Back-annealing of cold rolled steels through recovery and/or partial recrystallisation

Transcription

1 Back-annealing of cold rolled steels through recovery and/or partial recrystallisation R. K. Ray* 1, B. Hutchinson 2 and C. Ghosh 1 An evaluation of the back-annealing approach for the production of high strength steel sheet is presented. This approach is different from those used in conventional current practice, e.g. for rephosphorised, microalloyed, dual phase, multiphase and transformation induced plasticity steels. The principle is to retain as much as possible of the hardening due to cold rolling while gaining ductility by controlled low temperature heat treatment (back-annealing) through recovery and/or partial recrystallisation. Unlike other processes to achieve high strength, this approach offers the possibility of cost effective high speed production in continuous annealing lines. The important role of back-annealing or recovery annealing in the production of ultrafine grains in steels and in the development of higher strength ultra-thin strip cast products has also been widely recognised. In the present review, the fundamentals of the back-annealing process are discussed in terms of the constituent phenomena of recovery and recrystallisation and their kinetics. The challenges of back-annealing in steels, in terms of the production of sheets of thin gauges, are identified and possible methods to overcome these are suggested. Theoretically model approaches to the back-annealing process are also discussed. Keywords: High strength steels, Back-annealing, Ultrafine grained steels, Ultrathin strip cast products, Mechanical properties, Review Introduction Demand for the use of steels with higher and higher strength levels is increasing continuously. This has been particularly due to the increasing demand from automobile manufacturers for lighter and, therefore, more fuel efficient cars with improved corrosion resistance. Higher strength steels are desirable for better dent resistance, increased load bearing capacity, better crash worthiness and reduction in the amount of steel through a decrease in sheet metal thickness. Of course, such high strength steels must also possess adequate ductility for good formability and workability. In general, whenever a higher strength steel is substituted there is a reduction in the amount of steel that has to be produced and this means savings in energy during iron- and steelmaking, smaller amounts of waste products and smaller emissions of CO 2 into the atmosphere. The special ability of steels, in general, to be recycled and reprocessed without any deterioration adds to the environmental benefits derived from the usage of high strength steel. However, the addition of alloying elements to increase strength makes the selection of scrap more complicated in some cases, and several of these elements (e.g. Mn, Ti, Nb, Cr) are lost into the slag during remelting which implies a waste of natural resources. There is, therefore, an 1 Department of Research & Development, Tata Steel Ltd, Bistupur, Jamshedpur , India 2 Swerea KIMAB AB, SE Stockholm, Sweden *Corresponding author, rkray@iitk.ac.in argument in favour of producing high strength steel by using as low a level of alloying elements as possible in a low carbon steel base. Furthermore, steel processing operations that can be carried out at relatively low temperatures should be preferred since these will ensure saving of both energy and costs. A wide range of steels has been developed in recent years, having higher strength levels than conventional deep drawing and formable steels. However, it is well known that when a steel is strengthened, its ductility almost always decreases. There is accordingly a trade off between strength and ductility for different families of steel sheets as depicted in Fig. 1, taken from the work of Akisue and Hada. 1 The different classes of high strength steels shown in this figure achieve their mechanical properties through different strengthening mechanisms, such as grain refinement, solid solution hardening, precipitation hardening or phase transformation. Although every class shows the same tendency of reduced ductility as strength increases, they do not all coincide. Thus, different hardening mechanisms may give more or less favourable combinations of strength and formability. It is well known that metals are strengthened by cold working. During cold rolling which is a necessary stage in sheet production, the steel becomes intensely hardened but loses almost all its ductility. During conventional processing the cold rolled steel is fully recrystallised by annealing with the purpose of restoring ductility, but at the expense of strength. Then, if higher strength is demanded it is normally achieved by either ß 2011 Institute of Materials, Minerals and Mining and ASM International Published by Maney for the Institute and ASM International DOI / X International Materials Reviews 2011 VOL 56 NO 2 73

2 1 Relationship between strength and ductility for high strength cold rolled low carbon steels 1 alloying or by special heat treatments at higher temperatures. The above procedures entail wastage of energy and increased costs. On the other hand, by controlled low temperature annealing, it should be possible to retain as much as possible of the strength from cold rolling while at the same time restoring adequate ductility. This will yield a material with properties which are a compromise between the high strength low ductility of fully cold rolled sheet and the low strength high ductility of fully recrystallised sheet. Such a process has been variously termed as backannealing, partial annealing, recovery annealing, stress relief annealing, temper annealing or controlled incomplete annealing. Basically, the process consists of two stages, deformation and annealing. An alternative approach called temper rolling or rolling to temper, seeks to obtain the desired strength and ductility by plastically deforming the already fully annealed material, 2 while in back-annealing, a heavily deformed material is either subjected to recovery or partial recrystallisation the objectives in both the cases are the same. It has been claimed 3 5 that at all strength levels the ductility of the back-annealed material is always superior to that of temper rolled material. Furthermore, the former appears to be more economical than the latter 6 as it involves fewer process stages and lower annealing temperature. Although the present review is concerned with steels, it can be mentioned that back-annealing is a very widely used process in the aluminium industry for producing thin sheet and foil having strength and ductility that are optimised for different applications. 7,8 Many attempts have been made over the years to apply back-annealing to various steel products. Earlier work by Adams and Williams 3 has shown that backannealing can be used with advantage to produce tinplate of thinner gauge, which may compete effectively with other packaging materials. Toth and Lander 9 and Jefford 10 have also reported the application of partial annealing in the manufacture of high strength tinplate. It has been claimed that the tinplate so produced has equal or even better strength than that produced by double reduction, possesses superior ductility and at the same time costs much less. 6 Detailed investigations of the effect of partial annealing on the mechanical properties of low carbon mild steel sheets were undertaken by Devgun et al They also attempted to 2 Practical applications of high strength steel sheets in Japanese automobile sector 21 model the partial annealing process and to control the production of partially annealed material by continuous monitoring of the process using a non-destructive magnetic technique The effect of various process variables on mechanical properties after back-annealing of low carbon steels has been examined in work by Penning et al. 16 A project report on the Development of high strength steel sheet products by back-annealing was prepared by the European Commission in 2007, which embraces various aspects of the back-annealing process in details. 17 Production of microalloyed steels by recovery annealing of cold rolled steels with yield strengths in the range of MPa, has been reported in the USA. 18,19 In fact, the Society of Automotive Engineers has prescribed the use of recovery annealed cold reduced sheet steels in the MPa yield strength range for vehicle outer panels. 20 Figure 2 summarises the practical applications of high strength steels in the Japanese automobile sector, according to the strength elongation balance. 21 There, steels with tensile strength levels of MPa are used in the inner parts of automobiles, whereas ultrahigh strength steels with tensile strength levels of MPa are used in door impact beams and bumper reinforcement parts. This latter category of steel requires very high strength and only a moderate amount of ductility, and is therefore ideally suited for production by the back-annealing process. In fact, a possible use of such material in automobiles for structural and safety related parts, where only moderate formability is adequate, was earlier proposed by Mould. 22 Considerable success was achieved in the application of this type of material in door beam components during preliminary trials. 23 In all the above cases a restriction in the use of the back-annealing process stems from the difficulty of routinely obtaining partially recrystallised structures with consistent properties. It is for this reason that till now softening by recovery alone could be made use of in industrial practice. Probably the largest commercial production of back-annealed steel sheets concerns the sheet products made by Bluescope Steel in Australia where recovery annealing is combined with hot dip galvanising. Although the ductility of these sheets is low, their yield and tensile strengths lie typically in the vicinity of 700 MPa and they are well suited for roll forming applications. Some 5 Mton of these products have been manufactured during recent years International Materials Reviews 2011 VOL 56 NO 2

3 3 Proposed sketch of the back-annealing process suitable for an industrial continuous annealing line Apart from the level of properties that can be achieved in high strength sheet steel products, it is important that the mechanical properties should be uniform from band to band as well as along and across each sheet. This is especially important for higher strength sheet products due to the increased problems of controlling spring back during sheet forming operations. A basic understanding of the back-annealing process itself and its proper control by the choice of appropriate industrial practice are therefore vital for the realisation of the full potential and economic exploitation of this class of steels. More recently, however, another important role of back-annealing or recovery annealing has been recognised. Work by researchers all over the world, especially in Japan, has clearly shown that recovery annealing rather than recrystallisation of cold worked materials can lead to the formation of ultrafine grains in steels. Thus, the possibility of producing ultrafine grained (UFG) steels in the existing plants using normal manufacturing methods appears highly feasible, without utilising the so called severe plastic deformation (SPD) techniques. This realisation will undoubtedly lead to renewed interest and activity in the back-annealing process. This aspect, along with references to published literature, has been thoroughly described in the section on Back-annealing and production of UFG steels and ultrathin strip cast products. A schematic of the back-annealing process suitable for an industrial continuous annealing line is shown in Fig. 3. The Back-annealing process Annealing of a heavily cold worked material can be divided into three physically distinct but generally overlapping stages, i.e. recovery, recrystallisation and grain growth. 25 Recovery dominates at relatively low temperatures and includes the migration of vacancies and of dislocations introduced by cold deformation, leading to the annihilation or rearrangement of a certain portion of them. In more general terms, however, recovery relates to any modification of properties, during annealing, which occurs before the appearance of new strain free recrystallised grains. In other words, recovery does not involve migration of high angle grain boundaries (HAGBs). It is usual to find that during recovery both the mechanical and the physical properties of materials show some change from their values in the cold worked state. Normally the restoration of a mechanical property, such as hardness, yield strength or ductility to its fully annealed value, is only about one-fifth completed during recovery. Following the work of Drouard et al., 26 a parameter R may be defined which gives a measure of the fraction of the particular property recovered R~ X m{x (1) X m {X o where X o is the value of the property in the fully annealed material, X m is the value of the property in the cold worked material and X is the value of the property after a recovery annealing. For an industrial product such as steel, the recovery stage in the annealing process is marked by a slight decrease in strength and a slight increase in ductility, which makes the material suitable for certain applications requiring a small amount of formability. It is almost impossible to detect the progress of recovery from optical microstructures. However, a simple method, based on the X-ray line broadening data, can be utilised to follow the recovery process. 27 A line sharpening parameter R has been defined as R~ I min{i b (2) I max {I b where I min is the intensity minimum between K a1 and K a2 peaks, I max is the intensity maximum of the K a1 peak and I b is the background intensity. Although the parameter R does not give any direct information regarding either the cell size or residual stresses, it does give some general indication of the extent of the recovery process. In general, it has been observed that the rate of recovery of a property from its cold worked value depends on the instantaneous value of that property. 28 In other words, the rate of recovery decreases with time. If R is the fraction recovered at any instant, the rate of recovery can be written as 28 d ð dt 1{R Þ~f ð 1{R Þ (3) If f(12r) is simply proportional to (12R), the above equation may be integrated to give ln (1{R)~kt (4) This equation was first suggested by Cherian et al. 29 for certain types of recovery in aluminium. Another kinetic law which has been found to fit some recovery data 30,31 is known as the logarithmic or International Materials Reviews 2011 VOL 56 NO 2 75

4 hyperbolic law, expressed as (1{R)~b{a ln t (5) where b and a are both constants, dependent on the recovery temperature. A similar expression describing the kinetics of the recovery process was derived by Drouard et al. 26 from theirworkonsinglecrystalsofzinc.thiscanbewrittenas R~c ln t{ Q KT (6) where R is the fraction recovered in time t at a temperature T, c is a constant and Q is the activation energy for recovery. It is evident that Q canbeobtainedbyplotting ln t versus I/T for various fractions of recovery but it should be noted that there is generally no unique value for Q; the activation energy for recovery increases as recovery proceeds. In addition to the X-ray method described above, techniques such as thermoelectrical power (TEP), hardness and electrical resistivity measurements can also be used for monitoring recovery. Among these, TEP has shown to be a powerful technique for following the processes that occur during annealing of deformed materials. 32 Recrystallisation of a cold worked metal involves the nucleation and growth of new strain free grains, which are widely different in orientation from the surrounding matrix. The new grains grow at the expense of the cold worked matrix by high angle boundary migration. Recrystallisation usually occurs at intermediate temperatures, above those required for recovery, and is characterised by substantial decrease in yield strength, marked increase in the ductility and an increasing proportion of recrystallised grains in the optical microstructures. At any stage the progress of recrystallisation can be monitored by increasing the fraction of the area of microstructure occupied by the recrystallised grains which is numerically identical to the volume fraction of recrystallised structure. Recrystallisation essentially consists in the nucleation of strain free grains in the cold deformed matrix and their subsequent growth to the point of impingement. The kinetics of recrystallisation can be described in terms of the so called Johnson Mehl Avrami Kolmogorov (JMAK) model In this model, it is assumed that nuclei are formed at a rate N and that they grow into the deformed material at a linear rate G. The fraction of recrystallised material X will initially rise rapidly with time and, as the new grains eventually impinge on one another, the rate of recrystallisation then decreases, tending to zero as X approaches unity. Using this model, the fraction of recrystallised matrix X can be written as 4 Form of recrystallisation kinetics according to JMAK analysis with different values of the exponent n The essential feature of the JMAK approach is that the nucleation sites are assumed to be randomly distributed. According to this model, a plot of ln [ln 1/ (12X)] versus ln t should be a straight line, the slope of which should give the value of the JMAK exponent n. It is found that equation (7) usually conforms well to experimental recrystallisation kinetics. Theoretically, the exponent can be related to the nucleation behaviour and to the directionality of the growth process. 25 Values of y3 would be expected when this analysis is applied to steels but they are often found to be smaller, probably due to the over simplified assumptions of the model. There is a practical interest regarding the kinetic description since one desires recrystallisation to progress in a slow controlled manner so that it can be halted at the desired stage. As can be seen in Fig. 4, a low value of the exponent n should be preferable in this regard. Recrystallisation is effectively complete at higher temperatures which are marked by an equiaxed grain structure and by a steadily decreasing strength and slowly increasing ductility, ultimately leading to the grain growth stage. The changes in mechanical properties during the three stages of annealing of a cold worked metal are shown schematically in Fig. 5. Challenges of back-annealing of steel Back-annealing can be a successful industrial practice provided it produces consistent mechanical properties in X~1{ exp ð{bt n Þ (7) where n is the JMAK exponent, t is the time of recrystallisation annealing and b is a rate constant, which incorporates the nucleation rate N and growth rate G, and can be expressed as b~a exp { Q (8) RT where A is a constant and Q is the activation energy of recrystallisation. 5 Schematic diagram showing change in mechanical properties during annealing International Materials Reviews 2011 VOL 56 NO 2

5 the sheet steel in a routine manner. It is for this reason that, for production purpose, the operator must have the maximum flexibility in annealing temperature and time. This is either possible in the recovery annealed or the fully recrystallised regions of the annealing curve (Fig. 5), where the changes encountered in mechanical properties are the least over a range of annealing parameters. By contrast, in the partially recrystallised part of the curve, even a small change in temperature or time may result in large changes in properties. This emphasises the fact that not all of the desired combinations of strength and ductility will be achievable by back-annealing unless the possibility of using the partly recrystallised region of the annealing curve can be explored. This will, of course, be determined by the practical processing conditions. It has, however, been claimed by several workers 3,6,11,13,14 that good combination of strength and ductility was consistently achieved by annealing in the partially recrystallised range. It is quite understandable that the industrial practice can be made more flexible by providing a wider recovery annealing window or plateau in the annealing curve (Fig. 5). In plain carbon steel, increasing the amount of cold reduction is found to increase the rates of both recovery and recrystallisation. In other words, cold work makes the kinetics of both recovery and recrystallisation faster. Therefore, it may be quite difficult to provide a wide recovery window or plateau during the annealing of a heavily cold worked (>70%) steel sheet. Therefore, for producing recovery annealed high strength steel, cold reduction should be kept to a low value ((50%), provided that the targeted minimum yield strength is attained. This practice will effectively lengthen the recovery annealing plateau and provide a high operating flexibility. 36 However, because of this limitation on the amount of cold reduction and the present hot band thickness limitation (no less than y1?9 mm), the plain carbon recovery annealed grades are often produced only in gauges of 1?14 mm or thicker. 37 This can be easily understood by assuming a limiting hot band thickness of y1?9 mm and a maximum cold rolling reduction of 40% before backannealing. The resulting thickness of the sheet becomes y1?14 mm. This could be a factor that restricts the market potential for recovery annealed steel sheet. The above is especially true for the practice of batch annealing where temperature variations (y55uc) in the furnace may result in various degrees of annealing, from recovery to partial recrystallisation in different parts of the sheet product and, therefore, large variations in strength level. Methods that may be considered for permitting higher amounts of cold reduction, thereby making it possible to produce thinner gauges of recovery annealed steel sheet are: (i) microalloying the steel with elements such as V, Ti or Nb, which is expected to make the kinetics of the recovery and recrystallisation processes sluggish (ii) switching over to the continuous annealing practice with its closer control over the annealing parameters. In fact, the SAE has suggested that the high strength recovery annealed steels should contain additions of P, Mn or Si (used in high strength solution strengthened steels) as well as of V, Ti or Nb [used in high strength 6 Schematic diagram showing relative importance of parameters affecting the yield strength of HSLA cold rolled grades 38 low alloy (HSLA) steels]. 20 The addition of these solute elements will extend the recovery annealing plateau and coupled with the far better and tighter control over the annealing temperature, which is provided in continuous annealing practice and will ensure much better control over the properties, even for steels with relatively high amount of cold reductions (>70%). Optimisation of mechanical properties Back-annealing can be an efficient low cost process for developing new high strength steel sheets by providing the required amount of partial annealing after cold rolling. The optimisation of the mechanical properties by this processing method will involve close monitoring and control of the following three conditions: (i) steel chemistry (C, Mn, Si, Nb, V, Ti for low carbon steels and Cr, Ni, C, N for stainless (ii) steels) prior processing (hot rolling, coiling and cold reduction) (iii) final annealing (heating rate and annealing temperature, time of hold, cooling and overaging) The back-annealed strength of a given material is directly related to its as cold rolled strength. 23 For a given amount of cold reduction, the cold rolled strength is again a function of the strength of the hot band. The contributions to the strength after cold working arise from the total alloying element content (solid solution strengthening), hot band grain refinement by some of the alloying elements and cooling conditions, precipitation hardening due to the sizes and volume fractions of different dispersed phases and the dislocation density and subgrain size (due to cold work). Finally, the backannealing process will modify the properties of the cold worked material. The desired strength and ductility levels of a steel can therefore be attained by choosing the proper steel chemistry, hot rolling and coiling practice, amount of cold reduction and the extent of backannealing. The relative effects of the various processing parameters on the yield strength of the hot rolled, the cold rolled and of the cold rolled and annealed HSLA grades of steel are shown schematically in Fig This figure clearly shows that the high yield strength of the cold rolled steel is derived basically from the cold rolling process. The effects of alloying additions which strengthen the hot rolled steel, due to solid solution International Materials Reviews 2011 VOL 56 NO 2 77

6 7 Effects of alloying elements in solid solution on the yield strength of ferrite 39 and precipitation strengthening, are more or less transmitted to the subsequently cold rolled material with only some loss in intensity in the case of precipitation due to coarsening. However, the significant contribution of lower coiling temperature on the strength of the hot rolled steel is transmitted to the cold rolled product at a much decreased level. Recovery and/ or partial recrystallisation of the cold rolled steel will retain a reasonably high value of the yield strength (less than the cold rolled strength), along with some reasonable amount of ductility. The effects of the various compositional and processing parameters on the behaviour and mechanical properties of back-annealed steels will now be considered in the following subsections. Steel chemistry The effects of the different interstitial and substitutional alloying elements on the yield strength of ferrite are displayed in Fig The substantial solid solution strengthening effects of P, Si and Mn, as compared with other elements, are clearly shown here. This contribution to the hardening is considered to be essentially constant, irrespective of the process conditions. Nb (Cb), Ti and V are the three other elements whose addition to the compositions of the back-annealed steel has been strongly advocated. 20 These elements exert their influence on the property of the steels not only during back-annealing but also during their entire processing history from hot rolling (controlled thermomechanical processing), through cold rolling to the final back-annealing treatment. These elements, in fact, affect the properties in many ways, such as by increasing the strength (of both austenite c and ferrite a) through the precipitation of carbides, nitrides or carbonitrides; controlling the grain size (of both c and a phases); modifying the transformation temperature and kinetics for the c a transformation; as well as strongly influencing the kinetics of recovery and recrystallisation during the post-cold work annealing process, affecting both strength and ductility. Nb (Cb), Ti and V are strong carbide and nitride forming elements, even at low concentrations. Out of these three, Ti also has a high affinity towards oxygen and therefore can only be used as an alloying element only in Al killed steels. In addition, because of its strong affinity to sulphur, Ti can be used for sulphide shape control in steels. Since the crystal structures of the carbides and nitrides of these elements are isomorphous and soluble in each other, their carbides will inevitably contain some nitrogen, while the nitrides will also contain some carbon. Figure 8, taken from the work by Meyer et al., 39 shows the solubility isotherms of NbC, TiC and VN in steel at different temperatures. It can be seen from this diagram that these three compounds can dissolve and precipitate within the temperature range from 1200 to 900uC, which is required for hot rolling and heat treatment. These hard carbide and nitride particles can strengthen steels which exceed the precipitation strengthening effects caused by other phases. The roles played by the carbides, nitrides and carbonitrides of these three elements during different stages of the hot rolling process of steels have been thoroughly documented 39,40 and therefore their roles during hot rolling will be only briefly discussed, and emphasis will be given to their effects during cold rolling and subsequent annealing. The purpose of hot rolling (controlled thermomechanical treatment) in case of the high strength backannealed cold rolled steel is to produce a high strength hot band by the proper control of the finish rolling 8 Solubility isotherms of a NbC, b TiC and c VN in steel at different temperatures International Materials Reviews 2011 VOL 56 NO 2

7 9 Effects of Nb (Cb), Ti and V on the ferrite grain size of as rolled low carbon steel strip 39 temperature and coiling temperature, taking into consideration the maximum effects of precipitation of the carbides, nitrides or carbonitrides of Nb (Cb), Ti and V and grain refinement. The relative effects of these three elements on the ferrite grain size of as rolled low carbon steel strip are shown in Fig. 9. The correlation between the Nb/Ti/V content and increase in yield strength as a result of grain refinement and precipitation strengthening, in a hot rolled low carbon steel strip is shown schematically in Fig. 10. It is seen from this diagram that of the three elements, Nb has the most pronounced grain refinement and moderate precipitation hardening effects, while Ti has a strong precipitation strengthening but moderate grain refinement action. On the other hand, V produces moderate precipitation strengthening and relatively weak grain refinement. Figure 11 shows the effects of these three microalloying elements on the yield strength of a hot rolled steel strip. The base chemistry of this steel averaged 0?08%C and 0?3%Si and the strip thickness was 8 mm. The effects of thinner gauges, lower coiling temperatures or additional alloying are shown by the arrows in the diagram. Nb, Ti and to a lesser extent V retard both dynamic and static recrystallisation of austenite. 39,41,42 This is illustrated for the action of Nb (Cb) only in Fig. 12. Evidently, both Nb dissolved in solid solution in austenite and fine precipitated particles of Nb carbonitride effectively retard recrystallisation. 39 These effects may be very pronounced during the final reduction in a hot strip or plate mill, 40 to the extent that recrystallisation or partial recrystallisation of the austenite may be completely suppressed, resulting in an extremely fine grained and hard hot band with pronounced preferred orientation (texture). The effects of these microalloying elements in retarding recrystallisation (and recovery) persist during the back-annealing process of the cold rolled hot bands as illustrated in Figs. 13 and 14. Figure 13 depicts the variation of yield strength with annealing temperature for five steel compositions with varying Nb contents. 19 The C content in these steels varied between 0?08 and 0?12%, while the other alloying elements were Mn, Si, Al and Zr. It is clear from this figure that the increase in Nb content increases the recovery annealed yield strength plateau and retards the rate of recovery (and recrystallisation). Evidently, the presence of fine Nb carbonitride particles stabilises the as cold deformed structure. The superiority of Ti over Nb in this respect is illustrated in Fig. 14. When the C content of the steel is brought down to the level of an interstitial free (IF) composition Nb seems to act better than Ti in retarding recovery (and recrystallisation) during the annealing of cold worked steels (Fig. 15). The Al killed steel shows a much shorter recovery window as compared with the IF steels. The yield strength of cold rolled steel also depends on the presence of Ti or Nb in the composition. Figure 16 taken from the work of Meyer et al. 39 shows a comparison of the yield strengths of Ti and Nb containing steels, both in the hot rolled as well as in the cold rolled and box annealed conditions. Presumably, box annealing produced a completely recrystallised structure in both the steels in this work. The fact is that the dependence of the yield strength on the Nb and Ti levels after cold rolling and annealing is very similar to that observed in the hot band. Here again, the strengthening effect of Ti is considerably greater than that of Nb. In fact, yield strengths as high as 500 MPa or even more could be achieved in the Ti containing alloy in the cold rolled and annealed (recrystallised) condition. This value can further increase when the cold rolled steel is back-annealed, producing a recovered or only a partially recrystallised condition. 10 Schematic diagram showing the relationship between the a Nb (Cb), b Ti and c V content and increase in yield strength as a result of grain refinement and precipitation strengthening in hot rolled low carbon steel strip 39 International Materials Reviews 2011 VOL 56 NO 2 79

8 a Nb; b Ti; c V 11 The effect of microalloying content on the yield strength of hot rolled steel strip 39 a dynamic recrystallisation; b static recrystallisation 12 Effect of Nb (Cb) on the recrystallisation of austenite in a 0?05%C, 1?8%Mn steel Effect of Nb in retarding recovery and recrystallisation during back-annealing International Materials Reviews 2011 VOL 56 NO 2

9 14 Superiority of Ti over Nb in retarding recovery and recrystallisation during back-annealing Superiority of Nb (columbium) over Ti in retarding recovery and recrystallisation during back-annealing of a low carbon IF steel 39 a St 14 with Nb (Cb); b St 14 with Ti 16 Strengthening effect of Nb (Cb) and Ti in hot strip and cold rolled sheet (after temper rolling) 39 Finally a note may be added on the general effect of the microalloying elements on strengthening of steel through the refinement of grain size and subgrain size. Addition of Nb is known to strengthen steels due to the strain induced precipitation of Nb carbonitride particles during processing. 40 The magnitude of strengthening is proportional to the volume fraction of the precipitates. Presumably, the total amount of precipitated particles induced by deformation during hot rolling will very much depend on the finish rolling temperature, the amount of deformation and the Nb content of the steel. Mangonon and Heitmann, 40 who worked on two low carbon steels without any Nb addition and two others with Nb, observed that the Nb containing steels had finer ferrite grain sizes (ASTM 9 12) than the other two steels without any Nb (ASTM 6?5 10?5). The relationship of the lower International Materials Reviews 2011 VOL 56 NO 2 81

10 partial or complete recrystallisation with a very fine grain size. In this respect, heavy cold rolling reduction before annealing has been found to be advantageous. Back-annealed austenitic stainless steels also exhibit improved fatigue strength as compared with the conventional fully annealed condition Relationship between ferrite grain size and lower yield strength in several steels, with and without Nb addition 40 yield strengths of the steels with respective grain diameter, as determined by them, is shown in Fig. 17. In the Nb steels, the austenite may remain unrecrystallised when extensive hot deformation takes place at lower finishing temperatures and the transformed ferrite grains may contain small subgrains. Depending on the processing conditions, the strengthening effect of the subgrains may be nearly equal to that due to the precipitation of Nb carbonitrides. 36 Figure 18 also taken from this source clearly illustrates how strengthening due to subgrains may be related to the subgrain size. When the hot rolled and coiled microalloyed steels are cold rolled before the final back-annealing treatment, a part of their strength will presumably come from the initial subgrain structures. From the above it is quite clear that single phase low carbon steels are not amenable for processing by backannealing, since the loss in ductility in the partially recrystallised condition is rather high in relation to the increase in strength. On the other hand, microalloying with V and/or Nb leads to an improvement in the strength ductility balance after back-annealing and makes the heat treatment somewhat easier to control. Back-annealing can also be used with advantage in case of the more advanced multiphase steels such as the dual phase and the transformation induced plasticity steels. Dual phase steels with partially recrystallised ferrite matrix, produced by back-annealing, possess good bendability coupled with high strength. Similarly, transformation induced plasticity steels processed via back-annealing show very high strength levels together with significant amount of ductility. 17 Back-annealing of austenitic stainless steels can produce acceptable strength ductility combinations by low temperature annealing, which produces either Processing parameters The back-annealing of steel basically consists of two steps: production of a high strength steel strip by cold rolling of a suitably produced hot band and partial annealing of the cold rolled steel to induce higher ductility at the expense of strength. The hot rolling practice should utilise the maximum application of precipitation hardening by the microalloying elements, together with grain refinement to produce a high strength hot band. The basic requirements for the hot rolling practice have been briefly outlined in connection with describing the effects of Nb, Ti and V in the section on Steel chemistry. An important factor is the effect of the hot mill coiling temperature on the properties of back-annealed steels. The coiling temperature after hot rolling affects the size and distribution of precipitated particles and therefore the mechanical properties of the resulting hot band. Figure 19, taken from Meyer et al., 39 shows the effect of post-hot rolling annealing temperature on the strength of Nb and Ti hot rolled strips coiled at various temperatures. This diagram demonstrates that a lower 18 Relationship between subgrain size and strengthening in Nb containing steels 36 a tensile strength; b yield strength 19 Effect of annealing temperature on the strength of Nb (Cb) and Ti hot rolled strip coiled at various temperatures International Materials Reviews 2011 VOL 56 NO 2

11 20 Effect of hot mill coiling temperature on the recovery annealed yield strength of a 0?09%Nb steel, cold reduced 50% (Ref. 19) coiling temperature is beneficial in producing a higher strength hot band. A stress relief anneal of up to y650uc does not change the strength properties attained in the as hot rolled condition. The yield strength, in fact, shows even a slight increase up to this temperature for both the Nb and the Ti containing steel, indicating that complete precipitation strengthening by full use of the microalloying elements is not achieved by self-tempering in the coil. The beneficial effect of a lower coiling temperature persists so far as the strength of the subsequently cold rolled and finally back-annealed material is concerned. This is clearly illustrated in Fig. 20 which depicts the effect of coiling temperature on the recovery annealed yield strength of a steel containing y0?09%nb, after 50% cold reduction. 19 Although there is considerable scatter in the data shown in Fig. 20, this relationship may nevertheless be considered important. A favourable effect of low coiling temperature was also reported for plain low carbon steels subjected to recovery annealing treatments. 16 Perhaps one of the most important processing parameters for the production of high strength backannealed steels is the amount of cold rolling reduction which is applied to the hot band after coiling. Figure 21 depicts the effect of cold reduction on the tensile strengths of different grades of HSLA and deep drawing steels. 38 This figure shows that, irrespective of the initial tensile strength levels of the hot rolled materials, the work hardening is characterised by a nearly constant gradient of y50 MPa per 10% cold reduction. It is quite easy to understand from those results that the higher the tensile strength of the hot rolled material, the more difficult it will be to produce thinner gauges by cold rolling. On the other hand, a harder base composition is of special advantage if for some reasons the required amount of cold reduction cannot be realised. Figures 22 and 23 illustrate the effects of cold rolling reduction and annealing temperature on the yield strength and the per cent elongation for a 0?02%Nb containing steel. 19 The cold rolled steels were annealed at a constant heating rate and the annealing time at each temperature was 20 s. The above two diagrams clearly indicate that increasing cold reduction leads to an increase in the rates of both recovery and recrystallisation. This means that the recovery window or plateau is effectively shortened by an increase in the amount of cold reduction. The progress of recovery and recrystallisation after back-annealing was also followed by these authors by checking the microstructures at different stages of 21 Effect of cold reductions on the tensile strengths of HSLA and deep drawing steels 38 International Materials Reviews 2011 VOL 56 NO 2 83

12 22 Effect of cold reduction and annealing temperature on the yield strength of a 0?02%Nb steel 19 annealing. The above diagrams clearly indicate that provided the minimum target yield strength is attained, the best practice will be to keep the amount of cold reduction as low as possible. This will lengthen the useful recovery annealing window and ensure maximum operating flexibility. It is perhaps noteworthy that the same condition does not apply for aluminium alloys where a high degree of cold reduction is preferred for the best combination of strength and formability after back-annealing. 43 In this case the cold rolling deformation should preferably exceed y95% to achieve structural changes on annealing that have been described as continuous recrystallisation. In fact, such high reductions are seldom considered for steels due to their greater hardness and difficulty of rolling. There are, however, indications from the work on steels that a similar high deformation regime suitable for back-annealing of steels may also exist. 44 Keeping the amount of cold rolling reduction to a low value may not be always advantageous as it may effectively mean a limited market potential for the recovery annealed material in terms of sheet thickness as pointed out in section on Challenges of backannealing of steel. One possible way to circumvent this problem could be to use a higher level of microalloying element in the steel to allow a longer recovery annealing window during annealing after heavier cold rolling reductions. This will be amply clear from Figs which show that both Ti and Nb effectively make the recovery and recrystallisation processes sluggish that a higher amount of these elements has a greater effect in this direction. Figures 13, 22 and 23 illustrate the effects of annealing temperature on the mechanical properties of backannealed steels. Clearly, the alloy chemistry, amount of cold reduction and annealing temperature can be 23 Effect of cold reduction and annealing temperature on the per cent elongation of a 0?02%Nb steel International Materials Reviews 2011 VOL 56 NO 2

13 24 Effect of annealing time and temperature on the back-annealed yield strength of a 70% cold rolled 0?031%Nb steel 36 adjusted quite effectively to arrive at the right combinations of properties in these materials. The time of annealing, or soaking time, is another important processing parameter in these steels. Figures 24 and 25 from the unpublished work of Pradhan 36 show the variations of yield strength and total elongation respectively, in a 0?03%Nb steel with annealing time at five different annealing temperatures. Evidently, the initial (cold rolled) strength of the material is almost fully retained during recovery at the lower annealing temperatures from 800uF/427uC to 1000uF/538uC, over a long period of time, with very little improvement in total elongation. Recovery is followed by recrystallisation at a later stage in the 1100uF/593uC annealed material, showing an attendant substantial increase in the total elongation value. After annealing at 1200uF/650uC recrystallisation appears to start from the very beginning, leading to a substantially lower yield strength and considerably higher amount of total elongation. The interaction between the annealing temperature and time for a 50% cold rolled 0?02%Nb steel is very clearly shown in Fig As expected, shorter annealing times will allow the use of higher annealing temperatures to produce an equivalent annealed condition and therefore comparable mechanical properties. It is quite apparent from the above that relaxation processes such as recovery and recrystallisation are both time and temperature dependent and these latter parameters are by no means mutually independent. Larsen and Salmos 45 have factored these two variables into a single parameter M such that M~ ðtz273þðlog tz20þ 10 {3 (9) where T is in uc and t is in seconds. The above expression effectively states that different annealing treatments involving different T t combinations corresponding to the same M value will result in similar mechanical properties and microstructures. In fact, this can be clearly verified from the data presented in 25 Effect of annealing time and temperature on the per cent total elongation of a 70% cold rolled 0?031%Nb steel 36 International Materials Reviews 2011 VOL 56 NO 2 85

14 26 Effect of annealing temperature and time on the mechanical properties of a 50% cold rolled 0?02%Nb steel 19 Figs. 24 and 25 by determining the values of yield strength and total elongation for different T t combinations. Thus, the progress of recovery or recrystallisation during back-annealing of cold rolled steel can be expressed reasonably well by specifying the relevant M values for these processes, rather than stating the corresponding T t combinations. After analysing the data contained in Figs. 24 and 25 and then replotting them in Fig. 27 in terms of M values, Pradhan 36 came to the following conclusions (i) for recovery, M value lies between 16 and 19?3 (ii) recrystallisation starts from M value of y19?3 (this particular value of M has been referred to by him as M rs ) and proceeds between the M values of y19?3 and 20?5 (iii) values of M.20?5 correspond to the grain growth process after primary recrystallisation is complete. An implication of the above seems to be that by choosing the right kind of compositional and processing variables, the value of M rs can be increased and this will result in a wider recovery annealing window and therefore a higher flexibility of the operating conditions. This will indeed be the goal for the production of high strength back-annealed steel sheet. A comparison between the plots of the annealing data, as shown in Figs. 24 and 25 in one case and that in Fig. 26 in the other, reveals that plotting in terms of M values brings the different physical processes involved in annealing into much sharper focus. Thus, this kind of plots should be quite useful in planning for better process control and flexibility. In addition to this, the M plots offer other benefits in terms of understanding separately the effects of different factors such as alloy content and per cent cold reduction on the final properties. This can be seen from the set of diagrams shown in Figs These diagrams were redrawn from the figures given in the report by Pradhan. 36 Figures 28 and 29 depict the yield strength and per cent total elongation as a function of M for a plain C, a 0?031%Nb and 0?060%Nb steel cold reduced by 40 and 70% respectively. They clearly show that the effective recovery annealing window can be made wider, either by the addition of Nb to the steel or by decreasing the amount of cold reduction or by both. Both these factors will push the M rs to higher values. With the help of such diagrams the right M value, based on the desired combination of yield strength and elongation, can be easily determined. With the aid of the M value, appropriate annealing cycles, based on both temperature and time of annealing, can be designed. Figures 30 and 31 are diagrammatically similar to Figs. 28 and 29, but depict the effect of V additions on the mechanical 27 Effect of M value on the yield strength and per cent total elongation of a 70% cold reduced 0?031%Nb steel International Materials Reviews 2011 VOL 56 NO 2

15 28 Yield strength and per cent total elongation as a function of M for a plain C, 0?031%Nb and 0?060%Nb steels, 40% cold reduced 36 properties of back-annealed steel sheet. The effects of Nb and V appear to be rather similar, although Nb seems to be a somewhat more effective strengthener than V and also gives rise to better elongation values at 70% cold reduction level. On the other hand, at the 0?03% level, V is more effective than Nb in increasing the M rs value. It has been quite clear from the above that the microalloying elements, Ti, Nb and V are quite effective in widening the recovery plateau during annealing. However, the M value plots do not show any 30 Yield strength and per cent total elongation as a function of M for a plain C, 0?03%V/0?009%N and 0?10%V/ 0?009%N steels, 40% cold reduced 36 encouraging results on the effects of these elements on the slope of the recrystallisation portion of the mechanical property versus M value plots. However, the isothermal softening diagrams (Figs. 14 and 15) do indicate that the addition of Nb or Ti makes the slope of the recrystallisation portions of those diagrams somewhat gentler, which may provide a higher degree of operational flexibility. As has been already mentioned in section on Challenges of back-annealing of steel, back-annealing will be industrially viable, provided the annealing treatment is carried out in a continuous annealing line 29 Yield strength and per cent total elongation as a function of M for a plain C, 0?031%Nb and 0?060%Nb steels, 70% cold reduced Yield strength and per cent total elongation as a function of M for a plain C, 0?03%V/0?009%N and 0?10%V/ 0?009%N steels, 70% cold reduced 36 International Materials Reviews 2011 VOL 56 NO 2 87

16 rather than by batch annealing. If recovery annealing is to be carried out by the commercial batch recovery annealing process, a typical annealing temperature time combination will be y900uf/482uc and y12 hr. Pradhan 36 has calculated that these processing conditions correspond to an M value of 18?6 0?8, taking into account the variation of, at best, 50uF/28uC in the annealing temperature, with a potential for wider variations. This variation in M implies that the temperature is not uniform throughout the coil during annealing and therefore it may not be possible to avoid some portions of the coil undergoing partial recrystallisation and the attendant decrease in strength. By contrast, in a typical continuous annealing line, the processing parameters will be an annealing temperature of about 1100uF/593uC 25uF/14uC and an annealing time of about s and this will correspond to an M of 18?9 0?3. Thus, the variation in M value for continuous annealing is effectively less than half of that for batch annealing. As a result, batch annealing is bound to lead to a much wider variation in the mechanical properties than continuous annealing. Because of the much tighter control of the processing parameters (and hence the M value) during continuous annealing, higher cold reductions (>70%) can be tolerated and consistent mechanical properties can be achieved even in lighter gauges, which are not possible to be produced by batch annealing. Figure 32 shows the effect of annealing temperature on the longitudinal yield strength and total elongation of an Accu-Form 100 RK steel (Bethlehem Steel), which was designed to meet the minimum yield strength requirement of 100 ksi/ 685 MPa. 18 This steel was cold reduced 70% before annealing in a continuous annealing line. It is clear from this figure that even the recovery annealing plateau or window has a definite slope and, therefore, to achieve satisfactory uniformity in mechanical properties, a tight control of the annealing temperature (within, say 15uF/8uC) is absolutely necessary, which is of course possible in a continuous annealing line. Under such conditions it is possible to achieve consistent mechanical properties, as illustrated in Fig Up till now it has been repeatedly emphasised that a restriction in the use of the back-annealing process for steels could be the difficulty in consistently obtaining partially recrystallised structures, which means that only softening by recovery can be made use of in practice. Recrystallisation leads to rather sudden softening and, therefore does not provide sufficient opportunity or flexibility for the operator in obtaining consistent mechanical properties. These difficulties can be understood from Fig. 34 which illustrates the variation of mechanical properties and microstructure during the recrystallisation of an IF steel. 46 A significant effect which can be seen from Fig. 34 is the sharp increase in ductility that accompanies the last stage of recrystallisation despite the relatively small drop in strength level. The reason for this is not yet apparent but it does imply that a back-annealing treatment that would result in a large degree of recrystallisation would not be of practical interest. Nevertheless, the scope and utility of the back-annealing process will be enhanced if the process could be extended some way into the partially recrystallised region. However, that will require much more stringent control on the microstructural changes, since things happen rather fast beyond the recovery stage. It would, therefore, be definitely advantageous to 32 a yield strength and b total elongation versus annealing temperature for Accu-Form 100 RK steel 18 a yield strength; b total elongation 33 Uniformity of tensile properties achievable for Accu- Form 100 RK steel by continuous annealing International Materials Reviews 2011 VOL 56 NO 2

17 34 Microstructures and mechanical properties resulting from partial recrystallisation annealing of cold rolled IF steel 46 have a means of monitoring and assessing the condition of the microstructure continuously online during the annealing treatment. The proper integration of such an arrangement with the overall process control system may be the ultimate answer to the question of using partially recrystallised material for optimum combinations of mechanical properties, consistently obtainable by the back-annealing process. Various non-contact and non-destructive structural characterisation techniques of materials, based on magnetic or ultrasonic principles have been investigated so far, although these have not yet found acceptance for industrial application. A literature review on these techniques is available. 47 Devgun et al. 12 experimented with a non-contacting technique based on measuring magnetic permeability. These measurements correlated quite well with softening but required separate calibrations for different cold rolling reductions. More advanced techniques have been developed recently 48 which combine 35 Schematic diagram of apparatus for making non-contact ultrasonic measurements on steel at elevated temperatures using a laser source 53 many different magnetic signals including eddy currents and Barkhausen noise, incremental permeability and upper harmonics. They appear to offer a potential, at least for measurements on the sheet after cooling. In recent times some specific non-destructive techniques for the measurement of the extent of recovery and recrystallisation in steels, based on magnetic or ultrasonic principles, have been reported None of these, however, is a noncontact technique, capable of being used online. An approach using lasers for both the generation and detection of ultrasonic pulses offers many advantages over earlier techniques. 52 In particular, it can be applied even at elevated temperatures. A schematic diagram of such a non-contact measuring apparatus is shown in Fig It has been found that the generation of ultrasound with a pulsed laser is constant over a large temperature range. Laser based ultrasonics have been used to monitor recrystallisation of cold rolled low carbon steel and IF steel. 54 A change in the attenuation and scattering of the ultrasound waves between the as rolled and annealed samples takes place due to the formation of small, defect free recrystallised grains. In addition, there is a change in the wave velocity as the crystallographic texture changes during recrystallisation. Laser based ultrasonic transducers (interferometers) give sufficiently rapid response to follow the metallurgical changes that occur and maintain their performance at typical materials processing temperatures. 52 It will definitely be worthwhile to develop this system further and evaluate its ability to control the extent of partial recrystallisation during back-annealing. As mentioned earlier, TEP measurement has been found to be highly sensitive for monitoring the progress International Materials Reviews 2011 VOL 56 NO 2 89

18 36 Typical mechanical property combinations achieved by back-annealing and partial and full recrystallisation in steel sheets: (i), Ref. 44; (ii), Ref. 18; (iii), Ref. 4; (iv), Ref. 3; (v), Ref. 15; (vi), Ref. 22 of the recovery phenomenon. This is due to the fact that at lower annealing temperatures, diffusion of carbon in solid solution takes place towards dislocations, thereby leading to an increase in the TEP value. A schematic representation of the TEP apparatus can be found in Ref. 55. High resolution dilatometry and method based on the image quality of electron backscattered diffraction (EBSD) patterns can be used for the measurement of partial recrystallisation. The details of all the above measurement techniques can be found in Ref. 17. Comparison of some mechanical properties Figure 36 shows typical combinations of yield strength/total elongation data obtained by various workers 3,15,19,20,36 for back-annealed steels. In the same diagram similar mechanical property plots for both batch annealed and continuous annealed (fully recrystallised) cold rolled HSLA steels from Nippon Steel Corporation (Tokyo, Japan) 56 have been included. The values of the tensile strengths for the latter have also been indicated. This figure amply demonstrates that the back-annealing process can indeed yield steel sheets with a wide range of strength/ductility combinations, depending on the steel chemistry and processing conditions. Back-annealing and production of UFG steels and ultrathin strip cast products Of late, quite a bit of effort has been going on to produce high strength UFG steels with minimum alloy costs. These steels have an average grain size,1 mm Ultrafine grained steels with relatively simple chemical compositions and strengthened primarily by grain refinement appear to have great potential for replacing some conventional HSLA steels. At present the techniques to produce UFG steels belong to two main groups. The first group of methods comprises the different SPD techniques, such as equal channel angular pressing, 57,58 accumulative roll bonding, high pressure torsion, etc. The true logarithmic strains required in the above processes for obtaining submicrometre sized grains vary within Again, adopting these processes in industry will involve considerable modification to existing practices and large investment. The second group of methods come under what is known as advanced thermomechanical processing. These essentially involve development of innovative thermomechanical rolling and annealing practices, and as such these are highly amenable to ready adaptation by the existing plants. The true logarithmic strains involved in these processes lie in the range of 2?2 3?6. 65 Among the advanced thermomechanical processes, recovery annealing of warm rolled ferrite has been suggested as a very promising technique. 66 Studies have revealed that pronounced recovery instead of primary recrystallisation is very much required to obtain large fractions of HAGBs as a prerequisite for the development of ultrafine grains during warm deformation. 65,67 69 Primary recrystallisation has not been found generally beneficial in this context since it causes a drastic reduction in the total dislocation density, and also removes the substructure required for the gradual formation of subgrains, which finally convert into ultrafine grains, surrounded by HAGBs. In their work, Song et al. 66 produced steels with ultrafine ferrite grains and homogeneously distributed cementite particles by warm deformation and subsequent recovery annealing. Figure 37a d, from their work, shows the SEM image, EBSD map and TEM images of ultrafine ferrite grains obtained by this process. These ultrafine microstructures were found quite stable against coarsening. Okitsu et al. 70 investigated a low carbon steel sheet with a duplex microstructure composed of ferrite and martensite. The steel was cold rolled to 91% reduction in thickness (e52?8) and then annealed for a short time at uC. They obtained an almost equiaxed UFG microstructure which, according to them, evolved by continuous coarsening of the finely subdivided regions produced by cold work together with recovery. Figure 38a and b shows the relevant EBSD maps for the microstructures. Okitsu et al. 70 also claimed that one such UFG steel, with a grain size of 0?49 mm, exhibited high strength, coupled with total elongation of 8?4%, which is considerably larger than those of submicrometre ferrite single phase structures fabricated by SPD methods International Materials Reviews 2011 VOL 56 NO 2

19 37 a SEM image and b an EBSD map after large strain deformation (e51?6) and subsequent 2 h annealing at 823 K obtained for a plain C Mn steel: CD, compression direction; TD, transition direction; the black lines indicate grain boundary misorientations between 15 and 63u; white lines indicate grain boundary misorientations between 2 and 15u. c TEM image of an UFG steel after large strain warm deformation (e51?6 and 2 h at 823 K) with 0?74 mass-%mn: arrows 1 point at very fine cementite particles inside the ferrite grains; arrows 2 point at coarse cementite particles at ferrite grain boundaries. d corresponding TEM image for a steel with 1?52 mass-%mn (Ref. 66) Ultrafine grains were also produced by Ghosh et al. by cold rolling and recovery annealing of a duplex (ferritezmartensite) steel (Fig. 39). 72 High density of ultrafine grains with sizes between 50 and 250 nm was produced by this method. Refining the final grain size of a low carbon Ti IF steel has been made possible by stabilising the substructures of the heavily cold rolled steel by applying a recovery annealing treatment without allowing static recrystallisation to occur. 73 Formation of ultrafine grains in a 0?10%C, 1?98%Mn steel with a duplex ferrite martensite microstructure, using heavy cold rolling followed by recovery annealing has been reported by Okitsu et al. 70 Heavy cold rolling of a fully martensitic steel, followed by recovery and partial recrystallisation led to the development of ultrafine grains with sizes between 50 and 250 nm in a steel with the composition of 0?17C 1?6Mn 0?52Si 0?08Cr 0?13V 0?01S 0?018P (%). 72 Thus, in all the above cases, recovery annealing has been found to play a very important role in producing ultrafine grains in steels with various starting microstructures. Steels processed in this manner can have reasonably attractive strength elongation combination, for example 870 MPa 9% in a 0?13%C 0?37%Mn martensitic steel after 50% cold deformation followed by annealing at 550uC (Ref. 74) and 900 MPa 11% in a 0?17%C 1?6%Mn martensitic steel after 80% cold deformation followed by annealing at 500uC. 72 Nucor steel has established in their Crawfordsville, Indiana plant the world s first commercial installation for the production of ultrathin cast strip (UCS) via twin roll casting. 75 The strip thicknesses that have been produced regularly are in the range of 0?9 1?5 mm. This has obviously extended the strip thickness range for hot rolled steel products and will also allow substitution for cold rolled strip grades. The potential to further extend the thickness range for high strength structural quality grades to,0?9 mm has been explored by applying cold rolling and recovery annealing. It has been reported 76 that through minimal amounts of cold rolling followed by recovery annealing, production of high strength thin gauge steel sheet has been made possible. Laboratory experiments were conducted on ASTM A1011MSS grade 340 steel. Samples were cold rolled between 10 and 50% and then recovery annealed, simulating the conditions in the galvanising line at Nucor, Crawfordsville. 76 The peak metal temperature was varied from 500 to 850uC. The yield strength versus peak metal temperature and the yield strength versus elongation plots for the above steel grade are shown in Figs. 40 and 41 (Ref. 76) respectively. The above studies clearly demonstrated the potential for this grade to be developed into a high strength structural grade steel. Later mill trials also indicated that by using cold reduction levels,30% it was possible to achieve a strength elongation combination of greater than 480 MPa 10% by recovery annealing of the grade 340 steel (Fig. 42). 76 International Materials Reviews 2011 VOL 56 NO 2 91

20 40 Yield strength versus peak annealing temperature for recovery annealed UCS material, grade 340 feedstock Boundary misorientation maps obtained by EBSD measurement of the low carbon steel cold rolled to 91% reduction, annealed at a 620 and b 655uC for 120 s and cooled to room temperature in water: observed from TD; the bold lines show HAGBs and the narrow grey lines show low angle grain boundaries 70 Modelling of back-annealing process The back-annealing process will be an industrial success only when consistent mechanical properties, such as yield strength and total elongation, will be routinely obtained after annealing of cold rolled steel sheet. As mentioned earlier, many compositional and processing variables can affect the ultimate properties. Precise control of these parameters is essential for the optimisation of the back-annealing process. It is in this connection that there is a need to have a good physical model based on sound concepts, which will also have good predictive value. This will obviate the necessity of pursuing an empirical approach which has now been recognised as being of limited value, in view of the prohibitively expensive nature of industrial scale parametric investigations. Cold rolled metals and alloys, after annealing, undergo both recovery and recrystallisation. Although recrystallisation has been the subject of numerous modelling studies, much less attention has been paid towards recovery. In aluminium alloys, a distinct recovery stage is normally observed before the onset of recrystallisation. In fact, a substantial portion of the stored energy of cold work in Al is released during recovery. Recovery annealing or backannealing of Al based alloys has now become a standard industrial practice for achieving favourable combinations of mechanical properties. In recent years, several attempts have been made to model the recovery kinetics in Al alloys. The simple recovery models by Cottrell and Aytekin 30 and by Friedel 77 were successfully employed to explain the recovery kinetics in aluminium alloys. 78,79 In later years Nes and co-workers attempted to model recovery of mechanical properties during annealing of cold worked metals, with particular reference to Al alloys. Their approach is based on the assumption that the microstructure after cold work and during the subsequent recovery annealing stage is made up of a composite structure, consisting of the subgrains (characterised by 39 Image (TEM) of 80% cold rolled steel annealed at 500uC (Ref. 72) 41 Elongation versus strength curves for recovery annealed UCS material, grade 340 feedstock International Materials Reviews 2011 VOL 56 NO 2