UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Mechanical properties of dilute zinc - titanium alloys 1970

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
UBC_1970_A1 W35.pdf [ 21.32MB ]
UBC_1970_A1 W35.pdf
Metadata
JSON: 1.0104032.json
JSON-LD: 1.0104032+ld.json
RDF/XML (Pretty): 1.0104032.xml
RDF/JSON: 1.0104032+rdf.json
Turtle: 1.0104032+rdf-turtle.txt
N-Triples: 1.0104032+rdf-ntriples.txt
Citation
1.0104032.ris

Full Text

THE MECHANICAL PROPERTIES OF DILUTE ZINC - TITANIUM ALLOYS by ROBERT JAMES WALDRON B.A,Sc., University of British Columbia, 1965 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in the Department of METALLURGY We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA Apr i l , 1970 I n p r e s e n t i n g t h i s t h e s i s in p a r t i a l f u l f i l m e n t o f t h e r e q u i r e m e n t s f o r an advanced degree a t t h e U n i v e r s i t y o f B r i t i s h C o l u m b i a , I a g r e e t h a t t h e L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r r e f e r e n c e and s t u d y . I f u r t h e r ag ree t h a p e r m i s s i o n f o r e x t e n s i v e c o p y i n g o f t h i s t h e s i s f o r s c h o l a r l y p u r p o s e s may be g r a n t e d by t h e Head o f my Depar tment o r by h i s r e p r e s e n t a t i v e s . I t i s u n d e r s t o o d t h a t c o p y i n g o r p u b l i c a t i o n o f t h i s t h e s i s f o r f i n a n c i a l g a i n s h a l l n o t be a l l o w e d w i t h o u t my w r i t t e n p e r m i s s i o n . Depar tment The U n i v e r s i t y o f B r i t i s h Co lumbia Vancouver 8, Canada i ABSTRACT Zinc-titanium alloys (0.07-0.6 wt.%Ti.) in the form of compacted powder and c h i l l castings have been extruded at temperatures between 150°C and 350°C. The mechanical properties of these alloys have been studied as a function of temperature, strain rate, grain size and intermetallic (Znj^Ti) distribution. Due to a high value of "k" in the Hall-Petch relationship, maximum strengthening is obtained by a reduction i n grain size. However because of an increasing amount of grain boundary shear, this potential is not realized. The operation of dynamic recovery mechanisms at 20°C and higher also results i n limitations upon the development of high strength. The use of powder metallurgical techniques gives rise to the formation of intermetallic distributions which inhibit these processes and results i n high strength (>60,000 p.s.i.) and low strain rate sensitivity (m ^ 0.02). The mechanical properties are not a function of i n i t i a l powder size. The properties obtained using c h i l l castings do not reach these levels due to the d i f f i c u l t y associated with forming a fine second phase on s o l i d i f i c a t i o n . Such a distribution i s required to obtain a small stable grain size during subsequent extrusion. To satisfy compatibility requirements deformation modes other than the two supplied by basal s l i p must be invoked. High strengths are observed when grain boundary shear and migration are inhibited by the distribution of the second phase or by orientation effects. Under such conditions, non basal s l i p and basal s l i p are the operative deformation mechanisms. Significantly lower strengths result i f grain boundary shear and basal s l i p satisfy the conditions necessary for ductile behaviour. The strain rate sensitivity parameter at 20 C l i e s i n the range 0.02-0.07. Varying amounts of grain boundary shear occur, nevertheless deformation is s l i p controlled. Increased strain rate sensitivities are observed at high temperatures, but failure by cavitation limits d u c t i l i t y . The strain rate sensitivity i s not a function of titanium concentration. Under constant fabrication conditions the strength generally increases with increased Zn-j^Ti content. The thermal st a b i l i t y of the intermetallic distribution prescrib the fabrication conditions which must be used to develop high strength, and the temperature to which the mechanical properties can be retained. The high strength microstructures appear to be stable up to at least 150°C for short periods of time. ACKNOWLEDGEMENTS The author i s grateful for the advice and encouragement given by his director, Dr. N.R. Risebrough. Many of the faculty, staff and students of the Department of Metallurgy contributed helpful suggestions. The author also wishes to thank two great families for their support - the Waldrons and the Cramonds. Financial assistance provided by the Aluminum Company of Canada and the National Research Council of Canada is gratefully ac- knowledged. XV TABLE OF CONTENTS PAGE 1. INTRODUCTION 1 2. PROCEDURE 7 2.1 Starting Materials 7 2.2 Melting Procedures 7 2.3 Atomizing 9 2.4 Construction of the Atomizer 11 2.5 Compaction 11 2.6 Extrusion 12 2.7 Tensile Specimen Preparation 15 2.8 Metallography 16 2.9 Testing Procedures 17 3. ZINC - TITANIUM PHASE DIAGRAM 21 4. RESULTS 25 4.1 Reproducibility 25 4.2 Effect of Extrusion Temperature 25 4.3 C h i l l Cast and Powder Microstructures 31 4.3.1 Microstructure of Cast B i l l e t s 31 4.3.2 Microstructure of Powder 34 4.3.3 Extent of Oxidation of Atomized Powder 38 4.4 Microstructural Characteristics of Extruded Alloys 42 4.4.1 Stringering During Extrusion of Powders 42 4.4.2 Cast Material 45 4.5 The Effect of Extrusion Ratio 52 4.6 Importance of Preferred Orientation 52 PAGE 4.7 Thermal S t a b i l i t y 54 4.7.1 Introduction 54 4.7.2 E f f e c t of Extrusion Temperature 55 4.7.3 E f f e c t of Annealing Temperature 59 4.7.3.1 Powder Ma t e r i a l 59 4.7.3.2 True Stress - S t r a i n Curves 65 i 4.7.3.3 Annealed Cast-Extruded Zn-Ti A l l o y s 69 4.8 The E f f e c t of Titanium Concentration 70 4.8.1 Powder Material 70 4.8.2 Cast M a t e r i a l 76 4.8.3 D u c t i l i t y 76 4.9 Deformation C h a r a c t e r i s t i c s 77 4.9.1 Introduction 77 4.9.2 Deformation Mechanisms 77 4.9.2.1 S l i p 77 4.9.2.2 Twinning 77 4.9.2.3 Grain Boundary Shear 79 4.9.3 Recovery Mechanisms 80 4.9.3.1 Introduction 80 4.9.3.2 Grain Boundary Migration 80 4.9.3.3 D i s l o c a t i o n Climb 82 4.9.4 The Occurrence and Importance of Substructure i n Zinc 82 4.10 Strain-Rate S e n s i t i v i t y 83 4.10.1 Introduction 83 4.10.2 Deformation of Pure Zinc 85 4.10.3 Deformation of Zinc-Titanium A l l o y s 88 4.10.3.1 Strain-Rate S e n s i t i v i t y 88 v i PAGE 4.10.3.2 Ductility 94 4.11 High Temperature Deformation Characteristics 95 4.11.1 Deformation of Zinc-Titanium and Zinc-Aluminum Alloys 95 4.11.1.1 Introduction 95 4.11.1.2 Effect of Temperature on the Flow Stress - Strain-Rate Relationship 96 4.11.1.3 Ductility and Fracture Mechanisms 99 4.11.1.4 Activation Energy Analysis 104 4.11.2 Yield Stress Dependence on Temperature 106 4.12 Hall-Petch Analysis 110 4.12.1 Introduction 110 4.12.2 Petch Equation and Hexagonal Metals 111 4.12.3 Modifications of the Petch Analysis 112 4.12.4 Petch Analysis and Zinc-Titanium Alloys 113 4.12.4.1 Flow at -100°C 113 4.12.4.2 Flow at 20°C 115 4.12.4.2.1 Stringered Alloys 115 4.12.4.2.2 Unstringered Alloys 116 4.12.5 Value of a 117 o 5. SUMMARY 119 5.1 Introduction ±19 5.2 Deformation Characteristics 119 5.2.1 Strength 119 5.2.2 Strain-Rate Sensitivity 121 5.2.3 Dynamic Recovery 122 5.2.4 Ductility and Fracture 123 5.3 Thermal Stability 5.4 Other Considerations CONCLUSIONS BIBLIOGRAPHY v i i i LIST OF FIGURES PAGE F i g . 1 Hypothetical phase diagram. 5 F i g . 2 Schematic representation of the atomizing apparatus. 10 F i g . 3 Schematic representation of the extrusion apparatus. 13 Fi g . 4 Standardized extrusion time - temperature conditions. 14 F i g . 5 Zn - 0.6 wt.% T i a l l o y extruded from -35 + 100 mesh powder at 150°C - transmission technique. 18 F i g . 6 Zn - 0.6 wt.% T i a l l o y extruded from -35 + 100 mesh powder at 150°C - r e p l i c a t i o n technique. 19 F i g . 7 Zinc - titanium phase diagrams. 23 F i g . 8 Zinc r i c h end of the zinc-titanium phase diagram. 24 F i g . 9 The e f f e c t of extrusion temperature on- the y i e l d strength of Zn - 0.6 wt.% T i a l l o y s . 28 F i g . 10 The e f f e c t of extrusion temperature on the d u c t i l i t y of Zn - 0.6 wt.% T i a l l o y s . 29 F i g . 11 Zn - 0.36 wt.% T i a l l o y c h i l l cast microstructure. 32 F i g . 12 (a) Electron microprobe absorbed electron image of Zn - 0.36 wt.% T i c h i l l cast microstructure. 33 (b) T i X-ray scan of Zn - 0.36 wt.% T i c h i l l cast microstructure. 33 F i g . 13 Mi c r o s t r u c t u r a l c h a r a c t e r i s t i c s of the eutectic i n a c h i l l cast Zn - o.36 wt.% T i a l l o y . 35 F i g . 14 C h i l l cast microstructure of Zn - 0.6 wt.% T i a l l o y . 36 F i g . 15 Electron micrograph of the eutectic structure i n -35 + 100 mesh Zn - 0.6 wt.% T i powder. 37 F i g . 16 (a) Absorbed electron image from an electron microprobe analysis of -35 + 100 mesh Zn - 0.6 wt.% T i powder. 39 (b) T i X-ray scan from an electron microprobe analysis of -35 + 100 mesh Zn - 0.6 wt.% T i powder. 39 ix PAGE Fig. 17 Electron microprobe analysis of Zn - 0.6 wt.% T i -35 + 100 mesh powder compact: (a) Absorbed electron image 40 (b) T i X-ray scan 40 (c) 0^ X-ray scan. 41 Fig. 18 Microstructure of a Zn - 0.6 wt.%'Ti'alloy extruded at 150°C using -35 + 100 mesh powder. 43 Fig. 19 Microstructure of a Zn - 0.6 wt.% T i alloy extruded at 250°C using -35 + 100 mesh powder. 43 Fig. 20 Microstructure of a Zn - 0.6 wt.% Ti alloy extruded at 350°C using -35 + 100 mesh powder. 44 Fig. 21 The origin of the stringered structure i n a Zn - 0.6 wt.% Ti alloy fabricated from -̂35 + 100 mesh powder at 175°C. 46 Fig. 22 The origin of the stringered structure in a Zn - 0.6 wt.% T i alloy fabricated from -35 + 100 mesh powder at 175°C. 47 Fig. 23 Electron microprobe analysis of Zn - 0.6 wt.% T i alloy extruded at 250°C using -35 + 100 mesh powder: (a) Absorbed electron image 48 (b) T i X-ray scan 48 (c) Zn X-ray scan. 49 (d) O2 X-ray scan. 49 Fig. 24 Microstructure of a c h i l l cast Zn - 0.6 wt.% T i alloy extruded at 150°C. 50 Fig. 25 Microstructure of a c h i l l cast Zn - 0.6 wt.% T i alloy extruded at 250°C. 50 Fig. 26 Microstructure of a c h i l l cast Zn - 0.6 wt.% Ti alloy extruded at 350°C. 51 Fig. 27 The effect of extrusion ratio on the yield stress of Zn - 0.6 wt.% Ti alloys extruded at 175°C using -35 + 100 mesh powder. 53 Fig. 28 The structure and mechanical properties of a Zn - 0.6 wt.% Ti alloy extruded at 175°C using -35 + 100 mesh powder. 57 Fig. 29 The effect of extrusion temperature on the yield stress of a thermally stabilized Zn - 0.6 wt.% Ti alloy. 58 Fig. 30 (a) The microstructure resulting from extrusion of a -100 + 200 mesh powder b i l l e t annealed for 1 hour prior to extrusion at 150°C. 60 (b) The microstructure resulting from extrusion of a -100 + 200 mesh powder b i l l e t annealed for 1 hour prior to extrusion at 350°C. 60 X PAGE Fig. 31 The effect of annealing temperature on the yield strength of Zn - 0.6 wt.% T i alloy at 20°C. 62 Fig. 32 The effect of annealing temperature on the du c t i l i t y of Zn - 0.6 wt.% T i alloy at 20°C. 63 Fig. 33 The microstructure resulting from a short term anneal at 180°C of a Zn - 0.6 wt.% T i alloy fabricated at 175°C using -35 + 100 mesh powder. 64 Fig. 34 The microstructure of a Zn - 0.6 wt.% T i alloy fabricated at 175°C from -35 + 100 mesh powder and annealed for 44 hours at 350°C. 66 Fig. 35 The microstructure of a Zn - 0.6 wt.% T i alloy fabricated at 175°C from -35 + 100 mesh powder and annealed for 94 hours at 400°C. 66 Fig. 36 The effect of annealing temperature on the true stress - strain relationship at 20°C for Zn - 0.6 wt.% T i alloys extruded at 175°C using -35 + 100 mesh powder. 67 Fig. 37 True stress - strain curves for zinc-titanium extrusions of -35 + 100 mesh powder at 175°C. 72 Fig. 38 True stress - strain curves for zinc-titanium extrusions of -35 + 100.mesh powder at 350°C. 73 Fig. 39 True stress - strain curves for zinc-titanium extrusions of c h i l l castings at 175°C. 74 Fig. 40 True stress - strain curves for zinc-titanium extrusions of c h i l l castings at 350°C. 75 Fig. 41 Metallographic evidence of non-basal sl i p in a c h i l l cast Zn - 0.32 wt.% T i alloy extruded at 175°C. 78 Fig. 42 Metallographic evidence of twinning in a c h i l l cast Zn - 0.16 wt.% T i alloy extruded at 175°C. 78 Fig. 43 Metallographic evidence of grain boundary shear and migration in a c h i l l cast Zn - 0.32 wt.% Ti alloy extruded at 350°C. 81 Fig. 44 Metallographic evidence of grain boundary shear and migration in a c h i l l cast Zn - 0.32 wt.% T i alloy extruded at 350°C. 81 Fig. 45 Metallographic evidence of substructure i n a c h i l l cast Zn - 0.16 wt.% Ti alloy extruded at 175°C. 84 XI PAGE Fig. 46 The flow stress - yield dependence on strain-rate of zinc fabricated from -325 mesh powder at room temperature. 86 Fig. 47 The variation of "m" with strain-rate for pure zinc fabricated from -325 mesh powder. 87 Fig. 48 The strain-rate sensitivity of zinc-titanium alloys fabricated from -35 + 100 mesh powder at 175°C. 89 Fig. 49 The strain-rate sensitivity of zinc-titanium alloys fabricated from -35 + 100 mesh powder at 350°C. 90 Fig. 50 The strain-rate sensitivity of zinc-titanium alloys fabricated from c h i l l castings at 175°C. 91 Fig. 51 The strain-rate sensitivity of zinc-titanium alloys fabricated from c h i l l castings at 350°C. 92 Fig. 52 The microstructure of a Zn - 0.2 wt.% Al alloy, cast and extruded at 150°C. 97 Fig. 53 The microstructure of a Zn - 0.07 wt.% Ti alloy extruded at 350°C using -35 + 100 mesh powder. 97 Fig. 54 The effect of temperature on the flow stress - strain-rate relationship for Zn - 0.07 wt.% Ti alloys extruded at 350°C using -35 + 100 mesh powder. 98 Fig. 55 Cavitation failure at 300°C in a Zn - 0.07 wt.% T i alloy extruded at 350°C using -35 + 100 mesh powder. 101 Fig. 56 Scanning electron micrograph of cavitation failure at 300°C in a Zn - 0,07 wt.% T i alloy extruded at 350°C using -35 + 100 mesh powder: (a) Reflected electrons (b) Secondary electrons. 102 Fig. 57 Schematic representation of cavitation due to the presence of second phase particles on a grain boundary. 103 Fig. 58 Arrhenius plot for Zn - 0.07 wt.% Ti alloy extruded at 350°C using -35 + 100 mesh powder. 107 Fig. 59 The dependence of yield stress on temperature of a thermally stabilized Zn - 0.6 wt.% T i alloy. 108 Fig. 60 The effect of temperature on the flow stress - strain-rate relationship of a thermally stabilized Zn - 0.6 wt.% T i alloy. 109 Fig. 61 The Hall-Petch plot for zinc-titanium alloys. 114 XXI LIST OF TABLES PAGE Table I. Composition of Extruded Materials. 8 Table II. Summary of the Experimental Investigations of the Zinc Rich End of the Zinc-Titanium Phase Diagram. 22 Table III. Reproducibility of Mechanical Properties Along Extrusions. 26 Table IV. The Effect of Powder Size on the Mechanical Properties of Zn - 0.6 wt.% Ti Alloys. 30 Table V. Mechanical Properties of -35 + 100 Mesh Zn - 0.6 wt.% Ti Powder Extruded at 175°C and Annealed at Various Temperatures.68 Table VI. The Effect of Titanium Concentration on Mechanical Properties. 71 Table VII. Summary of Strain-Rate Sensitivity Parameters and the Variables Affecting these Parameters for Zn - 0.16 wt.% Ti Alloys. 93 1. INTRODUCTION Considerable effort has been made to improve the mechanical properties of metals at elevated temperatures through the introduction of one or more secondary phases into the metal matrix. Research along these lines has led to the development of multiphase alloys (dispersion strengthened alloys) which retain their strength at elevated temperatures (1-3) and have superior creep rupture characteristics. The definition of dispersion strengthening varies somewhat in the literature. However a dispersion strengthened alloy is generally considered to consist of a structure of fine, non-coherent second phase particles which are randomly distributed in a pure metal matrix or solid solution. Thus conventional strengthening is usually defined in terms of retard- ation of dislocation motion. As a result, models developed to explain this type of hardening are based on the description of the size, amount, and spacing of obstacles to dislocation motion.^ This approach describes hardening quite adequately when reasonably large grain sizes are involved, or when the dispersoid spacing is small in comparison with the grain size, but i t leads to an incomplete analysis as the grain size i s reduced. This suggests that grain boundaries must be considered in order to explain the role of second phase additions. Although dispersion hardening has been discussed at length in the literature, l i t t l e i s known about the individual contribution of such variables as grain size, mechanical properties of secondary additions, volume fraction, size and distribution of the dispersoid, etc. The basic understanding of the effect of these variables i s thought to be the approach to the development of even more effective alloys. The dispersion strengthening of zinc presents some rather unique problems as a result of i t s low melting point and consequent high effective temperature at room temperature. (T = T/Tm = 0.42) H Dynamic recovery mechanisms such as dislocation climb and grain boundary migration occur readily at 20°C.^^ The result is that zinc does not work harden appreciably and has poor creep properties at normal temperatures. Although the mechanical properties have been enhanced markedly by means of alloying, conventional hardening techniques, which are successful in other metal systems, are often ineffective when applied to zinc. The development of wrought zinc alloys has been inspired by possible commercial competition with aluminum and copper alloys and has led to the industrial use of such alloys as Zn-Cu-Ti and Z n - T i . ^ ' ^ ' ^ L i t t l e i s known regarding the basic deformation and recovery character- i s t i c s of these alloys, or the role of the dispersoid. At low temperatures where the grain boundaries may be con- sidered "hard", the strength of zinc is expected to increase as the grain size is decreased. Essentially this behavior has been related to the operation of grain boundaries as effective barriers to dislocation movement. This effect is usually analyzed by means of the Hall-Petch relationship which relates the yield stress and grain size according to the expression: _x x = T + kd ^ y o where T is the yield stress, k and x are y ' o constants, and d is the grain size. The high "k" values observed for zinc suggest high strengthening potential. However at fine grain sizes, grain boundaries may become the reason for decreased strength rather than increased strength. As a consequence of the high homologous temperature, grain boundary shear can occur readily at ambient temp- eratures. Thus grain refinement w i l l result in the improvement of the mechanical properties of a metal or alloy only i f the boundaries remain immobile. This suggests that the development of high strength alloys is at least in part dependent on achieving a small! stable grain structure and a fine stable dispersoid. In terms of zinc alloys i t is thought that the microstructure should remain stable up to at least 75°C before the alloy could be of any commercial importance. Increased interest in high temperature creep properties and superplastic deformation has led to an effort to define more clearly the role and characteristics of grain boundary deformation. Many different approaches have been taken in order to (2) develop a small grain size and second phase dispersion. These include: a) Solid state transformations b) Gas-metal reactions c) Precipitation from liquids during s o l i d i f i c a t i o n d) Powder metallurgical techniques. The use of powder or particulate metallurgical techniques has proved to be applicable to zinc alloys. These techniques have been applied to many dilute zinc alloy compositions with a wide range of mechanical (8,9,25-27) properties. The development of alloys hardened by dispersed intermetallic compounds through the application of powder metallurgical procedures has been the most successful approach from the point of view of the develop- ment of high strength and desirable creep properties. The so l i d i f i c a t i o n 4 or atomization of suitable alloys from a hypothetical system such as is shown in Fig. 1 has been shown to be advantageous. Solidification of a dilute alloy from this type of system w i l l result in a structure consisting of a dispersion of intermetallic compound in a zinc matrix. Some of the more important features which must be considered when using this approach are: a) Solid solution effects - The effect of solid solution on recovery processes is not clearly established; however i t appears as though l i t t l e retardation of dynamic recovery processes occurs. b) Thermal s t a b i l i t y - The intermetallic compound must be stable at normal useful temperatures. c) Intermetallic composition - It is of distinct commercial advantage for the intermetallic to be of a high zinc content so that a large volume fraction of intermetallic is formed with only dilute alloy additions. d) Mechanical properties of the intermetallic compound - The intermetallic should remain "hard" relative to the matrix at ordinary temperatures; this is thought to be a prerequisite for grain boundary stabilization. The actual intermetallic dispersion w i l l be a sensitive function of cooling rate and subsequent mechanical working. Rapid cooling rates and large amounts of working promote a finer dispersion. Fragmentation of an alloy melt by means of atomization results in a very rapid so l i d i f i c a t i o n rate and usually a fine intermetallic dispersion and grain size. A system which fulfil's.-; these requirements and has received (8 9 27 28 29) some attention is the zinc-titanium system. ' ' ' ' The limited  amount of investigation which has been carried out using dilute zinc- titanium alloys suggests that high strength and excellent creep properties can be obtained. The work to date suggests that the major requirement is the development of a fine-grained alloy, the potential strength of which may be realized by the selective location of a fine second phase so that the grain boundaries are stabilized. The necessity that the second phase be preferentially located at grain boundaries is a unique requirement in comparison with conventional dispersion-strengthened alloys. The purpose of this work is to attempt to define the micro- structural c r i t e r i a for high strength and good creep resistance i n zinc alloys. The approach involved the correlation of microstructures and mechanical properties in terms of the mechanisms responsible for deformation and recovery. A wide range of sol i d i f i c a t i o n and extrusion conditions were used to produce differing combinations of grain size, intermetallic size, and intermetallic distribution with titanium concentrations of up to 0.6 wt.%. 7 2. PROCEDURE 2.1 Starting Materials A l l alloys used in this study were made by dilution of a zinc - 3 wt.% titanium master alloy supplied by Cominco Limited, Sheridan Park, Ontario. Adjustment to the desired alloy composition was carried out using SHG (Special High Grade) zinc. Since one of the purposes of this work was to obtain technical information which could be used in practical applications, this grade of zinc was chosen because of i t s wide industrial use. As is shown in Table I, the impurity levels in the extruded materials are low. It i s thought that the effect of these impurities on the mechanical properties of the alloys would be negligible in comparison with the effect of the titanium present. 2.2 Melting Procedures Prior to melting, both the master alloy and the zinc additives were precleaned in a dilute hydrochloric acid solution. The alloys were melted in ten pound batches. The pure zinc component was melted in a silic o n carbide crucible heated i n a gas-fired furnace. To prevent excessive oxidation of either the zinc or the titanium, a cover consisting of equal weights of potassium chloride and sodium chloride was used on the melt. Additions of the master alloy were made through the molten cover. This procedure proved to be reasonably effective and allowed close control of the f i n a l alloy composition. The nominal and f i n a l titanium concentrations of the alloys used in this study were: 8 Spec. Analysis (wt. %) Material Extrusion Temperature Ti Cu Ag Pb Fe (°C) -35 + 100 mesh powder 350 0.32 0.1* 0.1* 0.001 0.01 Chill Cast 350 0.32 0.1* 0.1* 0.001 0.001 Mg, Be, As, Cr, Ni, Si, Al, Mo, Sn, Sr, Ca, Co, Bi not detected* ̂ * X-ray Fluorescence Analysis indicates these values are high. The true values are less than 0.01 and probably less than 0.001 wt. %. TABLE I Composition of Extruded Materials^, 9 Nominal wt.% T i Analyzed wt.% T i 0.10 0.07 0.20 0.16 0.30 0.32 0.60 0.60 The melt was held at 700°C for a minimum of thirty minutes during which time i t was agitated at five minute intervals. After allowing the melt to cool to approximately 600°C and thus freeze the salt cover to prevent contamination during pouring, three c h i l l cast b i l l e t s were obtained by casting into graphite molds. The remainder of the melt was immediately transferred to a preheated atomizing crucible. Atomization was initiated when the melt stabilized at 600°C. 2.3 Atomizing The purpose of fragmenting the melt by this procedure was to provide a very rapid cooling rate so as to produce a fine grain size and intermetallic dispersion. problem was that l i t t l e control over the product size distribution could be maintained. The f i n a l design of the apparatus (Fig. 2) provided the necessary control and allowed for the production of an extremely wide range of particle sizes, ranging from essentially a l l shotted (+35 mesh) material to approximately 50% -325 mesh . concentric gas blast around the alloy surface. Argon was used for atom- ization so that oxidation of the product could be minimized. Several atomizer designs were tested; however the common Basically the nozzle fragments the melt by means of a 10 a plug b graphite crucible c argon inlet d metal carrier e gas carrier Fig. 2 Schematic representation of the atomizing apparatus. The product size was controlled by adjustment of one or more of the following: a) argon pressure b) size of the argon outlet c) size of the molten metal o r i f i c e d) location of the molten metal o r i f i c e with respect to the argon outlet. 2.4 Construction of the Atomizer The apparatus consists of a two piece stainless steel nozzle press f i t t e d into a graphite crucible. The crucible and nozzle were heated and controlled independently. The nozzle i t s e l f had two prime functions: one section served as a gas carrier and the other as a molten metal carrier. The inner core of metal-carrying section was lined with a graphite sleeve which was replaced after each atomization. The atomized melt was impinged onto a cold water bath to f a c i l i t a t e collection, and to ensure complete so l i d i f i c a t i o n at a maximum rate. The product was fil t e r e d and then dried for approximately 20 hours at 80°C prior to grading as to size. 2.5 Compaction Prior to extrusion, a l l powder was compacted into b i l l e t s suitable for extrusion. Hydrostatic (isostatic) compaction was used since a sound uncontaminated b i l l e t easily accommodated by the extrusion apparatus could be produced at the relatively low pressure of 30,000 p.s. 12 2.6 Extrusion Fully dense rod of varying diameters was produced by an indirect extrusion technique using the apparatus shown in Fig. 3. Unless otherwise specified, the data in this work pertains to an extrusion ratio of 25:1 and corresponds to an extrusion product 0.150 inches in diameter. The extrusion temperature varied between 150°C and 350°C. Because of the high extrusion pressures which were required (up to 330,000 p.s.i.), i t was considered impractical to place a thermo- couple close to the b i l l e t cavity as any suitable well would structur- ally weaken the apparatus and thus put an unnecessary limitation on the operating pressure. It was also desirable to minimize the soaking time of the b i l l e t in order to prevent excessive coarsening of the grain size and intermetallic compound. Therefore a series of preliminary experiments were undertaken using a dummy b i l l e t containing a thermocouple. The procedure involved plotting the b i l l e t temperature after insertion into the preheated extrusion cavity. A series of standard curves for the desired extrusion temperature were plotted in order to establish the minimum soaking time required for the b i l l e t to reach the required temperature. An example of such a curve is shown as Fig. 4. It can be seen from this graph that for extrusion at 250°C, the b i l l e t should be soaked for at least 7 minutes, and further that the extrusion should be completed within 15 minutes. The purpose of this procedure was to standardize the fabrication conditions. Complex temperature gradients undoubtedly exist in the extrusion apparatus and b i l l e t ; thus accurate extrusion temperatures per se are d i f f i c u l t to obtain. Quoted extrusion pressures are of limited quantitative meaning as these pressures depend to a great extent on the condition of 13 Fig. 3 Schematic representation of the extrusion apparatus. Fig. 4 Standardized extrusion time - temperature conditions. the die and extrusion cavity. However the trend was for the pressures to be roughly proportional to the tensile strength of the product, with the highest pressures required when extruding the highest alloy powders at the lowest temperatures. Another consequence of the high extrusion pressures was that the extrusion rates were necessarily kept low in order to keep from exceeding the maximum loading capabilities of the extrusion apparatus. Attempts were made to lower the required extrusion pressures with the use of lubricants. Liquid lubricants tended to impregnate the powder b i l l e t s , leading to an unsound product. Cladding the b i l l e t with lead was helpful to a limited extent. However at high pressures (> 200,000 p.s.i.), the lead tended to pre-extrude the alloy, leading to lead depletion and/or pinching-off of the zinc extrusion. It was therefore decided to carry out a l l subsequent extrusions without lubrication. Four inches of either end of the extrusion were not used for mechanical testing as i t was thought that these portions of the extrusion might not be structurally representative of the product. 2.7 Tensile Specimen Preparation A l l tensile specimens were machined from extruded rod on a jeweller's lathe to a reduced diameter of approximately 0.075 inches and a gauge length of approximately 0.750 inches. As the grain size and intermetallic compound dispersion were stable at room temperature in both the heat-treated and as-extruded condition, i t was not necessary to employ cold machining techniques or to electropolish the surface prior to testing. 2.8 Metallography Electropolishing was found to be the most successful and convenient method of preparation for metallographic examination. Electropolishing was preceded by mechanically polishing the specimen on kerosene-wetted emery paper down to 4/0 grit followed by lapping with 5 micron diamond paste. Several polishing solutions were used, their success varying somewhat with the nature of the microstructure. The most d i f f i c u l t structures to prepare were those having the finest grain size and second phase dispersion. This was thought to be the result of break up of the thin polishing layer adjacent to the specimen surface by the intermetallic. The most successful electrolyte was found to be: 25 gm. chromium trioxide 133 c c . glacial acetic acid 7 c c . d i s t i l l e d water For coarser structures the following solution was found to be effective and reliable: 800 ml. ethyl alcohol 50 ml. butylcellusolve 60 gm. sodium thiocyanate 20 ml. d i s t i l l e d water For optical metallography using polarized light, no further preparation was necessary. However, because of the fine microstructures involved, optical microscopy was not particularly useful. Most metallography was accomplished using electron microscopy. Transmission microscopy was used only to verify the fact that replicas were representative of the structure. Replica techniques have a distinct advantage i n that definition of the second phase often is much better, and the technique is rather simple and reliable as 17 compared with transmission procedures. In a l l instances a two stage carbon-chromium replica technique was employed. Figures 5 and 6 show electron micrographs of similar microstructures using both techniques in order to i l l u s t r a t e the validity of the replication technique. Electropolishing did not define the grain boundaries sufficiently to allow a grain size determination. To define the grain boundaries more clearly, a l l metallographic specimens were deformed slightly by bending prior to replication. The amount of deformation required was dependent on the nature of the microstructure. Microstructural parameters were determined using a line intercept method. In cases where the structure showed a directional tendency such as intermetallic particle stringering, the grain size was determined by a line intercept at 45° to the extrusion direction. 2.9 Testing Procedures A Floor Model Instron was used for a l l tensile tests. Both constant and variable crosshead speed tests were used, u t i l i z i n g -4 speedsfrom 2 x 10 in/min to 2 in/min. During variable crosshead speed tests (used to evaluate strain-rate sensitivity), constant speed was maintained un t i l the specimen had deformed a minimum of 1% at the steady state stress before a strain-rate change was made. Tensile tests were performed in the following media: -196°C liquid nitrogen -140°C to -100°C liquid nitrogen cooled petroleum ether -100°C to +20°C liquid nitrogen cooled ethyl alcohol +20°C to +100°C heated water +100°C to +250°C heated cooking o i l +250°C and above molten salt F j-g- 5 Z n - 0.6 wt.% T i a l l o y extruded from -35 + 100 mesh powder at 150°C Transmission technique. (x25000) 19 F i § - 6 Z n ~ °- 6 w t - % T i a l l o y extruded from -35 + 100 mesh powder at 150°C. R e p l i c a t i o n technique. (x25000) True stress, strain, and strain-rate were calculated on the assumption that deformation had occurred homogeneously throughout the gauge length. True stress - strain curves presented in this work are valid up to the maximum stress. The apparent decrease in true stress at higher strain values is due to necking of the specimen. 3. ZINC - TITANIUM PHASE DIAGRAM Several Independent investigations have been made on the zinc rich portion of the zinc-titanium phase diagram. There is consid- erable disagreement as to the composition of the eutectic and the number and nature of peritectic reactions. The eutectic composition appears to l i e in the range 0.10 wt.% Ti to 0.50 wt.% Ti. Table II summarizes the information regarding the eutectic which has been reported to date. A significant feature of this data is that the eutectic evaluations other than those of Gebhardt, Cominco, and to a certain degree Rennhack, have a l l been based on metallographic examination. It is probable that large variations of the eutectic morphology and composition can occur depending upon the casting history. Thus i t would appear that determination of phase relations by means of analysis of microstructures alone may be somewhat unreliable. It is of interest to note that determinations made by means of differential cooling techniques yield results which are in close agreement. Figures 7 and 8 show the phase diagrams as determined by these investigations. Some effort has been made to establish the crystal structure (9) of the two zinc rich intermediate phases which are produced. It is generally agreed that Zn^Ti is the product of peritectic reaction of the remaining liquid with ZnigTi. The zinc rich compound (Zn^Ti) has an orthorhombic structure with the following composition and lattice parameters: o 4.5 wt.% Ti a = 3.87 A b = 5.69 A c = 11.87 A 22 Table II. Summary of the Experimental Investigations of the Zinc Rich End of the Zinc-Titanium Phase Diagram. Reference Year Eutectic Composition (wt.% Ti) Eutectic Temperature (°C) Solid Solubility (wt.% Ti) Anderson 1944 0.12 418.5 <.015 (300°C) P e l z e l < 3 4 ) 1961 0.18 (32) Heine and Zwicker 1962 0.23 <.0004 (400°C) ( 3 3 ) ~ J Rennhack 1966 0.19 418.6 (31) Gebhardtv J 1941 0.45 418 <.02 (400°C) „ • (34) Comxnco 1964 0.46 418 Fig. 7 Zinc - WT % Ti titanium phase diagrams. (after Heine and Zwicker)  4. RESULTS Since this work is based on both powder and cast material and since the procedures used in their fabrication have an inherently- large number of possible variables, i n i t i a l investigations were limited to a fixed composition in order that the effect of some of these variable could be established and optimized. The composition chosen was Zn - 0.6 wt.% T i . 4.1 Reproducibility Quality control tensile tests were performed on the i n i t i a l extrusions of both cast and powder materials i n order to establish i f a gradient in properties existed between the i n i t i a l and f i n a l portions of the extrusion. The results of these tests are shown in Table III; they indicate that the mechanical properties do not appear to change through- out the length of the extrusion. Therefore for a l l subsequent extrusions i t was assumed that the product was structurally consistent throughout. However as an added precaution, a l l tensile specimens were labelled according to the section of the extrusion from which they were obtained in the event that some structural or mechanical discontinuity was encountered. 4.2 Effect of Extrusion Temperature As -100 + 200 mesh powder was a medium size fraction obtained by the atomization process, this size was used to determine the effect of extrusion temperature on the mechanical properties of the material. Material Tx (°C) Specimen Location 0.2% O.Y.S. (103 psi) 0.5% O.Y.S. (10 3 psi) U.T.S. (10 3 psi) Total Elongation cast 150 cast 150 cast 200 cast 200 cast 250 cast 250 cast 300 cast 300 cast 350 cast 350 -100 + 200 mesh 150 -100 + 200 mesh 150 -100 + 200 mesh 250 -100 + 200 mesh 250 I F I F I F I F I F I F I F 24.4 24.2 25.1 25.4 30.3 30.2 29.7 29.9 24.6 24.5 62. 60. 28. 29. 26.0 25.8 26.6 26.8 32.2 31.7 31.2 31.7 26.0 27.2 63.7 63.4 31.1 31.5 27.5 26.7 27.7 27.6 33.0 32.3 31.9 32.5 28.8 29.7 63.7 63.5 32.1 32.6 24.8 22.6 14.8 19.7 14.4 12.4 12.8 14.4 14.9 18.0 6.8 6.0 16.3 14.6 8- ro H H M ro xi i-f o a. c o H* M H" rt O l-h s ro o 3* W , 3 H-o 03 -100 + 200 mesh 350 I -100 + 200 mesh 350 F 28. 30. 30.3 32.4 31.4 33.5 18.5 11.0 XI i-i O xt ro if rt- H- ro 05 o 3 09 M rt n c CO H- O 3 CO Tx = Extrusion Temperature I = I n i t i a l specimen from extrusion F = Final specimen from extrusion O.Y.S. = Offset Yield Stress Note: The composition of a l l extrusions is 0.6 wt.% T i . to A simultaneous investigation was carried out using cast material. The results of this study are shown in Figs. 9 and 10. There is no significant difference in the strength of specimens manufactured using powder and those extruded from the as-cast b i l l e t s when an extrusion temperature greater than 250°C is used. When the starting material is a cast b i l l e t there is only a slight change in yield strength as the extrusion temperature is lowered. A maximum in strength occurs at an extrusion temperature of 250°C. This can be attributed to a grain size effect and i t s relationship to the distribution of the second phase. The significance of this relationship is discussed in detail later. For extruded powder material, a rapid increase i n strength is obtained as the extrusion temperature is lowered below 250°C. The minimum extrusion temperature used was 150°C, and was determined by equipment limitations. Cast and powder products possess equivalent d u c t i l i t i e s when extruded above 250°C. Below this temperature, the ductility of the powder material decreases whereas the ductility of the cast material increases. Tensile properties of Zn - 0.6 wt.% T i alloys extruded at 175°C using three different powder sizes are shown in Table IV. These results indicate that no significant increase in strengthening is realized through the use of finer powders. The most significant feature of Fig. 9 is the strength differential between cast and powder extrusions at low fabrication temperatures. This must arise due to microstructural differences in the original material. The decrease in strength of powder extrusions which EXTRUSION T E M P E R A T U R E ( ° C ) Fig. 9 The effect of extrusion temperature on the yield strength of Zn - 0.6 wt.% Ti alloys. — s oo EXTRUSION TEMPERATURE ( ° C) Fig. 10 The effect of extrusion temperature on the ductility of Zn - 0.6 wt.% T i alloys. Table IV. The E f f e c t of Powder Size on the Mechanical Properties of Zn - 0.6 wt.% T i A l l o y s . Powder Size 0.2% O.Y.S. 0.5% O.Y.S. U.T.S. Total Elong- ,.J .x / i n 3 , .v / 1 r t3 .x ation (10 p s i ) (10 p s i ) (10 p s i ) -35 + 100 mesh 60.1 64.0 65.0 2 -100 + 200 mesh 60.0 63.7 . 64.0 4 -325 mesh 60.7 67.0 75.0 2 Note: e ̂  .0028 min. Tx = 175°C Extrusion Ratio = 25:1 31 occurs at higher extrusion temperatures must also be associated with microstructural modifications. 4.3 C h i l l Cast and Powder Microstructures 4.3.1 Microstructure of Cast B i l l e t s Most of the previous investigations of cast zinc- titanium alloys have been carried out at low titanium concentrations (26 29 34) (<0.2 wt.%). ' ' This corresponds to a composition which i s hypoeutectic according to most of the published phase diagrams. The microstructure of a c h i l l cast hypereutectic alloy is expected to consist of primary particles of intermetallic (Zn^Ti) surrounded by a Zn - ZnisTi eutectic. Similarly, in a hypoeutectic alloy primary zinc dendrites and eutectic are expected to make up the microstructure. Figures 11(a) and 11(b) indicate the microstructure of a c h i l l cast b i l l e t containing 0.36 wt.% T i . This titanium concentration is considerably higher than most of the published eutectic compositions (M3.2 wt.% T i ) . The structure consists of a zinc dendritic structure with large areas of eutectic. Electron microprobe analysis revealed that the primary dendrites were titanium depleted and that the titanium was essentially present only in the eutectic structures. Figures 12(a) and 12(b) show the absorbed electron image of a primary particle along with a correspond- ing T i X-ray scan. Quantitative measurements obtained by scanning both the primary particles and eutectic indicate no detectable titanium in the primary zinc dendrites. Fig* 11 Zn - 0.36 wt.% T i a l l o y c h i l l cast microstructure. 33 F i g . 12(b) T i X-ray scan of Zn - 0.36 wt% T i c h i l l cast microstructure. (xl500) The average titanium concentration in the eutectic structure was calculated from uncorrected microprobe data to be 0.45 wt.% T i . Applying absorption and fluorescence corrections as developed by B i r k ^ ^ increased this value to 0.49 wt.% T i . Considering the sensitivity of the microprobe analysis, this would suggest that the eutectic composition is in the range 0.4 - 0.6 wt.% T i . Clearly the s o l i d i f i c a t i o n rates used in this study are non-equilibrium. Therefore this value should not be related to values obtained from phase diagrams. Figures 13(a) and 13(b) indicate the morphology of the eutectic observed in a c h i l l cast Zn - 0.36 wt.% T i alloy. The microstructure of a c h i l l cast Zn - 0.6 wt.% T i alloy (Fig. 14) indicates large areas of primary zinc as well as primary particles of ZnisTi. This observation is not consistent with the expected structure. It is possible that with supercooling, the intermetallic particles may act as preferential nucleating centres for zinc. This would lead to the premature so l i d i f i c a t i o n of the zinc rich phase and could explain the observed microstructure. Basically these micro- structures indicate that at these s o l i d i f i c a t i o n rates, the apparent eutectic composition is between 0.36 wt.% T i and 0.6 wt.% T i . 4.3.2 Microstructure of Powder In order to investigate the structure in the powder particles, metallographic and microprobe studies were made using powder b i l l e t s compacted at pressures sufficiently high to provide a completely dense specimen. These analyses were carried out using -35+100 mesh powder containing 0.6 wt.% T i . Figures 15(a) and 15(b) are electron micrographs of the Fig. 13 Mlcrostructural characteristics of the eutectic in a c h i l l cast Zn - 0.36 wt.% T i alloy. F i g . 14 C h i l l cast microstructure of Zn - 0.6 wt% T i a l l o y . (x625) 3 7 eutectic structure which was present. The morphology of the eutectic was highly variable, a feature also observed in c h i l l cast structures. Although not shown i n these electron micrographs, primary intermetallic particles were also present. What appears to be an oxide surface coating on the pwder particles remained relatively intact. This suggests that compaction of the powder did not modify the as-solidified microstructure to any significant degree. Microprobe analysis confirmed that the continuous phase was titanium rich and that the non-continuous phase was titanium depleted (Figs. 16(a) and 16(b)). The eutectic has so l i d i f i e d dend- r i t i c a l l y and i s much finer than that observed in cast b i l l e t s . 4.3.3 Extent of Oxidation of Atomized Powder The study of the microstructure of the powder also provided the opportunity to evaluate the extent of surface oxidation which occurred during atomization. There is an apparent surface coating on the powder as is shown i n Fig. 15(b). However this could not be confirmed to be oxide as i t was too thin to be resolved in the microprobe No oxide was found in either powder compacts or extruded product. This suggests that the powder sol i d i f i e d within the argon atmosphere used during atomization. This further indicates that the titanium i s present as intermetallic and not as an oxide. In fact the microprobe analysis suggests that the surface.of the particles may be oxygen depleted (Figs. 17(a), (b), and (c)). The zinc matrix oxidizes to a limited extent during preparation for metallographic or microprobe analysis. The intermetallic compound is not expected to oxidize. Thus areas rich in intermetallic would tend to exhibit an apparent lower oxygen concentration. Fig. 16(b) Ti X-ray scan from an electron microprobe analysis of -35 + 100 mesh Zn - 0.6 wt.% Ti powder. (b) (x!500) Fig. 17 Electron microprobe analysis of Zn - 0.6 wt.% T i -35 + 100 mesh powder compact: (a) Absorbed electron image (b) T i X-ray scan (c) 0 X-ray scan . 41 (c) (xl500) 42 4.4 Microstruetural Characteristics of Extruded Alloys A basic difference between the microstructures of cast b i l l e t s and powder i s the decrease in dendrite and lamellar spacing which occurs with more rapid s o l i d i f i c a t i o n rates. The nature of the micro- structure formed during extrusion is dependent on the characteristics of the starting material and on the fabrication conditions. Figures 18, 19 and 20 show the effect of extrusion temperature on the microstructure of extruded powders. At low extrusion temperatures alignment of the intermetallic occurs to give the microstructure a "stringered" appearance in which the grain size is equivalent to the stringer spacing. As the fabrication temperature is increased, inter- metallic coarsening and grain growth occur resulting in a loss of this type of structure. There is a corresponding decrease in strength as stringering becomes more poorly defined. During this process the grain size and interparticle spacing remain approximately equivalent. 4.4.1 Stringering During Extrusion of Powders Since the microstructure of material manufactured u t i l i z i n g low extrusion temperatures and a high extrusion ratio consists of a well- defined stringered structure, an extrusion was carried out at an extrusion ratio sufficiently low so as to reduce the amount of plastic flow of the matrix to the point at which not a l l powder particles would deform. An extrusion ratio of 10:1 was found to be sufficiently low such that some particles were in a "hard" orientation, or in areas of the extrusion in which flow was minimized. As a result, the development of the stringered structure could be followed. 43 F i-g- 19 M i c r o s t r u c t u r e of a Zn - 0.6 wt.% T i a l l o y extruded at 250°C using -35 + 100 mesh powder. (x4000) Fig. 20 Microstructure of a Zn - 0.6 wt.% T i al l o y extruded at 350°C using -35 + 100 mesh powder. (x4000) 45 As can be seen from Figs. 21 and 22, the stringered structure is a direct result of the structure which i s produced during so l i d i f i c a t i o n of the powder. The continuous network of intermetallic surrounding the zinc phase is directed in the extrusion direction. Any intermetallic which i s perpendicular to the extrusion direction i s displaced by the plastic flow of the zinc down the "pipes" of intermetallic to give an overall stringered appearance. The contribution of the oxide to the overall stringering can also be seen in these electron micrographs. It appears reasonable to assume that the amount present has a negligible effect, since only a small percentage of the stringering i s due to oxide and the distance between oxide stringers i s generally several grain diameters (partially the result of the use of coarse powders). As a result of the short diffusion paths involved, high temperature soaking prior to extrusion results i n a considerable degree of intermetallic coarsening. Because of the ease of flow of the matrix at these temperatures, effective second phase alignment i s also minimized. Primary particles formed during s o l i d i f i c a t i o n appear as coarse stringers in the extruded product (Figs.23(a),(b) and (c)). High titanium concentrations in such areas may give the impression that other areas of the alloy have intermetallic concentrations somewhat lower than is consistent with the analysis. 4.4.2 Cast Material The coarse microstructures of c h i l l cast b i l l e t s result in rather coarse extruded microstructures, as i s shown in Figs. 24, 25 and 26. However lower extrusion temperatures promote fragmentation and displacement 46 Fig. 21 The origin of the stringered structure in a Zn - fabricated from -35 + 100 mesh powder at 175°C. 0.6 wt.% Ti alloy (x4000) 47 8 ml \ t £ ' « \ V ip -,'7^.''"/,: >̂ > rV F i§- 2 2 T h e origin of the stringered structure in a Zn - 0.6 wt.% Ti alloy fabricated from -35 + 100 powder at 175°C. (x4000) 48 F i g . 23 Electron microprobe analysis of Zn - 0.6 wt.% T i a l l o y extruded at 250°C using -35 + 100 mesh powder (specimens annealed 44 hours at 350°C): (a) Absorbed electron image (b) T i X-ray scan (c) Zn X-ray scan. (d) O2 X-ray scan.  Fig- 25 Microstructure of a chill cast Zn - extruded at 250°C. (x4000) 0.6 wt.% Ti alloy Fig- 26 Microstructure of a chill cast Zn - 0.6 wt.% Ti alloy extruded at 350°C. (x4000) 52 of the intermetallic particles to give the structure a more directional appearance. 4.5 The Effect of Extrusion Ratio Qualitatively, a reduction i n extrusion ratio should decrease the degree of deformation during fabrication and ultimately should result in a coarser, softer microstructure. An indication of the effect of the extrusion ratio is shown in Fig. 27. The data included in this plot were obtained under conditions i n which microstructural coarsening due to thermal effects is minimal. Although the data do not allow a definitive statement to be made regarding an optimum extrusion ratio, they indicate qualitatively that the mechanical properties are indeed a function of the extrusion ratio. A l l subsequent extrusions were carried out with a 25:1 ratio. 4.6 Importance of Preferred Orientation Investigations of extruded and rolled cast zinc-titanium (28 35) alloys have shown the tendency for a preferred orientation to develop. ' These studies were confined to titanium concentrations less that 0.2 wt.%. Cominco carried out an extensive study of textures developed during (34) extrusion of cast Zn - 0.16 wt.% T i alloys. Two distinct textures were observed, one being characteristic of the surface and another charac- t e r i s t i c of the bulk material. It was concluded that the hardness did not depend significantly on the development of a texture. Cast and rolled Zn - 0.16 wt.% T i alloys also tend to develop ( 28) a well-defined texture. Apparently the formation of dendrites permits an extension of the growth process during which favourably oriented grains (36) absorb those less favourably oriented. Continued ro l l i n g of cast 0.2% OFFSET YIELD STRESS ( I O 3 PSI) 4̂  O O O structures tended to produce an increasingly more random structure with only a slight texture being encountered at 50% deformation. The breakdown of the cast texture was promoted by increasing the titanium concentration. Data suitable for pole figure plots were obtained using a highly stringered material formed by low temperature extrusion of - 3 5 + 1 0 0 mesh powder containing 0 . 6 wt.% T i . No preferred orientation was detected in either the longitudinal or transverse directions. This result suggests that the development of high strength can not be associated with preferred orientation. The lack of texture in extruded rod has been observed in cases in which a reasonably high concentration of undeformable second phase exists in a zinc matrix. It is thought that the second phase tends to cause "turbulent" plastic flow of the matrix, resulting in ( 4 6 ) randomness. Thus previous work has shown that the existence of a texture ha's relatively l i t t l e effect on the creep properties or hardness of zinc-titanium alloys. The properties of the extrusion resulting from the directional nature of the dispersoid would overshadow effects due to the development of a texture. On this basis i t was thought that further efforts in this direction would provide l i t t l e knowledge as to the understanding of the mechanical properties of these alloys. 4 . 7 Thermal Stability 4 . 7 . 1 Introduction As has been indicated earlier, the major factor limiting the use of zinc or zinc alloys in commercial applications is the poor creep behaviour resulting from the operation of recovery mechanisms and grain boundary shear at room temperature (T^ = 0.42). The development of a commercially useful zinc alloy depends upon the development of a microstructure which inherently retards the operation of these processes and thus preserves reasonable creep properties up to at least 75°C. Therefore i f this type of microstructure is contingent on a particular dispersion of a second phase such as is present i n extruded zinc-titanium alloys, then creep and tensile properties at elevated temperatures w i l l be determined by the thermal st a b i l i t y of the intermetallic compound Zn 1 5 T i . 4.7.2 Effect of Extrusion Temperature Consideration of the effect of extrusion temperature shown in Figs. 9 and 10 suggests that coarsening of the intermetallic occurs at temperatures above 200°C. This results in essentially equivalent mechanical properties for extruded castings and extruded powders. There- fore to develop the potential strength obtainable through the use of powder metallurgical techniques, fabrication should be carried out at temperatures less than 200°C. In a l l instances, high temperature extrusion of powders results in the production of an equiaxed grain structure and a spheroidized second phase. Quoted extrusion temperatures are somewhat misleading in as much as the actual temperature at the die - b i l l e t interface can be expected to be considerably higher. For this reason i t is d i f f i c u l t to define the thermal s t a b i l i t y of a structure in terms of fabrication temperature. From these results i t i s apparent that a fine second phase dispersion in the fabricated product is contingent on retaining the fine intermetallic structure formed during atomization. It has been seen that low temperature fabrication is not an effective method of fragmenting and dispersing larger primary intermetallic particles to the extent that substantial enhancement of the mechanical properties i s realized. This has already been demonstrated in Figs. 24, 25 and 26, which indicate the degree to which fracture and dispersion of large second phase particles occurs as the extrusion temperature is lowered. From the previous discussion, i t is apparent that prior thermal treatment of the powder should have an intermetallic agglomeration effect similar to that obtained by extruding at elevated temperatures. This was substantiated by annealing a b i l l e t of -35 + 100 mesh 0.6 wt.% T i powder at 300°C for 1/2 hour prior to extrusion at 175°C. The mechanical properties which result are substantially the same as those of cast and extruded material or powders fabricated at higher temperatures. As can be seen from Fig. 28, no evidence of intermetallic fragmentation was observed after low temperature extrusion, further demonstrating that the second phase remains hard relative to the matrix. Thus thermal treatment prior to or during deformation must be avoided in order to ensure development of the potential strength of the alloys. The role of extrusion temperature and i t s effect on the dispesion of spheroidized or thermally treated powder b i l l e t s was >\ investigated by varying the extrusion temperature of -100 + 200 b i l l e t s containing 0.6 wt.% T i which had undergone a one hour anneal at 350°C. Thermal treatment under these conditions results in a reasonably coarse, spheroidized intermetallic structure. Figure 29 indicates the mechanical properties of the alloy produced under these conditions. Higher extrusion temperatures tend to coarsen the inter- metallic slightly. This is thought to be a function of the availability of thermal energy for diffusion during the pre-extrusion soak. As is 5 7 F ig- 2 8. T h e structure and mechanical properties of a Zn - 0.6 wt.% Ti alloy extruded at 175°C using -35 + 100 mesh powder. (The billet was annealed for 1/2 hour at 300°C prior to extrusion.) STRAIN RATE (MIN"') Fig. 29 The effect of extrusion temperature on the yield stress of a thermally stabilized Zn - 0.6 wt.% Ti alloy. oo to be expected, the coarsening of the second phase also results in a slight increase in grain size. The overall result is a slight loss in strength. However i t should not be assumed that the observed strength- ening effect at the lower fabrication temperatures is the result of a change in grain size alone. As has been discussed br i e f l y , grain boundary effects are important at 20°C. Therefore the role and properties of a dispersed second phase and i t s relation to grain boundaries are important factors which must be considered in a discussion of mechanical properties at this temperature. Figs. 30(a) and (b) show the structures which typically result from changes in extrusion temperature. From these micrographs i t can be seen that only a slight amount of coarsening of the intermetallic and grain size has resulted from extrusion at 350°C. However some alignment of the second phase in the extrusion direction has occurred at 250°C, giving rise to microstructures which are somewhat different in appearance. The degree of directionality of the microstructure is related to the relative ease of flow of the matrix at the temperature under consideration. The development of such a microstructure i s significant when considering the effect of the dispersion of the operation of grain boundary shear as an important deformation mechanism. 4.7.3 Effect of Annealing Temperature 4.7.3.1 Powder Material The use of powder metallurgical techniques to obtain enhanced mechanical properties is especially attractive from a commercial point of view i f the desired properties can be developed through the use of relatively F i g . 30(a) The microstructure r e s u l t i n g from extrusion of a -100 + 200 mesh powder b i l l e t annealed for 1 hour p r i o r to extrusion at 250°C. (x4000) F i g . 30(b) The m i c r o s t r u c t u r e r e s u l t i n g from e x t r u s i o n of a -100 + 200 mesh powder b i l l e t annealed f o r 1 hour p r i o r to e x t r u s i o n at 350°C. (x4000) coarse powders. This is due to the high cost of production of fine powders in large volumes. Thus the high strength levels achieved using -35 + 100 mesh zinc-titanium powder may be of some commercial interest, providing thermal s t a b i l i t y can be achieved. A simple and meaningful method of establishing the s t a b i l i t y of as-extruded microstructures is a simple annealing study. The results of such a study on extruded -35 + 100 mesh powder are shown in Figs. 31 and 32. It can be seen that for relatively short times, the inter- metallic appears to remain stable up to approximately 160°C. At annealing temperatures above 200°C, the strength is reduced to the same level as that obtained by extruding powders at high temperatures or by extrusion of cast b i l l e t s . A narrow range exists (175° - 200°C) in which the microstructure is very temperature sensitive. Thus the i n i t i a l structure which consists of well developed intermetallic stringers and an associated columnar-grained zinc matrix is stable to reasonably high temperatures. At slightly more elevated temperatures, some agglomeration of the intermetallic occurs. This coarsening of the second phase does not appear to take place generally throughout the material but rather occurs preferentially in isolated colonies. Changes in grain shape and some minor grain growth accompany this coarsening. The result is that a transition structure is formed consisting of colonies of equiaxed grains and spheroidized intermetallic. An example of this type of structure is shown in Fig. 33. Above 200°C, the stringered structure has been completely lost, with extensive modifications of the grain shape and second phase distribution. Higher annealing temperatures result in a slight increase in the amount of second phase coarsening and an increase in grain size. A N N E A L I N G T E M P E R A T U R E ( ° C ) Fig. 31 The effect of annealing temperature on the yield strength of Zn - 0.6 wt.% Ti alloy A N N E A L I N G T E M P E R A T U R E ( ° C ) Fig. 32 The effect of annealing temperature on the ductility of Zn - 0.6 wt.% Ti alloy at 20°C. ON 00 64 F i g . 33 The m i c r o s t r u c t u r e r e s u l t i n g from a short term anneal at 180°C of a Zn - 0.6 wt.% T i a l l o y f a b r i c a t e d at 175°C using -35 + 100 mesh powder. (x4000) 65 The o v e r a l l e f f e c t on the mechanical properties i s small, and the microstructure can be considered to be "thermally stable". Long term - high temperature anneals simply allow an extension of the coarsening process with a slow decrease i n strength accompanying increased agglomeration. The structure and properties which r e s u l t a f t e r such thermal treatments are shown i n Fi g s . 34 and 35. 4.7.3.2 True S t r e s s - S t r a i n Curves Figure 36 shows true s t r e s s - s t r a i n curves f o r material annealed at various temperatures. I t can be seen that the work hardening c h a r a c t e r i s t i c s change from parabolic hardening f o r as-extruded material and extrusions annealed at low and high temperatures to,a high degree of l i n e a r hardening i f the material i s thermally treated at intermediate temperatures. Fibre strengthened materials often show analogous behaviour. However the microstructures involved here do not lend them- selves to in t e r p r e t a i o n i n terms of f i b r e composite theories since no strength can be at t r i b u t e d to the discontinuous i n t e r m e t a l l i c s t r i n g e r s . Table V tabulates the t e n s i l e properties and work hardening rates obtained from these curves (Fig. 36) and other curves which have not been presented. From these r e s u l t s i t i s apparent that though the nature of the hardening i s modified as the annealing temperature i s increased, the work hardening rate at 0.5% s t r a i n i s constant up to annealing temperatures of approximately 200°C. At higher temperatures, the hardening rate decreases. I t i s also i n t e r e s t i n g to note that the amount of work hardening i s maximized when l i n e a r hardening occurs. From analysis of the true s t r e s s - s t r a i n curves, i t appears that material thermally treated at intermediate temperatures (160°C and Fig. 34 The microstructure of a Zn - 0.6 wt.% Ti alloy fabricated at 175°C from -35 + 100 mesh powder and annealed for 44 hours at 350°C. (x4000) O.Y.S. (t M).0028 min. ) - 22,500 p . s . i . Total elongation = 12% Fig. 35 The microstructure of a Zn - o.6 wt.% Ti alloy fabricated at 175°C from -35 + 100 mesh powder and annealed for 94 hours at 400°C. (x4000) O.Y.S. (e ̂ 0.0028 min."1) = 17,800 Total elongation = 14% T 1 1 1 1 1 r 80 h E ^ 0.0028 min. 3 or J I I I I I L 1 2 3 4 5 6 7 S T R A I N (%) F i § - 36 The effect of annealing temperature on the true stress - strain relationship at 20°C for Zn - o.6 wt.% Ti alloys extruded at 175°C using -35 + 100 mesh powder. *(See page 20). ON — I Table V. Mechanical Properties of -35 + 100 Mesh Zn - 0.6 wt.% Ti Powder Extruded at 175°C and Annealed at Various Temperatures. 68 Annealing Conditions Work Hardening Rate 0.2% O.Y.S. U.T.S. (1/2 hr at temperature) at 0.5% Strain .3 „3 r , (10 psi) (10 psi) (10 psi) -extruded 2.0 60.2 64.5 100°C 2.7 61.9 67.7 120°C 4.2 59.6 69.3 140°C 4.2 59.9 66.5 160°C 4.1 29.0 73.5 180°C 4.2 31.7 57.1 200°C 2.9 21.4 35.7 300°C 0.45 24.6 27.4 Note: The high observed hardening rates are not true rates as such. The values are indicative of a pseudo-elastic effect. 180°C) undergoes "quasi-yielding" at very low stress levels and low strains. The stress at which this type of yielding occurs corresponds closely to yielding in softer extrusions, i.e. extruded material heat- treated at higher temperatures. As has already been indicated, the zinc-titanium intermetallic appears to be thermally stable up to approximately 160°C when held for 30 minutes at temperature. The transition structure which i s the result of intermediate thermal treatments can be thought of as consisting of both hard and soft regions, the hard regions being areas in which the microstructure remains stringered. Deformation within the soft region gives rise to an apparent low yield stress ("quasi-yield"). For general deformation to occur, some stringered areas must also deform. This gives rise to a high ultimate tensile strength. If the material i s annealed at a sufficiently high temperature coarsening of the complete microstructure w i l l take place, resulting in both low yield and low ultimate tensile stresses. 4.7.3.3 Annealed Cast-Extruded Zn-Ti Alloys Short period thermal treatment has virtua l l y no effect on the mechanical properties of cast and extruded material. If any sig- nificant effect were possible i t was thought i t would occur in low temp- erature extrusions in which the microstructure is somewhat finer. The experimental results obtained by thermal treatment of a cast 0.6 wt.% Ti alloy extruded at 150°C shown in Figs. 31 and 32 indicate that microstructural modifications are minor. 70 4.8 The Effect of Titanium Concentration The mechanical properties of zinc-titanium alloys are determined by the nature of the microstructure, which in turn is determined by the fabrication processes used. It has been shown that the extrusion temperature has a significant influence on the mechanical properties of material fabricated from powder and a minor effect on the properties of cast material. Thus i t is useful to discuss the effects of the second phase content in conjunction with both high and low fabrication temperatures. Table VI indicates the effect of titanium concentration on the mechanical properties of alloys investigated in this study. These data show that the concentration effect is significant when extrusion is carried out at a low temperature. This effect is marginal when an extrusion temperature of 350°C is used. Figures 37 through 40 i l l u s t r a t e the true stress-strain curves obtained for these materials. 4.8.1 Powder Material Extrusion of powder at low temperatures results in the formation of a stringered structure in which the stringer spacing and thus the grain size are reduced as the titanium concentration is increased. Thus signif- icant increases in strength are observed in alloys of higher titanium content. At 350°C, short diffusion paths give rise to coarsening of the fine as-solidified intermetallic structure, thereby leading to the development of rather coarse extruded microstructures with no stringered appearance. Low strengths are associated with such microstructures. T i Concentration Bil l e t (wt.%) Tx 0.2% O.Y.S. (°C) (10 3psi) U.T.S. Elongation to Reduction in Total Elongation Maximum Stress Cross-sectional (10 psi) (%) Area (%) (%) 0.07 -35 + 100 175 37.2 40.8 1.2 15 4.5. mesh powder 0.16 " 175 44.9 49.8 1.4 3.1 0.32 " 175 45.7 49.2 1.0 10 2.8 0.60 " 175 60.1 64.4 0.8 3.5 0.07 cast 175 17.0 . 19.5 10.0 >90 30.0 0.16 cast 175 18.9 22.1 7.0 21.0 0.32 cast 175 23.9 26.2 3.0 >90 16.0 0.60 cast 175 23.1 25.7 6.5 26.0 0.07 -35 + 100 350 22.7 24.3 6.0 75 20.3 0.16 -35 + 100 350 21.3 23.4 2.0 17.5 0.32 -35 + 100 350 23.7 25.7 2.0 65 11.7 0.60 -100 + 200 350 28.1 29.5 1.8 8.8 0.07 cast 350 18.2 22.8 9.0 >90 21,0 0.16 cast 350 23.4 26.4 8.0 20.5 0.32 cast 350 24.6 27.8 8.0 >90 19.0 0.60 cast 350 21.1 25.7 9.0 16.3 Table VI. The Effect of Titanium Concentration on Mechanical Properties. (e ^ 0.0028 min."1) 72 20 10 1 Zn - 0.6 wt.% Tl 2 Zn - 0.32 wt.% Ti 3 Zn - 0.07 wt.% Ti £ ̂  0.0028 min -1 -L 1 2 3 S T R A I N (%) Fig. 37 True stress - strain curves for zinc - titanium extrusions of -35 + 100 mesh powder at 175°C. * (See page 20). S T R A I N (%) Fig- 38 True stress - s t r a i n curves for zinc - titanium extrusions of -35 + 100 mesh-powder at 350°C. * (See page 20). S T R A I N (%) Fig. 39 True stress - strain curves for zinc - titanium extrusions of chill castings at 175°C. * (See page 20). -p- Fig- 40 True stress - strain curves for zinc - titanium extrusions of c h i l l castings at 350 * (See page 20). 4.8.2 Cast Material Some grain refinement and rather coarse intermetallic string- ering are characteristic of low temperature cast extrusions. Increasing titanium concentrations give rise to more extensive stringering and slightly greater strength. The mechanical properties are insensitive to concentration when extrusion is carried out at 350°C. These properties must be discussed in terms of the second phase distribution and size, the grain size, and grain shape. This discussion appears in a later section. 4.8.3 Ductility As can be seen from Table VI, fabrication history is the major factor controlling du c t i l i t y , with second phase content playing a minor role. In the absence of a stringered structure, reduction i n area values are consistently above 50% with the highest values being obtained for cast and extruded material. Total elongation varies between 20% and 30% for cast and extruded material. The elongation values are slightly lower for powders extruded at 350°C. As expected, the lowest total elongation values were observed with low temperature powder extrusions. Similar trends were exhibited in elongation to maximum stress values. Any analysis of mechanical properties at constant strain-rate as a function of titanium concentration is not complete since i t must be assumed that the strain-rate sensitivity is the same for a l l compositions. Therefore to rigorously analyze trends in the mechanical properties, the strain-rate sensitivity must be established and related to the operative deformation mechanisms. 77 4.9 Deformation Characteristics 4.9.1 Introduction It i s generally thought that five independent deformation modes must be operative before a polycrystalline metal can exhibit any significant ductility. Slip, twinning, and grain boundary shear may f u l f i l l or partially f u l f i l l these requirements. The yield stress i s determined by the number of deformation modes available and by the ease of activation of these mechanisms. 4.9.2 Deformation Mechanisms 4.9.2.1 Slip The predominant slip system in zinc is the basal system {0001} <1120>. Two independent modes can be attributed to s l i p on these systems. The only non-basal s l i p system observed to be operative in zinc is the second order pyramidal system {1122} <1123>. The operation of non-basal s l i p i t s e l f is sufficient for homogeneous deformation in as much as i t provides five independent sl i p systems. The stress levels associated with the activation of non-basal sl i p are significantly higher than those required for basal s l i p . ^ ^ In zinc-titanium alloys, both basal and non-basal sl i p (Fig. 41) were observed. Basal sl i p traces were generally well-defined, straight and continuous, while non-basal traces were often wavy and discontinuous. 4.9.2.2 Twinning Deformation twinning is a common occurrence i n large-grained Fig. 42 Metallographic evidence of twinning in a c h i l l cast Zn - 0.16 wt.% Ti alloy extruded at 175°C. (x6000) 79 hexagonal metals. The importance of this mechanism is a function of temperature, grain size, and the presence and nature of a dispersed second phase. Twinning rarely occurs in fine-grained zinc alloys produced by powder metallurgical techniques. The literature contains no information regarding the effect of a fine grain size on twinning. However the lack of i t s presence suggests that the c r i t i c a l stress required for nucleation may be related to a c r i t i c a l dislocation pile-up length, the size of which may not be present in fine-grained material. Because of the limited occurrence of twinning, i t is not considered to affect significantly the mechanical properties of alloys investigated i n this study. Figure 42 shows an example of a twin in a Zn - 0.16 wt.% T i alloy. 4.9.2.3 Grain Boundary Shear A deformation process which is of particular importance when considering fine-grained materials at high effective temperatures is grain boundary shear. Numerous theories as to the actual mechanisms (14) involved in this process have been proposed, but as yet the mechanism is not completely understood. However the effect of temperature, strain- rate and presence of a second phase are areas in which extensive invest- igations are being carried out, particularly in the area of superplasticity. The number of independent modes which can be attributed to shear is a matter of controversy. It is generally agreed, however, that this process may occur at very low stress levels. A large percentage of metallographical information is based on surface observations. It is possible that the nature and degree of grain boundary shear within the material might differ significantly from that suggested by observations of the surface. Thus accurate estimates of the contribution of shear to the overall deformation of a material are d i f f i c u l t to make and involve sophisticated experimental procedures. It i s important to realize that a grain boundary deformation mechanism must operate in conjunction with a diffusion mechanism or with a mechanism whereby intragranular deformation occurs in order to f u l f i l l accommodation requirements. Figure 43 shows grain boundary shear operating in conjunction with grain boundary migration. 4.9.3 Recovery Mechanisms 4.9.3.1 Introduction In view of the high effective temperature of zinc at room temperature, dynamic recovery effects may have a substantial influence on the observed mechanical properties. In the broad sense, dynamic recovery may be defined as the continuous loss of dislocations through diffusion controlled processes. Cross s l i p is not considered an important recovery mechanism in zinc due to the lack of a cross slip system. 4.9.3.2 Grain Boundary Migration An important function of grain boundary migration is the partial fulfilment of accommodation requirements at phase boundaries and tri p l e points. This process is considered a dynamic recovery mechanism i f migration results in a net decrease in dislocation density at or near the boundary. Grain boundary shear results in the formation of a shear zone during sliding. The generation of dislocations within this zone acts Fig- 44 Metallographic evidence of grain boundary shear and migration in a c h i l l cast Zn - 0.32 wt.% Ti alloy extruded at 350°C. (x6000) as the driving force for local migration. ' ' Further dislocation loss may occur through the annihilation of pile-ups by a migrating boundary. Figure 44 shows an example of grain boundary migration. The presence of a second phase on a boundary may retard migration. The dispersoid spacing w i l l determine the degree to which migration occurs. 4.9.3.3 Dislocation Climb The climb of dislocations around obstacles or out of dislocation pile-ups at grain boundaries i s thought to be the most important recovery mechanism at T = .4 or higher. This mechanism has received much attention in the general theories of recovery creep where the rate of recovery is thought to be controlled by the rate and distance (44 45) of dislocation climb. ' Recovery by climb may be discussed quantitatively when considering steady state creep conditions, but only qualitative observations may be made when dealing with tensile data. It is expected that recovery by climb w i l l show a temperature dependence related to bulk self-diffusion of the material. 4.9.4 The Occurrence and Importance of Substructure in Zino Complete recrystallization of the zinc matrix occurs during the extrusion of dilute zinc-titanium alloys. This i s confirmed by the low dislocation density observed through transmission microscopy (Fig. 5). Thus no significant substructure should be present in the as-extruded condition. In view of this, the effect of substructure on the yield strength w i l l be negligible since the formation of a substructure w i l l be the result of deformation; in other words any substructure which is produced w i l l be formed after the fact. Some glide polygonization can be expected to occur (not a thermally activated process). Dislocation glide in zinc occurs primarily on a single plane i.e. basal plane and thus glide polygonization w i l l result in the formation of low angle boundaries perpendicular to the basal plane. This process can not be considered as a hardening mechanism but rather should be thought of as a softening mechanism, i n as much as i t tends to accommodate strain in a direction perpendicular to the basal plane. A sub-boundary generally extends completely across a grain and can move under the influence of a s t r e s s . ( F i g . 45) 4.10 Strain-Rate Sensitivity 4.10.1 Introduction A deformation characteristic which is excluded from many studies but which must be considered in any investigation of zinc or zinc alloy is the strain-rate sensitivity of the flow stress. Studies of this nature must be made before any qualitative appraisal of the potential creep prop- erties can be given. Low rate sensitivity is generally associated with superior creep behavior. For high temperature deformation (T u > 0.4), the flow stress n A - • - n i , A x. ,u u- (19,20,22,47,48) and strain-rate are usually related by the equation: T, *m T = Ke or m = dim 9£ne The rate sensitivity parameter "m" commonly has values between 0.15 and 0.2 for high temperature deformation (recovery creep) where Fig- 45 Metallographic evidence of substructure in a c h i l l cast Zn - 0.16 wt.% T i alloy extruded at 175 °C. (x6000) deformation may be associated with some grain boundary shear but is nevertheless thought to be controlled by s l i p . At very fine grain sizes or much lower strain-rates m may increase to approximately 0.5 and under such conditions the material may be superplastic. Increasing m values are thought to be associated with an increasing degree to which grain boundary shear contributes to the overall deformation of the material. 4.10.2 Deformation of Pure Zinc (49) Turner employed powder metallurgical techniques identical to those used in this work to obtain zinc extrusions with a grain size of A subsequent deformation study indicated that the flow or yield stress versus strain-rate plot was "S"-shaped in nature as i s shown in Fig. 46. The variation of m with strain-rate i s shown i n Fig.47; i t can be noted that m varies between 0.1 and 0.2 as expected for a "hot" deformation process. However i t must be appreciated that a significant amount of zinc oxide was present as a result of the formation of an oxide layer during powder production. This inert second phase can be expected to have a stabilizing effect on the grain boundaries, limiting both shear and migration somewhat, thus preventing the material from exhibiting (25) superplastic behavior. The results of the work by Tromans and Lund have demonstrated the effect of oxide on the boundaries on the mechanical properties of zinc. It is probably accurate to assume that the changes in the range of m = 0.1 to 0.2 were related to the degree of grain boundary shear even though the deformation was always slip controlled. 20 10 u O 0.2% Offset Yield Stress Extrusion Temperature = 150°C Grain Size = 2u 10 -4 10 10 10 -1 10" S T R A I N R A T E ( M I N - I ) F 1 g - 4 6 T h e f l o w stress - yield stress dependence on strain-rate of zinc fabricated from -325 mesh powder at roo.„ temperature. (after Turner R A T E S E N S I T I V I T Y P A R A M E T E R " m o Ti H- TO • H cr Hi ro rt ro <J H Cu i-t H H- C OJ H rt 3 H* fD o H 3 /—\ -P- O Hi —̂' s: rt 3* co rt i-t 01 H- 3 (13 rt ro Hi o H T3 C H 0) N H- 3 o Hi cr H H* O Cu rt ro Cu Hi H O 3 I LO a ro cn cr 13 o Cu ro H o o LO in H JO 70 > H m o i O o M rt i-i 1= CD H- O 3 H o ro i-i g CU •a H» ro 3 i-i cu 00 rt H* c N I-t ro II ro II i—1 Ln O o O f - K3 O o Z8 88 4.10.3 Deformation of Zinc-Titanium Alloys 4.10.3.1 Strain-Rate Sensitivity The strain-rate sensitivity of dilute zinc-titanium alloys at 20°C is shown in Figs. 48 through 51. These data indicate that the titanium concentration has virtually no effect on the strain-rate sensitiv- i t y . The nature of the second phase distribution as determined by the fabrication conditions i s the important parameter. In view of this, Table VII summarizes the strain-rate sensitivity parameters and grain sizes resulting from the fabrication conditions used with a single composition (Zn - 0.16 wt.% T i ) . The rate sensitivity for pure zinc was variable (m = 0.1 - 0.2) over the experimental strain-rate range. However for zinc-titanium alloys, the m values are constant over the same range. Thus the presence of the intermetallic effectively extends the low sensitivity region of the log T - log e curve to much lower strain-rates, resulting in a linear relationship. The observed m values are very low, indicating the Zn^ T i distribution is effective in stopping grain boundary shear and migration. Although the deformation is sli p controlled, some grain boundary deformation w i l l occur and w i l l be reflected by small increases in the value of m. In the case of zinc-titanium alloys the increase in m is small (0.02 to 0.07), but the corresponding decrease in strength is considerable. The ease with which a boundary may shear or migrate w i l l be determined by the grain size, grain boundary orientation, and distribution of the second phase along the boundary. The lowest m value was observed for material fabricated from C O C L IO O 100 co co UJ et: i - CO 50 Q _J LU 9 e Q - . • - -A - • •A • UJ to U. O 20 m =.02 O 0.6 wt.% Ti O 0.32 wt.% T i A 0.16 wt.% T i ° 0.07 wt.% Ti S T R A I N R A T E ( M I N - " ) Fig^__48 The strain-rate senqnt-iirif,, ~e S T R A I N R A T E ( M I N " ' ) a t e 3 5 ? c ' n " r a t e S e n S l t i V i t y ° f Z i n C ~ t i t a n i u m a l l o y s f a b r i c a t e d from -35 + 100 mesh powder S T R A I N R A T E ( M I N - 1 ) Fig. 50 The strain-rate sensitivity of zinc - titanium alloys fabricated from c h i l l castings at 175°C. vO CO CL ro O CO CO UJ or co o _l LU > h- LU CO U_ U. O C O a o 100 50 20 © 0.6 wt.% Ti • 0.32 wt.% Ti A 0.16 wt.% Ti O 0.07 wt.% T i 10 10 10 _L ,-2 X 10 " 10 S T R A I N R A T E ( M I N - 1 ) - l 1(T F i § - 5 1 T h e strain-rate sensitivity of zinc - titanium alloys fabricated from c h i l l castings at 350°C. VO Table VII. Summary of Strain-Rate Sensitivity Parameters and the Variables Affecting these Parameters for Zn - 0.16 wt.% T i alloys. Material Fabrication Temperature m Grain Size -35 + 100 powder 175°C 0.02 1.0 y -35 + 100 powder 350°C 0.07 2.1 y cast 175°C 0.07 4.1 y cast 350°C 0.04 8.1 y 94 powder at 175°C (m = 0.02). The fine grain size is conducive to shear. However most boundaries are oriented such that there is no resolved shear stress along them. The fine intermetallic stringers also provide a high degree of stabilization. More equiaxed grain structures have a higher percentage of boundaries oriented favourably for shear. The fine grain sizes and coarsened intermetallic present in powders extruded at 350°C lead to higher m values (m = 0.07). The coarse intermetallic distributions in material fabricated from cast b i l l e t s provided the lowest degree, of boundary stabilization. However somewhat larger grain sizes are formed. Material extruded at 175°C has a smaller grain size thus increasing the degree to which shear and migration may occur; such material has higher m values than material fabricated at 350°C. Qualitatively, good creep properties are related to high stress levels and low m values. The strain-rate data obtained in this study suggest that the properties of these alloys should be excellent. However i t is d i f f i c u l t to predict whether the rate sensitivity w i l l remain this low at very low (creep) strain-rates. 4.10.3.2 Ductility As expected, higher strength materials in general show lower % elongation and % reduction in area values. This is probably due to a condition of instability resulting from stress levels higher than the work hardening rate. Thus high strength extrusions begin to neck immediately following yielding. In this case, % elongation values are actually a reflection of the degree to which necking occurs. These values w i l l be governed by the nature of microstructural modification (probably grain boundary migration), which occurs within the necked region. The stress levels reached in cast and extruded material are much lower. As a result, necking i s initiated at higher strains. Grain boundary migration occurs more readily in these materials and ultimately results, in higher reductions of cross-sectional area. Boundary migration is probably retarded by fine intermetallic distribution and by oxide distribution present i n extrusions fabricated from powders. These effects are indicated by the data shown i n Table VI. (See p. 71) 4.11 High Temperature Deformation Characteristics 4.11.1 Deformation of Zinc-Titanium and Zinc-Aluminum Alloys 4.11.1.1 Introduction The extent and nature of the distribution of the second phase w i l l determine the amount of grain boundary deformation which w i l l occur in zinc-titanium alloys. Therefore a maximum degree of shear and migration and consequently the higher strain-rate sensitivity w i l l be observed under conditions in which the minimum amount of second phase (required for grain refinement) is present. Superplasticity has been observed in dilute zinc-aluminum alloys with grain sizes comparable to those of the zinc-titanium alloys used in this study. The important difference is that the properties of the second phases on these alloys are different at 20°C. The intermetallic formed in zinc-titanium alloys is "hard" relative to the zinc matrix but the second phase formed in zinc-aluminum alloys is "soft". Using Zn - 0.2 wt.% Al alloys, Cook^"^ obtained m values of approximately 0.5 and elongations of over 500% at 20°C. The low m values and d u c t i l i t i e s associated with zinc-titanium alloys are attributed to 96 the mechanical properties of the intermetallic compound. Higher m values and extended ductility can be expected i f the mechanical properties of the intermetallic become comparable to the matrix at elevated temperatures. 4.11.1.2 Effect of Temperature on the Flow Stress - Strain-Rate Relationship The microstructure of Zn - 0.07 wt.% T i alloys fabricated from -35 + 100 mesh powder at 350°C closely resembles that of the superplastic Zn - 0.2 wt.% Al alloys used by Cook (Figs. 52 and 53). The grain sizes of both alloys are fine and equiaxed, with a large percentage of the grain boundary area apparently free of second phase. The effect of temperature on the flow stress - strain-rate curves of the Zn - 0.07 wt.% Ti alloy is shown in Fig. 54. The rate sensitivity is a function of temperature and varies between 0.07 and 0.35. These values reflect the ease at which grain boundary deformation occurs at elevated temperatures. However the strain-rate sensitivity parameter remains below that required for superplastic behaviour. In the region of higher rate sensitivity there was a tendency for the material to resist necking and spread deformation more uniformly, but extended ductility was not obtained. Metallography indicated that a slight degree, of grain growth and intermetallic coarsening did occur during high temperature deformation. Agglomeration of the second phase is effectively the same as reducing the concentration present. For unstringered alloys such as this, the flow stress dependence on grain size i s reasonably small, thus the slight degree of coarsening which resulted w i l l have a minor effect on the observed flow stress. The grain size - flow stress relationship is discussed later. In Zn - 0.2 wt.% Al alloys, grain boundary shear is thought to be the dominant mechanism under conditions in which higher m values and superplastic behavior are observed. Metallographic evidence indicates F i § - 53 The microstructure of a Zn - 0.07 wt.% T i a l l o y extruded at 350°C using -35 + 100 mesh powder. (x4000) STRAIN RATE ( M I N - I ) F iS- 5 4 T h e effect of temperature on the flow stress - strain-rate relationship for Zn - 0.07 wt% Ti alloys extruded at 350°C, using -35 + 100 mesh powder. vo oo 99 that sliding occurs both at phase boundaries and at grain boundaries of similar phases. The operation of a grain boundary sliding mechanism in a two phase system must be accompanied either by deformation of both phases or by diffusional mechanisms which lead to fulfillment of accommod- ation requirements; otherwise extended duct i l i t y is not observed. Even though grain boundary shear occurs more readily at elevated temperatures i n zinc-titanium alloys, i t does not appear to become rate controlling. 4.11.1.3 Ductility and Fracture Mechanisms The explanation for the rather low amount of ducti l i t y observed is found i n an analysis of the mode of fracture of this material. As indicated previously, l i t t l e tendency toward necking was observed on specimens exhibiting larger m values. The fracture i t s e l f from a macroscopic point of view appeared quite b r i t t l e in nature. A large number of crack systems in various stages of development were apparent on the surface of the tensile specimens. The occurrence of these cracks increased significantly as the testing temperature was increased. These observations were made on specimens tested in two separate chemically inert testing media. Independent and isolated crack systems were also observed within the interior of the tensile specimen. On this basis, i t is assumed that the origin of these cracks is not due to the operation of a corrosion or stress corrosion mechanism. Since testing was carried out in liquid media, prior metallograph- i c a l preparation of the surface of the tensile specimen could not be made. Excessive mechanical and/or electrochemical polishing subsequent to testing tended to either smear or enlarge the microcracks. Thus metallography 100 was restricted to the interior of the specimen with preparation consisting of light mechanical polishing followed by light electrochemical polishing. A well-developed crack network is shown in Fig. 55. Figure 56 indicates that the origin of the cracks is at the intermetallic-zinc phase boundary. These cracks or voids gradually grow unt i l they combine with other cracks propagating along the same or another grain boundary. This process is extended unt i l a continuous network is formed and failure of the specimen occurs. When precipitates exist along a grain boundary, tr i p l e points are formed on each side of the second phase particle. If grain boundary or interphase boundary shear does occur, then i n order to f u l f i l accommodation constraints, deformation of the two phases by s l i p or grain boundary migration must take place. If accommodation conditions are not f u l f i l l e d at the triple points, the result i s the formation of (62 6 a void or microcrack and consequent premature failure due to cavitation. ' A schematic representation of this type of mechanism is shown in Fig.57 (after McLean^16"*). Since for the alloy under consideration variable m values are observed, more than one deformation process is suggested. The relative contribution of these processes is variable and is dependent on the strain- rate and temperature.. Grain boundary shear may result in high elongations and low strength in cases in which both phases of a two phase alloy are "soft". In the instance in which one phase is "hard" relative to the other, the mech- anism of grain boundary shear can be the mechanism which leads to cavitation and premature failure. Fig- 55 Cavitation f a i l u r e at 300°C i n a Zn - 0.07 wt.% T i a l l o y extruded at 350°C using -35 + 100 mesh powder. (x950) 102 F i g . 56 Scanning e l e c t r o n micrograph of c a v i t a t i o n f a i l u r e at 300°C i n a Zn - 0.07 wt.% T i a l l o y extruded at 350 °C using -35 + 100 mesh powder: (a) R e f l e c t e d e l e c t r o n s (b) Secondary e l e c t r o n s . F i g . 57 Schematic representation of c a v i t a t i o n due to the presence ot second phase p a r t i c l e s on a grain boundary. (after McLean 104 4.11.1.4 Activation Energy Analysis A deformation study establishing the effect of temperature on the yield or flow stress of the alloy under certain conditions may yield data suitable for an activation energy analysis. Generally speaking, when a wide range of temperatures and stress levels are involved, more than one process may be thermally activated. That i s , when considering log stress - log strain-rate curves in which the strain-rate sensitivity parameter is not constant, a number of thermally activated processes may be involved. The rate controlling process may not be the same throughout the complete strain-rate range. Processes may be considered to be acting i n either series or parallel. The suggestion that processes act in parallel implies that the mechanisms are interrelated and that their operation has a mutual dependence. Processes are considered to act in series i f in fact they operate independently of one another. Arrenhius plots of series processes generally have distinct regions and thus are not linear. However, i f limited temperature ranges are considered or i f one mechanism is much more dominant then a linear plot could result. The u t i l i z a t i o n of tensile data in which the strain-rate i s fixed is based upon the assumption that the magnitude of the yield stress and/or flow stress is a direct result of the rate controlling process or processes involved. In other words, at a constant stress level, the strain- rate is a direct measure of the rate of deformation. If deformation is assumed to be controlled by a single rate controlling process ( i.e. parallel model) then the strain-rate is related to the Gibb's Free Energy by the following: ^ ' - ^ £ = A e - A G / R T 105 Similarly, i f two processes are considered to be acting in series: i - B e " A G l / R T + Ce- A G2 / R T However, i f one process predominates in the series model, then the expressions for the strain-rate relationship are similar; that i s , one term becomes negligible. If i t i s further assumed that the entropy change is not sig- nificant with respect to the overall free energy change, then the equation may be modified to: ., -AH/RT e = A e where AH is referred to as the activation energy of the rate controlling process. Thus an analysis of this nature must be carried out in the region of the flow stress - strain-rate curve in which the strain-rate sensitivity parameter is constant at a l l temperatures. Under these conditions the relative contributions attributed to differing mechanisms should remain relatively constant. For an accurate determination of the apparent activation energy i t is a necessary condition that the microstructure remain constant. Slight coarsening of the grain size and intermetallic did occur during this study, however for a microstructure of this nature i t w i l l be shown that the mechanical properties are rather structure insensitive, and therefore the minor microstructural changes which did take place w i l l not have a great effect on the activation energy analysis. Data for this type of analysis is conveniently obtained using instantaneous strain-rate change tests previously described. Such procedures 106 were employed using a Zn - 0.07 wt.% T i alloy fabricated from -35 + 100 mesh powder at 350°C. The data obtained has already been shown as Fig. 54. Using a constant stress level of 6000 psi, the Arrhenius plot shown as Fig. 58 was made. The apparent activation energy of the rate controlling process was calculated to be 24.5 + 3 kcal/mole. This value corresponds closely to published results for the bulk self diffusion of zinc.^ Thus the controlling process is diffusion controlled and is probably that of dislocation climb. 4.11.2 Yield Stress Dependence on Temperature Determination of the yield stress dependence on temperature for fine stringered microstructures is prevented by the thermal instability of the intermetallic compound. It has been seen that a thermally stable material can be produced by annealing these structures at a sufficiently high temperature. (See Fig. 31) Thus such a determination can be made i f a stabilized structure is employed. Few microstructural modifications w i l l occur i f testing is completed in a relatively short time and i f the test temperature does not exceed that which was used during the thermal treat- ment. Such an investigation.was carried out using a 0.6 wt.% T i alloy produced by extrusion at 175°C of -35 + 100 mesh powder. The stabilizing thermal treatment involved annealing the specimens at 250°C for 1/2 hour prior to testing. The results of this study are shown in Figs. 59 and 60. The yield stress - temperature plot indicates that a discontinuity occurs at approximately 100°C. The existence of this discontinuity was verified by duplicate tests. Figure 60 indicates that above 100°C a significant increase in the strain-rate sensitivity from 0.06 to 0.18 is observed. Thus i t is 107 1 1 r 0.0015 0.0019 0.0023 0.0027 JL ( ° K - I ) T Fig- 58 Arrhenius plot for Zn - 0.07 wt.% T i alloy 350°C using -35 + 100 mesh powder. extruded at TEMPERATURE ( ° K ) Fig. 59 The dependence of yield stress on temperature of a thermally stabilized Zn - 0.6 wt.% Ti alloy. o oo F i g - 6 0 T h e e f f e c t o f t e m p e r a t u r e on t h e f l o w s t r e s s - s t r a i n - r a t e r e l a t i o n s h i p o f a t h e r m a l l y s t a b i l i z e d Zn - 0 . 6 w t . % T i a l l o y . 110 reasonable to correlate the discontinuity in the yield stress - temperature relationship with an increase in m with temperature. As has been discussed, an increase in strain-rate sensitivity is generally associated with an increase in the amount of grain boundary shear which occurs during deformation. Thus i t is not probable that this change in slope can be explained in terms of a change in the mode of deformation; but rather in terms of a mode which operates to a significant degree at higher temperatures. Such behaviour may be correlated with the "equicohesive" temperature.^"^ Essentially this is a temperature at which the grain boundaries are considered to become soft relative to the matrix. Although the equicohesive temperature is a function of the nature of the microstructure, i t is generally around 0.5 T (0.5 T„ for zinc is approximately 75°C or 350°K). If this indeed H H is the case, a significant increase in grain boundary shear can be expected to occur above this temperature. 4.12 Hall-Petch Analysis 4.12.1 Introduction The expression proposed by Hall and Petch relating the yield or flow stress, a, and the mean grain size, d, is o = a + kd~ 1 / 2 o where O q and k are constants. Although the Petch relation is often criticized because of an inherent normalizing effect on experimental data which consequently fits data mathematically which may have a poor correlation, i t does allow a large volume of results to be subjected to a simple analysis. I l l The f u l l significance of the grain size, slope, and intercept of the resulting plots is not fu l l y understood. Nevertheless such an analysis is useful in qualitatively describing the relative importance of grain size and grain boundary effects and their relationship to the flow stress. Originally the equation was formulated by the necessity to preserve continuity across grain boundaries in a single phase polycrystalline material in which plastic deformation occurred exclusively by s l i p . However modifications of the Petch analysis have been made in order to apply the (25 27 52) relationship to two phase materials. ' ' Successful application of this type of analysis depends upon the physical significance attached to the constants a and k. o 4.12.2 Petch Equation and Hexagonal Metals The constant is thought to be a form of internal f r i c t i o n stress related to the energy required to move a dislocation on a sli p plane in the absence of the influence of grain boundaries. If such is the case, a w i l l depend on variables such as solid solution constituents, i n i t i a l dislocation structure, etc. In the second term of the equation, k has been associated with (53) the "strength" of the grain boundary. The actual physical significance attached to the term "strength" of the boundary is not clearly defined, but i t is generally assumed to be associated with the ab i l i t y of a boundary to act as a barrier to dislocation motion and to transmit the stress associated with the resulting dislocation pile-up to an adjacent grain in order to induce plastic flow in that grain. It has also been associated (53) with the ab i l i t y of a boundary to act as a dislocation source. As the number of available s l i p systems in hexagonal metals is very low in comparison with cubic metals, the values of k for hexagonal metals 112 are generally much higher. This suggests that orientation conditions across a grain boundary w i l l affect the observed value of k substantially, as w i l l any mechanism which affects the formation of dislocation pile-ups. Thus when considering zinc or zinc alloys, the Petch relation may be considered as giving the stress required for the formation of dislocation pile-ups with resulting stress concentrations of a sufficient magnitude to activate a sufficient number of s l i p systems to satisfy the c r i t e r i a for plastic flow. This definition is made under the assumption that grain boundary shear dqes not contribute to a significant degree to the overall plastic deformation of the material. It follows that i f a sig- nificant amount of grain boundary shear does indeed occur, then a change in the Petch slope corresponding to the degree of shear should be observed. 4.12.3 Modifications of the Petch Analysis (25) Lund and Tromans modified the Petch equation to: . . .-1/2 . .,-1 CT0.2 = CTo + k l b " k 2 d The purpose of the d term was to include a provision for deformation resulting from grain boundary shear and/or migration. It was assumed that second phase particles were as effective as barriers to dislocation movement as were grain boundaries. The validity of including a function to account for the extent to which grain boundary sliding and migration influence the flow stress is questionable when a multiphase material is under consideration. The amount of grain boundary shear which contributes to the overall plastic strain can not be considered to be proportional to the grain boundary area per unit volume unless a l l boundaries are free of secondary phases. The amount of sliding and/or migration w i l l be a function of the amount, size, dispersion 113 and mechanical properties of the second phase particles. 4.12.4 Petch Analysis and Zinc-Titanium Alloys A Petch analysis was carried out employing data obtained from a series of zinc-titanium alloys of variable composition, intermetallic size and dispersion. The analysis was done at 20°C and at -100°C. Grain sizes were determined by a line intercept method. The procedure was thought to be reasonably accurate when the grains were equiaxed. However the grain size of material which was stringered and therefore consisted of block-shaped grains was established by taking line intercept measurements at 45° to the extrusion (and tensile) direction. It was thought that since the Petch analysis requires the mean grain diameter as a measure of the average possible sl i p distance, this dimension would most closely approximate the required distance. The results of this analysis are shown in Fig. 61. 4.12.4.1 Flow at -100°C At -100°C, grain boundaries in zinc or zinc alloys w i l l be "hard" and thus plastic deformation w i l l be the result of s l i p . Lund and Tromans modified the Petch relation to suggest that the presence of dislocation barriers other than grain boundaries w i l l affect the size and density of boundary pile-ups. This effect should be manifested by a change in the observed value of k when data are obtained under conditions such that deformation occurs exclusively by s l i p . (27) However Risebrough and Lund report that at -100°C Zn-Ti, Zn-Cr, Zn-ZnO, Zn-Cr-Ti and pure Zn exhibit the same flow stress - grain size dependence and Petch obeyance is observed. (GRAIN S I Z E ) " " 2 ( i | " , / 2 ) Fig. 61 The Hall-Petch plot for zinc-titanium alloys. study: At this temperature, the value of the constants were: a Q * 13,000 psi k * 50,000 psi (y)V2 These values compare reasonably with the values obtained in this 1/2 aQ ^ 20,000 psi k ^ 47,000 psi (y) The size, dispersion, and concentration of the intermetallic do not appear to have a significant effect on the mechanical properties of these alloys at this temperature. Therefore the strength is primarily a function of the grain size. The flow stress values obtained under these conditions can be considered to be an indication of the ultimate strengthening of zinc which could be accomplished with complete immobilization of the grain boundaries. 4.12.4.2 Flow at 20°C 4.12.4.2.1 Stringered Alloys Plastic flow of the zinc-titanium alloys studied is characterized by two significantly different k values at 20°C. The flow stress of material fabricated from powder by extrusion at a low temperature (175°C) follows a linear^, relationship when.plotted against the reciprocal square root of the mean grain size. It is then apparent that the flow stress is a function of the grain size only and is quite insensitive ; to the amount of intermetallic compound present. As has been previously discussed, extruding at 175°C gives rise to a well-defined stringered structure in which the grains are small and columnar in shape. Lowering the titanium content results in an increased stringer spacing and width and a slightly enlarged grain size. The columnar grain 116 shape is generally maintained although small colonies of more equiaxed grains were observed at lower intermetallic concentrations. The Zn{gTi stringers in a l l cases consist of extremely fine particles effectively pinning the grain boundaries oriented parallel to the extrusion ( and tensile) direction. Perpendicular boundaries are devoid of second phase. Because of the high degree of pinning of boundaries and because of the unsuitable orientation of both pinned and "clean" boundaries, shear or migration is not considered to play a major role in overall plastic deformation of the material. Since room temperature corresponds to a high effective temperature (T = 0.42), grain boundary shear and/or migration should be expected to have a great influence on the flow st ress. Therefore curtailing grain boundary effects by the formation of a suitable microstructure should result in an increase in flow stress similar to the result of decreasing the effective temperature, and consequently rather high values of k are expected. For material with this type of microstructure, a ^ 20,000 psi and k 'v* 28,000 psi (u) 4.12.4.2.2 Unstringered Alloys Extruding or thermally treating alloys with a fine intermetallic structure at temperatures at which the Zn^sTi is unstable, results in rapid changes in the nature of the microstructure. Intermetallic coarsening appears to occur very rapidly, forming spheroidized particles. As coarsening proceeds, a corresponding increase in grain size and change in grain shape occurs. The larger particles of second phase remain located at the grain boundaries to a large degree with only a small percentage being situated intragranularly. Thus the grain size is controlled to a large extent by the nature of the intermetallic dispersion. The grain sizes of material extruded using cast b i l l e t s i s moderately large in the as-extruded condition. A considerable degree of scatter does exist on the Petch plot for unstringered material. This type of scatter i s to be expected and is presented as such to demonstrate that a loose obeyance of the Petch -1/2 relationship and a significantly reduced slope (k ^ 8000 psi (u ) )o'ccurs when a stringered microstructure is not maintained. The degree of grain boundary pinning which can be associated with the intermetallic and to a minor extent with zinc oxide is variable and depends on grain size, grain shape, composition, size and dispersion of ZnisTi, which in turn are dependent upon the thermal treatment and/or the fabrication temperature. Generally the amount of grain boundary shear and migration (and their contribution to the flow stress) w i l l be proportional to the unpinned grain boundary area per unit volume and the grain size. Thus as the relative importance of boundary effects is increased, the degree of conventional Petch obeyance w i l l be reduced. At larger grain sizes, the flow stress tends to become less dependent on the presence of the hard second phase. 4.12.5 Value of a o An important feature of Fig. 61 is that the extrapolated value of appears to be constant, independent of both temperature and structure. As the grain size becomes increasingly larger, boundary deformation or recovery mechanisms become decreasingly important. Approaching the limiting case in which grain boundaries are absent, the results suggest that the 118 stress required to move dislocations along a slip plane must be relatively temperature insensitive, thus giving rise to a constant value of a . o 5. SUMMARY 5.1 Introduction A high homologous temperature at 20°C results in low strength and poor creep properties for most zinc alloys. More specifically, dislo cation climb, grain boundary shear and migration results i n limitations upon the development of the desired mechanical properties. Thus the development of a useful zinc alloy is contingent upon the formation of a multiphase microstructure which inherently inhibits these processes. These requirements suggest that zinc is somewhat unique i n that conven- tional dispersion hardening c r i t e r i a are not s t r i c t l y applicable. Addi- tions of titanium and the u s e of powder metallurgical techniques have led to alloys which have improved strength and creep properties. 5.2 Deformation Characteristics 5.2.1 Strength If deformation is carried out under conditions such that the grain boundaries are "hard", high stress levels are reached. The expla- nation for this strengthening effect necessarily involves the mechanisms operative during deformation. Low temperature fabrication of powder material results in a stringered intermetallic distribution and a small, block-shaped grain size. The boundaries are pinned or oriented such that a significant degree of grain boundary shear cannot occur. At these grain sizes, twinning is not observed, thus deformation must involve the operation of non-basal s l i p i n order to satisfy comptability conditions. 120 The stress levels required to activate second order pyra- midal s l i p are much higher than those required for either basal s l i p or grain boundary shear. The necessity to activate s l i p other than basal s l i p gives rise to high strengths and large values of "k" i n the Hall- Petch Relationship. Intermetallic distributions such as those which result during high temperature extrusion of power or extrusion of cast castings lead to microstructures i n which a significant amount of grain boundary shear and migration may occur. Basal sl i p and grain boundary shear then supply the number of deformation modes required for plastic flow, and lower strengths are observed. In the case in which grain boundary effects contribute significantly to the overall deformation, the slope of the Hall-Petch +1/2 curve is low. (>8000 psi u ). It is possible that the c r i t i c a l dislo- cation pile-up size required to i n i t i a t e non-basal sli p i n neighbouring grains is never reached due to annihilation by local boundary migration. The effect of titanium concentration on the strength can be considered as a variable of secondary importance. Lower intermetallic concentrations w i l l lead to larger stringer spacings and a reduction i n strength of powders extruded at low temperatures. The effect of com- position i s less pronounced i f the intermetallic distribution i s coarse. In general, a larger grain size is associated with a coarser intermetallic distribution. This suggests that the relative proportion of stabilized and unstabilized grain boundary area remains almost constant. The prerequisite for high strength is a fine intermetallic distribution in the as-solidified material. The fabrication must be such that this distribution must be retained in the f i n a l product. Lower temperature fabrication does not effectively fragment and distribute 121 coarse intermetallic particles formed at lower s o l i d i f i c a t i o n rates or with thermal treatments. The degree of "stringaring" in the product is not a function of the i n i t i a l powder size. The s o l i d i f i c a t i o n rates are such that an equivalent fine microstructure is produced independent of powder size? Less severe deformation during fabrication does result i n a more poorly defined stringered structure and consequently lower strength. The degree to which strengths are maintained at elevated temperatures is dependent on the thermal s t a b i l i t y of the intermettalic (Zn 1 5Ti). 5.2.2 Strain-Rate Sensitivity Pure zinc and zinc-aluminum alloys are associated with an S-shaped, log stress-log strain rate relationship at room temperature. Variables i n "m" values are related to either a change in the controlling deformation mechanism or to a change in the degree of operation of several mechanisms. This type of relationship is observed with zinc-titanium alloys at elevated temperatures and low second phase concentrations. At room temperature, the log stress-log strain rate relation- ships were linear. The changes in strain rate sensitivity observed between alloys are very small (m = 0.02-0.07). These values are indica- tive of sl i p controlled deformation. In alloys other than those fabrica- ted from powder at low temperatures, grain boundary shear may make a sub- stantial contribution to the overall deformation of the material. The strain rate sensitivity can be considered as a reflection of the controll- ing deformation process. High stress levels and low strain rate sensitivities suggest 122 the creep properties of stringered material may be excellent. However i t is d i f f i c u l t to predict the behaviour of these alloys at much lower (creep) strain rates. The strain rate sensitivity i s insensitive to titanium con- centration but rather is determined by the nature of the second phase distribution which in turn is determined by the fabrication conditions. At elevated temperatures, the rate sensitivity increases significantly. The observed "m" values are a function of both the microstructure and testing conditions. Experimental evidence indicates the existence of a c r i t i c a l temperature (^0.5TH) above which "m" increases markedly. Thus above this c r i t i c a l temperature, grain boundary shear operates to a considerable degree. Although high "m" values (up to 0.35) were obtained, extended duct i l i t y was not observed. This is due to the nature of the failure which occurred under these conditions. The activa- tion energy suggested that deformation is controlled by a process assoc- iated with the bulk diffusion of zinc — probably dislocation climb. 5.2.3 Dynamic Recovery The suppression of dynamic recovery results in enhanced strengths and limited, d u c t i l i t i e s . Due to the fabrication procedures, the intermetallic particles tend to be preferentially situated on grain boundaries. This in effect limits the amount of grain boundary shear and migration which may occur. Coarsening of the second phase has an effect similar to lowering the titanium concentration and leads to large inter- particle spacings along boundaries. Local migration occurs more readily as this spacing is increased as i t is a function of the unpinned grain boundary area. 123 Since the interparticle spacing is comparable to the grain size in a l l alloys, i t is doubtful i f the intermetallic has any effect on dislocation climb within the grains. 5.2.4 Ductility and Fracture In general higher strength material exhibits lower d u c t i l i t i e s . This arises from stress levels i n excess of the degree of work hardening which results in a condition of instability which is present at or immediately following yield. Thus in high strength material, necking is initiated at very low strains. The stress levels reached with lower strength products lead to neck formation at higher strain values. Thus % elongation values are very low but are associated with rather large % reduction in area values. This is particularly true for extruded c h i l l castings in which the reduc- tion of area values were >90%. The degree to which necking occurs is governed by the microstructural modifications (grain boundary migration) which take place within the necked region. Boundary migration is probably retarded by the intermetallic and oxide distribution in products fabr i - cated from powder. Enhanced boundary migration and shear which occur at eleva- ted temperatures can lead to extended ductility under conditions in which both phases in a two-phase system have equivalent mechanical pro- perties (e.g. Zn-Al alloys). The second phase (Zn^^Ti) in zinc-titanium alloys appears to remain hard relative to the zinc matric at a l l tempera- tures. Grain boundary shear leads to premature failure due to cavitation. Cracking is initiated at the intermetallic-matrix interphase boundary. 124 5.3 Thermal Stability Experimental results suggest that fabrication alone can not develop the desired intermetallic dispersion. The i n i t i a l fine micro- structure present in as-atomized power must be preserved during fabrica- tion i n order that the product has appreciable strength and reasonable strain rate sensitivity. A certain degree of fragmentation and relocation of large primary intermetallic particles does occur during low temperature extrusion. However the degree to which this occurs does not give rise to efficient grain refinement or stabilization. Further proof of this i s obtained by thermally treating powder b i l l e t s prior to low temperature fabrication. Under these conditions, the mechanical properties of the resulting product approach those obtained by extrusion of cast b i l l e t s . The short diffusion paths which exist in either powder or pro- ducts obtained by low temperature fabrication of powders give rise to rather rapid microstructural coarsening at moderate temperatures. The diffusion paths in cast material are sufficiently long that no significant softening occurs unless severe thermal treatments are used. These consider- ations place a limitation of the fabrication conditions which must be used in order to maintain a fine microstructure. Thermal treatments are of course dependent on both time and temperature. However the fine intermetallic dispersion obtained appears to be thermally stable up to at least 150°C for short periods of time. 5.4 Other Considerations There is evidence to suggest that i t may be possible to develop a fine continuous second phase structure in cast b i l l e t s . Colonies of 125 fine eutectic were observed to form during chill casting. An extensive solidification study of dilute zinc-titanium alloys should establish criteria for the formation of such a microstructure. This would be of considerable advantage as i t may be possible to develop high strength and superior creep properties without resorting to powder metallurgical techniques. There appears to be some controversy as to the nature of the zinc-rich portion of the zinc-titanium phase diagram. It is reasonable to assume that reconsideration of the phase relations would have to be made in conjunction with any solidification study. 126 6. CONCLUSIONS The conclusions that can be made from the interpretation of the metallorgaphic and mechanical characteristics of dilute zinc- titanium alloys are as follows: (1) The eutectic composition appears to be in the range 0.4 - 0.6 wt. % Ti at the solidification rates corresponding to c h i l l casting and atomizing. (2) Reduced lamellar and interdendritic spacings are formed at rapid solidification rates. Generally coarser solidification structures are observed in chill castings, however isolated areas of fine eutectic were present. (3) Extruded microstructures with high strength and low strain rate sensitivity are the direct result of a fine as-solidified starting material. The structure is stringered with small grain size (0.5 - 1 . 5 u ) . (4) High strengths are observed when grain boundary shear and migration are inhibited by grain boundary stabilization. The degree of stabiliz- ation is associated with the orientation of the boundary with respect to the tensile axis and the distribution of second phase. Deformation occurs by both basal and non-basal slip. Lower strengths are observed when basal slip and grain boundary shear are the operative deformation mechanisms. (5) The strain rate sensitivity parameter of dilute zinc-titanium alloys lies in the range 0.02 - 0.07. Varying amounts of grain boundary shear occur, nevertheless deformation is slip controlled. 127 (6) Increased strain rate sensitivities (m up to 0.35) are observed at elevated temperatures. However failure due to cavitation limits d u c t i l i t y . The activation energy corresponds to bulk self-diffusion of zinc. (7) The strain rate sensitivity is not sensitive to the titanium concen- tration i n the range 0.07 - 0.60 wt. % T i . Under constant fabrication conditions, the strength generally increases with increased intermetallic content. (8) Low % elongations and high % reduction in area values are generally observed at 20°C. These values are a function of microstructural modification within the necked region. (9) The tensile results obey the Hall-Petch relationship. The observed "k" values are a function of the microstructure which w i l l result from the fabrication conditions +1/2 k - 28,000 p.s.i.y T powder extruded at 175°C +1/2 k - 8,000 p.s.i . u a l l other alloys. (10) The intermetallic compound is thermally stable up to 150°C for short periodsof time. (11) The mechanical properties are not a function of i n i t i a l powder size. 1 2 8 7 . BIBLIOGRAPHY 1 . A. Kelly and R. B. Nicholson, "Progress in Materials Science", 1 0 , (Pergamon Press), 1 4 9 , ( 1 9 6 3 ) . 2 . N. J. Grant, "The Strengthening of Metals", (Reinhold, New York), 1 6 3 , ( 1 9 6 4 ) . 3 . C. G. Goetzel and R. F. Bunshah, "Powder Metallurgy", (Interscience, New York), 2 5 3 , ( 1 9 6 1 ) . 4 . M. Ashby, Z. fur Metallkunde, 5 5 , 5 , ( 1 9 6 4 ) . 5 . G. S. Ansell and F. V. Lenel, Acta Met., 8, 6 1 2 , ( 1 9 6 0 ) . 6 . N . R. Risebrough, Ph.D. Thesis, University of British Columbia, ( 1 9 6 5 ) . 7 . G . H. Laurie and J. R. Wellington, "Zinc Extrusion", presented at Automotive Engineering Congress, Detroit, Mich., ( 1 9 6 6 ) . 8 . J. A. Lund, D. Tromans, and S. F. Radtke, Inter. J. of Powder Met., 4_^ 4 1 , ( 1 9 6 8 ) . 9 . J. A. Lund et a l . , "Fundamental Investigation of Zinc Powder Metallurgy", for I.L.Z.R.O., 1 0 , 1 4 , 1 8 , ( 1 0 6 5 - 6 8 ) . 1 0 . P. Chaudhari, Sc. and Tech., 8 1 , 4 2 , ( 1 9 6 8 ) . 1 1 . P. Chaudhari, I.B.M. Research Rep., RC - 1 9 4 6 , ( 1 9 6 7 ) . 1 2 . E. Underwood, J. of Met., 1 4 , 9 1 5 , ( 1 9 6 2 ) . 1 3 . C. M. Packer and 0. D. Sherby, A.S.M. Trans. Quart., 6 0 , 2 1 , ( 1 9 6 7 ) . 1 4 . R. N. Stevens, Met. Rev., 1 1 , 1 2 9 , ( 1 9 6 6 ) . 1 5 . R. C. Gifkins, J. of Am. Cer. Soc, 5 1 , 6 9 , ( 1 9 6 8 ) . 1 6 . D. McLean, Rep. on Prog, in Physics, 29^, 1 , ( 1 9 6 6 ) . 1 7 . D. McLean, Trans. A.I.M.E., 2 4 2 , 1 1 9 3 , ( 1 9 6 8 ) . 1 8 . J. J. Jonas, C. M. Sellars, and W. J. McG. Tegart, Met. Rev., 1 3 0 , ( 1 9 6 8 ) . 1 9 . D. L. Holt, Trans. A.I.M.E., 2 4 2 , 2 5 , ( 1 9 6 8 ) . 2 0 . D. H. Avery and W. Backofen, A.S.M. Trans. Quart., 5iB, 5 5 1 , ( 1 9 6 5 ) . 2 1 . J. L. Walter and H. E. Cline, Trans. A.I.M.E., 2 4 2 , 1 8 2 3 , ( 1 9 6 8 ) . 129 22. T. H. Alden and H. E. C l i n e , Trans. A.I.M.E., 239,710, (1967). 23. T. H. Alden and H. W. Schadler, Trans. A.I.M.E., 242, 825, (1968). 24. W. H. McCarthy, J . C. Schyne and 0. D. Sherby, Trans. A.S.M., j62, 117, (1969). 25. J . A. Lund and D. Tromans, A.S.M. Trans. Quart., j)9, 672, (1966). 26. A. K e l l y , "Fibre Reinforcement of Metals", Her Majesty's Stationery O f f i c e , London, (1965). 27. N. R. Risebrough and J . A. Lund, Trans. A.S.M., 61, 722, (1968). 28. E. H. Rennhack and G. P. Conard, Trans. A.I.M.E., 236, 694, (1966). 29. E. H. Rennhack and G. P. Conard, Trans. A.I.M.E., 236, 1441, (1966). 30. E. A. Anderson, E. J . Boyle and P. W. Ramsey, Trans. A,I.M.E., 156, 278, (1944). 31. E. Gebhardt, Z. Metallk., 33, 335, (1941). 32. W. Heine and U. Zwicker, Z. Metallk., 53, 380, (1962). 33. E. H. Rennhack, Trans. A.I.M.E., 236, 941, (1966). 34. T. W. Watson and G. H. Laurie, Cominco Internal Report, (1964). 35. Cominco, Internal Report, "The E f f e c t s of Extrusion Temperature on the Extrusion Textures of a Zn-0.16 wt.% Titanium A l l o y " , (1963). 36. J . L. Walker, "Liquid Metals and S o l i d i f i c a t i o n " , (A.S.M. Cleveland, Ohio), (1958). 37. G. W. Groves and;A. K e l l y , P h i l . Mag., 8^ 877:, (1963). 38. G. W. Groves, P h i l . Mag., 18, 977, (1969). 39. U. F. Kocks and D. G. Westlake, Trans. A.I.M.E., 239, 1107, (1967). 40. W. J . McG. Tegart, P h i l . Mag., 9, 339, (1964). 41. U. F. Kocks, Acta Met., 6, 85, (1958). 42. K. C. Donaldson, pr i v a t e communication. 43. K. E. Put t i c k , Acta. Met., 13, 1043, (1964). .. 44. T. H. Alden, Acta. Met., 17, 1435, (1964). 45. F. Garofalo, "Fundamentals of Creep and Creep Rupture i n Metals", (MacMillan, New York), 184, (1965). 130 46. C. S. Barrett and T. B. Massalski, "Structure of Metals", (McGraw- H i l l , New York), 47, (1966). 47. G. Woodford, Trans. A.S.M., 62, 291, (1969). 48. C. R. D'Antonio et a l . , Trans. A.I.M.E., 242, 2295, (1968). 49. D. M. Turner, unpublished. 50. H. Conrad, J. of Met., 16, 582, (1964). 51. J. D. Lubahn and R. P. Felgar, "Plasticity and Creep of Metals", (Wiley, New York), 291, (1961). 52. R. Armstrong et a l . , Phil. Mag., ]_, 45, (1962). 53. J.C.M. L i , Trans. A.I.M.E., 227, 239, (1963). 54. D. V. Wilson and J. A. Chapman, Phil. Mag., 8, 1543, (1963). 55. R. C. Cook, M.A.Sc, Thesis, University of British Columbia, (1969). 56. W. A. Backofen and D. S. Fields, Trans. A.S.M., 51, 946, (1959). 57. R. L. Bell and R. W. Calm, Phil. Mag., 239, 494, (1957). 58. H. Gleiter, Acta. Met., 17, 565, (1969). 59. M. F. Ashby et a l . , Trans. A.I.M.E., 245, 413, (1969). 60. H. J. McQueen, J. of Met., 20, 31, (1968). : 61. J. J. Jonas, et a l . , Met. Rev., 130, 1, (1969). 62. H. E. Evans and R. P. Skelton, Met. Sc. J., _3» 1 5 2> (1969). 63. R. C. Gifkins et a l . , J. Aust. J. of Met., 11, 275, (1966). 64. R. Lagneborg, J. of Mat. Sc., _3> 596, (1968). 65. D. A. C. Williams et a l . , Sricp. Met., 3.> 1 1 7> 1969. 66. L. Birk, "Electron Probe Microanalysis", (Interscience, New York), (1963). 67. T. H. Alden, Acta Met., 15, 469, (1967).

Cite

Citation Scheme:

    

Usage Statistics

Country Views Downloads
France 9 0
United States 4 0
Germany 4 7
India 1 0
China 1 47
City Views Downloads
Unknown 14 7
Redmond 2 0
Mumbai 1 0
Beijing 1 1
Ashburn 1 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}

Share

Share to:

Comment

Related Items