and 20 at % Nb alloys, but it has been observed that with an increase to 25 at %, aged phase. It has also been demonstrated that microstructural development i n these alloys is very sensitive to composition and processing parameters, such as heating and cooling rates. Cooling rate after solution treatment, as well as TiAl is vital to the + Towards the goal of designing a high + established that microstructural control can be accomplished in alloys containing 15 25 at % Nb through targeted chemistry and processi ng controls.

PAGE 112

112 Elevated temperature compression testing was conducted on alloys with 15, 20, + 45Al 15Nb 5Cr 1Mo at 700 and 800C shows good high temperature strength and large compressiv e strain (40%) without sustaining macroscopic fracture. Strain rate sensitivity and phase are the predominant deformation mechanisms at 800C, especia lly at slower strain rates. Microstructural analysis of deformed samples supports pparent and higher strength and work hardening rates may be attributed to the lack of recrystallization behavior in this temperature regime. The impact of dynamic recrystallization behavior at lower strain rates warrants further study in order to determin It was found that increased Nb content produces alloys with higher strength at phase present in the microstructure. Accompanying the increase in st rength was a decrease in compressive strain to failure, especially at 700C and highest strain rates. All samples strained significantly past yielding exhibited flow softening behavior, which can be attributed to dynamic recovery and recrystallization mec hanisms. Deformation of the 25 at % Nb alloy resulted in a decrease in strength between 800C and 900C that was much larger than the lower Nb phase content and connectivity, which undergoes a softening transition a round 900C, indicating that the high temperature deformation

PAGE 113

113 phase properties than that of alloys with a finer more disconnected precipitate. From the standpoint of microstructure development and mechanical b ehavior at elevated temperature, the 20 at% Nb alloy produces both optimal microstructure and properties for the goals of this work. A fine, disconnected phase morphology was produced through heat treatment, and high temperature yield strengths of appro ximately 600 1000 MPa were achieved when tested at strain rates of 10 4 to 10 2 s 1 This combination of strength with desired microstructure makes this alloy the best of the three studied in this work, though further characterization and alloy development is required in order to fully determine if this alloy is most suitable for mechanical applications. One important finding from tensile testing is that cracking and failure is controlled phase in this coarse alloy microstructure, rather phase volume fraction is not high enough for it to phase, which accommodates strain such that stresses are never high en phase phase. phase is potentially beneficial to high phase dis persion to redistribute stresses and prevent the formation of microcracks. In order to phase on fracture in these alloys, it is desirable to produce a single phase before aging the alloy to p +

PAGE 114

114 grains that would serve to initiate crack propagation through the alloy.

PAGE 115

115 CHAPTER 8 FUTURE WORK At this stag e in the research into high Nb containing Ti Al Nb based alloys, investigations have encompassed computational and experimental thermodynamic assessment, study of nucleation sequences, phase equilibria, and phase reactions, as well as room temperature comp ression, elevated temperature compression, and initial exploration into fracture toughness and tensile properties at room temperature [ 1 10 14 16 23 33 38 40 42 65 66 69 70 82 ] The present work explores the microstructure development and mechanical properties of alloys with varying Nb concentration as a first step towards a more full scientific understanding of alloys in this system, and there is ample opportunity for fur ther study. As a direct continuation of this work, there are several avenues that can be explored. The first of these is the improvement in room and high temperature mechanical behavior, namely in tension. This work showed that refinement of the microstr ucture is very important for improvement in mechanical properties grains are a weak point in tension due to localized strain and opening of cracks. Future work could be completed to overcome this challenge in a few ways. Study of alloying additions for grain boundary engineering would offer the grain boundaries. If the chemistry of alloys in this system can be altered such that there is reduced s grains, this could improve tensile behavior. This could be accomplished by beginning with first principles calculations [ 91 92 ] and can be experimentally verified by measuring the relative grains after

PAGE 116

116 solution treatment and quench via electron backscatter diffraction (EBSD) mapping. The texturing of each alloy can be related to the strength of the material in tension, but in order for this to be an appropriate measure, grain size must be controlled phase content and morphology. Alternatively if alloying additions can be found t hat completely inhibit the formation phase grain bou ndaries upon quenching resulting in a uniformly ultrafine microstructure, this would serve to reduce the amount of strain localization and cracking in the phase that ultimately led to failure in this study. Another approach is the fabrication of single phase samples that can then be heat treated to grain boundaries in these alloys provides a location for heterogeneous nucleation of phase upon quenching. The driving force for nucleation of phase calculated in Chapter 4 only reflects homogeneous nucleation, but in the experimental alloys it was seen that phase was comparatively easier grain boundaries, and therefore lack of coarse phase grains, single crystal samples did not experience strain localization. A uniformly fine microstructure would be able to ac commodate more strain before failure through the distribution of strain and the presence of phase to reduce crack length and alter the local stress state in the surrounding phase. A nother option would be to explore chemistry, thermo mechanical proce ssi ng, or other routes phase grains can be

PAGE 117

117 refined to a significant extent there would be less strain incompatibility between them, and if they are sufficiently small could even limit the scale. In the c phase grain size is on the order of 1 3 mm, phase grains could be refined to 10 would represent polycrystalli ne behavior rather than the bond strength of one to a grain boundaries. grain size may make the prevention of phase nucleation upon quenching unnecessary if thermal processing can be determined to nucleate and grow the phase homogenously from phase, such as very long term aging. In addition to characterization and improvement of tensile properties, it would be of interest to investigate fatigue behavior in these alloys, as well as the relative fracture toughness of the various microstructure scales and morphologies that can be produced. The investigation into fracture toughness and fatigue would allow for the determination of whether crack nucleation or propagation is most important in tensile failure of these alloys. In order for toughness to be appropriately tested, sampl es must be made according to ASTM standard E1820, and it should be noted that indentation methods are not suitable to calculate fracture toughness of these alloys due to the presence of the connected phase matrix, which is too ductile to produce accurate readings by this method. Finally, if these alloys are intended for use in high temperature applications, it would be relevant to investigate their creep behavior. Previous research has studied creep of alloys with high phase content [ 1 12 ] so it

PAGE 118

118 would be appropriate to characterize the creep behavior of low content alloys to see if creep behavior is dependent on the phase as previo usly established [ 1 12 ] or more directly related to the phase content and contiguity as is the compressive and tensile deformation In order to further explore deformation mechanisms in these alloys, transmission electron microscopy (TEM) could be conducted in order to gain insight into dislocation structures and interfaces The first aspects of interest would be the character of the / inte rface. This could be done by using focused ion beam (FIB) milling in order to extract samples that contain this interface. TEM analysis of the interface could then be conducted to analyze for coherency, and would add to the understanding of how phase p recipitates behave with phase matrix and if coherency changes through the course of long term aging Also of great interest would be TEM analysis of interrupted high temperature testing. This would be done with the goal of determining if dislocation recovery mechanisms such as dynamic recrystallization are occurring. Evidence may include the presence of recrystallized, dislocation free grains and the overall decrease in dislocation density of samples interrupte d near the trough of recrystallization oscillations versus t hose interrupted near the peaks. In order to accomplish such characterization, a high temperature mechanical testing system must be set up which has capability for quenching directly from the hot test stage. Ideally, the testing would still be conducted under vacuum in order to minimize the effect of surface oxidation of alloy properties. Alternatively, in situ TEM mechanical testing could be used to

PAGE 119

119 determine the dislocation behavior in very sm all samples. However, this may not be representative of overall alloy characteristics since the scale of the microstructure is not nanoscale, the stress states of the TEM foil would necessarily be different than those of bulk alloy samples. Another approa ch to studying dynamic recrystallization behavior in these alloys would be to conduct in situ high temperature mechanical testing while performing neutron scattering. This could be accomplished using systems such as the VULCAN at Oak Ridge National Labora tory in which the sample can be induction heated while loading and would also allow for in situ studies of texture changes, stress development, and damage [ 93 ]

PAGE 120

120 APPENDIX A CHEMICAL ANALYSIS OF ALLOYS Table A 1: Summary of EMPA results of 15Nb alloy from Certificate of analysis; Sophisticated Alloys, Inc., nationally certified external laboratory. Element at% at wt wt/mol alloy wt% ICP (wt%) ICP (at%) Ti 34 47.88 16.28 35.45 35.14 33.70 Al 45 26.98 12.14 26.44 26.67 45.39 Nb 15 92.90 13.93 30.35 30.13 14.89 Cr 5 51.99 2.59 5.66 5.87 5.18 Mo 1 95.94 0.96 2.09 2.19 1.05 Total 100 45.91 100 100 100.21 Table A 2 : Summary of EMPA results of 20 Nb alloy from Certificate of analysis; Sophisticated Alloys, Inc., nationally certified external laboratory. Element at% at wt wt/mol alloy wt% ICP (wt%) ICP (at%) Ti 29 47.88 13.89 28.83 29.06 29.23 Al 45 26.98 12.14 25.21 25.38 45.31 Nb 20 92.90 18.58 38.58 37.88 19.64 Cr 5 51.99 2.59 5.40 5.50 5.10 Mo 1 95.94 0.96 1.99 2.18 1.09 Total 100 48.17 100 100 100.37 Table A 3: Summary of EMPA results of 25Nb alloy from Certificate of Analysis; Sophisticated Alloys, Inc., nationally certified external laboratory. Element at% at wt wt/mol alloy wt% ICP (wt%) ICP (at%) Ti 24 47.88 11.49 22.79 22.70 23.90 Al 45 26.98 12.14 24.08 23.89 44.64 Nb 25 92.91 23.22 46.07 46.37 25.16 Cr 5 51.99 2.60 5.16 5.02 4.87 Mo 1 95.94 0.96 1.90 2.02 1.06 Total 100 50.41 100 100 99.64

PAGE 121

121 APPENDIX B THERMAL ANALYSIS (DSC) Upon initial cycling of the as cast or quenched alloys, there are some metastable phase transformations that occur. In the as cast microstructure, there is some phase phases that would be expected to form upon solidif ication. This results in non equilibrium transformations that are seen in the first cycle, but disappear in subsequent thermal cycling. In the case of the 15 at% Nb alloy, phase into the phase. Upon subsequent cycling, this peak turns into a single peak, phases that formed upon equilibrium cooling of the alloy begin to dissolve at the same temperature. Similarly, there is a double peak upon heating of 25 at% Nb alloy near 800C phase, as seen in Figure 4 6 This double phase nucleation from the non equilibrium cast or quenched microstructure. As such, there is an extra dissolution peak corresponding to the dissolution of this non equilibrium phase, seen a shoulder in the endothermic peak near 1350 C. After cycling these alloys, the metastable phase transformations do not appe ar since the alloy has sufficient during phases phases. Additionally, with thermal cycling of all of the alloys there is a suppression of any exothermic peaks occurring below about 1000C. That is to say that the non microstructure in the second and third cycles, rather than one containing the metastable phase from with the two ph ase microstruc ture can nucleate.

PAGE 122

122 Figure B 1: DSC of as cast 15Nb alloy, cycled three times. Note stability of transformation temperatures between cycles. Dark, solid curve indicates initial heating (top curve), melting and cooling, with subsequent cycles settling to similar heat flow levels after initial alloy melting and solidification Figure B 2: DSC of as cast 20Nb alloy, cycled three times. Note stability of transformation temperatures upon cooling, and change in nature of transformation upon heating with increa sing cycles.

PAGE 123

123 Figure B 3: DSC of as cast 25Nb alloy, cycled three times. Note stability of transformation temperatures upon cooling and heating, with lack of low temperature peaks upon second and third heating. Figure B 4: DSC curve upon heating of sol ution treated and quenched 15Nb alloy

PAGE 124

124 Figure B 5: DSC curve upon heating of solution treated and quenched 20Nb alloy Figure B 6: DSC curve upon heating of solution treated and quenched 25Nb alloy

PAGE 125

125 Figure B 7: Comparison of DSC curves produced upon ini tial heating of solution treated and quenched 15, 20, and 25 at% Nb alloys

PAGE 126

126 APPENDIX C MECHANICAL TESTING DATA Ti 45Al 15 Nb 5Cr 1Mo Figure C 1: Compression testing of Ti 45Al 15Nb 5Cr 1Mo alloy at 700C Figure C 2 : Compression testing of Ti 45 Al 15Nb 5Cr 1Mo alloy at 8 00C

PAGE 127

127 Ti 45Al 20 Nb 5Cr 1Mo Figure C 3 : Compression testing summary of Ti 45Al 20 Nb 5Cr 1Mo alloy at 700C at strain rates of 10 2 10 3 and 10 4 s 1 Figure C 4 : Compression testing summary of Ti 45Al 20 Nb 5Cr 1M o alloy at 8 00C at strain rates of 10 2 10 3 and 10 4 s 1

PAGE 128

128 Figure C 5 : Compression testing summary of Ti 45Al 20 Nb 5Cr 1M o alloy at 9 00C at strain rates of 10 2 10 3 and 1 0 4 s 1 Ti 45Al 2 5Nb 5Cr 1Mo Figure C 6 : Compression testing summary of Ti 45Al 25 Nb 5Cr 1M o alloy at 7 00C at strain rates of 10 2 10 3 and 10 4 s 1

PAGE 129

129 Figure C 7 : Compression testing summary of Ti 45Al 25 Nb 5Cr 1M o alloy at 8 00C at strain rates of 10 2 10 3 and 10 4 s 1 Figure C 8 : Compression testing summary of Ti 45Al 25 Nb 5Cr 1M o alloy at 9 00C at strain rates of 10 3 and 10 4 s 1

PAGE 130

130 Figure C 9 : Strain rate dependence of strength for 15Nb alloy, showing stress exponent determination Figure C 10 : Strain rate dependence of strength for 20Nb alloy, showing stress exponent determination

PAGE 131

131 Figure C 11 : Strain rate dependence of strength for 25Nb alloy, showing stress exponent determination. Note: for fitting at 90 0C, two lines of different slope were fitted to determine n the following table. Table C 1: Calculated values for n, Q for compression testing of alloys at high temperature. Note change in stress exponent for 25Nb alloy at 900C; apparent change in mechanism leads to differences in n, Q Alloy T ( C) n Q T range ( C) 15 Nb 700 8.18 800 5.56 36.9 ( 700 800 ) 1000 3.82 59.2 ( 800 1000 ) 20 Nb 700 8.28 800 6.33 34.3 ( 700 800 ) 900 5.03 47.4 ( 800 900 ) 25 Nb 700 11.93 800 8.37 23.9 ( 700 800 ) 900 7.24 [high] 31.7 ( 800 900 [high] ) 2.68 [low] 49.1 ( 800 900 [low] )

PAGE 132

132 LIST OF REFERENCES [1] B.J.G. deAragao, F. Ebrahimi, High temperature deformation of Nb Ti Al alloys with sigma+gamma microstructure, Materials Science and Engineering a Structural Materials Properties Microstructure and Processing, 208 (1996) 37 46. [2] F. Appel, R. Wagner, Microstructure and defo rmation of two phase gamma titanium aluminides, Materials Science & Engineering R Reports, 22 (1998) 187 268. [3] D.M. Dimiduk, Gamma titanium aluminide alloys an assessment within the competition of aerospace structural materials, Materials Science and Engineering a Structural Materials Properties Microstructure and Processing, 263 (1999) 281 288. [4] F. Appel, M. Oehring, R. Wagner, Novel design concepts for gamma base titanium aluminide alloys, Intermetallics, 8 (2000) 1283 1312. [5] G. E. Aviation, The GEnx Engine Family, in, General Electric Company, 2012. [6] J.P. Lin, X.J. Xu, Y.L. Wang, S.F. He, Y. Zhang, X.P. Song, G.L. Chen, High temperature deformation behaviors of a high Nb containing TiAl alloy, Intermetallics, 15 (2007) 668 674. [7] E. Hamzah, M. Kanniah, M. Harun, Effect of chromium addition on microstructure and creep strength of as cast Ti 48Al alloy, Materials Science and Engineering a Structural Materials Properties Microstructure and Processing, 483 84 (2008) 555 559. [8] J.P. Quast, C.J. Boehlert, Comparison of the microstructure, tensile, and creep behavior for Ti 24Al 17Nb 0.66Mo (atomic percent) and Ti 24Al 17Nb 2.3Mo (atomic percent) alloys, Metall. Mater. Trans. A Phys. Metall. Mater. Sci., 38A (2007) 529 536. [9] Q. Liu, P. Nash, The effect of Ruthenium addition on the microstructure and mechanical properties of TiAl alloys, Intermetallics, 19 (2011) 1282 1290. [10] M.S. Kesler, S. Goyel, O. Rios, D.M. Cupid, H.J. Seifert, F. Ebrahimi, A study of phase transformation in a TiAlNb al loy and the effect of Cr addition, Materials Science and Engineering a Structural Materials Properties Microstructure and Processing, 527 (2010) 2857 2863. [11] F. Appel, J.D.H. Paul, M. Oehring, U. Frobel, U. Lorenz, Creep behavior of TiAl alloys with enh anced high temperature capability, Metall. Mater. Trans. A Phys. Metall. Mater. Sci., 34A (2003) 2149 2164. [12] B.J.G. DeAragao, F. Ebrahimi, High Temperature Deformation of a Nb Ti Al Alloy with sigma+gamma Microstructure, in, MRS Proceedings, 1995, pp. 1247 1252.

PAGE 133

133 [13] F. Ebrahimi, B.J.G. Dearagao, J.R. Castillogomez, Evaluation of Fracture Toughness of Duplex microstructures by an Indentation Technique, Materials Science and Engineering a Structural Materials Properties Microstructure and Processing, 177 (1994) L7 L10. [14] F. Ebrahimi, D.T. Hoelzer, J.R. Castillogomez, Fracture toughness of sigma+X microstructures in the Nb Ti Al system, Materials Science and Engineering a Structural Materials Properties Microstructure and Processing, 171 (1993) 35 45. [ 15] O. Rios, S. Goyel, M.S. Kesler, D.M. Cupid, H.J. Seifert, F. Ebrahimi, An evaluation of high temperature phase stability in the Ti Al Nb system, Scr. Mater., 60 (2009) 156 159. [16] M.S. Kesler, Deformation and Fracture Behavior of Ti Al Nb (Cr,Mo) All oys with a Gamma + Sigma Microstructure at Ambient Temperature, in: Materials Science and Engineering Department, Universidy of Florida, Gainesville, FL, 2011, pp. 237. [17] X.J. Wu, A.K. Koul, Grain Boundary Sliding in the Presence of Grain Boundary Prec ipitates during Transient Creep, Metall. Mater. Trans. A Phys. Metall. Mater. Sci., 26 (1995) 905 914. [18] H. Tanaka, M. Murata, F. Abe, K. Yagi, The effect of carbide distributions on long term creep rupture strength of SUS321H and SUS347H stainless stee ls, Materials Science and Engineering a Structural Materials Properties Microstructure and Processing, 234 (1997) 1049 1052. [19] G.E. Deiter, Mechanical Metallurgy, McGraw Hill, London, 1988. [20] a.J.L. John Price Hirth, Theory of Dislocations, McGraw Hi ll, Inc., 1968. [21] R. Reed Hill, R. Abbaschian, Physical Metallurgy Principles, 3rd Edition ed., PWS Publishing Company, Boston, 1994. [22] G. Bean, F. Ebrahimi, H.J. Seifert, M. Manuel, High Temperature Deformation of Ti Al Nb Cr Mo Alloys, in, TMS 2012 : Annual Meeting and Exhibition, Orlando, FL, 2012. [23] S. Goyel, O. Rios, M.S. Kesler, F. Ebrahimi, Two step nucleation of the gamma phase in a Ti 45Al 18Nb alloy, Intermetallics, 18 (2010) 1491 1497. [24] J.P. Lin, L.L. Zhao, G.Y. Li, L.Q. Zhang, X.P. S ong, F. Ye, G.L. Chen, Effect of Nb on oxidation behavior of high Nb containing TiAl alloys, Intermetallics, 19 (2011) 131 136. [25] E.L. Hall, S. C. Huang, Microstructures of rapidly solidified binary TiAl alloys, Acta Metall. Mater., 38 (1990) 539 549. [ 26] Y. TiAl alloys, Materials Science and Engineering: A, 192 193, Part 2 (1995) 519 533.

PAGE 134

134 [27] Y. W. Kim, D.M. Dimiduk, Progress in the understanding of gamma titanium aluminides, JOM, 43 (1991) 40 47. [28] S. C. Huang, E.L. Hall, Characterization of the effect of vanadium additions to TiAl base alloys, Acta Metall. Mater., 39 (1991) 1053 1060. [29] Y. W. Kim, Ordered intermetallic alloys, part III: Gamma titanium aluminides, JOM, 46 (19 94) 30 39. [30] Y.W. Kim, A. Rosenberger, D.M. Dimiduk, Microstructural changes and estimated strengthening contributions in a gamma alloy Ti 45Al 5Nb pack rolled sheet, Intermetallics, 17 (2009) 1017 1027. [31] H. Lipsitt, D. Shechtman, R. Schafrik, The d eformation and fracture of TiAl at elevated temperatures, Metallurgical Transactions A, 6 (1975) 1991 1996. [32] M. Yamaguchi, Y. Umakoshi, The deformation behaviour of intermetallic superlattice compounds, Progress in Materials Science, 34 (1990) 1 148. [ 33] J.R. Castillo Gomez, Development of High Temperature Niobium Based Alloys, in: Materials Science and Engineering, Univerisity of Florida, 1990. [34] D. Vojtech, T. Popela, J. Hamacek, J. Kutzendorfer, The influence of tantalum on the high temperature characteristics of lamellar gamma + alpha 2 titanium aluminide, Materials Science and Engineering a Structural Materials Properties Microstructure and Processing, 528 (2011) 8557 8564. [35] C.G. Wilson, F.J. Spooner, Order in NbAl and TaAl type Sigma Phases, Journal of Materials Science, 12 (1977) 1653 1658. [36] B.T. Solomon, An investigation into the correlation of fracture toughness and microstructure for the sigma+beta structure in the Nb Ti Al ternary system, in, University of Florida, 1991. [37] S. Das, J.M. Howe, J.H. Perepezko, A high resolution transmission electron Al Mo alloy, Metallurgical and Materials Transactions A, 27 (1996) 1623 1634. [38] D.M. Cupid, O. Fabrichnaya O. Rios, F. Ebrahimi, H.J. Seifert, Thermodynamic re assessment of the Ti Al Nb system, Int. J. Mater. Res., 100 (2009) 218 233. [39] S. Goyel, The Study of the Beta to Gamma Phase Transformation in High Nb Ti Al Nb Alloys, in: Materials Science and Eng ineering, University of Florida, Gainesville, FL, 2011. [40] F. Ebrahimi, Castillogomez, Jr., The Effect of Plastic Deformation on Fracture morphology of the Sigma Phase in the Nb Ti Al System, Acta Metall. Mater., 40 (1992) 1409 1416.

PAGE 135

135 [41] V. Recina, D. L undstron, B. Karlsson, Metallurgical and Materials Transactions A, 33A (2008) 2869 2881. [42] D.T. Hoelzer, in: Materials Sciencenand Engineering, University of Florida, Gainesville, FL, 1996. [43] J.B. Forsyth, J.P. Brown, Structure of Sigma Phase Nb2Al, Acta Crystallographica, 14 (1961) 362. [44] G. Bergman, D. Shoemaker, The Determination of the Crystal Structure of the Sigma Phase in the Iron Chromium and Iron Molybdenum Systems, Acta Crystallographica, 7 (1954) 857. [45] E.O. Hall, S.H. Algie, The Sig ma Phase, Metallurgical Reviews, 11 (1966). [46] T. Ishimasa, Y. Kitano, Y. Komura, Faulted Structure in the Fe Cr sigma Phase, Phys. Stat. Sol., 66 (1981) 703. [47] M.J. Marcinkowski, D.S. Miller, A study of defect sub structures in the Fe Cr sigma phase by means of transmission electron microscopy, Phil. Mag., 7 (1962). [48] D.S. Zhou, H.W. Ye, K.H. Kuo, An HREM study of the intergrowth structures of sigma related phases and the mu phase, Philosophical Magazine A, 57 (1988) 907 922. [49] A. Ohta, A. Noya, M. Takeyama, M. Taguchi, T. Sase, K. Sasaki, Preparation of Ta2Al intermetallic compound films and their application as diffusion barriers to Cu penetration, Thin Solid Films, 278 (1996) 6 11. [50] H. Zhu, e. al, Materials Science & Engineering A, 483 484 (2008) 533 536. [51] E.A. Ott, T.M. Pollock, Metallurgical and Materials Transactions A, 29A (1998) 965 978. [52] J.G.L. Ruizaparicio, F. Ebrahimi, Diffusivity in the Nb Al Binary Solid Solution, J. Alloy. Compd., 202 (1993) 117 123. [53] A.A. Bondar, V.T Witusiewicz, U. Hecht, M.V. Remez, V.M. Voblikov, N.I. Tsyganenko, Y.I. Yevich, Y.M. Podrezov, T.Y. Velikanova, S tructure and Properties of Ti Al Alloys Doped with Nb and Ta Powder Metallurgy and Metal Ceramics, 50 (2011) 397 415. [54] S. Azad, R.K. Mandal, A.K. Singh, Materials Science & Engineering A, 429 (2006) 219 224. [55] D. Banerjee, Progress in Materials Science, 42 (1997) 135 158. [56] S. Jung, H.S. Jang, M.H. Oh, J.H. Lee, D.M. Wee, Materials Science & Engineering A, 329 (2002) 13 18.

PAGE 136

136 [57] A .K. Singh, D. Banerjee, Metallurgical and Materials Transactions A, 28 (1997) 1735 1743. [58] H.Z. Niu, Y.Y. Chen, S.L. Xiao, L.J. Xu, Microstructure evolution and mechanical TiAl alloy, Intermetallics, 31 (2012) 225 231. [59] M.L. Weaver, M.J. Kaufman, P hase Relationships and Transformations in the Ternary Al Ti Ta System Acta Metall. Mater., 43 (1995) 2625 2640. [60] Y.S. Kim, K.R. Javed, G.J. Abbaschain, Final report on contract No. N00014 86 K 0178, Defense Advanced Researc h Projects Agency and Office of Naval Research, (1989). [61] W.J. Boettinger, A.J. Shapiro, J.P. Cline, F.W. Gayle, L.A. Bendersky, F.S. Biancaniello, Investigation of the phase constitution of Al2TiTa, Scr. Metall. Materialia, 25 (1991) 1993 1998. [62] V. T. Witusiewicz, A.A. Bondar, U. Hecht, V.M. Voblikov, O.S. Fomichov, V.M. Petyukh, S. Rex, Experimental study and thermodynamic modelling of the ternary Al Ta Ti system, Intermetallics, 19 (2011) 234 259. [63] C. McCullough, J.J. Valencia, C.G. Levi, R. Me hrabian, P eritectic Solidification of T i A l T a Alloys in the Region of Gamma TiAl Materials Science and Engineering a Structural Materials Properties Microstructure and Processing, 156 (1992) 153 166. [64] T. Velikanova, M. Turchanin, S. Ilyenko, G. Effenberg, Critical evaluation of the Aluminium Tantalum Titanium system, Calphad, 33 (2009) 192 199. [65] D.M. Cupid, O. Fabrichnaya, F. Ebrahimi, H.J. Seifert, Thermodynamic assessment of the Al Mo system and of the Ti Al Mo System from 0 to 20 at.% Ti, Intermetallics, 18 (2010) 1185 1196. [66] D.M. Cupid, M.J. Kriegel, O. Fabrichnaya, F. Ebrahimi, H.J. Seifert, Thermodynamic assessment of the Cr Ti and first assessment of the Al Cr Ti systems, Intermetallics, 19 (2011) 1222 1235. [67] M. Hatherly, F.J. H umphreys, Recrystallization and Related Annealing Phenomena, Elsevier Science, Ltd., Tarrytown, NY, 1996. [68] C.M. Sellars, M.J. Lutton, Dynamic Recrystallization in Ni and Ni Fe Alloys during High Temperature Deformation, Acta Metallurgica, 17 (1969) 103 3. [69] F. Ebrahimi, J.G.L. RuizAparicio, Diffusivity in the Nb Ti Al ternary solid solution, J. Alloy. Compd., 245 (1996) 1 9. [70] O. Rios, F. Ebrahimi, Spinodal decomposition of the gamma phase upon quenching in the Ti Al Nb ternary alloy system, Interm etallics, 19 (2011) 93 98.

PAGE 137

137 [71] J. Lankford, Journal of Materials Science Letters, 1 (1982) 493 495. [72] M.F. Ashby, Strengthening Methods in Crystals, Applied Science, (1971) 137 192. [73] X.Z. Liao, S. Cheng, E. Ma, Y.T. Zhu, Y.H. Zhao, Simultaneously I ncreasing the Ductility and Strength of Nanostructured Alloys, Adv. Mater., 18 (2006) 2280 2283. [74] M.H. F.J. Humphreys, Recrystallization and Related Annealing Phenomena, Elsevier Science Ltd., Tarrytown, NY, 1996. [75] W.G. Johnston, D.F. Stein, Stress dependence of dislocation velocity inferred from strain rate sensitivity, Acta Metallurgica, 11 (1963) 317 318. [76] F.J. Humphreys, M. Hatherly, Recrystallization and Related Annealing Phenomena, Elsevier Science Ltd, Tarrytown, NY, 1996. [77] R.W. Hertz berg, Deformation and Fracture Mechanics of Engineering Materials, 3rd Edition ed., John Wiley & Sons, In., 1989. [78] D. Hull, D.J. Bacon, Introduction to Dislocations, Robert Maxwell, M. C., 1984. [79] A.H. Committee, Volume 3, Alloy Phase Diagrams, ASM International, ASM Handbook Online, 1992. [80] A. International, Standard Test Methods for Tension Testing of Metallic Materials (E8/E8M), in, ASTM International, West Conshohocken, PA, 2009 (revised). [81] J. O. Anderson, T. Helander, L. Hoglund, P. Shi, B. Sundman, THERMO CALC & DICTRA, Computational Tools for Materials Science, Calphad, 26 (2002) 273 312. [82] G. Bean, F. Ebrahimi, M. Manuel, High temperature deformation of Ti Al Nb Cr Mo 137. [83] S. Neve, P. Masset, M. Shutze, H. Boumann, K. Bethge, H.E. Zschau, Characterization of the long time oxidation protection of fluorine implanted technical Ti Al alloys using ipn beam methods, Nuclear Instruments and methods in Physics Research B, 266 (2008) 2441 2445. [84] Y. Wu G.M. Song The Isothermal Oxidation Behavior of a Ti 3 Al Nb V Mo Alloy, The Open Corrosion Journal, 3 (2010) 7 15. [85] A.H. Committee, Volume 2, Properties and Selection: Nonferrous Alloys and Special Purpose Materials, AS M International, ASM Handbook Online, 1992. [86] M.H. Poech, H.F. Fischmeister, Deformation of Two Phae Materials: A Model Based on Strain Compatibility, Acta Metallurgica Et Materialia, 40 (1992) 487 494. [87] A.A. Griffith, Proc. First Int. Cong. App. Mech, 1 (1924) 55.

PAGE 138

138 [88] C.G. Sammis, M.F. Ashby, Acta Metallurgica, 34 (1986) 511. [89] M.F. Ashby, S.D. Hallam, Acta Metallurgica, 34 (1986) 498. [90] A.W. Thompson, J.C. Chestnutt, Identification of a Fracture Mode: The Tearing Topography Surface, Metallurgical Transactions A, 10A (1979) 1193 1196. [91] L.Q.C.e. al., MRS Bulletin, 26 (2001) 197. [92] M. Asta, V. Ozolins, C. Woodward, A First Principles Approach to Modeling Alloy Phase Equilibria, JOM, (2001) 16. [93] VULCAN Beam Line 7: The Engineering Materials Diffractometer at SNS, O.R.N. Lab (Ed.), ORNL, http://neutrons.ornl.gov/vulcan/

PAGE 139

139 BIOGRAPHICAL SKETCH Glenn obtained his BS in Mechanical Engineer ing from the University of Central Florida, while undertaking his research and Honors Undergraduate Thesis in Solid Oxide Fuel Cells under the direction of Dr Nina Orlovskaya During his undergraduate career, Glenn had the opportunity to work as an intern in the NASA Kennedy Space Center Materials Failure Analysis Laboratory learning about practical application and analysis of materials. As a graduate student at the University of Fl orida, Glenn began work under the advisement of Dr. Fereshteh Ebrahimi, and has cont inued to the completion of his m pursuing his Ph.D. conducting research on alloy design, microstructure development, and mechanical testing of high temperature alloys based on the Ti Al Nb system. Glenn has also recently begun work with aluminum alloys, acting as team leader and senior researcher on the design and development of Al based self healing metal matrix composites with shape me mory alloy reinforcement.