Light Alloys - Metallurgy of the Light Metals - PDF

Light Alloys  Light Alloys Metallurgy of the Light Metals Fifth Edition Ian Polmear David StJohn Jian-Feng Nie Ma Qian Butterworth-Heinemann is an imprint of Elsevier The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, United Kingdom 50...

Light Alloys  Light Alloys Metallurgy of the Light Metals Fifth Edition Ian Polmear David StJohn Jian-Feng Nie Ma Qian Butterworth-Heinemann is an imprint of Elsevier The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, United Kingdom 50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States Copyright © 2017 Ian Polmear, David StJohn, Jian-Feng Nie, Ma Qian. Published by Elsevier Ltd. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Details on how to seek permission, further information about the Publisher’s permissions policies and our arrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website: www.elsevier.com/permissions. This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may be noted herein). Notices Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understanding, changes in research methods, professional practices, or medical treatment may become necessary. Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any information, methods, compounds, or experiments described herein. In using such information or methods they should be mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility. To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in the material herein. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the Library of Congress ISBN: 978-0-08-099431-4 For Information on all Butterworth-Heinemann publications visit our website at https://www.elsevier.com/books-and-journals Publisher: Matthew Deans Acquisition Editor: Christina Gifford Editorial Project Manager: Ana C Glaudia Garcia Production Project Manager: Susan Li Designer: Mark Rogers Typeset by MPS Limited, Chennai, India  PREFACE TOTHE FIRST EDITION The fact that the light metals aluminium, magnesium and titanium have tradi- tionally been associated with the aerospace industries has tended to obscure their growing importance as general engineering materials. For example, alu- minium is now the second most widely used metal and production during the next two decades is predicted to expand at a rate greater than that for all other structural metals. Titanium, which has a unique combination of properties that have made its alloys vital for gas turbine engines, is now finding many applica- tions in aircraft structures and in the chemical industry. Light alloys have never been the subject of a single book. Moreover, although the general metallurgy of each class of light alloys has been covered in individual texts, the most recent published in English appeared some time ago—aluminium alloys in 1970, magnesium alloys in 1966 and titanium alloys in 1956. Many new developments have occurred in the intervening periods and important new applications are planned, particularly in transportation. Thus it is hoped that the appearance of this first text is timely. In preparing the book I have sought to cover the essential features of the metallurgy of the light alloys. Extraction of each metal is considered briefly in Chapter one, after which the casting characteristics, alloying behaviour, heat treatment, properties, fabrication and major applications are discussed in more detail. I have briefly reviewed the physical metallurgy of aluminium alloys in Chapter two although the general principles also apply to the other metals. Particular attention has been devoted to microstructure/property relationships and the role of individual alloying elements, which provides the central theme. Special features of light alloys and their place in general engineering are high- lighted although it will be appreciated that it has not been possible to pursue more than a few topics in depth. The book has been written primarily for students of metallurgy and engi- neering although I believe it will also serve as a useful guide to both producers and users of light alloys. For this reason, books and articles for further read- ing are listed at the end of each chapter and are augmented by the references included with many of the figures and tables. ix x PREFACE TO THE FIRST EDITION The book was commenced when I was on sabbatical leave at the Joint Department of Metallurgy at the University of Manchester Institute of Science and Technology and University of Manchester, so that thanks are due to Professor K. M. Entwistle and Professor E. Smith for the generous facili- ties placed at my disposal. I am also indebted for assistance given by the Aluminium Development Council of Australia and to many associates who have provided me with advice and information. In this regard, I wish partic- ularly to mention the late Dr E. Emley, formerly of The British Aluminium Company Ltd; Dr C. Hammond, The University of Leeds; Dr M. Jacobs; TI Research Laboratories; Dr D. Driver, Rolls-Royce Ltd; Dr J. King and Mr W. Unsworth, Magnesium Elektron Ltd; Mr R. Duncan, IMI Titanium; Dr D. Stratford, University of Birmingham; Dr C. Bennett, Comalco Australia Ltd; and my colleague Dr B. Parker, Monash University. Acknowledgement is also made to publishers, societies and individuals who have provided figures and diagrams which they have permitted to be reproduced in their original or modified form. Finally I must express my special gratitude to my secretary Miss P. O’Leary and to Mrs J. Colclough of the University of Manchester who typed the man- uscript and many drafts, as well as to Julie Fraser and Robert Alexander of Monash University who carefully produced most of the photographs and diagrams. IJP Melbourne 1980  PREFACETOTHE SECOND EDITION In this second edition, the overall format has been retained although some new sections have been included. For the most part, the revision takes the form of additional material that has arisen through the development of new compo- sitions, processing methods, and applications of light alloys during the last 8 years. Most changes have occurred with aluminium alloys which, because of their widespread use and ease of handling, are often used to model new pro- cesses. Faced with increasing competition from fiber-reinforced plastics, the aluminium industry has developed a new range of lightweight alloys contain- ing lithium. These alloys are discussed in detail because they are expected to be important materials of construction for the next generation of passenger aircraft. More attention is given to the powder metallurgy route for fabricat- ing components made from aluminium and titanium alloys. Treatment of this topic includes an account of techniques of rapid solidification processing which are enabling new ranges of alloys to be produced having properties that are not attainable by conventional ingot metallurgy. Metal–matrix composites based on aluminium are also finding commercial applications because of the unique properties they offer and similar magnesium alloys are being developed. New methods of processing range from methods such as squeeze casting through to advances in superplastic forming. In preparing this new edition, I have again paid particular attention to microstructure/property relationships and to the special features of light alloys that lead to their widespread industrial use. In addition to an expanded text, the number of figures has been increased by some 40% and the lists of books and articles for further reading have been extended. Once more, the book is directed primarily at undergraduate and postgraduate students although I believe it will serve as a useful guide to producers and users of light alloys. I am again indebted for assistance given by colleagues and associates who have provided me with information. Acknowledgment is also made to publish- ers, societies, and individuals who have provided photographs and diagrams which they have permitted to be produced in their original or modified form. xi xii Preface to the Second Edition Finally I wish to express my gratitude to Mesdames J. Carrucan, C. Marich, and V. Palmer, who typed the manuscript, as well as to Julie Fraser, Alan Colenso, and Robert Alexander of Monash University who carefully produced most of the photographs and diagrams. Melbourne, 1988  PREFACE TO THE THIRD EDITION The central theme of the first two editions was microstructure/property relation- ships in which special attention was given to the roles of the various alloying elements present in light alloys. This general theme has been maintained in the third edition although some significant changes have been made to the format and content. As before, much of this revision involves the inclusion of new material which, in this case, has arisen from developments during the seven years since the second edition was published. The most notable change in format has been to group together, into a new chapter, information on what have been called new materials and processing methods. Examples are metal matrix and other com- posites, structural intermetallic compounds, nanophase and amorphous alloys. Interest in these and other novel light alloys has increased considerably during the last decade because of the unceasing demands for improvements in the prop- erties of engineering materials. Since light alloys have been at the forefront of many of these developments, the opportunity has been taken to review this area which has been the focus of so much recent research in materials science. Another feature of the third edition is the greater attention given to applica- tions of light alloys and their place in engineering. More case studies have been included, such as the use of light alloys in aircraft and motor cars. Economic factors associated with materials selection are also discussed in more detail. Moreover, since the light metals are often placed at a competitive disadvan- tage because of the high costs associated with their extraction from minerals, more attention has been given to these processes. This has led to a considerable increase in the size of the first chapter. Joining processes are described in more detail and, once again, service performance of light alloys is discussed with particular regard to mechanical behaviour and corrosion resistance. As a result of these various changes, the text has been expanded and the number of figures has been increased by a further 20%. Lists of books and articles for further reading have been updated. While the book continues to be directed primarily at senior undergraduate and postgraduate students, I believe it will again serve as a useful guide to the producers and users of light alloys. xiii xiv PREFACE TO THE THIRD EDITION I am again indebted for assistance given by colleagues and associates who have provided me with information and helpful discussions. General acknowl- edgement is made to publishers, societies and individuals who have responded to requests for photographs and diagrams that have been reproduced in their original or modified form. Finally I wish to express my gratitude to my wife Margaret for her constant encouragement, to Carol Marich and Pam Hermansen who typed the manuscript and to Julie Fraser and Robert Alexander who once again carefully produced so many of the photographs and diagrams. I. J. Polmear Melbourne 1995  PREFACE TO THE FOURTH EDITION Since the third edition of Light Alloys appeared in 1995, developments with new alloys and processes have continued at an escalating rate. Competition between different materials, metallic and nonmetallic, has increased as produc- ers seek both to defend their traditional markets and to penetrate the markets of others. New compositions of aluminium, magnesium, and titanium alloy have been formulated, and increasing attention has been given to the development of novel and more economical processing methods. Because of their ease of han- dling, aluminium alloys in particular have been used as experimental models for many of the changes. Recently, potential automotive applications have led to a resurgence of interest in cast and wrought magnesium alloys. The central theme of earlier editions was microstructure/property relation- ships, and particular attention was given to the roles of the various alloying ele- ments present in light alloys. This general theme has been maintained in the fourth edition although further significant changes have been made to format and content. Special consideration has again been given to the physical metallurgy of aluminium alloys and many of the general principles also apply to magnesium and titanium alloys. The description of changes occurring during the process- ing of the major class of nonheat-treatable aluminium alloys has been extended. Although a century has now elapsed since the discovery of age hardening by Alfred Wilm, new observations are still being made as the latest experimental techniques reveal more details of the actual atomic processes involved. As exam- ples, more information is now available about the role of solute and vacancy clusters during the early stages of aging, as well as other phenomena such as secondary hardening. Some success has been achieved with the modeling of pre- cipitation processes. Precipitation hardening was hailed as the first nanotechnol- ogy and now it is possible to develop fine-scale microstructures in a much wider range of alloys through the use of novel processing methods. Some new topics in this fourth edition are strip and slab casting, creep forming, joining technologies such as friction stir and laser welding, metallic foams, quasicrystals, and the production of nanophase materials. Economic factors associated with the production and selection of light metals and alloys xv xvi PREFACE TO THE FOURTH EDITION are considered in more detail and information on recycling has been included. Sections dealing with the commercial applications of light alloys and their gen- eral place in engineering have also been expanded. This applies particularly to transportation as aluminium alloys face increasing competition from fiber-rein- forced polymers for aircraft structures, and where higher fuel costs make both aluminium and magnesium alloys more attractive for reducing the weight of motor vehicles. Because of these and other developments, the text has been updated and expanded by about 20%. A further 50 figures and several new tables have been added. References to original sources of information are shown with most fig- ures and tables but are not included in the general text. Relevant articles and books for further reading have been revised and are listed at the end of each chapter. As originally intended, the book is directed primarily at senior under- graduate and postgraduate students, but it is also believed it will to serve as a useful general guide to producers and users of light alloys. I am again indebted for the assistance given by colleagues and associates who have provided me with information and advice. In this regard, special mention should be made of J. Griffiths, E. Grosjean, J. Jorstad, R. Lumley, J.-F. Nie, R.R. Sanders, H. Shercliff, J. Taylor, and the Australian Aluminium Council. General acknowledgment is again made to publishers, societies, and individuals who have responded to requests for photographs and diagrams. Facilities provided by the Department of Materials Engineering at Monash University have been much appreciated and, as with all other editions, many of the figures have been skillfully reproduced in their original or modified form by Julie Fraser. Finally I wish to express my gratitude to my wife Margaret for her constant support. I. J. Polmear Melbourne, 2005  PREFACE TO THE FIFTH EDITION When the first edition of Light Alloys was published in England in 1981, the metallurgy of aluminium, magnesium, and titanium alloys had not been reviewed in a single book. Since then three more editions have been prepared, the fourth one having appeared in 2006. During the last decade, more light alloys have been developed, new processing methods have been adopted, and applications for these materials have expanded. To update these interesting developments, I have been delighted to welcome as co-authors three Australian colleagues with research interests in light alloys, namely Prof. David StJohn, University of Queensland, Prof. Jian-Feng Nie, Monash University, and Prof. Ma Qian, RMIT University. I thank them all for their valuable contributions to this fifth edition. The central theme of the earlier editions has been microstructure/property relationships in which special attention was given to the specific roles of the various alloying elements present in aluminium, magnesium, and titanium alloys. This theme has been maintained in this fifth edition and various top- ics have been introduced or expanded. As an introduction to the three metals, Chapter 1 describes their characteristics, availability, methods of production, and global importance. The physical metallurgy of aluminium alloys is then reviewed and these general principles also apply to the alloys of the other two light metals. Examples are given to the remarkable ability of modern micro- scopic techniques to reveal the actual atomic events occurring during heat treat- ment processes such as precipitation hardening. A new chapter has been added to outline the theory and practice of casting processes used to make products from aluminium and magnesium alloys. Wrought aluminium alloys remains the largest chapter since approximately 70% of all aluminium is used for manufac- turing these products. Chapters devoted to magnesium and titanium alloys have both been expanded by one-third and a section on additive manufacturing by 3D printing has been included in a final chapter that is concerned with novel materials and their processing methods. xvii xviii Preface to the Fifth Edition Because of these various developments, the text for this fifth edition has been expanded by about 25% and more than 80 new or modified figures have been included. Tables have also been updated. Relevant articles and books for further reading are listed at the end of each chapter. As with earlier editions, the book is directed primarily at senior undergraduate and postgraduate students, although it will serve as a useful guide for producers and users of light alloys. Melbourne, 2016  1 THE LIGHT METALS 1.1 GENERAL INTRODUCTION The term “light metals” has traditionally been given to both aluminium and magnesium, because they are frequently used to reduce the weight of compo- nents and structures. On this basis, titanium also qualifies and beryllium should be included although it is little used and will only be mentioned briefly later. These four metals have relative densities ranging from 1.7 (magnesium) to 4.5 (titanium) which compare with 7.9 and 8.9 for the older structural metals, iron, and copper, and 22.6 for osmium, the heaviest of all metals. Ten other elements that are classified as metals are lighter than titanium but, with the exception of boron in the form of strong fibers embedded in a suitable matrix, none is used as a base material for structural purposes. The alkali metals lithium, potassium, sodium, rubidium, and cesium, and the alkaline earth metals calcium and stron- tium are too reactive, whereas yttrium and scandium are comparatively rare. 1.1.1 Characteristics of light metals and alloys The property of lightness translates directly to material property enhance- ment for many products since by far the greatest weight reduction is achieved by a decrease in density (Fig. 1.1). This is an obvious reason why light met- als have been associated with transportation, notably aerospace, which has pro- vided great stimulus to the development of light alloys during the last century. Strength:weight ratios have also been a dominant consideration and the central positions of the light alloys based on aluminium, magnesium, and titanium with respect both to other engineering alloys and to all materials are represented in an Ashby diagram in Fig. 1.2. The advantages of decreased density become even more important in engineering design when parameters such as stiffness and resistance to buckling are involved. For example, the stiffness of a simple rectangular beam is directly proportional to the product of the elastic modulus Light Alloys. DOI: http://dx.doi.org/10.1016/B978-0-08-099431-4.00001-4 Copyright © 2017 Ian Polmear, David StJohn, Jian-Feng Nie, Ma Qian. 1 Published by Elsevier Ltd. All rights reserved. 2 CHAPTER 1  The light metals 25 Density Structural weight change 20 Tensile strength 15 Modulus 10 Compressive 5 yield strength 0 20 40 60 80 100 Parameter change Figure 1.1 Effect of property improvement on structural weight. Courtesy from Lockheed Corporation. and the cube of the thickness. The significance of this relationship is illustrated by the nomograph shown in Fig. 1.3 which allows the weights of similar beams of different metals and alloys to be estimated for equal values of stiffness. An iron (or steel) beam weighing 10 kg will have the same stiffness as beams of equal width and length weighing 7 kg in titanium, 4.9 kg in aluminium, 3.8 kg in magnesium, and only 2.2 kg in beryllium. The Mg–Li alloy is included because it is the lightest (relative density 1.35) structural alloy that is available com- mercially. Comparative stiffnesses for equal weights of a similar beam increase in the ratios 1:2.9:8.2:18.9 for steel, titanium, aluminium, and magnesium respectively. Concern with aspects of weight saving should not obscure the fact that light metals possess other properties of considerable technological importance. Examples are the generally high corrosion resistance and high electrical and thermal conductivities of aluminium, the castability and machinability of mag- nesium, and extreme corrosion resistance of titanium. Comparisons of some physical properties are made in Table 1.1. 1.1.2 Beryllium This element was discovered by Vauquelin in France in 1798 as the oxide in the mineral beryl (beryllium aluminium silicate) and in emerald. It was first isolated independently by Wöhler and Bussy in 1828 who reduced the chloride with potassium. Beryl has traditionally been a by-product of emerald mining and was until recently the major source of beryllium metal. Currently more beryllium is extracted from the closely associated mineral bertrandite (beryl- lium silicate hydroxide). Beryllium has some remarkable properties (Table 1.1). Its stiffness, as measured by specific elastic modulus, is nearly an order  1.1 General introduction 3 10,000 Strength-density Engineering SiC Diamond Metal and polymers: yield strength ceramics Sialons Engineering Ceramics and glasses: compressive strength alloys Elastomers: tensile tear strength B Cermets Composites: tensile failure Glasses Sl CFRP 1000 GFRP UNIPLY Steels Engineering KFRP Pottery Ti W alloys CFRP alloys composites Mo alloys GFRP laminates Cast KFRP irons Nl alloys Al alloys Cu alloys Mg Stone. alloys Zn rock alloys 100 ASH OAK Nylons PMMA Engineering Strength σf (MPa) Pine Fir PP MEL alloys Parallel PVC To grain PS Wood Epoxies Lead products Polyesters Cemeni alloys Balsa Hope concrete ASH PTFE Woods OAK Pine Porous Fir PU ceramics 10 Perpendicular to grain Engineering Lope Silicone polymers Soft Balsa butyl Elastomers Polymers foams Cork 1 0.1 0.1 0.3 1 3 10 30 Density ρ (g/cm3) Figure 1.2 The strength:density ratios for light alloys and other engineering materials. Note that yield strength is used as the measure of strength for metals and polymers, com- pressive strength for ceramics, tear strength for elastomers, and tensile strength for com- posites. Courtesy from M. F. Ashby. of magnitude greater than that for the other light metals, or for the commonly used heavier metals iron, copper, and nickel. This has led to its use in gyro- scopes and in inertial guidance systems. It has a relatively high melting point, and its capture cross section (i.e., permeability) for neutrons is lower than for any other metal. These properties have stimulated much interest by the aero- space and nuclear industries. For example, a design study specifying beryl- lium as the major structural material for a supersonic transport aircraft has indicated possible weight savings of up to 50% for components for which it could be used. However, its structural uses have been confined largely to com- ponents for spacecraft and for applications such as satellite antenna booms. 4 CHAPTER 1  The light metals X Berylium Magnesium-lithium Magnesium Aluminium Titanium Vanadium Steel Zirconium Monel 1 2 3 4 5 6 7 8 9 10 Comparative weight — equal stiffness Figure 1.3 Nomograph allowing the comparative weights of different metals or alloys to be compared for equal levels of stiffness. These values can be obtained from the intercepts that lines drawn from point X make with lines representing the different metals or alloys. Courtesy from Brooks and Perkins Inc. In nuclear engineering it has had potential for use as a fuel element can in power reactors. Another unique property of beryllium is its high specific heat which is approximately twice that of aluminium and magnesium, and four times that of titanium. This inherent capacity to absorb heat, when combined with its low density, led to the selection of beryllium as the basis for the reentry heat shield of the Mercury capsule used for the first manned spacecraft devel- oped in the United States. In a more general application, it has served as a heat sink when inserted in the center of composite disk brakes used in the landing gear of a large military transport aircraft. Beryllium also shows outstanding optical reflectivity, particularly in the infrared, which has led to its combat use in target acquisition systems as well as in space telescopes. Despite much research in several countries, wider use has not been made of beryllium because it is costly to mine and extract, it has an inherently low ductility at ambient temperatures, and the fact that the powdered oxide is extremely toxic to some people. The problem of low ductility arises because of the dimensions of the close-packed hexagonal crystal structure of beryllium. The c/a ratio of the unit cell is 1.567 which is the lowest and most removed of all metals from the ideal value of 1.633. One result of this is a high degree of anisotropy between mechanical properties in the a and c crystallographic direc- tions. At room temperature, slip is limited and only possible on the basal plane, which also happens to be the plane along which cleavage occurs. Furthermore, there has also been little opportunity to improve properties by alloying because Table 1.1 Some physical properties of pure metals Property Unit Al Mg Ti Be Fe Cu Atomic number − 13 12 22 4 26 29 Relative atomic mass (C= 12.000) − 26.982 24.305 47.90 9.012 55.847 63.546 Crystal structure − fcc cph cph cph bcc fcc a nm 0.4041 0.3203 0.2950 0.2286 0.2866 0.3615 c nm − 0.5199 0.4653 0.3583 − − Melting point °C 660 650 1678 1289 1535 1083 Boiling point °C 2520 1090 3289 2472 2862 2563 Relative density (d) − 2.70 1.74 4.51 1.85 7.87 8.96 Elastic modulus (E) GPa 70 45 120 295 211 130 Specific modulus (E/d) − 26 26 26 160 27 14 Mean specific heat 0–100°C J kg−1 K−1 917 1038 528 2052 456 386 Thermal conductivity 20–100°C W m−1 K−1 238 156 26 194 78 397 Coefficient of thermal expansion 0–100°C 10−6 K−1 23.5 26.0 8.9 12.0 12.1 17.0 Electrical resistivity at 20°C μ ohm cm 2.67 4.2 54 3.3 10.1 1.69 From Lide, DR (Ed.): Handbook of Chemistry & Physics, 72nd Ed., CRC Press, Boca Raton, FL, USA, 1991–92; Metals Handbook, Volume 2, 10th Ed., ASM International, Metals Park, OH, USA, 1990. Note: Conversion factors for S1 and Imperial units are given in the Appendix. 6 CHAPTER 1  The light metals the small size of the beryllium atom severely restricts its solubility for other elements. One exception is the eutectic composition Be–38Al in which some useful ductility has been achieved. This alloy was developed by the Lockheed Aircraft Company in 1976 and became known as Lockalloy. Because beryl- lium and aluminium have little mutual solid solubility in each other, the alloy is essentially a composite material with a microstructure comprising stiff beryl- lium particles in a softer aluminium matrix. Lightweight (specific gravity 2.09) extrusions and sheet have found limited aerospace applications. Beryllium is now prepared mainly by powder metallurgy methods. Metal extracted from the minerals beryl or bertrandite is vacuum melted and then either cast into small ingots, machined into chips and impact ground, or directly inert gas atomized to produce powders. The powders are usually con- solidated by hot isostatic pressing and the resulting billets have properties that are more isotropic than are obtained with cast ingots. Tensile properties depend on the levels of retained BeO (usually 1–2%) and impurities (iron, alu- minium, and silicon) and ductilities usually range from 3% to 5%. The billets can then be hot worked by forging, rolled to sheet, or extruded to produce bar or tube. Lockalloy (now also known as AlBeMet™ 162) is now also manufac- tured by inert gas atomization of molten prealloyed mixtures and the resulting powders are consolidated and hot worked as described earlier. 1.1.3 Relative abundance The estimated crustal abundance of the major chemical elements is given in Table 1.2 which shows that the light metals aluminium, magnesium, and tita- nium are first, third, and fourth in order of occurrence of the structural metals. Table 1.2 Crustal abundance of major chemical elements Element % by weight Oxygen 45.2 Silicon 27.2 Aluminium 8.0 Iron 5.8 Calcium 5.06 Magnesium 2.77 Sodium 2.32 Potassium 1.68 Titanium 0.86 Hydrogen 0.14 Manganese 0.10 Phosphorus 0.10 Total 99.23 From Stanner, RJL: Am. Sci., 64, 258, 1976.  1.1 General introduction 7 It can also be seen that the traditional metals copper, lead, and zinc are each present in amounts 5%) Spherical GP zones GP zones solvus below room temperature if 3:1. a = 0.496 nm (0001)η′//(111)α; η′//[112 ]αα. Semi- coherent. Disk shaped. c = 1.405 nm a//αα, c//α. Composition close to MgZn (e.g., Fig. 2.14). η (or M) hexagonal MgZn2 Forms at or from η′, may have one of nine a = 0.521 nm orientation relationships with matrix. Most c = 0.860 nm common are: (10 10)η//(001)α; (0001)η// (110)α and (0001)η//(1 11)αα; (10 10)η// (110)α. T′ hexagonal, probably Semi-coherent. May form instead of η in Mg32 (Al, Zn)49 alloys with high Mg:Zn ratios. a = 1.388 nm (0001)T′//(111)α; (10 11)T′//(112 )αα. c = 2.752 nm T cubic Mg32 (Al, Zn)49 May form from η if ageing temperature a = 1.416 nm >190°C, or from T′ in alloy with high Mg:Zn ratios. (100)T//(111)α; T//[112 ]α. Al–Li–Mg δ′ cubic Al3Li Metastable coherent precipitate with a = 0.404 to 0.401 nm ordered Cu3Au(L12) type superlattice (Fig. 4.35). Low misfit. Al2LiMg cubic Forms as coarse rods with growth a = 1.99 nm directions in alloys with ≥ 2%Mg. (110)p//( 110)αα; [1 1 0]p//α. Al–Li–Cu δ′ cubic Al3Li As for Al–Li and Al–Li–Mg alloys. δ cubic AlLi Nucleates heterogeneousy, mainly in grain boundaries. a = 0.637 nm (011)δ//( 111)αα;(0 11)δ//(1 12)αα. T1 hexagonal Al2LiCu Thin hexagonal-shaped plates with {111}α habit plane. a = 0.497 nm (0001)T1//{111}α; TT1//< 110>αα 1 (Figs. 2.27 and 4.39). c = 0.934 nm θ″, θ′ Phases present in binary Al–Cu alloys may also form at low Li:Cu ratios.  2.2 PRINCIPLES OF AGE HARDENING 63 Modeling concepts are now permitting further refinements to be made in the understanding of the microstructural design of high-strength aluminium alloys. While it is agreed that the desired microstructure to obtain high strength combined with a high resistance to fracture is one that consists of a small vol- ume fraction of very fine, hard particles, it is also recognized that a common feature in high-strength aluminium alloys is the presence of shear-resistant, plate-shaped precipitates that form on the {100}α or {111}α matrix planes, or rod-shaped precipitates that form in the α directions. Less attention has been paid to a quantitative analysis of the effects of particle shape and orien- tation because of a lack of appropriate versions of the Orowan equation that relate the critical resolved shear stress due to dispersion hardening to precipi- tate characteristics. The version of the Orowan equation currently accepted for spherical par- ticles is: Gb 1 πd ∆τ = ⋅ ⋅ ln 2π 1 − ν λ 4b where Δτ is increment in critical resolved shear stress due to dispersion strength- ening, ν is Poisson’s ratio, G is shear modulus, b is Burgers vector of gliding dislocations, λ is effective interparticle spacing, and d is diameter of precipitate particles. Within this equation, it is λ that varies with shape, orientation, and dis- tribution of the particles, and derivation of appropriate versions of the Orowan equation require the calculation of λ for different particle arrays. If it is assumed that {100}α precipitate plates are circular disks of diameter d and thickness t dis- tributed at the center of each surface of a cubic volume of the matrix (Fig. 2.24A), then the intersection of these plates with the {111}α slip plane in the matrix will have a triangular distribution on this slip plane (Fig. 2.24B). Calculations show that the effective planar interparticle spacing for the {100}α plates is given by λ = 0.931(0.306πdt/f)1/2 – πd/8 – 1.061d, where f is volume fraction of particles. Similar calculations for {111}α precipitate plates show λ = 0.931(0.265πdt/f)1/2 – πd/8 – 0.919t, and for α rods λ = 1.075d (0.433π/f)1/2 – (1.732d)1/2. Substitution of these expressions for λ in the Orowan equation shown earlier enables the critical resolved shear stresses to be deter- mined for model alloys containing these three different types of precipitates. This analysis shows, quantitatively, that plate-shaped precipitates are more effective barriers to gliding dislocations than either rods or spherical precipi- tates. Furthermore, the increment of strengthening produced by {111}α plates is invariably larger than that produced by {100}α plates and, for both orientations, this increment becomes progressively larger as the aspect ratio (length to thick- ness) increases, Fig. 2.24C. Fig. 2.25 shows a comparison of phase field simulations of dislocation glid- ing in a forest of particles that have same number density, volume fraction, and distribution of precipitates. The dislocation bows when it approaches the 64 CHAPTER 2   PHYSICAL METALLURGY OF ALUMINIUM ALLOYS Figure 2.24 (A) Circular {100}α precipitate plates in a cubic volume of the aluminium matrix and (B) projection of intersected {100}α precipitate plates on a {111}α plane of the matrix. (C) Variation in ratio of critical resolved shear stress Δτ (plate, rod)/Δτ (sphere) with aspect ratio for Orowan strengthening attributable to {111}α and {100}α precipitate plates and α precipitate rods. Volume fraction of precipitates is 0.05. precipitates and bypasses the particles if the applied shear stress is sufficiently large. While a shear stress of 55 MPa is required for the dislocation to glide through the entire forest of spherical particles, a 50% higher stress is needed for the plate-shaped particles. Above a critical value of aspect ratio, plates on either set of planes form what is essentially a closed network that entraps gliding dislocations. In such situation, it is inevitable that precipitate shearing occurs. The contribution of shearable precipitates to the critical resolved shear stress is:  2  1  F 3 / 2 ∆τ =      b Γ  L  2  p   2.2 PRINCIPLES OF AGE HARDENING 65 Figure 2.25 Effects of precipitate shape on critical resolved shear stress required for a dislocation to glide through precipitate forest. The number density and volume fraction of precipitates are identical in (A) and (B). Courtesy H. Liu. where Γ is the dislocation line tension in the matrix phase, Lp is the mean pla- nar center-to-center interprecipitate spacing, and force F is a measure of the resistance of the precipitates to dislocation shearing. For {111}α precipitate plates, the dihedral angle between their habit plane and the {111}α slip plane is 70.53°. Assuming the slip plane is (1 11)αα, shearing of a (111)α precipitate plate occurs in [10 1 ]αα, α, or α directions on the (1 11)αα slip plane (Fig. 2.26A). When a (111)α plate is sheared in the α direction, the energy of the created particle/matrix interface is approximately 2γidpb sin 60°, and the pre- 2d p bγi sin60° 1.282dbγi cipitate strength is: F = = , where γi is the specific tp t interfacial energy of the newly created particle/matrix interface, dp and tp are the mean planar radius and planar thickness of the precipitate plate, respec- tively. Combining this expression and that for Lp, the interfacial strengthening increment in aluminium alloys containing {111}α precipitate plates is given as 1/ 2 1.211d γ 3 / 2  bf  ∆τi =   . For {100}α precipitate plates, the dihedral angle t2 Γ between their habit plane and the {111}α slip plane is 54.74°. To shear the {100}α plates with an additional particle/matrix interface area of 2dpb sin 60°, 2 d p bγi sin60° 1.110dbγi the required shear force F = =. Combining this tp t equation and that for Lp, the interfacial strengthening increment in critical 1/ 2 0.908d γ 3 / 2  bf  resolved shear stress is given as ∆τi =   . For a given value t2 Γ 66 CHAPTER 2   PHYSICAL METALLURGY OF ALUMINIUM ALLOYS Figure 2.26 (A) Schematic diagrams showing shearing of a circular precipitate plate and projection of a sheared precipitate plate. (B) Variation of ratio ∆τ(plate/rod)/∆τ(sphere) with aspect ratio for {111}α and {100}α plates and α precipitate rods, calculated assuming interfacial strengthening of sheared particles. of dislocation line tension, the variations in the ratio ∆τi(plate)/∆τi(sphere) with plate aspect ratio for various orientations of plates are also shown in Fig. 2.26B. Unlike the results for spherical or rod particles, it is evi- dent that the contribution due to interfacial strengthening may become significant when particles take, in particular, a plate shape. For identical vol- ume fractions and number densities of precipitates per unit volume, the yield stress increments produced by {111}α and {100}α precipitate plates are orders of magnitude larger than those produced by α precipi- tate rods and by spherical particles. The increments in CRSS produced by {111}α and {100}α plates increase substantially with an increase in plate aspect ratio and are up to three orders of magnitude larger than that pro- duced by spheres, when the plate aspect ratio is in the range of 5:1 to 95:1. Traditionally, attempts to improving alloy strength involve efforts to increase nucleation rate and hence number density of precipitates. Modeling  2.2 PRINCIPLES OF AGE HARDENING 67 Figure 2.27 Electron micrographs of the alloy Al–5.3Cu–1.3Li–0.4Mg–0.4Ag–0.16Zr: (A) quenched and aged 8 h at 160°C showing finely dispersed, coherent θ″ particles and occasional plates of the T1 phase. Hardness 146 DPN; (B) quenched, cold worked 6% and aged 8 h 160°C showing a much coarser but uniform dispersion of semi-coherent T1 plates. Hardness 200 DPN. Electron beam is parallel to α. (B) Courtesy S. P. Ringer. results from both shear resistant and shearable precipitates indicate that an alternative approach to strengthening is to modify precipitate shape and orienta- tion. Any further enhancement in strength may be achieved if precipitate plates of large aspect ratio could be formed in the alloy, probably via the addition of microalloying elements. The influences of precipitate orientation and shape are in accord with the observed behavior of high-strength aluminium alloys. One example of high- strength alloys hardened by precipitates that form on the {111}α planes is those based on the Al–Zn–Mg–Cu system (Table 2.3). Some of these commer- cial alloys develop yield strengths exceeding 600 MPa which are significantly higher than the yield strengths possible with alloys based on the Al–Cu system in which the precipitates form on the {100}α planes. Another example of a pre- cipitate which forms on the {111}α planes is the T1 phase (Al2CuLi) that was mentioned earlier. This phase has a particularly high aspect ratio and its ability to promote greater hardening than a much higher density of zones of the finer, shearable phase θ″ formed on the {100}α planes is illustrated in an Al–Cu–Li– Mg–Ag–Zr alloy in Fig. 2.27. In this regard, it may be noted that yield stresses exceeding 700 MPa have been recorded for this alloy which are close to the the- oretical upper limit for aluminium (∼900 MPa). 68 CHAPTER 2   PHYSICAL METALLURGY OF ALUMINIUM ALLOYS 2.3 AGEING PROCESSES 2.3.1 Precipitation sequences As mentioned earlier, several aluminium alloys display a marked response to age hardening. By suitable alloying and heat treatment, it is possible to increase the yield stress of high-purity aluminium by as much as 50 times. Details of the precipitates that may be present in alloy systems having commercial signifi- cance are given in Table 2.3. The actual precipitate or precipitates that form in a particular alloy during ageing depends mainly on the ageing temperature. For example, for GP zones to form, ageing must be carried out below the relevant GP zones solvus temperature as mentioned in the Section 2.2.2. If intermedi- ate precipitates are formed, they may nucleate from pre-existing GP zones, at the sites of these zones, or independently depending on the alloy concerned. At some ageing temperatures, both GP zones and an intermediate precipitate may be present together. Cold work prior to ageing increases the density of dislo- cations which may provide sites for the heterogeneous nucleation of specific precipitates. A partial phase diagram for the Al–Cu system was shown in Fig. 2.1. Al– Mg–Si alloys can be represented as a pseudo-binary Al–Mg2Si system (Fig. 2.28) and sections of the ternary phase diagrams for the important Al–Cu–Mg and Al–Zn–Mg systems are shown in Figs. 2.29 and 2.30. Most commercial alloys based on these systems have additional alloying elements present that modify the respective ternary diagrams and, in Fig. 2.31, an example is shown Figure 2.28 Pseudo-binary phase diagram for Al–Mg2Si.  2.3 AGEING PROCESSES 69 Figure 2.29 Section of ternary Al–Cu–Mg phase diagram at 460°C and 190°C (estimated). θ = Al2Cu, S = Al2CuMg, T = Al6CuMg4. for the section at 460°C for Al–Zn–Mg alloys containing 1.5% copper. Since this is close to the usual solution treatment temperature for alloys of this type, it should be noted that some quaternary compositions will not be single phase prior to quenching. 2.3.2 Clustering phenomena Although the random clustering of solute atoms prior to precipitation in quenched and aged aluminium alloys was detected by small angle X-ray dif- fraction many years ago, the effects of this phenomenon on subsequent ageing processes have been little understood. Now there is evidence that clustering events may promote formation of existing precipitates in an alloy, stimulate nucleation of new precipitates as was discussed in the Section 2.2.4, and con- tribute to the actual age hardening of certain alloys. In the Al–Mg–Si system in which ageing processes are particularly com- plex, atom probe studies have shown that the formation of GP zones may be preceded initially by the appearance of individual clusters of magnesium and silicon atoms, followed by the formation of co-clusters of these elements. This behavior is demonstrated in Fig. 2.32 which shows atom probe concentration profiles after ageing the alloy Al–1Mg–0.6Si for (A) 0.5 h and (B) 8 h at 70°C. These profiles are developed by collecting, counting, and identifying ions as they are evaporated from the tip of an alloy specimen in a field ion microscope. The clusters will, for example, form during a delay at ambient temperature after quenching and before artificial ageing. As described in the Section 4.4.3, their 70 CHAPTER 2   PHYSICAL METALLURGY OF ALUMINIUM ALLOYS Figure 2.30 Section of ternary Al–Zn–Mg phase diagrams at 200°C. M = MgZn2, T = Al32(Mg,Zn)49. Figure 2.31 Section of Al–Zn–Mg–Cu phase diagram (1.5% Cu) at 460°C. S = Al2CuMg, T = Al6CuMg4 + Al32(Mg,Zn)49, M = MgZn2 + AlCuMg. presence may alter the dispersion of precipitates that form on subsequent age- ing such that the response to hardening is reduced. It is well known that age hardening in most alloys based on the Al–Cu–Mg system occurs in two distinct stages over a wide temperature range (~100°C to 240°C). The first stage, which may account for 60–70% of the total hardening response, is characteristically and uniquely rapid, and may be completed within 60 s. This behavior is then followed by what, for some compositions, can be a pro- longed period (e.g., 100 h) during which the hardness shows little or no change, and after which there is a second rise to a peak value, Fig. 2.33. Initially, this early hardening phenomenon was attributed to the rapid formation of GP(Cu,Mg) zones  2.3 AGEING PROCESSES 71 Figure 2.32 Atom probe profiles showing evidence of clustering of magnesium and silicon atoms in the alloy 6061 (Al–1Mg–0.6Si) quenched and aged (A) 0.5 h and (B) 8 h at 70°C. From Edwards, GA et al.: Applied Surf. Sci., 76/77, 219, 1994. (also known as GPB zones). However, later studies using high-resolution elec- tron microscopy and electron diffraction did not detect evidence of these zones until alloys were aged for times that placed them further along the hardness pla- teau. Observations made from one-dimensional APFIM revealed the presence of a high density (e.g., 1019 cm−3) of small, disordered clusters of atoms immediately after rapid early hardening is completed, and the phenomenon was termed “cluster hardening” to distinguish it from normal precipitation reactions. However, subse- quent work made by three-dimensional atom probe revealed little or no evidence of Cu–Mg co-clusters in the aluminium matrix. Recently, single pairs of copper and magnesium atoms were proposed to form in Al–Cu–Mg alloys based on thermo- dynamic calculations, but these pairs are yet to be confirmed experimentally. The formation of solute pairs or clusters might be responsible for the rapid hardening phenomenon in Fig. 2.33, also the mechanism by which they can cause hardening remains uncertain. One possibility is that rapid solute/dislocation interactions occur in which the relatively small copper atoms and large magnesium atoms immobilize edge dislocations by segregating preferentially to the respective compression and tension regions. Another suggestion is that hardening may arise because of differ- ences in elastic moduli between aluminium and the Cu–Mg clusters. 72 CHAPTER 2   PHYSICAL METALLURGY OF ALUMINIUM ALLOYS Figure 2.33 Rapid hardening phenomenon in Al–2.5Cu–1.5Mg and Al–2.5Cu–1.5Mg–0.5Ag alloys during isothermal ageing at (A) 150°C and (B) 200°C. From Vietz, JT and Polmear, IJ: J. Inst. Met., 94, 410, 1966.  2.3 AGEING PROCESSES 73 2.3.3 GPB zones in Al–Cu–Mg alloys GPB zones have been observed towards the end of the hardness plateau shown in Fig. 2.33. It is at this stage that further ageing causes a second increase in hardness which reaches a maximum value when a critical distri- bution of the GPB zones has formed. The structure of these zones remains uncertain but a recent study using atomic-resolution Z-contrast scanning transmission electron microscopy has revealed that at least two distinctly different structures may form (Fig. 2.34): one structure is comprised of an agglomeration of structural units with a translational periodicity along a sin- gle α direction, while the other structure has a core having a hexagonal structure and a shell comprising structural units (Fig. 2.34). Figure 2.34 Cross sections of GPB zones in Al–Cu–Mg alloys. From Kovarik, L et al.: Acta Mater., 56, 4804, 2008. 74 CHAPTER 2   PHYSICAL METALLURGY OF ALUMINIUM ALLOYS 2.3.4 Intermediate precipitates The intermediate precipitates are often the key strengthening phases in com- mercial aluminium alloys. For some of these precipitates, there is now strong evidence to show that microalloying elements segregate to the interface between the precipitates and the aluminium matrix, see for example Fig. 4.17. The interfacial solute segregation phenomenon is even found in θ′ precipitate plates in simpler binary Al–Cu alloys. A recent study using atomic-resolution Z-contrast scanning transmission electron microcopy has revealed the pres- ence of excessive Cu atoms in the θ′/matrix coherent interface during early stages of the θ′ precipitation (Fig. 2.35A) which is distinctly different from that previously assumed for the θ′/matrix interface. This single layer of cop- per atoms is similar or identical to GP zones, and its presence in the interface appears to be critical in the early stages of thickening of θ′. In the late growth stage, the θ′/matrix interface becomes less enriched in copper. Apart from the copper segregation in the broad surface of θ′ plates, the end facet of θ′, or the semicoherent interface, is found to have a complex but well-defined structure (θ″) that enables the progression from aluminium structure to θ′, Fig. 2.35B. This structure allows interface to migrate through a series of individual atomic movements. Solute segregation in precipitate/matrix interfaces is also found to occur in Al–Mg–Si–Cu alloys (Fig. 2.36) and in Al–Sc–Mg alloys where magnesium segregates in the Al3Sc/matrix interface. This is similar to the co-segregation Figure 2.35 HAADF-STEM images showing (A) segregation of Cu atoms in the broad face of θ′ precipitates, and (B, C) θ′ precipitate end facet exhibiting a thin layer of θ″. From Bourgeois, L et al.: Acta Mater., 59, 7043, 2011 and Phys. Rev. Lett., 111, 046102, 2013.  2.3 AGEING PROCESSES 75 Figure 2.36 Segregation of copper (white contrast) in Q/matrix interface in an Al–Mg–Si– Cu alloy. From Fiawoo, M et al.: Scripta Mater., 88, 53, 2012. of silver and magnesium in interfaces associated with Ω and T1 (Figs. 4.16 and 4.39). Different explanations have been proposed to account for the sol- ute segregation, but the most commonly accepted one is interfacial energy minimization. It has generally been accepted that most intermediate precipitates which form in aged aluminium alloys have compositions and crystal structures that differ only slightly from those of the respective equilibrium precipitates. In fact, for Al–Cu–Mg alloys in which the equilibrium precipitate is S(Al2CuMg), the intermediate precipitate S′ differs so little in its crystallographic dimensions that it is sometimes ignored. However, studies of some alloys using one-dimen- sional and three-dimensional APFIM have revealed some unexpected composi- tional variations between other intermediate and equilibrium precipitates. One example is the Al–Mg–Si system in which the compositions of the intermediate precipitates β″ and β′ were assumed to be the same as the equi- librium precipitate β (Mg2Si) (Table 2.3). Because of this, the compositions of some commercial alloys have been designed deliberately to have a balanced (2:1) atomic ratio of magnesium and silicon in order to maximize precipita- tion of β″ and β′ during ageing. Now there is strong experimental evidence that the actual Mg:Si ratios of these intermediate precipitates are close to 1:1. As mentioned in Section 4.4.3, this has opened up the prospect of producing a new range of Al–Mg–Si alloys in which the magnesium content has been reduced to improve their hot working characteristics. Al–Zn–Mg–(Cu) alloys are others in 76 CHAPTER 2   PHYSICAL METALLURGY OF ALUMINIUM ALLOYS which atom probe studies have shown that the Mg:Zn ratio for the intermedi- ate precipitate η′ differs substantially from that of the equilibrium precipitate η (MgZn2). In this case, the Mg:Zn ratio appears to lie in the range 1:1 to 1:1.15 rather than the expected 1:2. This suggests that the composition of η′ is linked more to the preexisting GP zones than to the equilibrium precipitate η and sup- ports the suggestion that η′ can nucleate directly from these zones. These new observations about the compositions of some intermediate precipitates means that a substantial number of atom positions in their unit cells must still be occu- pied by aluminium atoms rather than the respective solute atoms. 2.3.5 Secondary precipitation For many years there was an implicit acceptance that, once an alloy had been aged at an elevated temperature, its mechanical properties remained stable on exposure for an indefinite time at a significantly lower temperature. However, it was found that highly saturated Al–Zn alloys aged at 180°C will continue to age and undergo what has been termed “secondary precipitation” if cooled and then held at ambient temperature. Similar behavior has also been observed in highly saturated lithium-containing aluminium alloys aged first at 170°C and then exposed at temperatures in the range 60–130°C. In this case, there is a progressive increase in hardness and mechanical strength accompanied by an unacceptable decrease in ductility and toughness that is attributed to secondary precipitation of the finely dispersed δ′ throughout the matrix. More recently, observations on a wide range of aluminium alloys have shown that secondary precipitation is, in fact, a more general phenomenon. This conclusion is sup- ported by results obtained using the technique of positron annihilation spectros- copy which have indicated that vacancies may be retained and remain mobile at ambient temperatures after aged aluminium alloys are cooled from a higher ageing temperature. Positron annihilation spectroscopy is proving to be a powerful tool for studying the role of lattice defects, such as vacancies, in the decomposition kinetics of aged alloys. This technique involves measurement of the lifetimes of positrons emitted from a radioactive source before there are annihilated by interacting with electrons in the alloy. The annihilation process may be thought of as a chemical reaction that has a rate which is directly proportional to the local electron density. Vacant lattice sites can be detected because they are open volume defects and offer temporary “shelter” to incident positrons, thereby slowing their annihilation rate. As an example, Fig. 2.37 shows the evolution of positron lifetimes during ageing the alloy Al–4Cu–0.3Mg at 20°C after first being solution treated, quenched, and aged for various times at 180°C. The increases in positron lifetimes, which are comparatively large for alloys initially aged for short times (30 and 120 s) at 180°C, but still significant for longer times of 1 and 10 h, are all taken to indicate that retained vacancies (and solute atoms) are mobile at 20°C thereby allowing further ageing to occur.  2.3 AGEING PROCESSES 77 Figure 2.37 Positron lifetimes during secondary ageing Al–4Cu–0.3Mg at 20°C after solu- tion treatment at 520°C, quenching at 0°C and first ageing 30 s, 120 s, 1 h, or 10 h at 180°C. From Somoza, A et al.: Phys. Rev. B, 61, 14454, 2000. Figure 2.38 Hardness–time curves for the Al–Zn–Mg–Cu alloy 7075 aged at 130°C (solid line), and underaged 0.5 h at 130°C, quenched to 25°C, and held either at this temperature or at 65°C. From Lumley, RN et al.: Mater. Sci. Tech., 22, 1025, 2005. Detailed studies of secondary precipitation in a wide range of aluminium alloys have revealed that the levels of residual or “free” solute atoms that remain in solid solution until the alloys are in the overaged condition is higher than has generally been accepted. However, as expected from Fig. 2.37, the response to secondary precipitation at the lower ageing temperature is greater if an alloy is first artificially aged for a short time (i.e., underaged). The behav- ior is illustrated in Fig. 2.38 for the Al–Zn–Mg–Cu alloy 7075 which normally 78 CHAPTER 2   PHYSICAL METALLURGY OF ALUMINIUM ALLOYS reaches a peak hardness of 195 DPN if aged at 130°C for 24 h (T6 temper). If this alloy is first underaged for 0.5 h at 130°C and quenched to 25°C, the hardness is 150 DPN. During prolonged secondary ageing at this lower tem- perature, the hardness gradually increases and it becomes equal to the T6 value. Alternately, if 7075 is held at the slightly higher temperature of 65°C after quenching from 130°C, the hardness increases faster and reaches the higher value of 225 DPN after 10,000 h. During secondary ageing at the lower temperature, GP zone formation usu- ally occurs. If, after a dwell period, ageing at the initial elevated temperature is resumed, then the microstructure at peak hardness is refined so that a greater overall response to hardening can be achieved than is possible using a single- stage artificial ageing treatment. This effect is shown for the alloy Al–4%Cu in Fig. 2.39 in which the multistage ageing schedule is given the designation Figure 2.39 (A) Differences in the hardness curves for Al–4%Cu artificially aged at 150°C, with and without an interrupted period of secondary ageing at 65°C. The inset plot shows the hardness change during this dwell period at 65°C. (B) Transmission electron micro- graphs in the α direction showing dispersions of the θ′ precipitate plates and minor amounts of the θ″ phase in these two aged conditions. T6 temper, 100 h at 150°C; T6I6 tem- per, 3 h at 150°C, quench, 500 h at 65°C, and 50 h at 150°C. From Lumley, RN et al.: Mater. Sci.Tech., 19, 1453, 2003.  2.4 CORROSION 79 “T6I6,” where “I” means that artificial ageing at an elevated temperature has been interrupted. As given in Table 4.7, experimental interrupted ageing cycles have been developed that enable simultaneous increases to be achieved in ten- sile and fracture toughness properties of a wide range of aluminium alloys. 2.4 CORROSION 2.4.1 Surface oxide film Aluminium is an active metal which will oxidize readily under the influence of the high free energy of the reaction whenever the necessary conditions for oxidation prevail. Nevertheless, aluminium and its alloys are relatively stable in most environments due to the rapid formation of a natural oxide film of alu- mina on the surface that inhibits the bulk reaction predicted from thermody- namic data. Moreover, if the surface of aluminium is scratched sufficiently to remove the oxide film, a new film quickly reforms in most environments. As a general rule, the protective film is stable in aqueous solutions of the pH range 4.5–8.5, whereas it is soluble in strong acids or alkalis, leading to rapid attack of the aluminium. Exceptions are concentrated nitric acid, glacial acetic acid, and ammonium hydroxide. The oxide film formed on freshly rolled aluminium exposed to air is very thin and has been measured as 2.5 nm. It may continue to grow at a decreasing rate for several years to reach a thickness of some tens of nanometers. The rate of film growth becomes more rapid at higher temperatures and higher relative humidities, so in water it is many times that occurring in dry air. In aqueous solutions, it has been suggested that the initial corrosion product is aluminium hydroxide, which changes with time to become a hydrated aluminium oxide. The main difference between this film and that formed in air is that it is less adherent and so is far less protective. Much thicker surface oxide films that give enhanced corrosion resistance to aluminium and its alloys can be produced by various chemical and elec- trochemical treatments. The natural film can be thickened some 500 times, to say 1–2 μm, by immersion of components in certain hot acid or alkaline solu- tions. Although the films produced are mainly Al2O3, they also contain chemi- cals such as chromates that are collected from the bath to render them more corrosion resistant. A number of proprietary solutions are available and the films they produce are known generally as conversion coatings. Even thicker, e.g., 10–20 μm, surface films are produced by the more commonly used treat- ment known as anodizing. In this case the component is made the anode in an electrolyte, such as an aqueous solution containing 15% sulfuric acid, which produces a porous Al2O3 film that is subsequently sealed, i.e., rendered nonpo- rous, by boiling in water. Both conversion and anodic coatings can be dyed to give attractive colors and the latter process is widely applied to architectural products. 80 CHAPTER 2   PHYSICAL METALLURGY OF ALUMINIUM ALLOYS It should be noted that chromate conversion coatings are widely used in cor- rosion protection schemes for aluminium alloys in aircraft structures and other applications. However, it has been recognized recently that chromates may present a health hazard which has led to an interest in other, nontoxic, coat- ing processes. Promising results have been reported for cerium-rich coatings which can be applied by several methods. A durable cerium oxide/hydroxide film replaces natural Al2O3 and protection is afforded by partial or complete suppression of the reduction of oxygen at cathodic sites which normally occurs during electrolytic corrosion. Chemicals known as inhibitors may be added to potentially corrosive liq- uid environments for the purpose of minimizing or preventing corrosion of alu- minium and its alloys. Inhibitors may be classified as being anodic, cathodic, or mixed depending on whether they mainly affect the anodic, cathodic, or both anodic and cathodic corrosion processes. Anodic inhibitors stifle the anodic reaction, usually by depositing on the surface sparingly soluble substances as a direct anodic product. There is often no change in the surface appearance. Chromates are commonly used for this purpose and are generally effective. However, they can cause problems if present in insufficient amounts because they may decrease the surface area under attack without decreasing appreciably the amount of metal dissolution. Such a situation may lead to intensified local attack, e.g., by pitting. Cathodic inhibitors are safer in this respect. They serve to stifle the cathodic reaction either by restricting access to oxygen or by “poi- soning” local spots favorable for cathodic hydrogen evolution. They form a vis- ible film on the aluminium and are usually less efficient than anodic inhibitors as they do not completely prevent attack. Examples are phosphates, silicates, and soluble oils. The choice and concentration of an inhibitor depends on sev- eral factors such as the compositions of the alloy and the liquid environment to which it is to be exposed, the temperature, and the rate of movement of the liquid. An inhibitor that offers protection in one environment may increase it in another. 2.4.2 Contact with dissimilar metals The electrode potential of aluminium with respect to other metals becomes particularly important when considering galvanic effects arising from dissimi- lar metal contact. Comparisons must be made by taking measurements in the same solution and Table 2.4 provides the electrode potentials with respect to the 0.1 M calomel electrode (Hg–HgCl2, 0.1 M KCl) for various metals and alloys immersed in an aqueous solution of 1 M NaCl and 0.1 M H2O2. The value for aluminium is −0.85 V whereas aluminium alloys range from −0.69 to −0.99 V. Magnesium which has an electrode potential of −1.73 V is more active than aluminium whereas mild steel is cathodic having a value of −0.58 V.  2.4 CORROSION 81 Table 2.4 Electrode potentials of various metals and alloys with respect to the 0.1 M calomel electrode in aqueous solutions of 53 g l−1 NaCl and 3 g l−1 H2O2 at 25°C Metal or alloy Potential (V) Magnesium −1.73 Zinc −1.10 Alclad 6061, Alclad 7075 −0.99 5456, 5083 −0.87 Aluminium (99.95%), 5052, 5086 −0.85 3004, 1060, 5050 aluminium −0.84 1100, 3003, 6063, 6061, Alclad 2024 alloysa −0.83 2014–T4 −0.69 Cadmium −0.82 Mild steel −0.58 Lead −0.55 Tin −0.49 Copper −0.20 Stainless steel (3xx series) −0.09 Nickel −0.07 Chromium −0.49 to +0.18 From Metals Handbook, Vol. 1, ASM, Cleveland, OH, USA, 1961. a Compositions corresponding to the numbers are given in Tables 4.2 and 4.4. Table 2.4 suggests that sacrificial attack of aluminium and its alloys will occur when they are in contact with most other metals in a corrosive environ- ment. However, it should be noted that electrode potentials serve only as a guide to the possibility of galvanic corrosion. The actual magnitude of the gal- vanic corrosion current is determined not only by the difference in electrode potentials between the particular dissimilar metals but also by the total electri- cal resistance, or polarization, of the galvanic circuit. Polarization itself is influ- enced by the nature of the metal/liquid interface and more particularly by the oxides formed on metal surfaces. For example, contact between aluminium and stainless steels usually results in less electrolytic attack than might be expected from the relatively large difference in the electrode potentials, whereas contact with copper causes severe galvanic corrosion of aluminium even though this difference is less. Galvanic corrosion of aluminium and its alloys may be minimized in several ways. If contact with other metals cannot be avoided, these should be chosen so that they have electrode potentials close to aluminium; alternatively it may be 82 CHAPTER 2   PHYSICAL METALLURGY OF ALUMINIUM ALLOYS possible to locate a dissimilar metal joint away from the corrosive environment. If not then complete electrical isolation of aluminium and the other metal must be arranged by using nonconducting washers, sleeves, or gaskets. When paint is used for protection, it should be applied to the cathodic metal and not the aluminium. This practice is required because pinholes that may form in a paint film on aluminium can lead to pitting attack because of the large cathode to anode area ratio. In a closed loop system, such as used for cooling automobile engines, a mixed anodic and cathodic inhibitors should be added to the cooling water. 2.4.3 Influence of alloying elements and impurities Alloying elements may be present as solid solutions with aluminium, or as microconstituents comprising the element itself, e.g., silicon, a compound between one or more elements and aluminium (e.g., Al2CuMg) or as a com- pound between one or more elements (e.g., Mg2Si). Any or all of the above conditions may exist in a commercial alloy. Table 2.5 provides values of the electrode potentials of some aluminium solid solutions and micro-constituents. In general, a solid solution is the most corrosion-resistant form in which an alloy may exist. Magnesium dissolved in aluminium renders it more anodic although dilute Al–Mg alloys retain a relatively high resistance to corrosion, particularly to seawater and alkaline solutions. Chromium, silicon, and zinc Table 2.5 Electrode potentials of aluminium solid solutions and microconstituents with respect to the 0.1 M calomel electrode in aqueous solutions of 53 g l−1 NaCl and 3 g l−1 H2O2 at 25°C. Solid solution or microconstituent Potential (V) Mg5Al8 −1.24 Al−Zn−Mg solid solution (4% MgZn2) −1.07 MgZn2 −1.05 Al2CuMg −1.00 Al−5% Mg solid solution −0.88 MnAl6 −0.85 Aluminium (99.95%) −0.85 Al−Mg−Si solid solution (1% Mg2Si) −0.83 Al−1% Si solid solution −0.81 Al−2% Cu SSSS −0.75 Al−4% Cu SSSS −0.69 FeAl3 −0.56 CuAl2 −0.53 NiAl3 −0.52 Si −0.26 From Metals Handbook, Vol. 1, ASM, Cleveland, OH, USA, 1961.  2.4 CORROSION 83 in solid solution in aluminium have only minor effects on corrosion resistance although zinc does cause a significant increase in the electrode potential. As a result, Al–Zn alloys are used as clad coatings for certain aluminium alloys (see Section 4.1.5) and as galvanic anodes for the cathodic protection of steel struc- tures in seawater. Copper reduces the corrosion resistance of aluminium more than any other alloying element and this arises mainly because of its presence in micro-constituents. However, it should be noted that when added in small amounts (0.05–0.2%), corrosion of aluminium and its alloys tends to become more general and pitting attack is reduced. Thus, although under corrosive con- ditions, the overall weight loss is greater, perforation by pitting is retarded. Microconstituents are usually the source of most problems with electro- chemical corrosion as they lead to nonuniform attack at specific areas of the alloy surface. Pitting and intergranular corrosion are examples of localized attack (Fig. 2.40), and an extreme example of this is that components with a marked directionality of grain structure show exfoliation (layer) corrosion (Fig. 2.41). In exfoliation corrosion, delamination of surface grains or layers occurs under forces exerted by the voluminous corrosion products. Figure 2.40 Microsection of surface pits in a high-strength aluminium alloy. Note that intergranular SCC are propagating from the base of these pits (× 100). 84 CHAPTER 2   PHYSICAL METALLURGY OF ALUMINIUM ALLOYS Figure 2.41 Microsection showing exfoliation (layer) corrosion of an aluminium alloy plate (× 100). Iron and silicon occur as impurities and form compounds most of which are cathodic with respect to aluminium. For example, the compound Al3Fe provides points at which the surface oxide film is weak, thereby promoting electrochemi- cal attack. The rate of general corrosion of high-purity aluminium is much less than that of the commercial-purity grades which is attributed to the smaller size and number of these cathodic constituents throughout the grains. However, it should be noted that this may be a disadvantage in some environments as attack of high-purity aluminium may be concentrated in grain boundaries. Nickel and titanium also form cathodic phases although nickel is present in very few alloys. Titanium, which forms Al3Ti, is commonly added to refine grain size (Sections 4.1.3 and 3.3) but the amount is too small to have a significant effect on corro- sion resistance. Manganese and aluminium form Al6Mn, which has almost the same electrode potential as aluminium, and this compound is capable of dis- solving iron which reduces the detrimental effect of this element. Magnesium in excess of that in solid solution in binary aluminium alloys tends to form the strongly anodic phase Mg5Al8 which precipitates in grain boundaries and pro- motes intercrystalline attack. However, magnesium and silicon, when together in the atomic ratio 2:1, form the phase Mg2Si which has a similar electrode potential to aluminium. Where a basic alloy is vulnerable to corrosive attack, it is possible to pro- vide surface protection for wrought products such as sheet, plate, and, to a lesser extent, tube and wire by means of metallurgically bonded thin layers of pure aluminium or an aluminium alloy. Such alloys are commonly those based on the Al–Cu–Mg and Al–Zn–Mg–Cu systems and products are said to be alclad (Fig. 4.8). In a corrosive environment, the cladding will anodic with respect to the core and provide sustained electrochemical protection at abraded areas and exposed edges.  2.4 CORROSION 85 2.4.4 Crevice corrosion If an electrolyte penetrates a crevice formed between two aluminium surfaces in contact, or between an aluminium surface and a nonmetallic material such as a gasket or washer, localized corrosion may occur by etching or pitting. The oxygen content of the liquid in the crevice is consumed by the reaction at the aluminium surfaces and corrosion will be inhibited if replenishment of oxygen by diffusion into the crevice is slow. However, if oxygen remains plentiful at the mouth of the crevice, a localized electrolytic cell will be created in which the oxygen-depleted region becomes the anode. Furthermore, once crevice attack has been initiated, this anodic area becomes acidic and the larger external cathodic area becomes alkaline. These changes further enhance local cell action and more corrosion occurs in the crevice, particularly in a submerged situation. A common example of crevice corrosion occurs when water is present in the restricted space between layers of aluminium sheets or foil in close con- tact in stacks or coils. This may take place by condensation during storage if the metal temperature falls below the dew point, or by the ingress of rain when being transported. Irregular stain patches may form which impair the surface appearance. In severe cases, the corrosion product may cement two surfaces together and make separation difficult. A special form of crevice corrosion may occur on an aluminium surface that is covered by an organic coating. It takes the form of tracks of thread-like fila- ments and has the name filiform corrosion. The tracks proceed from one or more places where the coating is breached for some reason and the corrosion products raise bulges in the surface. The amount of aluminium consumed is small and fili- form corrosion only assumes practical importance if the metal is of thin cross section. Filiform corrosion occurs only in the atmosphere and relative humidity is the key factor. It has been observed on lacquered aluminium surfaces in air- craft exposed to marine or high-humidity environments and may be controlled by anodizing, chemical conversion coatings or by using chromate-containing prim- ers prior to painting. 2.4.5 Cavitation corrosion Protective films on the surfaces of aluminium and its alloys may be removed by mechanical actions of many sorts such as turbulent effects arising from mov- ing fluid. Electrolytic reactions may then occur which can proceed without inhibition. If voids (gas bubbles) form in the turbulent liquid because the pres- sure falls below the vapor pressure, then cavitation corrosion may take place. Collapse of these voids at the metal surface allows the sudden release the latent heat of vaporization, which may dislodge a protective film during service and even alter the state of work hardening of the metal at the surface. Cavitation corrosion therefore combines electrochemical action with mechanical damage, the relative proportion of each being controlled by the severity of the turbulence and the aggressiveness of the environment. 86 CHAPTER 2   PHYSICAL METALLURGY OF ALUMINIUM ALLOYS Weight loss in standard tests on aluminium alloys has been found to decrease as strength (and hardness) increases. However, compared with other nonferrous metals and alloys, aluminium and its alloys do not perform well under cavitation conditions. For example, common wrought aluminium alloys, that are considered to be relatively resistant to corrosion, have been found to suffer weight losses 100–200 times greater than the copper alloy, aluminium bronze, under cavitation conditions in fresh water. 2.4.6 Waterline corrosion This form of corrosion can affect semisubmerged structures, such as ships, whereby the zone very close to the air/water boundary can suffer differential corrosion that is sometimes severe. With aluminium alloys, waterline corrosion may arise because of a difference in the chloride level between the seawater at the air/water parting line and that contained in the meniscus formed by cap- illary action in which the chlorides become concentrated by evaporation. This effect is weak in water that is in motion because the meniscus is constantly being renewed. Although aluminium alloys used for the hulls of ships and other semisubmerged structures are not very sensitive to waterline corrosion, this region should be painted to avoid the risk of attack. If the water is stagnant, painting is essential. 2.4.7 Metallurgical and thermal treatments Treatments that are carried out to change the shape and achieve a desired level of mechanical properties in aluminium alloys may also modify corrosion resistance, largely through their effects on both the quantity and the distribu- tion of microconstituents. In this regard, the complex changes associated with ageing or tempering treatments are on a fine scale and these are considered in Chapter 4. Both mechanical and thermal treatments can introduce residual stresses into components which may contribute to the phenomenon of stress- corrosion cracking and this is discussed in Section 2.5.4. If one portion of an alloy surface receives a thermal treatment different from the remainder of the alloy, differences in potential between these regions can result. Welding processes provide an extreme example of such an effect and dif- ferences of up to 0.1 V may exist between the weld bead, heat-affected zones, and the remainder of the parent alloy. Most wrought products do not undergo bulk recrystallization during sub- sequent heat treatment so that the elongated grain structure resulting from mechanical working is retained. Three principal directions are recognized: longitudinal, transverse (or long transverse), and short transverse, and these are represented in Fig. 2.42. This directionality of grain structure is signifi- cant in components when corrosion processes involve intercrystalline attack  2.4 CORROSION 87 Figure 2.42 The three principal directions with respect to the grain structure in a wrought aluminium alloy. Note the appearance of cracks that may form when stressing in these three directions. From Speidel, MO and Hyatt, MV: Advances in Corrosion Science and Technology, Plenum Press, New York, NY, USA, 1972. as has been illustrated by exfoliation corrosion. It is particularly important in regard to SCC, which is discussed in Section 2.5.4. In certain products such as extrusions and die forgings, working is non- uniform and a mixture of unrecrystallized and recrystallized grain structures may form between which potential differences may exist. Large, recrystallized grains normally occur at the surface (see Fig. 4.7) and these are usually slightly cathodic with respect to the underlying, unrecrystallized grains. Preferential attack may occur if the relatively more anodic internal grains are partly exposed as may occur by machining. 88 CHAPTER 2   PHYSICAL METALLURGY OF ALUMINIUM ALLOYS 2.5 MECHANICAL BEHAVIOUR The principal microstructural features that control the mechanical properties of aluminium alloys are as follows: 1. Coarse intermetallic compounds (often called constituent particles) that form interdendritically by eutectic decomposition during ingot solidifica- tion. One group comprises virtually insoluble compounds that usually con- tain the impurity elements iron or silicon and examples are Al6(Fe,Mn), Al3Fe, α-Al(Fe,Mn,Si), and Al7Cu2Fe. The second group, which are known as the soluble constituents, consists of equilibrium intermetallic compounds of the major alloying elements. Typical examples are Al2Cu, Al2CuMg, and Mg2Si. Both types of particles form as lacy networks sur- rounding the cast grains and one purpose of the process referred to as pre- heating or ingot homogenization (Section 4.1.5) is to dissolve the soluble constituents. During subsequent fabrication of the cast ingots, the largest of the remaining particles usually fracture, which reduces their sizes to the range 0.5–10 μm and causes them to become aligned as stringers in the direction of working or metal flow (Fig. 2.43). Constituent particles serve no useful function in high-strength wrought alloys and they are tolerated in most commercial compositions because their removal would necessitate a significant cost increase. They do, however, serve a useful purpose in certain alloys such as those used for canstock (Section 4.6.5). 2. Smaller submicron particles, or dispersoids (typically 0.05–0.5 μm) that form during homogenization of the ingots by solid-state precipitation of compounds containing elements which have modest solubility and which diffuse slowly in solid aluminium. Once formed, these particles resist either dissolution or coarsening. The compounds usually contain one of the transi- tion metals and examples are Al20Mn3Cu2, Al12Mg2Cr, and Al3Zr. They serve to retard recrystallization and grain growth during processing and heat treat- ment of the alloys concerned. Moreover, they may also exert an important influence on certain mechanical properties through their effects both on the response of some alloys to ageing treatments and on dislocation substruc- tures formed as a result of plastic deformation. 3. Fine precipitates (up to 0.1 μm) which form during age hardening and nor- mally have by far the largest effect on strengthening of alloys that respond to such treatments. 4. Grain size and shape. The most significant microstructural feature that dif- ferentiates wrought products such as sheet from plate, forgings, and extru- sions is the degree of recrystallization. Aluminium dynamically recovers during hot deformation producing a network of subgrains and this charac- teristic is attributed to its relatively high stacking-fault energy. However, thick sections, which experience less deformation, usually do not undergo bulk recrystallization during processing so that an elongated grain structure is retained (Fig. 2.42).  2.5 MECHANICAL BEHAVIOUR 89 Figure 2.43 Aligned stringers of coarse intermetallic compounds in a rolled aluminium alloy (× 250). 5. Dislocation substructure, notably that caused by cold working of those alloys which do not respond to age hardening, and that developed due to ser- vice stresses. 6. Crystallographic textures that form as a result of working and annealing, particularly in rolled products. They have a marked effect on formability (Section 2.1.4) and lead to anisotropic mechanical properties. Each of these features may be influenced by the various stages involved in the solidification and processing of wrought and cast alloys and these are dis- cussed in detail in Chapters 4 and 5. Here it is relevant to consider how these features influence mechanical behavior. 2.5.1 Tensile properties Aluminium alloys may be divided into two groups depending upon whether or not they respond to precipitation hardening. The tensile properties of commer- cial wrought and cast compositions are considered in Chapters 4 and 5. Here it is the finely dispersed precipitates that have the dominant effect in inhibiting 90 CHAPTER 2   PHYSICAL METALLURGY OF ALUMINIUM ALLOYS dislocation motion, thereby raising yield and tensile strengths. For the other group, the dislocation substructure produced by cold working in the case of wrought alloys and the grain size of cast alloys are of prim