UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Ferrite grain refinement in dual phase steels Hazra, Sujoy S. 2006

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2006-0048.pdf [ 11.53MB ]
Metadata
JSON: 831-1.0078761.json
JSON-LD: 831-1.0078761-ld.json
RDF/XML (Pretty): 831-1.0078761-rdf.xml
RDF/JSON: 831-1.0078761-rdf.json
Turtle: 831-1.0078761-turtle.txt
N-Triples: 831-1.0078761-rdf-ntriples.txt
Original Record: 831-1.0078761-source.json
Full Text
831-1.0078761-fulltext.txt
Citation
831-1.0078761.ris

Full Text

FERRITE GRAIN REFINEMENT IN DUAL PHASE STEELS by Sujoy S. Hazra B . Eng., National Institute of Technology, Durgapur, India, 1998 M . Tech., Indian Institute of Technology, Kharagpur, India, 2000 A THESIS S U B M I T T E D I N P A R T I A L F U L F I L L M E N T O F T H E R E Q U I R E M E N T F O R T H E D E G R E E O F M A S T E R OF A P P L I E D S C I E N C E in T H E F A C U L T Y O F G R A D U A T E S T U D I E S ( M A T E R I A L S E N G I N E E R I N G ) T H E U N I V E R S I T Y OF B R I T I S H C O L U M B I A November 2005 © Sujoy S. Hazra, 2005 11 A B S T R A C T Dual phase steels with microstructure of hard martensite islands dispersed in ductile ferrite matrix, possess an optimum balance of strength and ductility alongwith a high work hardening rate. With this attractive combination of mechanical properties, dual phase steels are the candidate materials for structural and reinforcement auto-body parts, where refining the ferrite-martensite microstructure is expected to further improve the properties. In order to produce fine-grained dual phase steels, this study examines refinement of ferrite grain size using Deformation Induced Ferrite Transformation (DIFT) technique in two low carbon microalloyed steels, one with combined addition of Nb & M o and the other with only Nb . In this thermo-mechanical processing technique, the steels have been rapidly cooled from austenitization temperature to approximately 25-50 ° C above austenite to ferrite transformation start temperature without prestrain, to produce highly undercooled austenite, followed by heavy deformation, and subsequent rapid cooling thereby facilitating transformation to dual-phase microstructure consisting of fine grained ferrite with martensite. The effect of prior austenite grain size, degree o f undercooling, amount of deformation and microalloying additions on the final microstructure of the steels has been studied with tests performed on a Gleeble 3500 thermomechanical simulator. Microstructures have been quantitatively characterized with Scanning Electron Microscopy and to a limited extent by E B S D technique. The maximum ferrite grain refinement has been observed at the highest amount of deformation employed with a true strain of 0.6 for austenitization temperature of 950 ° C . I l l Here, the microstructure consists o f at least 60% ultrafine ferrite (UFF) grains with a mean grain size of 1-2 um and rest martensite. The results indicate that increase in undercooling leads to an increase in U F F fraction, although the effect of the same on U F F grain size is insignificant. Strain rate seems to have a secondary effect on the DIFT potential. Further, though the post deformation isothermal hold after highest employed deformation at the highest undercooling did not register any significant change in terms of U F F grain size and fraction in the final dual phase microstructure, transition from a quasi-polygonal to polygonal U F F morphology has been observed. The combined addition of M o and N b leads to enhanced retardation of the austenite to ferrite transformation. Hence the single addition of N b seems to be more beneficial in terms of obtaining optimum U F F fraction in fine grained D P steel. iv Table of Contents Abstract i i Table of Contents iv List o f Tables v i i i List o f Figures xiv Lis t o f Abbreviations xv List o f Symbols xv i i Acknowledgements x ix Chapter 1 Introduction 1 Chapter 2 Literature Review 4 2.1 Grain refinement in steels 4 2.2 Thermomechanical processing of austenite and resulting ferrite grain size 5 2.2.1 Strain free austenite 6 2.2.2 Work hardened austenite 8 2.3 Various routes o f producing Ultrafine Ferrite (UFF) 12 2.3.1 Equal Channel Angular Pressing ( E C A P ) ...13 2.3.2 Accumulative R o l l Bonding ( A R B ) 15 2.3.3 Deformation and annealing of Martensite start microstructure 17 2.3.4 Asymmetric rolling 18 V 2.3.5 Deformation Induced Ferrite Transformation (DIFT) : 19 2.4 Factors affecting D I F T 21 2.4.1 Strain level : 21 2.4.2 Prior austenite grain size 24 2.4.3 Degree of undercooling 26 2.4.4 Strain rate 29 2.4.5 Post deformation cooling rate 30 2.4.6 Post deformation isothermal holding 30 2.4.7 Steel chemistry 31 2.5 Mechanical properties of U F F steels 32 2.5.1 Balance between strength, ductility and work hardening rate 32 Chapter 3 Research Objectives 37 Chapter 4 Experimental Methods 39 4.1 Steels investigated 39 4.2 Experimental equipment 39 4.2.1 Gleeble 3500 Thermomechanical Simulator 39 4.3 Experimental methodology 40 4.3.1 Prior Austenite Grain Sizes (PA G S) 40 4.3.2 Detection o f T s 41 4.3.3 Static isothermal ferrite transformation tests 43 4.3.4 DIFT tests 43 4.4 Microstructural investigation 45 4.4.1 Optical microscopy 45 v i 4.4.2 Scanning electron microscopy 47 4.4.3 Orientation imaging microscopy 48 4.4.4 Scanning transmission electron microscopy 48 4.5 Hardness measurements 49 Chapter 5 Results 50 5.1 Initial austenite grain size measurements 50 5.2 Precipitates in the austenitized condition 52 5.3 Continuous Cooling Transformation (CCT) test results 54 5.3.1 Determination of 5% transformation start temperatures 54 5.3.2 Microstructures obtained after C C T tests 56 5.4 Ferrite grain size obtained after static isothermal transformation tests 58 5.5 DIFT tests for producing fine grained dual phase steels 59 5.5.1 Effect o f strain 61 5.5.2 Effect of P A G S 63 5.5.3 Effect o f undercooling 65 5.5.4 Effect of strain rate 69 5.5.5 Effect o f post deformation isothermal hold 71 5.6 Microstructural investigation using Orientation imaging microscopy 72 5.7 Relation between ferrite grain size, fraction of second phase and hardness 78 v i i Chapter 6 Discussion 79 6.1 Refinement of ferrite grain size in dual phase steels 79 6.1.1 Effect of P A G S . . . . , 80 6.1.2 Effect o f deformation condition 81 6.1.3 Effect of steel chemistry 84 6.2 Grain refining ability of DIFT technique 84 6.3 Microstructural refinement and concurrent increase in strength 85 Chapter 7 Conclusions and Future work 87 7.1 Conclusions 86 7.2 Future work 88 References 90 v i i i LIST OF TABLES Table Title Page Nos. 2.1 The microstructural and tensile characteristics of dual phase steels 36 4.1 Chemical compositions of steels investigated (wt. %) 39 5.1 Heat treatment schedules and measured austenite grain sizes 50 5.2 Selected austenitizing temperatures and measured 5 % transformation 55 start temperatures 5.3 Ferrite grain size and fraction after isothermal transformation tests 59 5.4 Deformation Induced Ferrite Transformation (DIFT) test matrix for N b + 60 M o steel 5.5 Deformation Induced Ferrite Transformation (DIFT) test matrix for Nb 60 steel 5.6 Final ferrite grain size and fraction resulted after DIFT test with e = 0.6 67 5.7 Final ferrite grain size and fraction resulted after DIFT tests with ds / dt = 70 10 s - 1 5.8 Final ferrite grain size and fraction resulted after DIFT test with highest 72 strain (0.6) and post deformation isothermal hold 5.9 Prediction of 8% flow stress based on Tabor's relation 78 i x LIST OF FIGURES Figure Title Page Nos. 2.1 Relationships between the ferrite grain size and the strength and 5 toughness of ferritic steels as processed via different routes 2.2 Relationship between the conversion ratio of y/a and austenite grain size 7 in hot rolled and reheated plain carbon steels 2.3 (a) Relationship between the ferrite grain size and cooling rate with 10 increasing levels of retained strain (b) Evolution of ferrite grain size with accumulated strain at two cooling rates 2.4 Relationship between the ferrite grain size and austenite grain size at 11 progressively higher cooling rates at various levels of retained strains 2.5 Ferrite fraction vs. cooling rate for dr (vol.) of (a) 16 urn and (b) 32 jam, 12 respectively 2.6 Schematic diagram of an E C A P die 14 2.7 Schematic illustration of A R B process 16 2.8 Schematic illustration of DIFT (same as DSIT) at isothermal deformation 20 temperature 2.9 Schematic representation of the variation of the transformation start curve 22 as a function of strain 2.10 (a) Volume fraction of U F F as a function of strain at austenitization and 24 deformation temperature of 1200 and 775 ° C respectively for a C - M n - V steel (b) Ferrite grain size as a function of strain at austenitization temperatures of 1200 (Coarse grain) & 900 ° C (Fine grain) and deformation temperature of 775 ° C for the same steel as in (a) 2.11 Effect o f deformation temperature on U F F grain size obtained in low C 27 steels through DIFT processing 2.12 Variation of ferrite volume fraction with degree of undercooling for 28 different amounts o f applied strain in a plain low carbon steel 2.13 Variation of critical amount of reduction with degree of undercooling for 28 plain low carbon steel 2.14 Variation of ferrite volume fraction with isothermal post deformation 31 holding time 2.15 Grain size dependence of strength for an Ti-IF steel processed through 33 A R B and annealing route 2.16 Dependence of yield and tensile strength on grain size for industrially 33 processed steel strips 2.17 Stress-strain curves for industrially processed steel strips with different 34 grain sizes XI 4.1 Schematic of C C T test set-up for detecting the transformation start 41 temperature: (a) full jaw set; (b) set-up of specimen 4.2 Schematic of a typical dilation response during continuous cooling 42 transformation 4.3 Specimen geometry for Deformation Induced Ferrite Transformation 43 (DIFT) tests 4.4 Schematic thermal path used for the DIFT tests 44 5.1 Austenite microstructure after reheating to (a, c) 950 ° C and (b, d) 1100 51 ° C , with a holding time of 120 s and water quenching, for Nb + M o and Nb Steel respectively 5.2 ( a )TEM replica micrograph and (b) E D X spectrum of Nb-rich precipitate 52 at 950 ° C in the Nb + M o steel 5.3 (a) T E M replica micrograph and (b) E D X spectrum, showing co- 53 precipitation of Nb-rich precipitate at austenite grain boundary onto A1N precipitate at 950 ° C in the Nb + M o steel 5.4 (a) T E M replica micrograph and (b) E D X spectrum, of Nb-rich 54 precipitate at 950 ° C in the Nb steel 5.5 Detection of T o.os temperatures for selected prior austenite grain sizes in 55 N b + M o and Nb steels with a cooling rate of 40 °C/s 5.6 Microstructure after continuous cooling tests for Nb + M o steel for 56 austenite grain sizes o f : (a) 10 um and (b) 26 jam respectively x i i 5.7 Microstructure after continuous cooling tests for Nb steel for austenite 57 grain sizes o f : (a) 11 urn and (b) 25 u.m respectively 5.8 Microstructure after isothermal holding at 627 ° C for 20 mins. and 58 subsequent natural cooling for Nb + M o steel with austenite grain size of 10 urn 5.9 Microstructure after isothermal holding at 672 ° C for 20 mins. and 58 subsequent natural cooling for N b steel with austenite grain size of 11 urn 5.10 Scanning Electron micrographs after DIFT tests (strain rate = Is"1) of N b 61 + M o steel for austenite grain size of 10 um at deformation temperature of 627 ° C at strain level o f : (a) 0.2, (b) 0.4 and (c) 0.6 respectively 5.11 Scanning Electron micrographs after DIFT tests (strain rate = Is"1) of Nb 62 steel for austenite grain size of 11 urn at deformation temperature of 672 ° C at strain level o f : (a) 0.2, (b) 0.4 and (c) 0.6 respectively 5.12 Fraction of U F F as a function of applied strain in Nb + M o and N b steels 63 respectively for smaller P A G S and the highest undercooling 5.13 Scanning electron and optical micrographs respectively, after DIFT tests 63 of N b + M o steel at deformation temperatures of 627 ° C and 553 ° C for austenite grain sizes o f : (a) 10 and (b) 26 um , at s = 0.2 5.14 Scanning electron and optical micrographs respectively, after DIFT tests 64 of Nb steel at deformation temperatures of 672 ° C and 529 ° C for austenite grain sizes o f : (a) 11 and (b) 25 urn , at s = 0.2 X l l l 5.15 Scanning electron micrographs after DIFT tests of Nb + M o steel at 64 deformation temperatures of 627 ° C and 553 ° C for austenite grain sizes o f : (a) 10 and (b) 26 urn respectively, at s = 0.6 5.16 Scanning electron micrographs after DIFT tests of Nb steel at 65 deformation temperatures of 672 ° C and 529 ° C for austenite grain sizes o f : (a) 11 and (b) 25 urn respectively, at s = 0.6 5.17 Scanning Electron micrograph after DIFT tests o f N b + M o (a, b) and N b 66 steel (c, d) respectively, at undercooling of: (a) 200 ° C , (b) 175 ° C , (c) 1 5 4 ° C a n d ( d ) 129 ° C 5.18 Ferrite grain size distribution after DIFT tests with d Y = 10 um, 8 = 0.6, 68 ds / dt = 1 s"1 and corresponding Log-Normal fit for Nb + M o steel at the following conditions: (a) T d e f = 627 ° C , A v g . Ferrite grain size = 1.9 ± 0.6 ^m; (b) T def = 652 ° C , A v g . Ferrite grain size = 2.2 ± 0.6 um 5.19 Ferrite grain size distribution after DIFT tests with d Y = 11 um, e = 0.6, 68 ds / dt = 1 s"1 and corresponding Log-Normal fit for N b steel at the following conditions: (a) T d e f = 672 ° C , A v g . Ferrite grain size = 1.3 ± 0.5 um; (b) T d e f = 697 ° C , A v g . Ferrite grain size = 1.1 ± 0.6 um 5.20 Scanning Electron micrograph after DIFT tests (ds / dt = 10s"1) of N b + 70 M o (a, b) and Nb steel (c, d) respectively, at undercooling of: (a) 200 ° C , (b) 175 ° C , (c) 154 ° C and (d) 129 ° C 5.21 Scanning Electron micrograph after DIFT tests with small austenite grain 71 xiv size employing a strain of 0.6 and post deformation isothermal hold for (a) Nb + M o steel and (b) Nb steel, at 627 ° C and 672 ° C , respectively 5.22 (a) Band contrast map from an area of 20 X 20 um represented by Fig . 74 5.17 (a) for Nb + M o steel (dY = 10 um, T d e f = 627 °C , ds / dt = Is" 1 ), (b) Misorientation angle distribution of the same map as in (a) 5.23 (a) Band contrast map from an area of 20 X 20 um represented by Fig . 75 5.17 (b) for Nb + M o steel (dy = 10 um, T d e f = 652 ° C , de / dt = Is"1), (b) Misorientation angle distribution of the same map as in (a) 5.24 (a) Band contrast map from an area of 20 X 20 urn represented by Fig . 76 5.17 (c) for N b steel (dy = 11 um, T d e f = 672 °C , de / dt = Is"1), (b) Misorientation angle distribution o f the same map as in (a) 5.25 (a) Band contrast map from an area o f 20 X 20 um represented by Fig . 77 5.17 (d) for Nb steel (dy = 11 um, T d e f = 697 ° C , ds / dt = Is"1), (b) Misorientation angle distribution of the same map as in (a) 6.1 Variation of % U F F formed in the final microstructure with the 83 undercooling employed for smaller austenite grain size at strain of 0.6 and strain rate of Is"1 XV LIST OF ABBREVIATIONS A R B Accumulative roll bonding A S T M American Society for Testing o f Materials B C C Body Centered Cubic - a type of crystal structure B H Bake Hardenable C A N M E T - Canada Center for Mineral & Energy Technology Center- Materials M T L Technology Laboratory, Ottawa, Canada C C T Continuous Cooling Transformation C R M Centre for Research in Metallurgy, Belgium C S M Centro Sviluppo Materiali S.p.A., Italy C W C Curtain wal l cooling DIFT Deformation Induced Ferrite Transformation D P Dual Phase D R X Dynamic recrystallization DSIT Dynamic Strain Induced Transformation (same as DIFT) % E l Elongation occurred in a tensile test expressed in percentage E B S D Electron Backscattered Diffraction E C A P Equal Channel Angular Pressing E D X Energy dispersive X-ray analysis x v i E Q A D Equivalent area diameter F C C Face Centered Cubic - a type of crystal structure H I P E R S 21 Development of High Performance Structural Steels for 21st century, South Korea H S L A High Strength L o w A l l o y I E H K - Dept. of Ferrous Metallurgy, R W T H , Aachen University, Germany R W T H IF-HS Solution strengthened Interstitial Free P A G S Prior austenite gain size R O T Run out table S E M Secondary electron microscopy S P D Severe plastic deformation S R D D Single roll drive with different diameter S T E M Scanning transmission electron microscopy T E M Transmission electron microscopy U F G Ultrafine grained U T S Ultimate tensile strength X R D X ray diffraction Y S Y i e l d strength XVII LIST OF SYMBOLS r j y Y ie ld strength of the steel ao Friction stress representing the overall resistance of the crystal lattice to the dislocation movement k Strengthening coefficient d Average grain size da Average ferrite grain size dy Average austenite grain size y Symbol denoting austenite a Symbol denoting ferrite s True strain s c Critical true strain required for initiation of dynamic austenite to ferrite transformation EC.DIFT Critical strain for initiation of DIFT phenomenon SC,UFF Critical strain for prevalence of U F F in final microstructure (higher than £C,DIFT) AT3 Austenite to ferrite transformation temperature while cooling Ae3 Austenite & ferrite equilibrium temperature T s Transformation start temperature corresponding to 5 % transformed x v m fraction when austenite is continuously cooled at a cooling rate of 40 °C/s X (T) Fraction transformed dp (T) Extrapolated dilation from the product region dm (T) Measured dilation dy (T) Extrapolated dilation from the austenite region ay thermal expansion coefficient for austenite phase ap thermal expansion coefficient for product phase Tf Transformation frinish temperature corresponding to 95 % transformed fraction when austenite is continuously cooled at a particular cooling rate Tdef Deformation temperature in the DIFT schedule A T Difference between A e 3 and Tdefi.e.Ae3-Tdef xix A C K N O W L E D G E M E N T S I would like to express sincere gratitude to my supervisor, Dr. Matthias Mil i tzer for his keen insight, wise guidance and encouragement throughout the course of this work. I wish to thank Dr. W . J. Poole and Dr. C. W. Sinclair, respectively, for making me conversant with the Gleeble 3500 equipment and the E B S D technique. I would also like to thank Mrs. M . Mager for providing me the required training on T E M . Many thanks to Dr. F. Fazeli; it was a rewarding experience for the technical discussions I had with him. Heartfelt thanks are also due to Krishnendu Mukherjee for his help on innumerable instances. The financial support received from the National Centre of Excellence (NCE) A U T 0 2 1 and the Natural Sciences and Engineering Research Council o f Canada is acknowledged with gratitude. I would like to convey sincere appreciation to Tata Steel, my present employer, for granting me the required study leave. The sacrifice and endurance of my wife and parents are greatly treasured. This thesis is dedicated to my beloved maternal uncle, late Dr. Bijoy Sarkar. 1 Chapter 1 Introduction Iron is one of the oldest materials in the world with its first usage reportedly dating back to 4000 B C . Today steel (the carbon alloy of iron) finds application in every imaginable facet of our life. Steel is an irreplaceable structural material widely used for many decades for building constructions, bridges, automobiles, ships, and so on. Rapid advances in the steel making technologies in compliance with increased level o f environmental consciousness has resulted in the development of novel grades of steels, with superior combination of properties, such as higher strength, toughness, weldability etc. In particular for the automotive sector, the current production-ready steels include bake-hardenable (BH), rephosphorized (RP), and solution-strengthened interstitial-free (IF-HS) steel. Near term solutions include advanced high strength steels like dual-phase (DP), complex-phase (CP), and transformation-induced plasticity (TRIP) steels. The well-known inverse relationship between strength and grain size in polycrystalline aggregates has provided the impetus for grain refinement in steels. This has fuelled enormous Research and Development work worldwide to achieve microstructures with an ultrafine grain size especially in, Europe (Corus, Manchester, C R M , C S M , I E H K - R W T H ) , Japan (Ferrous super metal project and S T X 21 -Structural Materials for the 21st century), Korea (HIPERS 21-Development of High Performance Structural Steels for 21 s t century), Australia (Bluescope Steel-Deakin) and China (New Generation Steels Project). It is to be noted here that the aforementioned Chapter 1 INTRODUCTION 2 national projects are primarily focussed on plain carbon steels to be used in the construction sector. Although thermo-mechanical controlled processing has always been the general approach for grain refinement in steels, the recent advent of innovative and exotic severe plastic deformation techniques allows the grain size of steels refined to the sub-micrometer level. This has facilitated in achieving ultrahigh strength when compared to their coarse-grained counterparts. However, in general, the ultrafine grained steels suffer from an inherent drawback, i.e. a lack of strain hardening, which is undesirable for any structural application. Not only ferritic steels, but also ultrafine grained ferrite-pearlite steels are no exception to this demerit. In this regard dual phase (DP) steels with hard martensite islands dispersed in a soft ferrite matrix, are manifested by attractive mechanical properties like continuous yielding, low yield ratio, high initial work hardening rate and high uniform elongation. It is however noteworthy that most of the proposed processing routes for laboratory scale production of ultrafine grained steels have little commercial viability for industrial production due to the extremely high amount of deformation required {i.e. equivalent true strain > 5). Hence the present work is aimed at achieving a dual phase microstructure consisting of hard martensite islands dispersed in ultrafine ferrite (UFF) matrix following a deformation induced ferrite transformation (DIFT) processing route, employing an order of magnitude lesser amount of strain, in two low carbon microalloyed steels. The thesis consists of seven chapters. Chapter 2 is a review of literature pertaining to the refinement of ferrite grains in steels. The potential benefits of ultrafine microstructure have been emphasized. Starting Chapter 1 INTRODUCTION 3 with traditional thermo-mechanically controlled processing of steels, several commonly used methods to obtain ultra-refinement o f grains have been presented. One of the methods amongst these, namely, Deformation Induced Ferrite Transformation (DIFT) technique has been critically reviewed. Finally, limitations of ultrafine fine ferritic microstructure in terms of mechanical properties have been indicated and current research efforts in mitigating the same have been highlighted. In Chapter 3 the research objectives have been elucidated and subsequent strategies in achieving the same have been presented. In Chapter 4, experimental details such as steels used in this study and DIFT test procedures are provided. Optical and electron microscopy analysis o f the resulting microstructures have also been described. Chapter 5 contains the results from this investigation which indicates DIFT technique as a prospective method of producing fine grained dual phase steels. In Chapter 6, the results presented in the preceding chapter are analysed and discussed in detail in the light of dynamic transformation of austenite to ferrite which occurs at suitable thermo-mechanical conditions. The potential benefits of ferrite grain refinement in dual phase steels have been delineated. Chapter 7 draws some relevant conclusions from this study alongwith proposition for further work in this direction. 4 Chapter 2 Literature Review 2.1 Grain refinement in steels It is widely known that high strength steels with excellent impact toughness find many applications, especially in heavy structures, where they can offer economical benefits by reducing both material and fabrication costs. Among various strengthening mechanisms, grain size refinement is the only mechanism which concurrently improves strength and impact toughness. The Hall-Petch relation [1, 2] describes the effect of grain size on the yield strength, i.e. a y = a 0 + k d " 1 / 2 (2.1) where, a y is the yield strength o f the steel, ao is the "friction stress" and it represents the overall resistance of the crystal lattice to the dislocation movement, k is the strengthening coefficient, and d is the grain size. Reducing the ferrite grain size to about 1 jam (i.e. Ultra Fine Ferrite - U F F ) has been reported to increase the yield strength by almost 350 M P a , compared to a plain carbon steel with ferrite grain size of 4 - 5 um, which is regarded as the achievable grain size limit for the thermo-mechanically processed ( T M C P ) steels [3, 4] as schematically depicted in Fig . 2.1. i Chapter 2 LITERATURE REVIEW Ferrite Grain Size, d(ptm) Fig. 2.1 Relationships between the ferrite grain size with the yield strength and the ductile-to brittle transition temperatures of ferritic steels as processed via different routes [3] Although significant achievements have been attained at the laboratory / pilot m i l l scale by various processing techniques, mass scale production of ultrafine grained steel remains a challenge. However, it has been reported that in Japan, Nippon Steel has introduced a plate product to the market with a claimed 2-3 um grain size in the surface layer for superior crack arrest properties [5] and, Nakayama Steel Works is commercially producing a strip product of 2 mm thickness with a grain size of about 2 - 5 um [6, 7]. 2.2 Thermomechanical processing of austenite and resulting ferrite grain size Thermo-Mechanical Controlled Processing i.e. controlled rolling and accelerated cooling, has been utilized for more than two decades now for the manufacturing of High Strength L o w A l l o y steel ( H S L A ) plates and sheets with a yield strength up to 500 M P a Chapter 2 LITERATURE REVIEW 6 (sheets up to about 700 MPa) and excellent impact toughness as well as weldability. The microstructure of these steels is essentially polygonal ferrite with a grain size of 4 um at the minimum, and in some cases 2 - 3 um with quasi-polygonal grain structures. The controlling factors on ferrite grain size are well documented in the literature, and are linked with the austenite grain size in the reheating process, recrystallization during the rough rolling operation, accumulation of strain in the finishing rolling stage in the non-recrystallization regime o f austenite, and the austenite decomposition in the accelerated cooling stage on the Run Out Table (ROT). Similar phenomena are also believed to occur to some extent in the formation of U F F microstructures so that they are briefly discussed in the following. 2.2.1 Strain free austenite In the case of the transformation of ferrite from recrystallized, strain-free austenite (i.e. recrystallization controlled rolling), only prior austenite grain boundaries provide the nucleation sites for the ferrite phase. The transformed ferrite grain size (da) is affected by the cooling rate, the prior austenite grain size (dy), and the type of nucleation sites (grain corners, edges or boundary faces). However, as it can be understood, there exists a maximum for the cooling rate, for i f it is exceeded, the austenite w i l l transform into other transformation products along with ferrite, i.e. low temperature transformation products like, bainite and/ or martensite, desirable for novel advanced high strength steels. Although the ferrite grain size decreases with decreasing austenite grain size, it reaches a limiting value of around 10 um, when the austenite grain size becomes about Chapter 2 LITERATURE REVIEW 7 10 um, as shown in Fig. 2.2 [8]. Consequently, there is a limit in attaining the ferrite grain refinement through the y - » a transformation from recrystallized, strain-free austenite. y grain size (um) c 15 20 30 40 50 75 100 150 200 300 o .2 4 D c 3 o '(/> <u _ £ 2 o o ? . O I0 9 8 7 6 5 4 3 2 1 Y grain size (ASTM number) Fig . 2.2 Relationship between the conversion ratio of y/a and austenite grain size in hot rolled and reheated plain carbon steels [8] However, a variation of this process was developed by Jonas and co-workers [9] where the aim was to further refine the austenite by dynamic recrystallization. B y reducing the temperature and using typical strain rates of hot strip, rod and bar mills it appears possible to produce austenite that is much finer than that obtained by recrystallization controlled rolling [10]. This phenomenon has been judiciously used to the complete exclusion of static recrystallization by Subramanian et al. [11] for obtaining a claimed ultrafine ferrite microstructure with an average grain size o f 1.5 um. In their study on controlling the time evolution of strain induced precipitation of N b C and strain accumulation through precipitate interaction with recovery and recrystallization at short Chapter 2 LITERATURE REVIEW 8 inter-pass times characteristic o f strip rolling, the achievement o f ultrafine ferrite grains has been attributed to recovery followed by delayed strain induced precipitation due to higher M n content (1.8 wt%) in a C-Mn-Nb steel. Nevertheless, it is noteworthy that precise control of deformation and temperature limits the practical application of this technique. 2.2.1 Work hardened austenite The aforementioned limit in obtaining ferrite grain size can be extended to a considerably much lower value concomitantly with predominant ferritic microstructure, by utilizing the full potential o f the transformation from work hardened austenite to ferrite. For the plastic deformation in the non-recrystallization temperature regime, the effect o f resulting retained strain on the austenite to ferrite transformation kinetics is generally acknowledged as an accelerating effect. This has been observed in the different parameters that reflect the transformation kinetics. More specifically, in cooling experiments both the transformation start temperature and the transformation finish temperature have been observed to increase, the temperature range o f the transformation to contract, the final fraction of polygonal ferrite to increase and the final ferrite grain size to decrease. Umemoto et al. [12] postulated that three different mechanisms may affect ferrite nucleation, whilst a fourth mechanism affects ferrite grain growth. The mechanisms proposed by Umemoto et al. are: (1) accelerated nucleation kinetics as a result o f an increase in grain boundary area; (2) increased nucleation potency at grain boundaries via the introduction of grain boundary ledges leading to localized increases in Chapter 2 LITERATURE REVIEW 9 grain boundary energy; (3) additional nucleation sites due to a deformation substructure and, (4) accelerated growth rates arising from the stored energy o f deformation. The first three mechanisms help overcome the energy barrier for nucleation leading to a reduction of the undercooling required to bring about nucleation in a cooling experiment, and possibly also to an increase in the nucleation site density. These mechanisms thus provide plausible explanations for the observations of increased transformation start and finish temperature, and also for the refinement of microstructure. The stored energy that gives rise to mechanism (4) essentially increases the driving pressure for the transformation and may explain the contraction of the transformation range. In terms of microstructure, the stored energy o f deformation is not only a function of the dislocation density but is also dependent upon the nature of the substructure in which the dislocations are stored [13]. Although experimental observations which suggest that deformation affects the nucleation behaviour are commonly reported, there is little experimental evidence o f the influence of plastic deformation on the growth rates. Even though it is often assumed that the effect on growth kinetics is negligible it remains nevertheless ambiguous whether this is broadly true for all conditions of strain, undercooling and nucleation pattern [13]. The dependence of ferrite grain size on the increasing amount of retained strain in austenite and on the cooling rate is shown in Fig . 2.3 for similar prior austenite grain size [14, 15]. A s can be clearly noticed the amount of deformation below the non-recrystallization temperature has strong influence on the final microstructure. From Fig . 2.3 (a) it can be seen that for a constant cooling rate the ferrite grain size decreases with increasing amount of accumulated strain and similar to the earlier reported results of Mil i tzer et al. [16], the effect of retained strain is relatively small at higher cooling rates. Chapter 2 LITERATURE REVIEW 10 Nevertheless, it can be concluded from Fig. 2.3 (b) that for a constant cooling rate the effect of increasing strain is high in the range of lower level of accumulated strain and tends to saturate for higher strains. (a) 24 n 25 « • * n "IS era -1 s» t i 5" ce R" « 8 3 4 Austenite grain size = 30 Jim <m—™# Bainite Eret = 0 eret = 1.0 erPf = 1.5 j eret = 2.0 (b) 20 1 5 . . £ si 10 4 8 12 tS Cooling rate, K/s 2 0 • stasisitncft) o St&eJ S 2 (1»C/S) • Steel St (5°Cte) \ SI: C-Mti-O.OBiNb S2:C-Bfci-0.03lM»-0rj8W 0.5 1 Sacc 1,5 Fig . 2.3 (a) Relationship between the ferrite grain size and cooling rate with increasing levels o f retained strain [14] and, (b) Evolution of ferrite grain size with accumulated strain at two cooling rates [15] Moreover, it can be concluded that increasing the cooling rate, in addition results into finer ferrite grain size. The dependence of ferrite grain size on the austenite grain size under accelerated cooling conditions and with retained strain from one of the aforementioned studies [14] is shown in Fig . 2.4. Chapter 2 LITERATURE REVIEW 11 14 12 10 Ferrite grain 6 size, urn g 4 2 0 150 120 90 60 30 0 Austenite grain size, um Fig . 2.4 Relationship between the ferrite grain size and austenite grain size at progressively higher cooling rates at various levels of retained strains [14] It can be seen that the cooling rate has a significant influence while the retained strain must be high to obtain a fine ferrite grain size even from a fine-grained austenite. Further, at a constant cooling rate, the effect o f retained strain is more pronounced for coarser austenite grains, similar to findings of the earlier investigations on a leaner chemistry C - M n - N b steel by Gibbs et al. [17] and a richer chemistry C - M n - N b steel by Pandi et al. [18] and later on, by Nakata et al. [19]. The principle of grain size refinement by accelerated cooling can be understood by considering the effect o f transformation temperature on ferrite grain size in an isothermal transformation. During the continuous cooling transformation, that can essentially be considered as the sum of short-time isothermal holdings at successive temperatures, the ratio of nucleation to growth rate increases with the decrease in transformation temperature, leading to a decrease in size o f Chapter 2 LITERATURE REVIEW 12 the resulting ferrite grains. The dependence of ferrite fraction in the final microstructure follows a similar trend as can be seen in Fig . 2.5 (a) and (b) where recently reported results of a study on austenite decomposition kinetics in a M o bearing D P steel displayed [20]. 1.0 are (b) g 0 8 g 0.6 20 40 m Cooling rate, °G& 0.4 0.2 0.0 £=0.5 —ii \ \ e = 0 . 2 1 • ' 60 20 40 60 Cooling rate, °C/s Fig . 2.5 Ferrite fraction vs. cooling rate for dy (vol.) of (a) 16 um and (b) 32 um, respectively [20]. It is noteworthy that very recently ultrafine ferrite grain sizes of approx. 2 um after transformation from pancaked austenite have been reported by Mili tzer and co-workers [21 - 23] in various commercial H S L A steels. Due to the very high undercooling accompanying the transformation from coarse grain pancaked austenite in these cases the ferrite morphology is, at least in part, quasi-polygonal and acicular in nature. 2.3 Various routes of producing Ultrafine Ferrite (UFF) There are several ways that can refine the grain size of ferrite beyond what can be obtained in the conventional T M C P , i.e. to the ultrafine scale (viz. order of 1-2 um or less), even though most of them are rather exotic in nature and collectively termed as Chapter 2 LITERATURE REVIEW 13 Severe Plastic Deformation (SPD) techniques (with true strains > 10). In the following a concise description of these techniques is presented. For the sake of brevity, results reported only pertaining to low carbon steels will be discussed. At the end features of the Deformation Induced Ferrite Transformation (DIFT) technique which has been adopted in the present study is presented briefly; the details of interaction of various metallurgical parameters in this phenomenon have been dealt with in a following separate section. i f 2.3.1 Equal Channel Angular Pressing (ECAP) The equal channel angular pressing (ECAP) technique was developed about two decades back by Segal and co-workers [24] to introduce severe plastic strain into material without causing any change in the cross-sectional area and has extensively been used for producing ultrafine grain sizes especially, in A l and M g alloys in the last few years [25-28]. Severe plastic straining is achieved in ECAP by pressing the sample through a die as shown schematically in Fig. 2.6 [29]. The sample is machined to fit within a channel which passes through the die in an L -shaped configuration (various angles can be used). Passing through the angle, the sample undergoes straining in shear. Because of no change in the cross-sectional dimensions, the process can be repeated as many times as desired in order to achieve very high total strains. It is also possible to rotate the sample between consecutive pressing passes so that different shearing systems are activated. Chapter 2 LITERATURE REVIEW 14 die Fig. 2.6 Schematic diagram of an ECAP die [25]. Most of the work on steels applying this technique has emerged in recent years [30-35]. The ECAP technique has been applied to Ti-stabilized Interstitial Free (IF) steel [30] and 0.15 wt% C steel [31], Grain sizes of 0.2 um in the IF steel and 0.3 um in plain low carbon steel were obtained. Ultrafine ferrite grains with an average diameter of 0.3 um were obtained in plain low-carbon steel by imposing severe plastic strain close to 8 while rotating the sample by 180° between each pass at 350°C [31]. At the initial stages of deformation the rotation of the material between the subsequent passes produced severe interactions between the slip systems that are typical in BCC structures. This produced ultrafine subgrains with serrated and low angle boundaries. By systematically studying the deformed microstructures with the help of XRD and TEM techniques, Shin et al. [31] argued that the ultrafine subgrains created during the initial passes then rotate in the following passes to create ultrafine grains with high angle grain-boundaries. They contented that subgrain rotation is more favourable than intragranular strain for Chapter 2 LITERATURE REVIEW 15 accommodating a large amount of strain since serrated boundaries restrict dislocation movement causing intragranular strain. This study elaborated upon the strain accommodation process at large strains and concomitant mechanism of high angle boundary formation which appears to be consistent with the previously observed trend of the grain boundary misorientation distribution in Ti stabilized IF steel [30]. In terms of mechanical properties, after an annealing treatment of the same plain carbon steel at 480 °C for 24 hours, with 88 % ferrite and the rest pearlite, the grain refinement resulted in YS = 713 MPa and % ef = 18 with little strain hardening [32] . Very interestingly, Park and co-workers [35, 36] have presented recently first results on ultrafine grained Dual Phase (DP) steels produced by ECAP of low C-l.lwt% Mn steel at 500 °C with an accumulated true strain level of 4, and subsequent annealing at 740 °C for 10 mins. followed by water quench. This process has resulted in about 72 vol% ferrite with average an grain size of 1.4 urn and the rest martensite with an average island size of 1.3 um, leading to a very attractive combination of mechanical properties, i.e. YS = 581 MPa, UTS = 978 MPa and %eu = 9. 2.3.2 Accumulative Roll Bonding (ARB) A novel plastic straining process named Accumulative Roll Bonding (ARB) has recently been proposed by Saito and co-workers [36, 37]. In the accumulative roll-bonding process (ARB), as shown in Fig. 2.7, a strip is placed on another strip of the same thickness and rolled to a 50% reduction. To ensure a good bonding, the surfaces of the strips are cleaned both mechanically and chemically. The material is then sectioned into two halves in length, stacked and rolled together. The whole process can be repeated Chapter 2 LITERATURE REVIEW 16 again and again. The proposed A R B process is being touted as a promising technique compared to E C A P in terms of commercial production of high strength bulk materials. Surface treatment Degneaaiftg/1 ^Wire brushing Stacking Cutting Heating Fig . 2.7 Schematic illustration of A R B process [37]. The process is generally conducted at a warm rolling temperature in a ferrite regime below the recrystallization temperature to accumulate strain in the ferrite. It has been reported that ultrafine grains of 0.4 -1 um have been obtained in Ti-stabilized IF steels after rolling at 500 ° C with accumulated true strain of 5.6 [37]. Mechanical property-wise this resulted in U T S = 870 M P a with very low %E1, approximately 1 - 2%. Based on the T E M analysis, it has been suggested that the formation of ultrafine grains occurs by grain subdivision under intense plastic straining in the work-hardened ferrite and by the subsequent grain boundary formation with large misorientation. It is interesting to note that even after applying heavy accumulated strain, the resulting microstructure consisted in part of subgrain structure. Chapter 2 LITERATURE REVIEW 17 2.3.3 Deformation and annealing of Martensite start microstructure Recently, Ueji et al. [38 - 40] have invented a process route in which starting with a martensite starting microstructure in a plain carbon steel (Fe-0.13 wt% C) , when cold-rolled at reductions o f 25-75%, produced a U F F microstructure by annealing it at temperatures between 500-600°C. In their study a plain low-carbon steel sheet with a lath-martensite microstructure has been used as starting material. It has been cold-rolled by 50% (equivalent true strain 0.8) and then annealed at 500°C. The final microstructure composed of ultrafine equiaxed ferrite grains (mean grain size of 180 nm), uniformly precipitated nano-cementite and tempered martensite [39]. Lath martensite has a three-level hierarchy in its morphology. (I) Lath: single crystal of martensite including high density o f lattice defects, (II) Block: aggregation of laths with same crystallographic orientation, (III) Packet: aggregation of the blocks having the same habit plane (for example, {111 }y). According to Morito et al, 83% of the inter block boundaries and inter packet boundaries can be regarded as high angle boundaries [41]. Hence martensite can be considered as a kind o f fine-grained structure subdivided by a number of high-angle boundaries (inter block and inter packet boundaries). Fine grained martensite structure does not only possess a high density of dislocations but also, accelerates grain subdivision. The presence of dislocation substructure along with solute carbon atoms enhances inhomogeneous deformation, which in turn results in rapid sub-grain division. Further, solute carbon atoms precipitate uniformly as fine carbides during warm rolling which inhibits grain growth of the ultrafine ferrite. A l l these factors facilitated in attaining a nano-structured final microstructure without intense straining. The final product had both high strength ( Y S = Chapter 2 LITERATURE REVIEW 18 710 M P a , U T S = 870 MPa) and adequate ductility (8% uniform and 20% total elongation). The main advantage of this process is that it does not need intense straining and hence it w i l l be easier to adapt it to practical use without any new metal working facilities. 2.3.4 Asymmetric rolling In case of asymmetric rolling the peripheral speed of the upper and lower rolls differ from each other. This results in severe plastic deformation by simultaneous action o f two modes, namely compression and additional shear deformation in the rolled material. In case of hot rolled steel this severe plastic deformation results in ultrafine grained structure, whereas in case of cold-rolled-annealed steel, the energy stored during asymmetric rolling results in additional grain refinement during annealing. According to Cui et al. [42] the fine grain evolution during asymmetric rolling of aluminium is associated with the development of sub-boundaries into high angle grain boundaries promoted by the abovementioned severe plastic deformation. During asymmetric rolling this additional shear component has to be extended to the center o f the sheet by appropriately controlling the rolling conditions. Lee et al. [43] have investigated the dynamic transformation of austenite to ferrite with the help of asymmetric rolling in a C - M n - T i - B steel with a reduction ratio of 30 -56 %. The rolling was performed at temperatures between 650 - 700 ° C using an uneven roller with a roll diameter ratio of 1.5. With the application of shear strain of approximately 2.0 at 675 ° C , followed by water quenching the final microstructure was Chapter 2 LITERATURE REVIEW 19 composed of uniformly distributed and randomly oriented ferrite grains (mean ferrite grain size = 1.3 um) alongwith fine patches of twinned martensite. Recently, Morimoto et al. [7] reported industrial production of fine grained microstructure at the Nakayama Steel, Japan, with a grain size of 2-5 um, in a 2 mm thick plain carbon steel strip by applying this technique. This has been claimed as the first industrial production of ultrafine grained steels trips in the world. The two foremost enabling factors were: (a) high reduction rolling of 40 % or more by Single R o l l Drive with Different Diameter (SRDD) in the last three stands of a 6-stand hot strip finishing mi l l and, (b) the strong cooling by a Curtain Wal l Cooling ( C W C ) system just after rolling at each of the three latter stands. 2.3.5 Deformation Induced Ferrite Transformation (DIFT) In this thermo-mechanical processing technique, the steels are rapidly cooled from austenitization temperature to a deformation temperature which is below the A e 3 but above the Ar3 temperature without deformation. This procedure produces highly undercooled austenite that is subjected to heavy deformation, and subsequent rapid cooling. A s a result the ferrite w i l l be formed assisted by deformation, as illustrated in Fig . 2.7 [44]. This process is known as DIFT. Yada et a/. [45] claimed for the first time that 2-3 um U F F can be formed in a C -M n steel by deforming in the temperature range from A r i + 50 ° C to Ar3 + 100 ° C when Chapter 2 LITERATURE REVIEW 20 reductions of more than 50% have been applied in less than one second. There are evidences now that ferrite is formed dynamically from austenite and not during post Ae3 3 1 © ct ,1 A r 3 Fine grain. Coarse grain Low deform aiion y JS**^ d ^ 1 temperature* """"""" T d r 0 DSIT DSIT Straining A r 3 T a Fig . 2.8 Schematic illustration of DIFT (same as DSIT) at isothermal deformation temperature [44]. deformation treatment [46-51]. Though several groups of researchers throughout the world have now been successful in obtaining ferrite grains in the size range of 1-3 urn, the finer details of grain refinement during DIFT still lacks unambiguous acceptance. Amongst the various proposed mechanisms, deformation enhanced dynamic ferrite transformation, dynamic recovery and / or recrystallization of ferrite are widely accepted to occur either independently, simultaneously or interactively [45 - 66]. Although researchers working in this field are divided in their opinion for devising a term (e.g. Deformation Induced Ferrite Transformation, Strain Induced Transformation, Strain Induced Dynamic Transformation, Deformation Enhanced Ferrite Transformation etc.) to Chapter 2 LITERATURE REVIEW 21 best describe the phenomenon, for the sake o f brevity, Deformation Induced Ferrite Transformation, or in abbreviation " D I F T " is being used in this thesis. Apart from regular industrial production o f hot rolled fine grain steel strip by Nakayama Steel in Japan using asymmetric rolling, the concept of DIFT has reportedly been successfully employed by P O S C O in South Korea in the industrial trial production o f hot rolled ultrafine grained low carbon microalloyed steel strips with varied proportion o f martensite and / or bainite by selection of the appropriate cooling conditions [65], It is encouraging to note that unlike the utilization of special asymmetric rolling arrangements and curtain wall cooling as in Nakayama Steel, in P O S C O the DIFT phenomenon has been exploited in a conventional hot rolling mi l l with very high reduction capability and subsequent laminar flow cooling on the Run Out Table (ROT). 2.4 Factors affecting DIFT There are several processing parameters that can influence the final ferrite grain size produced v ia the D I F T technique. They are discussed in more detail in the following sub-sections. 2.4.1 Strain level Strain is one of the most important parameters for U F F formation through DIFT processing. From a fundamental point of view, deformation accelerates phase transformation, both nucleation (start) and growth (finish), as shown schematically in Fig . 2.9 [51]. I f imposed strain is so small (s = si) that transformation does not start until Chapter 2 LITERATURE REVIEW 22 the deformation is ceased, dynamic transformation w i l l not occur at all . If deformation is continued, however, so that transformation starts during strain is being imposed (s = s 2), then transformation wi l l dynamically start. Fig . 2.9 Schematic representation of the variation of the variation of the transformation start curve as a function of strain [51]. Therefore, it can be interpreted that the critical strain (s c) for DIFT lies between ei and 82. Choi et al. [51] argued that DIFT and D R X are similar phenomenon as both of them are dynamic softening mechanisms of deformed austenite. Consequently, they determined the critical strain for DIFT by applying the method of Poliak et al. [67] developed for dynamic recrystallization without any modification. In a similar fashion Beladi et al. [44] defined two strain conditions for DIFT. The strain value for the start o f D I F T is termed as the critical strain for DIFT (SC,DIFT) whereas the strain to produce an U F F final microstructure by DIFT processing is higher than SC.DIFT, and termed as critical strain for U F F formation (8C,UFF)- They argued that at strains Ferrite Start Time Chapter 2 LITERATURE REVIEW 23 higher than £C,UFF the final microstructure does not change significantly. With the increase in strain value not only the volume fraction of U F F increases but also the average ferrite grain size decreases as illustrated in Figs. 2.10 (a) and (b), respectively for a C - M n - V steel. This is in concurrence with the earlier results, among others, reported by Hurley et al. [68] by hot torsion DIFT experiments carried out on C - M n - M o steel. Apart from the effect of deformation per se it is of great significance to realize the role o f deformation mode in refining the microstructure. The deformation modes which have been extensively studied in connection with DIFT phenomenon are generally divided into two types, compressive and shear modes. Hodgson et al. [52] produced the U F F microstructure with only 30% deformation in the surface of a plate from coarse grained austenite, and one o f the important reasons cited was the presence of the high shear strain at the surface. Inoue et al: [69] studied the effect o f shear deformation on grain refinement and concluded that the ferrite structures transformed from the deformed austenite are affected by not only the strain but also the deformation mode. Hickson et al. [59] studied the evolution of U F F microstructure by applying three different schedules, specifically, laboratory rolling, hot compression and hot torsion; they concluded that the amount of shear strain had a dominant role in producing U F F grains in the surface layer of a C - M n steel strip. . Chapter 2 LITERATURE REVIEW 24 Fig . 2.10 (a) Volume fraction of U F F as a function of strain at austenitization and deformation temperature of 1200 and 775 ° C respectively for a C - M n - V steel; (b) Ferrite grain size as a function of strain at austenitization temperatures of 1200 (Coarse grain) & 900 ° C (Fine grain) and deformation temperature o f 775 ° C for the same steel as in (a) [44] Hurley et al. [ 68 ] studied the effect o f pure shear strain on the efficiency of refinement in DIFT, and found that it had a similar influence on the efficiency of DIFT as compressive strain. Inoue et al. [70] further studied the different strain modes in multidirectional deformation and proposed that a concept of equivalent plastic strain can be used to estimate the efficiency of DIFT. 2.4.2 Prior austenite grain size Hurley et al. [54] reported that the coarser the austenite grains, the finer w i l l be the deformation induced ferrite grain size, which was in conflict with the generally Chapter 2 LITERATURE REVIEW 25 accepted view that fine austenite size decreases the final average ferrite grain size. However, i f intragranular nucleation of ferrite is considered to be the most important factor contributing to the additional grain refinement observed when DIFT occurs, then this is encouraged by large austenite grain sizes and accelerated cooling, both of which suppress the formation of grain boundary proeutectoid ferrite. Based on a very systematic study for delineating the effects of various process parameters on D I F T mechanism, Beladi et al. [44, 66] concluded that ferrite formed entirely at the prior austenite grain boundaries during initial stages of deformation, later on propagated inside the grains on the intragranular defects at a critical strain concurrently with the initiation o f DIFT. The fine grained austenite does not only produce a more homogeneous distribution of U F F grains but also, it facilitates in reducing the level of critical strain required for the initiation of dynamic transformation of austenite to ferrite. However, surprisingly the final ferrite grain size at room temperature as wel l as the volume fraction of U F F in the as-cooled microstructure was almost similar for two different levels of prior austenite grain size considered. In their study, although Beladi et al. elaborated upon the effect of prior austenite grain size on activation of intragranular and grain boundary ferrite nucleation sites during the DIFT phenomenon, direct microstructural evidence is still required to confirm this mechanism. In contrary to the aforementioned experimental results reported by Beladi et al., Hong et al. [50] reported of obtaining finer ferrite grain size with finer prior austenite grain size and a concomitant increase in volume fraction o f U F F in the final microstructure. Hence the aforementioned salient observations can be summarized based on the fact that the primary effect o f reducing the austenite grain size is to increase the amount Chapter 2 LITERATURE REVIEW 26 of grain boundary surface, the number of grain edges, corners per unit volume and, therefore, to increase the number of favorable nucleation sites, thereby favoring DIFT. The secondary effect is to increase the strain homogeneity and the overall dislocation density and, therefore, to further accelerate the transformation. Finally, the AJ*3 temperature increases with refinement of the prior austenite grain size which reduces the required strain for DIFT at an early stage of transformation for a given deformation temperature. 2.4.3 Degree of undercooling U F F obtained through DIFT processing is stable between Ae3 - AX3, but unstable above Ae3 [71]. Deformation just above A r 3 can supply the maximum undercooling for DIFT; hence Ar3 is a crucial parameter. Hodgson and coworkers [52, 54] obtained U F F in plain carbon steels by single-pass deformation just above AT3 and they suggested that this deformation temperature is very important for DIFT. However, Priestner et al. [72] contended that U F F grains were also obtained at a deformation temperature up to 150°C above A r 3 for a fine grained austenite. Niikura et al. [73] showed the positive effect of lowering deformation temperature (or increasing the undercooling) on ferrite grain size at a strain level of 1.2 for a C-Mn-Nb steel compared to a C - M n - S i steel after reheating to a temperature of 950 ° C , as shown in Fig. 2 .11 . Chapter 2 LITERATURE REVIEW 27 400 500 600 700 BOO 900 1 000) D e f a m a t i o n tenfperature(*C) Fig . 2.11 Effect of deformation temperature on U F F grain size obtained in low C steels through DIFT processing [73]. Hong et al. [74] have also systematically studied the effect of undercooling and they reported that, when the strain is relatively small, increasing undercooling is very effective in increasing the amount of U F F , but at relative high strains, increasing the undercooling w i l l affect only slightly the amount of U F F , as shown in Fig . 2.12. They also contended that increase in undercooling has resulted in decrease in the critical reduction required for initiation of formation of U F F , as illustrated in Fig . 2.13. This trend is readily to be expected because o f the concurrent increase of driving pressure for transformation with increase in undercooling. Chapter 2 LITERATURE REVIEW 28 c . 9 ae J „ 9t , Degree of undercooling, ATCC) Fig . 2.12 Variation of ferrite volume fraction with degree of undercooling for different amounts of applied strain in a plain low carbon steel [74] SO - i so-ion, o 40-.1 3S" 30-i o as-20-' \ V \ * v V V 1 1 • 1 1 r "———r-——»»"' £ 0 100 120 140 1BO l £ q Degree of undercooling, ATCC) Fig . 2.13 Variation of critical amount of reduction with degree of undercooling for plain low carbon steel [74] Chapter 2 LITERATURE REVIEW 29 2.4.4 Strain rate A s discussed earlier the DIFT starts on the prior austenite grain boundaries and at a later stage ferrite nucleation propagates to the austenite grain interiors. Hence, it implies that when ferrite is formed by DIFT, the rate of intragranular defect formation w i l l be decreased in the austenite since ferrite is softer than the austenite at a given deformation condition. Amongst very few studies carried out on the effect strain rate on DIFT mechanism, Beladi et al. [44] experimentally observed that significant amount of deformation concentrated on grain boundary ferrite when the strain rate was kept at 0.1s"1 for a C - M n - V steel, giving rise to dynamic recovery. For a given strain the ferrite grains formed by DIFT have a smaller size for a strain rate of Is"1, than for 0.1s"1. Similar results in terms of higher volume fraction U F F as a result of DIFT with increasing strain rate by an order of magnitude has been reported by Hong et al. [50], although nothing has been indicated about the effect on refinement of ferrite grains. A s fine ferrite grains have enhanced work hardening rate compared to coarse ferrite grains, there is little difference in flow stress between austenite and DIFT grains at a strain rate of Is"1 [51]. Consequently deformation progresses into austenite grain interiors and activates the intra-granular nucleation sites. This results in large number of DIFT ferrite grains per unit area at the early stage of transformation. For lower strain rates, more austenite deformation energy is supposed to be released by dynamic recovery than that for higher strain rates. This should enhance DIFT as strain rate increases. But at the same time at higher strain rates there is lesser time during deformation for DIFT to occur. Considering these two opposing trends it has not been currently possible to propose general conclusions on the strain rate dependence of DIFT. Chapter 2 LITERATURE REVIEW 30 2.4.5 Post deformation cooling rate Beladi et al. [44] studied the effect of the post deformation cooling rate on the final grain size of ferrite formed by deformation induced transformation. After applying a strain of 2 at 775 ° C to a C - M n - V steel with a prior austenite grain size of 84 um, different cooling rates were implemented. The final ferrite grain size decreased from 3.3 um to 2.2 um as the cooling rate increased from l °C / s to 5 °C/s . The cooling rates higher than 5 °C/s affected the U F F volume fraction rather than the U F F grain size. Cooling rate is believed to act as a controlling factor on the grain growth rate. Yada et al. [75] noticed that a cooling rate of 20 °C/s is necessary to suppress the grain growth after hot working, and even a higher cooling rate is desirable. It is worth mentioning here that with the advent of an ultrafast cooling technology developed at C R M , Belgium [76], cooling rates of approximately 280 °C/s in the temperature range o f 650-400 ° C have been claimed to be attained at industrial production lines. With the aid of ultrafast cooling rates, the grain size of 2~3 um in hot rolled strips has been attained alongwith other non-equilibrium phases in various low plain C - M n and microalloyed steels. 2.4.6 Post deformation isothermal holding Hong et al. [50] studied the effect of post deformation isothermal holding time for a C-Mn-1.2 S i (wt%) and a similar V - N b microalloyed steel which were deformed 80% at 740 ° C and 755 ° C (AT3+10 °C) , respectively. Accordingly, the effect is shown in Fig. 2.14 for the C - M n - S i - V - N b steel. With increasing isothermal holding time both the ferrite volume fraction and the mean ferrite grain size increased. The reason attributed Chapter 2 LITERATURE REVIEW 31 for the increase in ferrite volume fraction was that, during the isothermal holding new ferrite formed in addition to earlier formed ferrite by deformation induced transformation. Also the growth of the deformation induced ferrite increased the final ferrite fraction as well as the final ferrite grain size. i«h o 2 « 70 2 60-o > a) so-il) 40 30 4.1/im 3.1/« | a grain size 'A.lm I 2.3m I I 1 I ' I ' 1 ' F ' I • 1 ' I ' V 1 • I < I ' I ' I •20-10 0 10 20 30 40 50 60 70 SO 90 100110120 Intercritical holding time(s) Fig . 2.14 Variation o f ferrite volume fraction with isothermal post deformation holding time [49] A n important implication of post deformation holding was that there was gradual change from being mostly low angle grain boundaries to mostly high angle grain boundaries. This behaviour was then subsequently linked to coalescence of sub-grains and formation of new isothermal static ferrite grains. 2.4.7 Steel chemistry Hickson et al. [77] found that the chemical composition of the steel can slightly influence the morphology and volume fraction of U F F grains formed in the surface layers of a strip in laboratory rolling mi l l trials. In plain carbon grades, the level of ferrite refinement increased slightly, when the carbon content increased. However, they contended that when the carbon concentration is below the solubility limit in ferrite, finer Chapter 2 LITERATURE REVIEW 32 and more U F F can be obtained. It was also concluded that austenite stabilizers such as M o and B do not enhance the ultrafine ferrite formation. Based on their experimental observations Choi et. al. [49] proposed that V addition to low carbon steels enhances not only the static but also the dynamic transformation. Hong al. [78] studied the effects o f addition of Nb on DIFT in C - M n steel. N b has been found to affect DIFT differently depending on whether it is in solid solution or it is in precipitate form (NbC). When Nb was dissolved DIFT was hardly observed even after reduction of 80%. This has been attributed to a reduction in the driving force for transformation by dynamic precipitation o f N b C and kinetic effects to delay transformation by solute drag. When N b C pre-existed, the DIFT kinetics was similar to that of C - M n steel. A s readily expected the ferrite grain size was finer than that in the C -M n steel. 2.5 Mechanical properties of steels with UFF microstructure 2.5.1 Balance between strength, ductility and work hardening rate The quantitative relationship proposed by Hal l and Petch between the yield strength and the grain size in metals was given in Eq. 2.1. In Fig. 2.15 the yield stress is shown as a function of grain size revealing that the Hal l - Petch relation to be valid to very fine grain sizes for a Ti-IF steel processed through A R B and annealing route as discussed earlier. Clearly, a very high strength level in excess of 800 M P a can be achieved with the decrease in grain size. Chikushi et al. [6] have recently compared the dependence of yield and tensile strength of commercially produced conventional, fine Chapter 2 LITERATURE REVIEW 33 and ultrafine grained (approx. 3 um) plain carbon steel strips as shown in Fig . 2. 16. Y ie ld strength of more than 600 M P a can be achieved with this refined grain size. two BOO b B O O £ 400 c Q t 200 Mean Grain Size, d I M m O i 0.3 I P S.O 2.0 1 . 0 yield stress • Yield stei^th o Tensile slrergth OS 1 1.5 2 Mean Grain Size, t^um"2 Fig . 2.15 Grain size dependence of strength for an Ti-IF steel processed through A R B and annealing route [79] 800 700 600 500 400 300 0.20 0.30 0.40 0.50 0.60 0.70 1/graln a l io , p m " * Fig . 2.16 Dependence of yield and tensile strength on grain size for industrially processed steel strips [6] Chapter 2 LITERATURE REVIEW 34 But this enormous increase in strength with decreasing grain size is concomitant to the decrease in ductility, especially when the grain size is reduced to or below 1 um. A s can be seen in Fig . 2.17 for the same steel with same processing route as in Fig . 2.15, at a very high level of refinement in grain size, it shows practically "zero" work hardening behaviour. Although the operative mechanism for this behaviour is presently a subject of active research, the possibility of producing ultrafine grained materials with second phase dispersion is actively being pursued in order to maintain a proper balance o f work hardening characteristic alongwith an increased level of strength. wo QUO tn f 65© n oo cn yi 2 TO os 2B0 too 0 0.06 0.1 CIS 0.2 0.25 0.3 0S5. 0.4 Engg. strain Fig . 2.17 Stress-strain curves for industrially processed steel strips with different grain sizes [6] O n the same lines, recently Ohmori et al. [80] investigated the strain-hardening behaviour of U F F low-carbon steels with dispersed cementite particles. They contented that strain-hardening and consequently uniform ductility can be improved with an increasing volume fraction of cementite particles, as a result of increasing density of geometrically necessary dislocations generated by local non-uniform plastic deformation. Chapter 2 L I T E R A T U R E R E V I E W 3 5 It is worth mentioning here that dual phase steel, manifested by its continuous yielding behaviour, low yield ratio and rapid initial work hardening, is a promising candidate for a steel intensive application like autobody. Moreover, it should be recalled here that these tensile characteristics of ferrite-martensite dual phase (DP) steel are typical for coarse grained microstructure. Hence i f the microstructure of the D P steels can be refined, following the logic of Hal l - Petch behaviour, strength can be expected to increase with minimum loss of ductility. This has been very recently demonstrated by Park and co-workers [34, 35] as the first reported results on ultrafine D P steels, obtained by applying E C A P route, discussed in a preceding section. The relevant microstructural and tensile characteristics are summarized in Table 2.1 for three different steel chemistries, along with the behaviour of a coarse grained D P steel produced by identical intercritical annealing without E C A P , for comparison. The DPO steel is a plain low carbon steel, whereas the steels DPI and DP2 contain 0.06 and 0.12 wt% V , respectively. Detailed analysis of the stress-strain response indicated that although two stage work hardening behaviour was observed in both coarse and ultrafine grained dual phase steels, faster rate of load transfer from ferrite to martensite was observed in the latter. However, in their study a detailed quantification was not provided regarding the effect of grain refinement on mechanical properties since the volume fraction of martensite in the ultrafine grained varieties was also increased as compared to the coarse grained reference D P steel. Nevertheless, this investigation clearly indicated the potential of U F G D P steels for excellent property combinations. Chapter 2 LITERATURE REVIEW 36 Table 2.1 The microstructural and tensile characteristics of dual phase steels [35] Steel ; ! ! Martensite j fraction j (%) | 1 Ferrite grain size, um Martensite island size, um Y S , M P a U T S , ' M P a i Su (%) j i S f ( % ) C G -DPO : 22 1 ! 1 19.4 i 9.8 510 843 : l i 9.8 : f 13.5 U F G - j DPO ! 28 j J 0.8 ; j 0.8 581 978 j i 9.3 | f ! 17.6 U F G -DP1 , 3 5 ! 1 f 0.9 i . i ! 540 ] i 1044 i 11.5 ; 18.1 | i U F G - : DP2 • 32 | 1.2 i . i 565 1015 ! i 10.4 i i . . j 16.6 3 7 Chapter 3 Research Objectives The present research emphasises on the evaluation of the potential of producing a novel class o f steel for automotive application based on ultrafine ferrite (UFF) dual-phase microstructures. Hence, the major goal of the present investigation is to examine ferrite grain refinement in two low carbon microalloyed steels, in order to produce dual phase (ferrite + martensite) microstructures by employing the novel thermo-mechanical processing technique of deformation induced ferrite transformation (DIFT). To achieve this objective the following strategy has been adopted: (a) a series of systematic DIFT tests to be carried out for two steel chemistries which are typical for dual-phase steels, one with N b and M o and the other with only N b addition. (b) various thermo-mechanical treatment paths have to be so employed, to study the effect o f the following processing parameters on the final microstructures: (i) prior austenite grain size (ii) degree of undercooling (iii) amount of strain (iv) strain rate (c) to evaluate ferrite grain refining ability of the DIFT processing route, average ferrite grain size obtained from static isothermal ferrite transformation tests are to be Chapter 3 RESEARCH OBJECTIVES 3 8 compared to otherwise similar dynamic ferrite transformation tests (i.e. D IFT tests) with post-deformation isothermal hold (d) to make a first order approximation of strength increment with the help of microhardness measurements Chapter 4 Experimental Methods 39 4.1 Steels investigated Experiments were carried out on two low carbon steels, one microalloyed with Nb and M o and the other microalloyed with only Nb. The chemical compositions (wt. %) are given in Table 4.1. The steels were laboratory cast at C A N M E T - M T L , hot rolled to approximately 25 mm thick plates at their pilot mi l l and then delivered to U B C . Table 4.1 Chemical compositions of steels investigated (wt. %) ! Grade I i ! C | M n 1 S i ! s P Nb 1 ; M o : A 1 ' N | N b + : M o i 0.058 1.84 0.086 1 I 0.005 0.007 0.045 j 0.145 1 i 1 0.063 0.005 i 1 I N b i 1 0.058 1.78 0.092 | 0.004 0.005 0.06 3 ; 0.036 i 0.006 Thermocalc™ software employing the Fe 2000 database has been used to determine the A e 3 temperatures, which are 827 and 826 ° C respectively for N b + M o and Nb only steel. 4.2 Experimental equipment 4.2.1 Gleeble 3500 thermomechanical simulator A Gleeble 3500 thermomechanical simulator has been employed to perform the austenitization, transformation and deformation tests. Temperature control of the specimen during thermal cycling was achieved by spot welding 0.25 mm diameter Chapter 4 EXPERIMENTAL METHODS 40 Chromel-Alumel (type K ) or Platinum-PlaunumlO%Rhodium (type S) thermocouple onto the central position of the outer surface of the specimen. Type K thermocouples were used for measuring temperatures upto 1100 ° C and for higher temperatures type S thermocouples were used. Temperature feedback control and direct resistive heating of the system provided accurate heating, soaking and cooling of the specimen. Natural cooling (heating has been shut off), He gas quenching and water quenching were applied as and when required. The chamber was evacuated to a pressure below 10"4 Pa for each test in order to minimize decarburization and oxidation of the specimen and back filled with A r gas when required. L o w force jaws were used to perform all the deformation and /or phase transformation tests, except for the high strain rate tests (10 s'1) where high force jaws were employed. Feedback control of thermal and mechanical systems in addition to data acquisition were performed by the computer interface software QuickSim. 4.3 Experimental methodology 4.3.1 Prior Austenite Grain Sizes (PAGS) Rectangular specimen geometry with a length of 15 mm, width of 6 mm and 3 mm thickness was used. A typical thermal treatment path was employed to establish two uniform initial austenite grain sizes for subsequent tests. Specimens were heated at 5 °C/s to various austenitizing temperatures between 950 to 1250 ° C and held for 120 s to obtain the desired austenite grain sizes and subsequently water-quenched to room temperature in order to obtain a martensitic microstructure. The specimens were then tempered at 550 ° C for 6 h in a tube furnace in order to aid in revealing the prior austenite grain boundaries during subsequent metallographic preparation. Chapter 4 EXPERIMENTAL METHODS 41 4.3.2 Detection ofTs Continuous cooling transformation (CCT) tests were conducted to dilatometrically detect the transformation start temperature at a cooling rate of 40 °C/s for two levels of prior austenite grain sizes. Tubular specimens, 20 mm in length with 8 mm outer diameter and 1mm wall thickness, were austenitized to the previously established conditions to obtain the desired austenite grain size. Following austenitization, the specimens were immediately cooled by employing resistive heating in combination with He gas for achieving a cooling rate of 40 °C/s . The cooling rate was calculated by taking an average at ± 20 ° C o f the A e 3 temperature. Temperature and mid-length dilation was recorded as a function of time for each specimen undergoing testing using a crosswise dilatometer. The schematic diagram o f a specimen held in place by tubular stainless steel holders and copper anvils inside stainless steel jaws is shown in Fig. 4.1. The specimen was allowed to expand during heating without deformation by placing a spring between the jaws and load cell. When austenite decomposes during cooling there is a measurable increase in atomic volume from austenite ( F C C crystal structure) to ferrite ( B C C crystal structure) and other product phases. This is a result of the different crystal structures e.g. the F C C crystal structure is close packed i.e. high density, whereas a B C C crystal structure has a lower density. Volume fraction transformed, X (T), was calculated using the dilation measurements, dm(T), and employing the Lever rule as follows, dm{T)-d(T) dp(T)-dr(T) Chapter 4 EXPERIMENTAL METHODS 42 (a) Copper Anvils Stainfess Steel Jaws Specimen Specimen Holier + 2Z[: Spring / n 'Load Cell Helium In Thermocouples Specimen / Specimen Holder Helium Out Strain Measurement by Duatometer Fig . 4.1 Schematic of C C T test set-up for detecting the transformation start temperature: (a) full jaw set; (b) set-up of specimen. where dy(T) = Iy +ayT and dp(T) = Ip+apT are the extrapolated dilations from the austenite and product phase regions, and ay and ap are the thermal expansion coefficients for the austenite and product phases, respectively. Fig . 4.2 shows a schematic of a dilation response as a function of temperature, subsequent austenite-to-ferrite transformation kinetic curves can be derived from this data. T s was calculated from X = 0.05 and Tf, the transformation finish temperature, from X = 0.95. Chapter 4 EXPERIMENTAL METHODS 43 4.3.3 Static isothermal ferrite transformation tests In addition to the C C T tests isothermal ferrite transformation tests were conducted at T s + 25 ° C . After rapidly cooling from the established austenitizing conditions with a cooling rate o f 40 °C/s , specimens were held isothermally at the aforementioned temperature for 20 mins. prior to natural cooling to room temperature. The initiation and completion of austenite decomposition were precisely detected by dilatometry as described earlier. 4.3.4 Deformation Induced Ferrite Transformation (DIFT) tests After having established the transformation behaviour without deformation D I F T tests were carried out. Specimens were 120 mm in length, with a diameter of 10 mm; the gauge length of the working zone was 10 mm in length with a 6 mm diameter; the compression axis being parallel to the rolling direction of the plate as shown in Fig . 4.3. This complex sample geometry unlike simple axisymmetric compression specimen was Chapter 4 EXPERIMENTAL METHODS 44 r 10 mm 16.5 mm (fa 6 mm k - H 10 mm 120 mm Fig . 4.3 Specimen geometry for Deformation Induced Ferrite Transformation (DIFT) tests chosen consistent to previously established findings regarding on optimized compromise in controlling the demands of high cooling rates and deformation [81]. The specimens were austenitized at the established conditions, cooled to the deformation temperature with a cooling rate of 40 °C/s enabling a high level of undercooling, immediately followed by deformation at the set strains and constant strain rate. The deformation temperatures were so chosen that two levels of undercooling are maintained i.e. T s + 25 ° C and T s + 50 ° C . Investigations were carried out for three levels of strain i.e. diametrical true strains o f approximately 0.2, 0.4 and 0.6 constant strain rates of 1 and 10 s"1, respectively. Immediately after deformation the specimens were He-quenched with cooling rates of approximately 100 °C/s . F ig . 4.4 depicts a schematic thermal path used for the DIFT tests. Helium gas cooling, in order to maintain a constant deformation temperature (± 2 ° C ) was utilized during the deformation. Cooling rates were calculated between the deformation temperature and 400 ° C . Chapter 4 EXPERIMENTAL METHODS 45 Austenitization (t = 120 s) d Y Tdef_= 6 2 7 - 6 9 7 ° C V j A T (25 & 50 °C) Helium quench (- 100 °C/s) 826 - 827 ° C A e 3 Ts Fig. 4.4 Schematic thermal path used for the DIFT tests Further, additional tests were carried out to characterize the role of an isothermal hold after the deformation at the highest level of undercooling with maximum amount of strain employed (i.e. 0.6) at a strain rate of Is"1. 4.4 Microstructural investigation Microstructural investigation was carried out following every thermal or thermo-mechanical treatment. The techniques employed and the associated sample preparation methods are presented in the following. 4.4.1 Optical microscopy The samples to be used for determination o f austenite grain size and metallographic examination o f continuously cooled transformation tests without deformation were cut at the welded thermocouple position using a silicon carbide cut-off wheel and spray coolant, to protect the specimens from overheating during cutting. The surfaces o f interest were cold mounted in either an acrylic resin or a polymer resin and then gradually ground using silicon carbide papers to 1200 grit finish followed by final Chapter 4 EXPERIMENTAL METHODS 46 polishing employing 6 urn and 1 um diamond suspensions on a Buehler (Phoenix 400) grinding and polishing machine. Metallographic examination of the as-received samples was subsequently performed by etching with 2 % Nital solution (2 ml . HNO3 + 98 ml . H2O). In order to facilitate in revealing the prior austenite grain boundaries, an additional 'tempering' step of isothermally holding the as-quenched samples after reheating treatment, at 550 ° C for 6 h in a tube furnace, was employed for both the steels. Following the tempering treatment the samples were etched by immersion at an elevated temperature of 40 ° C for varied duration between 10 s to 180 s as per the enhancement of stain contrast under optical microscope. The etchant used was based on Vi le l la ' s reagent [82] with 2 gm saturated Picric acid, 5 ml HC1 and 100 ml Ethanol. In order to reveal the as transformed microstructures after continuous cooling tests at 40 °C/s , a 2 % Nital etchant was employed with an etching time of 5-10 s by immersing the samples in the etchant. A n optical microscope was utilized to obtain micrographs of the sample's cross-sectional area. For quantitative analysis of mean austenite grain sizes, the grain boundaries were outlined using a thin felt-tip marker on transparencies that were laid over micrographs. The mean austenite grain sizes were calculated using Jeffries planimetric procedure as per the A S T M E l 12-96 (2004) standard [83]. Using this method, grains completely within the measurement field were counted once and grains intersecting the perimeter of the measurement field were counted as a half grain. For statistically relevant results, several random areas were chosen as measurement fields and at least 50 grains were present in each field. Also , for each specimen a minimum o f 300 grains were counted. The mean grain area was then used to calculate the mean equivalent Chapter 4 EXPERIMENTAL METHODS 47 area diameter ( E Q A D ) . The Clemex™ Image Analysis System was employed to facilitate the measurements. 4.4.2 Scanning electron microscopy Due to practical limitations of optical microscope for resolution, all the samples with U F F microstructure were quantitatively analyzed using a W-filament Hitachi S-3000N Scanning Electron Microscope (SEM) operated at 20 keV. After the DIFT tests samples were obtained by first cutting the gripped portions o f the tested piece with the help of a silicon carbide cut-off wheel as described earlier. Subsequently the final sample was obtained by cutting the tested piece at the welded thermocouple position by employing a slow speed wheel cutter (Leco V C - 50). Co ld mounting, mechanical grinding and polishing were followed as described earlier. A n additional final polishing step using 0.05 um colloidal silica suspension utilizing the Vibrometry technique (Vibromet 2, Buehler Ltd.) was followed before light etching with 2 % Ni ta l solution. The scanning electron micrographs were obtained from various positions around the centre of the sample. The image analysis software was then employed for determination of ferrite grain size, distribution, mean ferrite grain size and area fraction. Since the microstructure is composed of two phases i.e. ferrite and martensite, the total martensite area needed to be subtracted from the total measurement field area so that only the ferrite area was used for the grain size calculations. Ferrite fraction was quantified by dividing the measured ferrite area by the total measurement field area. Chapter 4 EXPERIMENTAL METHODS 48 4.4.3 Orientation imaging microscopy Few samples with U F F microstructure were analyzed using the orientation imaging microscopy technique by electron backscattered diffraction ( E B S D ) system attached to a W-filament Hitachi S570 S E M at an operating at a voltage of 20 keV with suitable calibration. In order to obtain a damage free surface after mechanical polishing, electro-polishing technique was employed. A l l the sample surfaces were electro-polished using an electrolyte of 95% Acetic acid + 5% Perchloric acid at 30 V and 0.3 A m p for approximately 120 s. The electro-polishing was carried out at room temperature with the provision for running cold water jacket outside electro-polishing vessel. Analysis area of approximately 20 X 20 um was covered on the centre of the sample surface in the beam scan mode with a step size of 0.1 um for approximately 10 h. For data acquisition and processing Channel 5 software of H K L Technology, Denmark was employed. 4.4.4 Scanning transmission electron microscopy In order to assess qualitatively the presence of carbide, nitride and/or carbo-nitride precipitates in the as-quenched microstructure after austenitizing at 950 ° C and 1100 ° C following previously established conditions, carbon extraction replica technique was employed. To prepare carbon extraction replicas, the specimens were first ground and polished as previously described and then lightly etched in 2% Nital solution. A thin fi lm of carbon, approximately 20 nm thick was subsequently deposited on the surface of each sample using a vacuum evaporator. The carbon coated surface was scored with a sharp razor blade into 2-3 mm squares. The extraction replicas were then stripped off by re-etching the samples in 5% Nital solution followed by placing them in distilled water. A l l electron microscopy work was performed at an operating voltage of 200 keV on a Chapter 4 EXPERIMENTAL METHODS 49 Hitachi H-800 transmission electron microscope equipped with H-8010 scanning system and an energy dispersive X-ray ( E D X ) analysis system. 4.5 Hardness measurements In order to make a first order approximation of the strength levels associated with the dual phase microstructures produced in these laboratory tests, microhardness measurements were made following the A S T M E384-99 standard [84]. The microhardness tester (Micromet 3, Buehler Ltd.) was used with 100 gm load on Vicker ' s scale with measurements taken in the same cross-sectional plane as the microstructural analysis was performed. A random selection of five measurement fields was made approximately at the centre of the sample. 50 Chapter 5 Results 5.1 Initial Austenite Grain Size Conditions Isothermal austenite grain growth tests were performed to establish suitable reheating conditions resulting in prior austenite grain sizes with a normal grain size distribution. A summary of the selected heat treatment cycles is specified in Table 5.1. Specimens were heated at 5 °C/s to the desired austenitizing temperature where they were held for 120 s and subsequently water quenched. The austenite grain size is reported Table 5.1 Heat treatment schedules and measured austenite grain sizes Steel \ i Holding Temp. (°C) i Measured Austenite Grain Size ( E Q A D ) , dEQAD (Mm) Estimated Volumetric Austenite Grain Size, d v o i (um) i i i i Nb + M o Steel i i i 1 i i 950 10 12 1000 } 12 14 1050 j 14 17 1100 j 26 31 1150 | 27 34 1200 36 43 • 1250 j 60 72 1 i Nb Steel I ! i r i 950 j 11 13 1000 13 16 1050 | • - J 4 J 17 1100 ' 25 | 30 as an equivalent area diameter. The resulting equivalent volume diameter which was obtained by multiplying the measured equivalent area diameter ( E Q A D ) by 1.2 i.e. dvoi = Chapter 5 RESULTS 51 1.2 dsQAD, as discussed by Giumell i et al. [85] has also been indicated. Austenite grain growth was observed to be uniform for temperatures upto 1100 ° C . Further increase in the austenitization temperature lead to non-homogeneous grain size distribution exhibiting fine grains embedded in regions of substantially larger grains. Hence, the two austenitizing conditions selected for subsequent continuous cooling, isothermal holding and deformation-transformation tests were at 950 C and 1100 ° C . The resulting austenite microstructures for these conditions are shown in Fig. 5.1. A n overall experimental error associated with the austenite grain size measurements was estimated to be approximately 10 % based on the standard deviation of the measurements. Fig . 5.1 Austenite microstructure after reheating to (a, c) 950 ° C and (b, d) 1100 ° C , with a holding time of 120 s and water quenching, for Nb + M o and Nb Steel respectively. Chapter 5 RESULTS 52 5.2 Precipitates in the austenitized condition Electron microscopy replica studies were carried out for the selected two austenitizing conditions, i.e. at 950 and 1100 ° C in order to investigate the possible presence of precipitates. A semi-quantitative analysis of particle composition was made for the precipitates through the use of energy-dispersive X-ray ( E D X ) technique. The precipitates observed were mainly composed o f niobium and near spherical in shape. Co-precipitation of Nb-rich precipitate onto A1N precipitate was also observed in the N b + M o steel austenitized at 950 ° C . The T E M replica micrographs for Nb + M o steel are shown in Fig. 5.2 and Fig . 5.3 alongwith the precipitate E D X spectrum. • o u n U BOO Fig . 5.2 (a )TEM replica micrograph and (b) E D X spectrum of Nb-rich precipitate at 950 o. C in the Nb + M o steel Chapter 5 RESULTS 53 u l 7 i l l I l l N b K A Fig . 5.3 (a) T E M replica micrograph and (b) E D X spectrum, showing co-precipitation of Nb-rich precipitate at austenite grain boundary onto A1N precipitate at 950 ° C in the Nb + M o steel The C u peaks in the E D X spectra resulted from the fine Cu-grids which were used for holding the carbon extraction replicas. It is interesting to note, however, that no M o peak was observed indicating that most of the M o content of the steel to be present in solution at 950 ° C . In case of Nb steel, a visibly increased number of Nb-rich precipitates were observed as shown in Fig. 5.4. A t 1100 ° C similar but visibly decreased number of Nb-rich precipitates was observed in both the steels. Chapter 5 RESULTS 54 J. WbKB I ll t i l l I — ini I Jmit, m Fig . 5.4 (a) T E M replica micrograph and (b) E D X spectrum, of Nb-rich precipitate at 950 C in the Nb steel 5.3 Continuous Cooling Transformation (CCT) test results 5.3.1 Determination of 5% transformation start temperatures Continuous cooling transformations tests were employed to detect the 5% transformation start temperatures at the pre-established austenitization conditions for a cooling rate o f 40 ° C /s. The measured transformation start temperatures without any Chapter 5 RESULTS 55 deformation alongwith the associated austenitizing conditions are presented in Table 5.2. The influence of austenite grain size on the continuous cooling transformation behaviour is illustrated in F ig . 5.5, without prestrain. Table 5.2 Selected austenitizing temperatures and measured 5 % transformation start temperatures Steel \ 1 Measured Austenite Grain Size ( E Q A D ) , dEQAD Measured T 0.05 , (°C) Nb + M o j Steel i 10 602 | 26 528 Nb Steel ! . _ J - . „ . " 1 647 25 1 504 ( O Nb + Mo, d =10 um • Nb + Mo, d =26 utn Nb,d=11 um Nb, d =25 um Temperature, °C Fig . 5.5 Continuous cooling transformation behaviour for selected prior austenite grain sizes in N b + M o and Nb steels with a cooling rate of 40 °C/s Chapter 5 RESULTS 56 It can be noted that there is a clear shift of the transformation to lower temperatures with an increase in the initial austenite grain size. A larger initial austenite grain size provides less boundary surface area per unit volume, thus reducing the number o f available nucleation sites for ferrite and as a result transformation occurs at lower temperatures. In addition, the carbon diffusion distance is then increased for a larger initial austenite grain size requiring additional time for growth, thereby lowering the transformation temperature. The transformation start temperature (T s) generally defined as 5% transformation and similarly, the transformation finish temperature (Tf) as 95% transformation, both show similar trend of decrease with the increase in austenite grain sizes. 5.3.2 Microstructures obtained after C C T tests Resulting microstructures from the C C T tests have been illustrated in Fig. 5.6 and Fig . 5.7 for Nb + M o and Nb steel, respectively. A mixture of quasi-polygonal ferrite, acicular ferrite and bainite was obtained for Nb + M o steel with smaller austenite grain F ig . 5.6 Microstructure after continuous cooling tests for Nb + M o steel for austenite grain sizes of: (a) 10 um and (b) 26 um respectively Chapter 5 RESULTS 57 Fig . 5.7 Microstructure after continuous cooling tests for Nb steel for austenite grain sizes o f : (a) 11 um and (b) 25 um respectively size whereas in case of Nb steel, similar level of austenite grain size resulted in predominantly quasi-polygonal microstructure. It is to be noted here that for small P A G S , N b + M o steel transforms at ~ 50 ° C lower temperature and thus only the Nb steel shows a fully ferritic structure. Fully bainitic microstructures were observed in both the steels when the austenite with a large grain size of approximately 25 um decomposed at the same cooling rate. For the smaller austenite grain size, the presence of non-polygonal transformation products indicates the complex solute drag like effects of M n , Nb and M o and their concurrent interactions with C, on the austenite decomposition phenomenon. For the larger austenite grain size, there is an increase in the diffusion distance and as a result an increase in the time required for the redistribution of carbon, resulting in a greater degree of undercooling and hence, encouraging the formation of non-polygonal transformation products like bainite, under the same cooling conditions. Chapter 5 RESULTS 58 5.4 Ferrite grain size obtained after static isothermal transformation tests Isothermal transformation tests were conducted at 25 ° C above the T s for both the steels for the smaller austenite grain size. These tests were performed to estimate the resulting ferrite grain size from highly undercooled austenite, but without any deformation. The tests further indicated precisely the duration of complete austenite decomposition, as measured by dilatometry. The ensuing microstructures are shown in F ig . 5.8 and F ig . 5.9 for N b + M o steel and N b steel, respectively. They were quantified F ig . 5.8 Microstructure after isothermal holding at 627 ° C for 20 mins. and subsequent natural cooling for Nb + M o steel with austenite grain size of 10 um Fig . 5.9 Microstructure after isothermal holding at 672 ° C for 20 mins. and subsequent natural cooling for Nb steel with austenite grain size of 11 um Chapter 5 RESULTS 59 in terms of ferrite grain size and ferrite fraction. The microstructures as can be observed were essentially consisted o f polygonal ferrite and pearlite. The average ferrite grain size and fraction are indicated in Table 5.3. The time for 95% completion of transformation was measured as 380 and 680 seconds respectively, for Nb + M o and Nb steels. Table 5.3 Ferrite grain size and fraction after isothermal transformation tests j Steel j ; 1 Isothermal Average ferrite Ferrite i 1 | transformation grain size, d a fraction I i temperature (°C) (um) (%) N b + M o Steel 1 627 J i 1 4.6 82 Nb Steel \ 672 4.9 88 5.5 DIFT tests for producing fine grained dual phase steels D I F T tests were employed to investigate the combined effect of deformation and accelerated cooling on the resulting ferrite grain refinement for producing fine grained dual phase steels. Details on the experimental conditions employed for the tests are given in Table 5.4 and Table 5.5 for Nb + M o and Nb steels, respectively. The effects of amount of strain, prior austenite grain size, degree of undercooling, strain rate and post deformation hold on the final microstructure were systematically investigated. Chapter 5 RESULTS 60 5.4 Deformation Induced Ferrite Transformation (DIFT) test matrix for N b + M o steel T * aus» PAGS T > ! i I j i Tdef AT Strain rate 1 s"1 10 s 1 Strain 950 °C, j 10 um ! I i 1 1 i ! 602 °C j 1 ! I i 627 °C i 1 j 25 °C ! i 0.2 0.4 ! 0-6 0.6 652 °C ! 50 °C | 0.2 0.4 . . . 0-6 1 0.6 i 1100°C, ; 26 um i j i 528 °C | i 1 . - — i 553 °C 25 °C | 0.2 I 0.4 j 0.6 | | 578 °C j 50 °C j 0.2 0.4 0.6 Table 5.5 Deformation Induced Ferrite Transformation (DIFT) test matrix for N b steel T 1 aus» i PAGS i T 1 s ! i j Tdef i 1 I 1 AT Strain rate 10 s"1 J Strain i i i 950 °C, ! 11 um t i ! i i 647 °C i 1 672 °C j | 25 °C 0.2 0.4 0.6 0.6 ! 697 °C i i 50 °C j 0.2 0.4 . M 0.6 1100°C, 25 um 1 i I i 504 °C ! i 1 1 I 529 °C | 1 25 °C 0.2 j 0.4 | 0.6 | 554 °C j i 50 °C | 1 0.2 | 0.4 I 0.6 i 1 Chapter 5 RESULTS 61 5.5.1 Effect of strain The evolution of the final microstructure with the application of strain upto 0.6 has been illustrated in Fig. 5.10 and Fig. 5.11 for Nb + M o and Nb steels, respectively, with the smaller austenite grain size. Ferrite grains and martensite islands as delineated on the basis of resultant Nital-etch stain contrast are also indicated in Fig . 5.10 (c) and Fig . 5.11 (c). IK, ' i •» <— -5 nm Martensite Ferrite 5 |im Fig. 5.10 Scanning Electron micrographs after DIFT tests (strain rate = Is"1) o f N b + M o steel for austenite grain size of 10 um at deformation temperature of 627 ° C at strain level o f : (a) 0.2, (b) 0.4 and (c) 0.6 respectively Chapter 5 RESULTS 62 (b) 5 um 5 nm (C) • Martensite • Ferrite 5 u m Fig . 5.11 Scanning Electron micrographs after DIFT tests (strain rate = Is - 1 ) o f Nb steel for austenite grain size of 11 urn at deformation temperature of 672 ° C at strain level of: (a) 0.2, (b) 0.4 and (c) 0.6 respectively With increasing level of strain from 0.2 to 0.6, intragranular nucleation sites are believed to have been activated, apart from grain boundary sites. Hence, the resultant effect was more homogeneous microstructure with an increased fraction of deformation induced ferrite, as illustrated in Fig . 5.12. For example, the U F F fraction increased from 0.12 to 0.70 in the N b + M o steel whereas in case of N b steel, it increased from 0.38 to 0.75 with the employment of increasing strain from 0.2 to 0.6. Chapter 5 RESULTS 63 80 7 0 60 3- 5 0 H J 40 o it 3 ° 20 10H Nb + Mo steel Nb steel 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 True strain, E Fig . 5.12 Fraction of U F F as a function of applied strain in Nb + M o and Nb steels respectively for smaller P A G S and the highest undercooling 5.5.2 Effect of PAGS The resulting microstructures after employing a strain of 0.2 for the selected two levels o f P A G S are illustrated in Fig. 5.13 and Fig. 5.14. It is evident that for both the steels finer austenite grain size has resulted in the reduction of critical strain for (a) I I (b) 5 um Fig . 5.13 Scanning' electron and optical micrographs respectively, after DIFT tests o f Nb + M o steel at deformation temperatures of 627 ° C and 553 ° C for austenite grain sizes o f : (a) 10 and (b) 26 um , at s = 0.2 Chapter 5 R E S U L T S 64 5 um Fig . 5.14 Scanning electron and optical micrographs respectively, after DIFT tests of Nb steel at deformation temperatures of 672 ° C and 529 ° C for austenite grain sizes o f : (a) 11 and (b) 25 um , at s = 0.2 Deformation Induced Ferrite Transformation (DIFT) to occur. With a coarser grain size, initiation of DIFT did not happen at s = 0.2 in either of the steels. The microstructures obtained after employing highest amount of strain i.e. 0.6, are similarly illustrated in Fig . 5.15 and Fig. 5.16. 5 nm 5 nm Fig . 5.15 Scanning electron micrographs after DIFT tests of Nb + M o steel at deformation temperatures of 627 ° C and 553 ° C for austenite grain sizes o f : (a) 10 and (b) 26 um respectively, at s = 0.6 Chapter 5 RESULTS 65 Fig. 5.16 Scanning electron micrographs after DIFT tests of Nb steel at deformation temperatures of 672 ° C and 529 ° C for austenite grain sizes o f : (a) 11 and (b) 25 um respectively, at s = 0.6 Significant amount of grain refinement alongwith presence of considerable proportion of U F F indicates that the finer an austenite grain, the higher is the volume fraction of DIF at a given strain due to increase of grain boundary area. The condition represented by Fig . 5.15 (a) corresponds to an average ferrite grain size of 1.9 um and percentage ferrite fraction of 70%. Likewise, the condition represented by F ig . 5.16 (a) corresponds to an average ferrite grain size of 1.3 um and percentage ferrite fraction of 75%. 5.5.3 Effect of undercooling The effect of undercooling on the final dual phase microstructure was investigated by selecting two deformation temperatures for both the steels, one at T s + 25 ° C and the other, T s + 50 ° C , with three levels of strain employed as represented in Tables 5.4 and 5.5. These selected deformation temperatures lead to undercooling of 200 ° C and 175 ° C for N b + M o steel and, 154 ° C and 129 ° C for Nb steel, respectively. The final Chapter 5 RESULTS 66 microstructures obtained after employing the highest amount of strain i.e. 0.6 to the smaller level of austenite grain size are illustrated in Fig. 5.17. The quantification of the microstructures is presented in Table 5.6. 5 um 5 nm Fig . 5.17 Scanning Electron micrograph after DIFT tests of Nb + M o (a, b) and Nb steel (c, d) respectively, at undercooling of: (a) 200 ° C , (b) 175 °C , (c) 154 ° C and (d) 129 ° C Chapter 5 RESULTS 67 Table 5.6 Final ferrite grain size and fraction resulted after DIFT test with s = 0.6 Steel Fig. Nos. Test conditions Ferrite grain size, Jim Ferrite fraction, % N b + M o 5.17(a) d Y = 10 um, T d ef =627 ° C , Undercooling = 200 ° C 1.9 70 1 5.17(b) i d y = 10 um, Tdef =652 ° C , Undercooling = 175 ° C 2.2 J 60 1 i Nb 5.17(c) • d Y = 11 um, Tdef =672 ° C , Undercooling = 154 ° C 1.3 75 5.17(d) d y = l l um, Tdef =697 ° C , Undercooling - 129 ° C 1.1 64 Variation in level of undercooling seems to have negligible effect on U F F fraction at small strain levels of 0.2 and 0.4 for both the steels. However, at an increased strain level of 0.6, as clearly indicated in Fig. 5.17, the consequence of higher undercooling was evident in terms of higher ferrite fraction as a result of higher driving force for ferrite transformation. Nonetheless, microstructural observation suggests that the effect of level of undercooling on the ferrite grain refinement is not apparent. To further illustrate this, the ferrite grain size distributions are shown in Fig . 5.18 and Fig . 5.19 for N b + M o and N b steels, respectively where the E Q A D is expressed in um. These distributions are obtained from measurements of a minimum of 1000 grains in each case. The Log-Normal fit for the respective distribution is also indicated. It is interesting to note that more than 90% of the grains fall within the spread of 0.5 - 0.6 um from the average values reported, with very few grains being in order of 4 - 5 um. Chapter 5 RESULTS 68 0.14 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 2.5 In (EQAD, um) -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 2.5 In (EQAD, nm) Fig . 5.18 Ferrite grain size distribution after D I F T tests with d Y = 10 um, e = 0.6, de / dt = 1 s"1 and corresponding Log-Normal fit for Nb + M o steel at the following conditions: (a) T d e f = 627 ° C , A v g . Ferrite grain size = 1.9 ± 0.6 um; (b) T def = 652 ° C , A v g . Ferrite grain size = 2.2 ± 0.6 um (a) -2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 (b) 0.16 0.15 0.14 0.13-0.12-0.11 -0.10-0.09 c •2 0.08 O JO 0.07 U " 0.06 0.05 0.04 0.03 H 0.02 0.01 0.00 V -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 In (EQAD, um) In (EQAD, nm) Fig . 5.19 Ferrite grain size distribution after DIFT tests with d y = 11 um, e = 0.6, ds / dt = 1 s"1 and corresponding Log-Normal fit for N b steel at the following conditions: (a) T d ef = 672 ° C , A v g . Ferrite grain size = 1.3 ± 0.5 um; (b) T d e f = 697 ° C , A v g . Ferrite grain size = 1.1 ± 0 . 6 um Chapter 5 RESULTS 69 In order to check the accuracy of the measurements of ferrite grain size and fraction, measurements were repeated on two sample cases one for each steel with specimens obtained from the identical DIFT tests. A variation of approximately 15 % in the measurement of mean ferrite grain size was noted, although the variation in the measurement o f U F F fraction was less than 5 %. A n important source in the variation of the ferrite grain size measurement is the difficulty that is associated with the demarcation of high and low angle grains boundaries based on the etch stain contrast. The error of the quantification of U F F fraction is mainly associated with the smaller size o f second phase islands (~ 1 urn) in the U F F - D P microstructure. 5.5.4 Effect of strain rate O n the premise of significant grain refinement alongwith considerable amount o f U F F in the final dual phase microstructure from the evident pre-requisites of small austenite grain size and high amount of strain, DIFT tests were carried out at a high strain rate o f 10 s"1 in order to investigate the effect of same level of undercooling as that in low strain rate tests. The resulting microstructures are given in Fig. 5.20. The quantification of the microstructures is presented in Table 5.7. Chapter 5 RESULTS 70 5 u m Sum Fig. 5.20 Scanning Electron micrograph after DIFT tests (ds / dt = 10s _ I) of Nb + M o (a, b) and Nb steel (c, d) respectively, at undercooling of: (a) 200 ° C , (b) 175 ° C , (c) 1 5 4 ° C a n d ( d ) 129 ° C Table 5.7 Final ferrite grain size and fraction resulted after DIFT tests with ds / dt = 10 s'1 Steel Fig. Nos. Test conditions Ferrite grain size, um Ferrite fraction, % Nb + M o 5.20 (a) d y = 1 0 u m , T d e f = 6 2 7 o C , Undercooling = 200 ° C 1.1 68 5.20 (b) d Y = 1 0 u m , T d e f = 6 5 2 ° C , Undercooling = 175 ° C 1.9 64 Nb 5.20 (c) d y = 11 u m . T d e f = 6 7 2 ° C , Undercooling = 154 ° C 1.8 71 5.20 (d) dy= 11 urn, T d e f =697 ° C , Undercooling = 129 ° C 1.4 62 Chapter 5 RESULTS 71 Comparing the grain sizes and ferrite fraction with those of the low strain rate tests shown in Table 5.6 it appears that the microstructural observations are in both cases very similar. Hence within the margin of error for the present investigated cases a change of strain rate from 1 to 10 s"1 can be considered to be negligible in terms of U F F formation. 5.5.5 Effect of post deformation isothermal hold Based on the static ferrite transformation tests as described earlier, two DIFT tests one for each steel, were carried out at the highest undercooling employing the highest amount of strain with post deformation isothermal holding prior to He-quench. The duration for isothermal holding at the deformation temperature after deformation was approximately 380 and 680 seconds for N b + M o and N b steels, respectively. These tests helped in delineating precisely the effect of deformation on ferrite grain refinement as registered in the DIFT schedule. The resulting microstructures are shown in Fig. 5.21. 5 Um 5 ^n , Fig. 5.21 Scanning Electron micrograph after DIFT tests with small austenite grain size employing a strain of 0.6 and post deformation isothermal hold for (a) Nb + M o steel and (b) Nb steel, at 627 ° C and 672 ° C , respectively Chapter 5 RESULTS 72 Comparing these with the microstructures obtained after He-quench without any holding (see Figs. 5.17 (a) and (c)) clearly indicates the transition from quasi-polygonal to polygonal ferrite microstructure. It is indeed interesting to note from Table 5.8 that the quantification of these microstructures in terms of ferrite grain size and fraction has indicated minimal effect for either of the steels when compared to microstructures obtained without any holding (see Table 5.6). Table 5.8 Final ferrite grain size and fraction resulted after DIFT test with highest strain (0.6), strain rate = Is"1 and post deformation isothermal hold Steel Fig. Nos. Test conditions Ferrite grain | size, urn j Ferrite fraction, % N b + M o 5.21 (a) ; d y = 10 um, T d e f =627 ° C , Undercooling = 200 ° C , Isothermal hold = 380 s 1.9 69 ' N b 5.21 (b) : d Y = 11 jam, T d e f = 672 ° C , Undercooling = 154 °C , Isothermal hold = 680 s 2.0 * ! 5.6 Microstructural investigation using Orientation imaging microscopy Orientation imaging microscopy which works on the principle of Electron Back Scattered Diffraction (EBSD) technique attached to a S E M , has been used to further analyze the most promising U F F microstructures in the dual phase steels. The results of the E B S D analysis in terms of the band contrast maps alongwith corresponding misorientation angle distribution for the selected conditions of the two steels are shown in Figs. 5.22, 5.23, 5.24, 5.25. The band contrast is a measure of the contrast in the diffraction pattern affected by their sharpness. The white area in these maps represents the unindexed data points. Though it is believed that diffraction response from the ferrite Chapter 5 RESULTS 73 and martensite phases would be apparently identical due to the similarity in the crystal structures, presumably the fine lath structure of the martensite alongwith the neighbourhood of the U F F boundary regions remained unindexed due to the limitation posed by the spatial resolution of present E B S D instrument. Hence, further data analysis obtained from the E B S D was kept limited only to the correctly indexed data points. Theoretical random distribution, popularly known as Mackenzie plots is represented in the figures by the bold line. Further, the distribution of the correlated misorientation angles has been shown. The correlated misorientation angles are the set of all misorientation angles calculated from grain pairs directly in contact. In all the present cases considered the misorientation angle distribution which is dominated by large angles (i.e. > 15 ° ) , clearly indicates towards the randomness of the comprising U F F microstructures. Further, it possibly indicates towards the fact that U F F grains so formed predominantly have high angle grain boundaries. Chapter 5 RESULTS 74 (b) Fig . 5.22 (a) Band contrast map from an area of 20 X 20 um represented by Fig . 5.17 (a) for Nb + M o steel (dy = 10 nm, T d e f = 627 °C, ds / dt = I s - 1 ) , (b) Misorientation angle distribution of the same map as in (a); the bold line indicates random distribution Chapter 5 RESULTS 75 Fig. 5.23 (a) Band contrast map from an area of 20 X 20 um represented by Fig. 5.17 (b) for N b + M o steel (dy = 10 um, T d e f = 652 ° C , ds / dt = Is"1), (b) Misorientation angle distribution of the same map as in (a); the bold line indicates random distribution Fig . 5.24 (a) Band contrast map from an area of 20 X 20 (am represented by Fig. 5.17 (c) for Nb steel (d y =11 um, T d e f = 672 ° C , de / dt = Is - 1), (b) Misorientation angle distribution of the same map as in (a); the bold line indicates random distribution Fig. 5.25 (a) Band contrast map from an area of 20 X 20 um represented by Fig. 5.17 (d) for Nb steel (d y =11 um, T d e f = 697 ° C , ds / dt = Is"1), (b) Misorientation angle distribution of the same map as in (a); the bold line indicates random distribution Chapter 5 RESULTS 78 5.7 Relation between ferrite grain size, fraction of second phase and hardness Since the specimen geometry of the DIFT test samples did not allow regular tensile test to characterize the mechanical properties, microhardness tests were employed to estimate the strength using Tabor's relation [84] as shown in Eq . 5.1. H y — A o.o8 cr 0.08 (5-1) where H v is the Vicker 's hardness number (Kg/mm 2 ) , A o.os is a constant with values ranging between 2.9 to 3.0 and <y 0.08 is the flow stress at 8% strain. The results of the microhardness tests and predicted strength using Eq . 5.1 with A o.os = 3.0 are presented in Table 5.9. Table 5.9 Prediction of flow stress at 8% strain based on Tabor's relation [86] Steel Average | ferrite grain 1 size, u,m j 2nd phase fraction, % i H v 0.1, Kg/mm2 Predicted a o.os, MPa N b + M o 1.9 ± 0 . 6 30 278 ± 6 909 ± 20 2.2 ± 0 . 6 j i i 40 286 ± 5 939 ± 16 N b 1.3 ± 0 . 5 25 248 ± 3 811 ± 10 1.1 ± 0 . 6 | i 36 254 ± 4 | 831 ± 13 79 Chapter 6 Discussion In the literature survey presented in Chapter 2 several ways of producing ultrafine grained steels were examined. DIFT route, the deformation induced ferrite transformation o f austenite, has been presented as a promising technique to obtain a U F F microstructure. In the present study the evaluation of the potential of producing a novel class of steel for automotive application based on ultrafine ferrite (UFF) dual-phase microstructures utilizing D I F T technique has been investigated in two low carbon microalloyed steels (C-1.8 Mn-0.045 Nb-0.145 M o and C-1.8 Mn-0.06 Nb). The effect of prior austenite grain size, degree of undercooling, amount o f strain, strain rate and post deformation isothermal hold have been examined on the final microstructure obtained. The experimental results presented in Chapter 5 have indicated that the controlling parameters in the DIFT route are: (a) refined prior austenite grain size i.e. low austenitization temperature (b) maximizing of undercooling i.e. low deformation temperature for a given austenite condition (c) high strain i.e. high reduction In all cases of the present study, heavy deformation with a true strain of at least 0. 6, i.e. the highest deformation amount employed in the Gleeble tests, to the specimens with small austenite grain size is required to obtain a predominant U F F microstructure, 1. e. U F F fraction of 0.60 or higher with average ferrite grain size of 1 -2 urn. Further, the Chapter 6 DISCUSSION 80 variation of strain rate in the range of 1 to 10s" 1 appears to be of secondary importance with respect to ferrite fraction and ferrite grain size; no significant effects have been found in the present study. Moreover, it is very interesting to note that though the post deformation isothermal hold after the highest employed deformation at the highest undercooling did not register any significant change in terms of U F F grain size and fraction in the final dual phase microstructure, transition from a quasi-polygonal to polygonal U F F morphology has been observed. 6.1.1 Effect of PAGS One of the most crucial parameters for obtaining the U F F microstructure in the present case is a small prior austenite grain size resulting from low austenitization temperature. For example, in the present study this was obtained by austenitizing at a temperature of 950 ° C . It is evident from Fig. 5.13 and Fig. 5.14 that for smaller P A G S the critical strain required for the initiation of formation of U F F grains is lower than (or equal to) 0.2 in both the steels investigated. This is not the case anymore with a higher P A G S irrespective of the two steels. For example, only the highest employed strain of 0.6 has resulted in U F F as illustrated in Fig. 5.15 (b) and Fig . 5.16 (b). It can be speculated from these two figures that at the initial stage of DIFT phenomenon the nucleation of ferrite grains started at the prior austenite grain boundaries, and subsequently progressed into the interior of the grains. These findings are in concurrence to those of Beladi et al. [44, 66] and Hong et al. [50] as discussed earlier. Further, the effective increase in potential ferrite nucleation sites and the strain homogeneity in case of small P A G S has Chapter 6 DISCUSSION 81 resulted in U F F fraction of 0.60 or higher with average ferrite grain size of 1 -2 um, as illustrated by Fig. 5.15 (a) and Fig . 5.16 (a). However, the aforementioned observations from the present study and results of Beladi et al. and Hong et al. are at odds with results of Hurley et al. [54] who reported an increased DIFT potential for increasing austenite grain size. These contradictory results may also be linked to the fact that a change in prior austenite grain size cannot be considered as an isolated parameter since it also affects the deformation temperature in the DIFT schedule which can become a dominant parameter as observed in the present study [87]. 6.1.2 Effect of deformation condition Perhaps the most important parameter in the DIFT schedule is the amount of deformation. The effect of increased amount of strain from 0.2 to 0.6 in terms of increased U F F fraction has already been illustrated in Figs. 5.10 and 5.11. It is to emphasize here that in the present study the highest amount o f deformation employed, i.e. true strain of 0.6 is at least required to obtain a U F F fraction of 0.6 or higher as illustrated in Fig. 5.12. For example, this resulted in U F F fraction of ~ 0.7 with average ferrite grain size of 1.9 um for the Nb + M o steel, and U F F fraction of ~ 0.75 with average ferrite grain size of 1.2 um for the Nb steel, at the highest undercooling and at strain rate of Is"1. The observations possibly indicate that the density of potential nucleation sites increases significantly with strain as a result of progressive substructure development that occurs within the finer prior austenite grains. Subsequently, it is speculated that ferrite nucleates preferentially at the defect substructure formed during deformation. The Chapter 6 DISCUSSION 82 increase in driving force due to deformation leads to a decrease in activation barrier for nucleation of ferrite. The other dominant parameter which is believed to have acted beneficially towards obtaining U F F microstructure is the very high degree of undercooling (i.e. T A e 3 -TDEF) employed in the present study resulting from the fact austenite was cooled fast to the deformation temperature from austenitization temperature, prior to deformation. However, varying this parameter by 25 ° C registered insignificant effect in terms of U F F grain size as depicted in Table 5.6. This finding is in concurrence with that of Niikura et al. [73]. The U F F fraction in the final microstructure decreases marginally when increasing the deformation temperature by 25 ° C . This is illustrated in Fig . 6.1. The trend is readily to be expected because of the concurrent increase in driving force for ferrite transformation with increase in undercooling i.e. decrease in deformation temperature. These findings are similar to the reported results of Hong et al. [74]. Chapter 6 DISCUSSION 83 80 75 H 70 65 60 H 55 1s"1 10s"1 o • Nb + Mo steel A A Nb steel O - i — • — i — 1 — i — 1 — i — 1 — i — 1 — i — ' — i — • — i — 1 120 130 140 150 160 170 180 190 200 210 Undercooling ( T ^ - T J , °C Fig . 6.1 Variation of % U F F formed in the final microstructure with the undercooling employed for smaller austenite grain size at strain of 0.6 and strain rate of Is"1 and 10s"1 A s discussed earlier in Chapter 2, the effect o f increasing strain rate in the formation of U F F by DIFT phenomenon is characterized by the occurrence o f two opposing factors. A n increased impetus for the release of deformation energy is counteracted by lesser time during deformation for DIFT to occur. Significant effects may perhaps be observed at higher levels of strain applied, but in the present study with the application of the highest level o f strain of 0.6, the average U F F grain size did not indicate any trend when comparison is made between otherwise similar conditions with strain rates of 1 and 10s"1 (See Tables 5.6 and 5.7). Chapter 6 DISCUSSION 84 6.1.3 Effect of Steel chemistry N b and M o are known to be both strong carbide forming elements. A s can be seen from Fig . 5.2 and Fig. 5.4, Nb rich precipitates pre-existed, but the entire M o is in solution at 950 ° C . Although there are considerable differences in the continuous cooling austenite-to-ferrite transformation kinetics due to Nb both the steels appear to have quite similar potential for grain refinement produced by DIFT when just the fine prior austenite grain sizes of 10-1 l u m , i.e. reheating at 950 °C , are considered (See Tables 5.6 and 5.7). However, a noticeable difference emerged when the U F F fractions formed in the final microstructures are compared. The effect of M o addition is evident in retardation of austenite to ferrite transformation. 6.2 Grain refining ability of DIFT technique The degree of ferrite grain refinement can be defined as per the following and the same has been calculated based on the results as presented in Table 5.6 and Table 5.8. Degree of refinement = d a R / dao where d a R is the ferrite grain size after DIFT test with true strain of 0.6 at strain rate of 1 s"1 at the highest undercooling with smaller austenite grain size and isothermal holding at the deformation temperature for approximately 380 and 680 seconds respectively, for N b + M o and N b steels, and dao is the ferrite grain size obtained after isothermal ferrite transformation test without any deformation Chapter 6 DISCUSSION 85 A s can be noticed from Table 6.1 the ferrite grain refinement efficiency by the application of DIFT technique in the investigated steels is similar. However, it is important to consider that an increased level o f ferrite grain refinement concomitant to a higher fraction of U F F alongwith martensite has been obtained in the Nb steel. Table 6.1 Degree of ferrite grain refinement by utilizing the technique of DIFT Steel i I i t ! d<xR, um j d a 0 , um Degree of refinement ( d t t R / d a 0 ) j N b + M o 1.9 4.6 i 2.4 N b i 2.o ; 4.9 ! „ , ,i 2.5 6.3 Microstructural refinement and concurrent in crease in strength The general aspects of the underlying mechanisms leading to formation of U F F grains by DIFT can be rationalized on the basis of an increased nucleation rate of ferrite grains in the highly undercooled dislocated austenite. The resulting ferrite grains show a random distribution of their orientations as is evident from Figs. 5.22 to 5.25. However, the presence of a considerable fraction of low angle grain boundaries is also indicated. Although it is a very preliminary investigation constrained by the limit o f spatial resolution in a W-filament S E M the current findings are similar to that reported by other researchers [44-51, 54-61]. However, further investigation is needed with the aid of high resolution E B S D (i.e. F E G S E M - E B S D ) to confirm these indications. Further, Chapter 6 DISCUSSION 86 microstructural evidence indicates U F F grains to be stable against coarsening even after prolonged isothermal holding prior to quenching for at least 6 mins. This stability may also be associated with the presence of precipitates that had been retained due to a rather low austenitizing temperature of 950 ° C . The pinning force of these precipitates had been sufficient to prevent austenite grain growth (see Table 5.1) and the pinning of ultra fine grained structures may be even more efficient as recently discussed by Brechet and Mil i tzer [88]. It is worth noting that though substantial ferrite grain refinement has been obtained in the dual phase steels considered in this study by application of DIFT technique, the morphology of ferrite grains in the as-quenched microstructure is distinctly different from that of the classical coarse grained hot rolled or cold rolled annealed (bare / coated) steels. In the present study in conjunction with the U F F grains formed at a low transformation temperature, the non-uniform distribution of martensite islands in the former is clearly observed. Hence, the observed very high strength in the investigated steels as reported in Table 5.9 may possibly be, partly attributable to transformation strengthening of low temperature transformation of austenite to ferrite, apart from the major part believed to be resulting from the significant ferrite grain refinement. 87 Chapter 7 Conclusions and Future Work 7.1 Conclusions In this investigation ferrite grain refinement in two low carbon microalloyed steels, i.e. one microalloyed with Nb & M o and the other microalloyed with only Nb , for producing dual phase (ferrite + martensite) microstructures has been systematically studied. The refinement of ferrite grains have been achieved by employing the novel thermo-mechanical processing technique of deformation induced ferrite transformation (DIFT). Based on the experimental results the following conclusions can be made: 1. The austenite decomposition is greatly influenced by the initial austenite grain size for a given cooling rate. A n increase in austenite grain size results in lower transformation start temperatures and non-polygonal microstructures. 2. Refining the prior austenite grain size (PAGS) highly enhances the ferrite grain refinement. The effective increase in potential ferrite nucleation sites and the strain homogeneity in case of small P A G S of approximately 10 um has resulted in U F F (Ultrafine ferrite) fraction of 0.60 or higher with average ferrite grain size of 1 -2 um. 3. In all the cases of the present study, heavy deformation with a true strain of at least 0.6, i.e. the highest deformation amount employed in the Gleeble tests, is Chapter 7 CONCLUSIONS AND FUTURE WORK 88 required to obtain a predominant U F F microstructure, i.e. U F F fraction of at least 0.60. 4. The effect of varying degree of undercooling is evident only in terms of U F F fraction i.e. increase in undercooling results in an increase in U F F fraction. The effect of the same on U F F grain size is insignificant. 5. Strain rate seems to have a secondary effect on the DIFT potential for the conditions that have been investigated. 6. The combined addition of M o and Nb leads to enhanced retardation of the austenite to ferrite transformation. Hence between the two steels studied, the single addition of Nb seems to be more beneficial in terms of obtaining optimum U F F fraction in fine grained D P steel. 7. The minimal growth registered for the U F F grains during post-deformation isothermal holding may indicate significant pinning due to the presence of carbides that were not dissolved during the austenitization treatment. 7.2 Future Work 1. The U F F microstructure obtained needs to be characterized by high resolution E B S D i.e. F E G S E M - E B S D as this might provide relevant information regarding their evolution with strain. 2. The concomitant work hardening characteristics with the increased level of strength in these dual phase steels need to be ascertained possibly by correlating with the result of "shear punch" tests or "automated ball indentation" tests. Chapter 7 CONCLUSIONS AND FUTURE WORK 89 3. It is proposed to further investigate the potential of DIFT phenomenon to produce U F F - D P steels with hot torsion tests that can simulate the entire processing o f a hot strip mi l l . Further, these studies would also allow quantifying the mechanical properties with a regular tensile test. 4. A l so the investigation of DIFT from pancaked austenite may be of interest, particularly for the Nb containing steels. 90 References [1] E . O. H a l l ; Proc. of Phys. Soc, B 64, Inst, of Physics, London,1951, p. 747 [2] N . J. Petch; Journal of Iron Steel Inst., 1953, 174, p. 25 [3] A . Sato, Research project on innovative steels in Japan (STX-21 Project), Proc. Ultra-Steel 2000 Intl. Workshop on the Innovative Structural Materials for Infrastructure in 21st Century, Tsukuba City, Japan, 2000, p. 1 [4] I. Tamura, C. Ouchi, T. Tanaka and H . Sekine; Thermomechanical Processing of High Strength Low Alloy Steels, Butterworths & Co. Ltd., London, 1988, p. 30. [5] T. Ishikawa, H . Yoshikawa, T. Inoue, Y . Nomiyama and S. Imai; Nippon Steel Tech. Rep. 1997,75, p. 31 [6] I. Chikushi, M . Kozaki , R. Kurahashi and M . Takahashi; AISE Steel Technology, 2002, 79, p. 42 [7] T. Morimoto, I. Chikushi, R. Kurahashi, and J. Yanagimoto; Proc. ofTMP 2004 : 2nd Intl. Conf on Thermomechanical Processing of Steels, Stahleisen Verlag G m b H , Dusseldorf, 2004, p. 415 [8] I. Tamura, C . Ouchi, T. Tanaka and H . Sekine; Thermomechanical Processing of High Strength Low Alloy Steels, Butterworths & Co. Ltd. , London, 1988, p. 80. [9] F. H . Samuel, S. Yue, J. J. Jonas and K . R. Barnes; ISIJIntl., 1990, 30, p. 216. [10] P. D . Hodgson, H . Beladi and G . L . Kel ly ; COM 2004, Intl. Symp. On Ultrafine Structured Steels, Metallurgical Society of C I M , Montreal, PQ, 2004, p. 19 [11] S. V . Subramanian, G . Zhu, C. Klinkenberg and K . Hulka; Mater. Sci. Forum, 2 0 0 5 , 4 7 5 - 4 7 9 , p. 141 REFERENCES 91 [12] M . Umemoto, A . Hiramatsu, A . Moriya , T. Watanbe, S. Nanba, N . Nakjima, G . Anan and Y . Higo; ISIJIntl, 1992, 32, p. 306 [13] D . N . Hanlon, J. Sietsma and S. van der Zwaag; ISIJIntl, 2001, 41, p. 1028 [14] R. Rriestner and A . K . Ibraheem; Mat. Sci. and Tech., 2000, 163, p. 1267 [15] R. Bengochea, B . Lopez and I. Gutierrez; ISIJIntl, 1999, p. 583 [16] M . Mili tzer, R. Pandi, E . B . Hawbolt and T.R. Meadowcroft; Accelerated Cooling / Direct Quenching of Steels, Ed . R. Asfahani, A S M , Materials Park, O H , 1997, p. 151 [17] R. K . Gibbs, B . A . Parker and P. D . Hodgson; Intl. Symp. on Low Carbon Steels for the 90's, Ed. R. Asfahani and G . Tither, T M S , 1993, p. 193 [18] R. Pandi, M . Militzer, E . B . Hawbolt and T.R. Meadowcroft; 37th Mechanical Working and Steel Processing Conf. Proc, ISS, Warrendale, P A , V o l . X X X I I I , 1996, p. 635 [19] N . Nakata and M . Mili tzer; 42nd Mechanical Working and Steel Processing Conf. Proc, ISS, Warrendale, P A , V o l . X X X V I I I , 2000, 38, p. 813. [20] P. Petkov; MASc Thesis, Univ . of British Columbia, Vancouver, Canada, 2004 [21] M . Mil i tzer and F. Fazeli; 44th Mechanical Working and Steel Processing Conf. Proc, ISS, Warrendale, P A , V o l . X L , 2002, p. 1129 [22] N . Nakata and M . Militzer; ISIJIntl, 2005,45, 1, p. 82 [23] R. Lottey; MASc Thesis, Univ. of British Columbia, Vancouver, Canada, 2004 [24] V . M . Segal, V . I. Reznikov, A . D . Drobyshevski and V . I. Kopylov; Russ. MetalL, 1981, l , p . 99 REFERENCES 92 [25] F. J. Humphreys, P. B . Prangnell, J. R Bowen, A . Gholinia and C. Harris; Philos. Trans. R. Soc .A, 1998, 357, p. 1663 [26] J. R. Bowen, P. B . Prangnell and F. J. Humphreys; Proc. 20th Risfilntl. Symp. on Materials Science, Ed . J. B . Bilde-S(|)rensen et al., Roskilde, Ris<t> National Lab., 1999, p. 269 [27] C. Harris, P. B . Prangnell and X . Duan; Proc. 6th Intl. Conf. on 'Aluminium alloys' (ICAA-6), Japan Institute of Metals, 1998, p. 583 [28] K . Matsubara, Y . Miyahara, Z . Horita and T. Langdon; Met. Trans. A, 2004, 35A, p. 1735 [29] R. Z Valiev, R. K . Islamgaliev and I. V . Alexandrov; Prog, in Mater. Sci., 2000, 45, p. 103 [30] A . K . Ibraheem, R. Priestner, J. R. Bowen, P. B . Prangnell and F. J. Humphreys; Ironmaking and Steelmaking, 2001, 28, p. 203 [31] D . H . Shin, I. K i m , J. K i m , and K . Park; Acta Mater., 2001, 49, p. 1285 [32] K . - T . Park, Y . S. K i m , J. G . Lee and D . H . Shin; Mater. Sci. & Eng. A, 2000, 293, p. 165 [33] D . H . Dobatkin; Intl. Symp. on Ultrafine Grained Steels (ISUGS), 2001, ISIJ, Tokyo, Japan, p. 122 [34] K . - T . Park, S. Y . Han, B . D . Ahn , D . H . Shin, Y . K . Lee and K . K . U m ; Scripta Mater, 2004, 51, p. 909 [35] Y . Son, Y . K . Lee, K . -T. Park, C . S. Lee and D . H . Shin; Acta Mater., 2005, 53, p. 3125 [36] Y . Saito, H . Utsunomiya, N . Tsuji and T. Sakai; Acta. Mater., 1999, 47, p. 579 REFERENCES 93 [37] N . Tsuji, Y . Saito, H . Utsonomiya, S. Tanigawa; Scripta Mater., 1999, 40, p. 795 [38] R. Ueji , N . Tsuji, Y . Minamino and Y . Koizumi ; Intl. Symp. on Ultrafine Grained Steels (ISUGS), ISIJ, Tokyo, Japan, 2001, p. 222 [39] R. Ueji , N . Tsuji, Y . Minamino and Y . Koizumi ; Acta. Mater., 2002, 50, p. 4177 [40] R. Ueji , N . Tsuji, Y . Minamino and Y . Koizumi ; Sci. and Tech. of Adv. Mater.; 2004, 5, p. 153 [41] S. Morito, H . Tanaka, T. Furuhara, and T. M a k i , Proceedings of the 4th international conference on recrystallization and related annealing phenomenon (Rex'99), Eds. T. Sakai, and H . G . Suzuki, Japan Institute of Metals, 1999, p. 295 [42] Q. Cu i , and K . Ohori, Mat. Sci and Tech., 2000, 16, p. 1095 [43] K . - M . Lee, J. - H . Kang and H . C. Lee, Proc. 5th Workshop on the Development of High Performance Structural Steels for 21st Century (HIPERS21), Korean Institute of Metals and Materials, Pohang, Korea, 2003, p. 212 [44] H . Beladi, G . L . Kel ly , A . Shokouhi and P. D . Hodgson; Mater. Sci. & Eng. A, 2004,367, p. 152 [45] H . Yada, Y . Matsumura and K , Nakazima; United States Patent No . - 4466842, August 21, 1984 [46] Y . Matsumura and H . Yada; Trans. ISIJ, 1987, 27, p. 492 [47] H . Yada and Y . Matsumura; Proc. Int. Conf. Physical Metallurgy of Thermomechanical Processing of Steel and Other Materials (Thermec '88). ISIJ, Tokyo, (1988), p. 200 [48] C. - M L i , H . Yada and H . Yamagata; Scripta Mater., 1998, 39, p. 963 REFERENCES 94 [49] J. - K . Choi , D . Suh, J. Kang, and H . C. Lee, Ultrafine grained materials II, Eds. Y . T. Zhu, T. G . Langdon, R. S. Mishra, S. L . Semiatin, M . J. Saran, and T. C . Lowe, T M S , 2002, p. 259 [50] S. C . Hong and K . S. Lee; Mater. Sci. & Eng. A, 2002, 323, p. 148 [51] J. - K . Choi , D . Seo, J. Lee, K . U m , and W . Choo; ISIJIntl, 2003, 43, p. 746 [52] P. D . Hodgson, M . R. Hickson and R. K . Gibbs; Scripta Mater., 1999, 40, p. 1179 [53] M . Niikura, M . Fujioka, Y . Adachi, A . Matsukura, T. Yokota, Y . Shirota and Y . Hagiwara; J. Mater. Proc. Tech., 2001, 117, p. 341 [54] P. J. Hurley and P. D . Hodgson; Mater. Sci. Eng. A, 2001, 302, p. 206 [55] S. Lee, D . Kwon , Y . K . Lee and O. K w o n ; Met. Trans. A, 1995, 26A, p. 1093 [56] K . Nagai; J. of Mater. Proc. & Tech., 2001, 117, p. 329 [57] Z . Q. Sun, W . Y . Yang, J. J. Q i and A . M . Hu; Mater. Sci. Eng. A, 2002, 334, p. 201 [58] Y . D . Huang and L . Froyen; J. Mater. Proc. & Tech., 2002, 124, p. 216 [59] M . R. Hickson, P. J. Hurley, R. K . Gibbs, G . L . Kel ly and P. D. Hodgson; Met. Trans. A, 2002, 33A, p. 1019 [60] L . Wei-juan, L . Cui-qin, W . Guo-dong and L . Xiang-hua; J. Iron & Steel Res. Intl., 2003, 10, p. 54 [61] H . Beladi, G . L . Ke l ly and P. D . Hodgson; Mater. Trans., 2004, 45, A , p. 2214 [62] K . Nagai; COM 2004, Intl. Symp. On Ultrafine Structured Steels, Metallurgical Society of C I M , Montreal, PQ, 2004, p. 3 REFERENCES 95 [63] D . - W . Suh, C, -S. Oh and S.-J. K i m ; 2005, Met. Trans. A, 2005, 36A, p. 1057 [64] S. V . S. Narayan Murty, S. Torizuka, K . Nagai, T. Ki ta i and Y . Kogo; Scripta Mater., 2005, 40, p. 1179 [65] J. S. Lee, J. - K . Choi and W . Y . Choo; ICASS 2004: Proc. of the 2nd Intl. Conf. on Advanced Structural Steels, The Chinese Society for Metals, Shanghai, China, A p r i l 1 4 - 1 6 , 2004, p. 172 [66] H . Beladi, G . L . Ke l ly , A . Shokouhi and P. D. Hodgson; Mater. Sci. & Eng. A, 2004, 371, p. 343 [67] E . I. Poliak and J. J. Jonas; Acta Mater, 1996, 44, p. 127 [68] P. J. Hurley and P. D . Hodgson; Mat. Sci. and Tech., 2001, 17, p. 1360 [69] T. Inoue, S. Torizuka and K . Nagai; Mat. Sci. and Tech., 2002, 18, p. 1007 [70] T. Inoue, S. Torizuka and K . Nagai; Mat. Sci. and Tech., 2001, 17, p. 1329 [71] Z . Yang and R. Wang; ISIJIntl, 2003, 43, p. 761 [72] R. Priestener, Y . M . Al-Horr and A . K . Ibraheem; Mat. Sci. and Tech., 2002, 18, p. 973 [73] M . Niikura, M . Fujioka, M , . Adachi, A . Matsukura, T. Yokota, Y . Shirota, Y . Hagiwara; J. Mater. Proc. & Tech., 2001,117, p. 341 [74] S. C. Hong, S. H . L i m , K . J . Lee, D . H . Shin, K . S. Lee; ISIJIntl, 2003, 43, p. 761 [75] H . Yada, C . - M L i and H . Yamagata; ISIJIntl, 2000, 40, p. 200 REFERENCES 96 [76] A . Lucas, P. Simon, G . Bourdon, J. - C . Herman, P. Riche, J. Neutjens and P. Harlet; Steel Res. Intl., 2004, 75, p. 139 [77] M . R. Hickson, R. K Gibbs and P. D . Hodgson; ISIJIntl., 1999, 39, p. 1176 [78] S. C . Hong, S. H . L i m , H . S. Hong, K . J. Lee, D . H . Shin, and K . S. Lee; Mater. Sci. and Eng. A, 2003, 55, p. 241 [79] N . Tsuji, Y . Ito, Y . Saito, Y . Minamino; Scripta Mater., 2002, 47, p. 893 [80] A . Ohmori, S. Torizuka and K . Nagai; ISIJIntl., 2004, 44, p. 1063 [81] K . Mukherjee, PhD Thesis Proposal; Univ . of British Columbia, Vancouver, Canada, 2005, p. 29 [82] G.F . Vander Voort; Metallography: Principles and practice, M c G r a w - H i l l Book Co. Ltd. , New York, 1984, p. 219 [83] Annual Book o f A S T M Standards, vol . 3.01, ASTM Standard Designation, E 112-96 (2004), Philadelphia Pa, 2004, p.227 [84] Annual Book of A S T M Standards, vol . 3.01, ASTM Standard Designation, E 3 84-99, Philadelphia Pa, 2000, p. 437 [85] A . Giumell i , M . Mili tzer and B . Hawbolt; ISIJIntl., 1999, 39, p. 271 [86] D . Tabor; The hardness of metals, Oxford, Clarendon Press, 1951 [87] K . Mukherjee, S. Hazra and M . Mil i tzer ; in Innovations in Steel Sheet Products and Processing, S A E - 0 6 M 114, S A E , Warrendale, P A , 2006, in press [88] Y . Brechet and M . Militzer; Scripta Mater., 2005, 52, p. 1299 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0078761/manifest

Comment

Related Items