The Effects of Microalloying Elements on Cracking

To ensure the appropriate quality in finished products, it is important that defects in continuously cast products are minimised. As the use of hot charging and thin slab rolling coupled with direct rolling becomes more common, it is increasingly important to produce defect free continuously cast product, as inspection and repair in these situations becomes more difficult.

Of the many types of defect in continuously cast products, only transverse surface cracking is strongly influenced by the presence of microalloying elements. Nb has a particularly strong detrimental effect, and Nb additions of as low as 0.01% can promote cracking. For V steels with <0.005%N, transverse cracking does not appear to occur, although at high levels of V and N (0.15%V, 0.02%N), transverse cracking has been reported. It is believed that transverse cracks form in the mould, and propagate later in the continuous
casting process, particularly during the straightening process. Microalloyed steels can exhibit low ductility over certain temperature ranges, and when the straightening process is carried out in this low ductility region, cracking can occur. In this respect, Nb has a strong effect in deepening the ductility trough, and extending it to higher temperatures. This behaviour is due to the presence of Nb(CN) precipitates, which promote low ductility failures, and retard recrystallisation. The effect of V on hot ductility is much less marked, and only at high levels of V and N does their ductility approach that found in Nb steels. V additions to Nb steels
appear to slightly improve hot ductility, by promoting coarser precipitates. The effects of Ti on hot ductility are complex and still not completely understood. Transverse cracking may be minimised by appropriate selection of steel composition, such as minimising Nb, replacing Nb by V and N combinations, or by making V additions to Nb steels. Machine operating conditions such as secondary cooling strategy are also important in avoiding transverse cracking. By selecting straightening temperatures, which are outside the
temperature range of low hot ductility, cracking can be reduced.

1. INTRODUCTION

During the production of continuously cast products, it is very important to avoid both surface and internal defects, as otherwise expensive and time consuming slab or bloom/billet repair operations are required, or defective final product may be produced. The production of defect free continuously cast products is becoming ever more important as the use of hot charging and direct rolling from thin slab casters increases. In these situations, inspection and repair of defects in continuously cast slab becomes more difficult, and the production of defect free continuously cast products is of vital importance. Some high strength, microalloyed steels are particularly prone to some types of continuous casting defect, and there have been a number of excellent reviews on the topics of hot ductility and defects in this type of continuously cast product. In such steels, it is found that the type of microalloying elements used, and the overall composition of the steel, is very
important in controlling the amount of defects.The objectives of this report are to review briefly the effects of the microalloying elements V,Nb and Ti on the formation of defects in continuously cast products, the mechanisms by which these microalloying elements influence defects, and to identify possible ways in which
defect-free continuously cast products can be produced.

2. DEFECTS IN CONTINUOUSLY CAST PRODUCTS

2.1 Influence of Composition on Defects in Continuously Cast Products

2.1.1 General

Of the many types of defect in continuously cast product only transverse surface cracks are known to be strongly influenced by the microalloying elements V, Nb and Ti. Some of the other types of surface defect, such as longitudinal surface cracks, are influenced by composition, particularly C (0.07-0.18% being prone to longitudinal cracking) S, P, and Mn/S ratio,5,6) increased P and S, and decreased Mn/S ratio leading to increased cracking. Internal crack formation is also influenced by composition, and again C, S and P are
particularly important.
2.2.2 Transverse Surface Cracks

Nb Steels

There are numerous reports in the literature stating that Nb additions promote the formation of transverse cracks in continuously cast slab. The amount of Nb required for transverse crack formation appears to be very low, and cracking has been reported to increase sharply for Nb additions of as little as 0.01%,  Most authors have reported that for Nb containing steels, increasing Al contents also produced increased cracking, However, other authors have reported no influence of Al on transverse cracking in Nb steels. also shows that other factors as well as composition influence transverse cracking; in this case, one casting machine was performing significantly better than another for the same composition. The influence of machine variables on transverse cracking will be discussed in subsequent sections.Increased N also promotes transverse cracking in Nb containing steels,9,14) but this can be minimised if N is kept below 0.004%.The C content has a very important influence on transverse cracking, and C contents within the range 0.10-0.17% are particularly prone to transverse cracking. reports states increased transverse cracking in Nb steels with higher S levels. However, at very low S levels (<0.005%), transverse cracking was reported to increase in Nb containing
steels. For Nb steels, additions of 0.2-0.3% Cu and Ni have been reported to promote transverse
cracking.There have been conflicting reports concerning the influence of Ca on transverse cracking in Nb steels. Ca additions have been reported to reduce transverse cracking Nb steels. However, calcium silicide has been associated with uneven oscillation marks, which in turn promotes cracking. Elements reported to reduce transverse cracking in Nb containing steels include Ti,Ce and Zr. Ti additions of 0.02-0.04% were required to reduce transverse cracks, but 0.15% Ti was required to completely eliminate the cracks.The previous results relate to conventional thick (i.e. >225mm) continuously cast slab.However, there are also reports of Nb leading to transverse cracking in thin slabs of 50mm thickness.

V Steels

In contrast to Nb steels, there are few references in the literature to V containing steels being associated with transverse cracking. At N < 0.005%, V has little effect on transverse cracking. However, at high N levels (0.02%), transverse cracking can occur in 0.15%V steels. However, when cast as thin, 50mm slab, VN steels are reported to have better surface quality than Nb steels.
Ti Steels

There are few reports on the effects of Ti alone on transverse cracking. it has reported that no slab scarfing was required for steels containing 0.01-0.06%Ti, suggesting that these levels of Ti did not result in any slab surface defects. However, for these steels, C contents were <0.09%, and Mn 0.3-0.8%.

Other Steels

Cryogenic steels containing 9% nickel have also been reported to be prone to transverse cracking.

2.3 Description of Transverse Cracks

Transverse cracks may be formed on the broad face, narrow face, or corner of continuously cast slab, but are not always apparent to visual inspection unless the slab surface is dressed. They are usually associated with the depression of oscillation marks, and are predominantly found on the top slab surface. The cracks themselves can be several 10s of mm in length in extreme cases, and generally follow austenite grain boundaries. The cracks are partially oxidised, but there is little decarbonisation, and little oxidation of the inner end of the crack.Examination of fracture surfaces of transverse cracks has indicated intergranular fracture surfaces, with ductile dimples initiating at a variety of particle types, but predominantly MnS
and AlN.There have also been reports of examinations of what are believed to be the earliest stages of
the initiation of transverse cracks. Examination of break-out shells has indicated small cracks below oscillation marks formed in the mould section. There have been other reports of fine subsurface cracks prior to straightening, and observations of internal cracks formed in segregated regions immediately below oscillation marks.
2.4 Summary

Of the many types of defect encountered in continuously cast products, only transverse surface cracks are strongly influence by the presence of microalloying elements. These cracks can form on the broad face, narrow face or at the corners of continuously cast slab, and can be many 10s of mm in length. The presence of Nb greatly promotes the formation of transverse cracks, but V at low N levels has no effect, although combinations of 0.15%V and 0.02%N have been reported to lead to transverse cracking. There are no reports of Ti alone leading to transverse cracks, and Ti additions to Nb steels can be beneficial in reducing transverse cracking.

3. FORMATION MECHANISMS FOR TRANSVERSE CRACKS

3.1 Introduction

For crack formation to occur there must be an applied stress combined with the inability of the material to support this stress. To understand the various crack formation processes in continuous casting therefore requires a knowledge of the sources of stress, and also high temperature properties of the material, particularly ductility. Also it should be noted that crack formation need not proceed uniformly; there may be distinct crack initiation and crack propagation phases.

3.2 Stresses and Strains during Continuous Casting

Stresses can arise from a large number of different causes during continuous casting, and the subject has been reviewed by Lankford.18) Stresses may arise due to transformation effects, thermal effects (variable heat transfer within the mould, temperature gradients within slabs, effects of cooling water sprays, contact with rollers, etc.), friction between strand and mould, bulging of the strand caused by ferrostatic pressure, mechanical effects due to misalignment of the casting machine, and straightening strains.

The observation of numerous, large transverse cracks in the final straightened slab, together with the fact that these are often most numerous on the top surface of the slab (i.e. the surface which is in tension during straightening) suggests that there is much crack propagation induced by the stresses experienced during the straightening process.
3.3 High Temperature Ductility

Cracking is much more likely to occur if regions of low ductility are present. Several techniques are available to study hot ductility under the conditions relevant to continuous casting, and these are discussed in Appendix 1. It is possible to identify 4 distinct regions of low ductility under test conditions relevant to continuous casting. These four regions are illustrated on a schematic hot ductility curve
Region I - Embrittlement by Incipient Melting

Region IIa - Embrittlement by Second Phase Particles - (Mn,Fe)S

Region IIb- Embrittlement by Second Phase Particles - Nb(CN), AlN, V(CN)

Region III - Embrittlement by Transformation

Region I occurs at high temperatures, typically 20-50°C below the mean solidus temperature.Fracture surfaces are characterised by inter-dendritic failure and the presence of particles such as MnS. This region of low ductility is associated with incipient melting at interdendritic and grain boundaries, and is important in the formation of many types of defect in continuously cast products, such as longitudinal surface cracking. The segregation of elements such as S to inter-dendritic regions during solidification is important to this type of
failure.
This region of low ductility may be responsible for the initiation phase of transverse surface cracks, as small subsurface cracks have been observed associated with oscillation marks.
The oscillation marks themselves are regions in which high degrees of segregation of elements such as S, P and Mn19) can occur. Heat transfer to the mould in the vicinity of the oscillation mark is also reduced, which will tend to keep temperatures high, and within the brittle zone.

Region II occurs over the approximate temperature range 1200-900°C, depending on composition and test conditions, and fracture surfaces are typically along austenite grain boundaries, and sometimes show the presence of second phase particles, with ductile dimples around these second phase particles. These low ductility regions are associated with precipitates - (Mn,Fe)S for Region IIa and Nb(CN), V(CN), Ti(CN) and AlN for Region IIb. The distinction between regions IIa and IIb is determined by the stability of the different particle types. Type IIa low ductility is only apparent at quite high strain rates; at lower strain
rates, or when there is an extended hold prior to testing, ductility is good.18,20,21) On the other hand, Type IIb ductility loss is worse as strain rate decreases, Type IIa ductility loss is strongly dependent on composition, particularly Mn/S ratio Figure 8. It hs also been suggested that IIa ductility loss is due to the precipitation of liquid FeS particles, and reduction of grain boundary decohesion due to S segregation. Transverse cracking is usually associated with high strength microalloyed steels, with high Mn contents, and therefore high Mn/S ratios. The strain rates during the processing of continuously cast slabs are also too slow for Type IIa ductility loss to occur, and this suggests that type IIa ductility loss is not responsible for transverse crack formation.

Type IIb ductility loss is initiated by austenite grain boundary sliding, which encourages crack formation at grain boundaries, and the presence of second phase particles such as Nb(CN), V(CN) or AlN. These particles have two major roles; they can delay the onset of recrystallisation, and they can reduce the strain required for fracture. The high temperature end of this ductility trough is believed to be associated with the onset of recrystallisation. If recrystallisation can occur prior to failure, any developing grain boundary cracks become isolated, and further propagation is not possible. It is well known that the microlloying elements Nb and V can delay recrystallisation, either in solution or as precipitates, and this retardation of recrystallisation is believed to be responsible for extending the Type IIb ductility trough to higher temperatures. However, in this respect, V is much less effective than Nb in delaying recrystallistion.

The presence of microalloy precipitates can also reduce the strain to fracture by a number of possible mechanisms: precipitate free zones are often observed adjacent to austenite grain boundaries, and this may lead to strain concentration at the grain boundary; the particles (or groups of particles) at the grain boundaries may act as crack initiation sites; or general matrix precipitation can lead to an increase in strength, and an overall reduction in ductility. The proposed mechanism for low ductility failures in the presence of Nb and V carbonitrides is illustrated.

Region III occurs over the approximate temperature range 900-600°C, depending on composition, and if Type II low ductility is present, these two ductility troughs can merge together. Fracture surfaces are characterised by intergranular failures, and the facets of the individual grains are often associated with void formation around second phase particles. It is believed that this region of low ductility is associated with the austenite to ferrite transformation. On cooling below the transformation temperature, ferrite formation
commences at austenite grain boundaries, leading to the formation of films of ferrite around the austenite grains. At temperatures within the transformation range, ferrite is softer than austenite and so when deformation commences, strain is concentrated within the ferrite at grain boundaries, and the processes of ductile failure, i.e. void nucleation at second phase particles, and the growth of these voids, continues within the ferrite film. Thus on a microscopic scale, fracture can be described as ductile, but overall the failure is brittle. The mechanism for this type of fracture is illustrated.The high temperature end of the ductility trough is associated with the start of transformation, and is thus determined by composition and processing conditions. There appears to be a good relationship between the temperature at ductility starts to fall and the Ar3 temperature, the transformation temperature measured during cooling. It has also been suggested that the temperature at which ductility starts to fall is very close to the equilibrium transformation temperature Ae3, rather than the Ar3, as the deformation process accelerates the transformation kinetics. Ductility recovers at lower temperatures because the volume fraction of ferrite is higher, and the strain distribution between austenite and ferrite becomes more uniform. At lower temperatures, the strength differential between austenite and ferrite is also less, which will again contribute to a more uniform distribution of strain between austenite and ferrite. For ductility to recover completely, it appears that approximately 50% of the austenite must have transformed to ferrite,Microalloying elements can influence the position of this type of ductility trough through their influence on transformation temperature. For example, the presence of Nb in solution
prior to transformation is known to reduce transformation temperatures, and has also been shown to reduce the temperature at which the type III ductility trough occurs,  In this respect V also has a lesser effect, as quite large additions of V are required to depress the transformation temperature significantly. Microalloys can also deepen this type of ductility trough when they are present in the form of precipitates. The precipitates may act as nucleation sites for voids within the thin ferrite films, or reduce the ductility of the thin ferrite films by retarding recovery processes.

3.4 Formation Mechanism of Transverse Cracks

It has been suggested that the earliest stages of transverse crack formation occur in the mould, and are associated with segregation in the vicinity of oscillation marks. The low melting point of these regions, coupled with higher temperatures due to reduced heat transfer to the mould, lead to hot tearing. Further evidence of the importance of events in the mould to transverse cracking is suggested by the strong effect of carbon on transverse cracking. When the C content is such that some peritectic solidification can occur, transverse cracking increases and it has been suggested that this is due to transformation strains during
solidification. There is no evidence to show that microalloying elements influence this stage of transverse crack formation. There have been reports suggesting that Nb additions refine the as-cast grain size, which should help to reduce transverse cracking, but these beneficial effects must be overshadowed by the detrimental effect of Nb on hot ductility.Although the early stages of transverse crack formation may be in the mould, there is evidence to suggest that these defects become larger and more numerous under the
application of stresses from various sources below the mould, particularly those encountered during slab straightening. When these stresses occur in the temperature range over which ductility is poor, transverse cracking is severe, and as hot ductility is strongly influenced by microalloys, this is the proposed mechanism by which microalloying elements effect transverse cracking. As well as having a role in the nucleation of transverse cracks,oscillation marks would also tend to favour the propagation of cracks, in that grain size may often be coarse beneath the oscillation mark, and the notch like geometry will also tend to concentrate stresses.

3. 5 Summary

In summary, there is evidence to suggest that transverse cracks initiate at high temperatures in the mould, and this initiation is associated with oscillation marks. However, subsequent propagation of these cracks continues at lower temperatures is a result of the application of further strains, particularly during slab straightening. When stresses are applied in the regions of low temperature ductility which occur due to the precipitation of microalloy carbides and nitrides, and the presence of the austenite to ferrite transformation, severe transverse cracking can occur.

4. THE INFLUENCE OF MICROALLOYING ELEMENTS ON HOT DUCTILITY

4.1 Introduction

In previous sections, the role of microalloying elements, and particularly Nb, in promoting transverse cracks was described, and the importance of high temperature properties, and particularly low levels of ductility, in leading to crack formation, was highlighted. In this section, the effect of the different microalloying elements on hot ductility will be compared, and the mechanisms by which they influence hot ductility will be discussed in more detail.

4.2 Composition Effects

V Steels

There have been several studies of the influence of V on hot ductility using hot tensile test and typical results are illustrated. The results from all the various studies are consistent in that V additions of up to 0.1% at low N contents (<0.005%) have only a very slight detrimental effect on hot ductility by broadening the ductility trough. At higher N levels, the effect of V additions becomes more marked, and the ductility trough
becomes deeper and broader. In fact, a good relationship can be constructed between the product VxN, and the depth and breadth of the ductility trough, Figure15. It should also be noted that in this example it is only at the highest VxN product, 0.1%Vx0.01%N, that hot ductility approaches that of a 0.028%Nb steel.
In other reports where direct comparisons between the hot ductility of V steels and Nb steels have been made under the same test conditions, the ductility of the V containing steels is superior to that of the Nb steels. For example, in the work of Fu,30) a steel containing 0.16%V and 0.011%N had superior hot ductility to a 0.039%Nb steel. The differences in hot ductility behaviour were attributed to differences in precipitation, in that the V steel exhibited little VN precipitation, whilst the Nb steel showed copious precipitation of NbCN, as a result of the different solubilities of the two precipitate types. Mintz28) has also shown that VN precipitates tend to be coarser than NbCN precipitates under processing conditions similar to those found in continuously cast slab, and hence less detrimental to hot ductility. However, when comparing V-Ti steels and Nb-Ti steels, it has been found that both steel types had similar hot ductility.

In a study using a hot bend test to simulate thin slab casting, additions of 0.1%V were found to have no effect on hot ductility for an N content of 0.007%. As the N level was increased to 0.02%, ductility did decrease, but not to the extent observed in a 0.04%Nb steel,

Nb steels

There have been a very large number of studies of the effect of Nb on hot ductility, and the activity in this field is probably related to the perceived detrimental effect of Nb on slab surface quality. These results can be summarised by saying that Nb additions deepen and broaden the ductility trough to extend to higher temperatures. Nb additions of as little as 0.017% had a detrimental effect, and ductility continues to deteriorate up to at least 0.074%. Typical results are shown. Al additions to Nb containing steels deepened and broadened the ductility trough, as did increasing N contents.24,29) When expressed in terms of AlxN, the combination of increased Al and/or N was also detrimental to the hot ductility of Nb containing steels.

There are conflicting reports as to the influence of P, with some workers reporting a slight beneficial effect of increased P levels, whilst others report no influence.When examining the influence of Ti additions on the hot ductility of Nb steels, care must be taken to ensure that the thermal cycle is appropriate to continuous casting conditions. It is common practice when performing hot ductility tests to carry out a solution treatment at a
high temperature, prior to cooling down to and testing at a lower temperature
1). Whilst this is acceptable for many steels, for steels containing Ti, it can lead to the formation of a fine austenite grain size, due to the grain boundary pinning effects of TiN precipitates which are stable to high temperatures. Thus there are several reports of the apparent benefits of Ti additions to hot ductility, but when the finer grain sizes of these steels are taken into account, the benefits are not as apparent,Thus for simulating
the continuous cast condition for Ti containing steels, it is particularly important that the test pieces are melted in-situ, as described . When only tests from samples melted in-situ are considered, Ti additions to Nb steels have only a very small beneficial effect on hot ductility, or even a detrimental effect, and it is believed that this is due to the influence of Ti:N ratio in controlling precipitate size. If compositions are carefully chosen to ensure that precipitate size is maximised, then good ductility can be achieved in Nb-Ti steels. The picture
is complicated further still be cooling rate effects; at relatively slow cooling rate of 25°C/min, Ti can give a large improvement the hot ductility of Nb steels, again due to the formation of coarse precipitates.As with Ti, the effect of S on the hot ductility of Nb steels also depends upon the thermal cycle used in the test. For steels reheated to a solution temperature prior to testing, S has little effect on hot ductility.39) However, for in-situ melted test pieces increasing S levels have a detrimental effect on hot ductility, as more S is taken into solution to precipitate on grain boundaries.There are several reports that V additions to Nb steels improve hot ductility, compared with a steel containing only Nb, and this may be due to the formation of coarser (V,Nb)(C,N) precipitates in this type of steels.
Ti Steels

As discussed in the previous section, when examining the influence of Ti on hot ductility, it is important to consider the austenite grain size. In some reports, the apparent benefit to hot ductility of Ti additions is due to a refinement of grain size. It is most appropriate to evaluate the influence of Ti on hot ductility using samples melted in-situ, as this technique produces approximately similar grain sizes for Ti and Ti free steels.There are relatively few reports looking at the influence of Ti additions to C-Mn-Al steels after in-situ melting, and relationship between Ti and hot ductility appears to be complex. In situations where large TiN precipitates can form, such as at slow cooling rates or high values of TixN, hot ductility may be slightly improved by Ti additions, but for conditions which generate large volume fractions of fine TiN particles, such as a stoichiometric ratio of Ti:N in low N steels, hot ductility can deteriorate with Ti additions,

4.3 The Influence of Microalloy Precipitation on Hot Ductility

From the discussions in section 3.3 on the occurrence of ductility troughs, and the previous section on the influence of composition on these ductility troughs, it is clear that the precipitation of microalloy carbides and nitrides has a crucial role to play in determining the depth, position and width of the ductility trough, through their influence on dynamic recrystallisation, strain to fracture and transformation. The influence of precipitates on dynamic recrystallisation is dependent on their size and volume fraction, large volume fractions of fine particles having the greatest effect on delaying recrystallisation. Similarly, large volume fractions of fine precipitates are likely to increase strength, and hence reduce ductility.

Equilibrium volume fractions of carbonitride precipitates have been calculated for the steels, and compared with reduction of area values at 850°C in a hot tensile test. Testing was carried out under the same conditions for all the steels, and hence the results are directly comparable. There is a decrease in ductility as the equilibrium volume fraction of precipitates increases, but at a given volume fraction of precipitates, ductility is much lower for Nb steels in comparison with V and V-Nb steels. However, it is difficult to measure experimentally particle volume fractions, and it is unclear whether equilibrium precipitate volume fractions are achieved in hot tensile tests, or indeed under continuous casting conditions. The imposition of strain, either during a tensile test or during continuous casting, rapidly increases the rate of precipitation, and this type of precipitation occurring simultaneously with deformation is known as dynamic precipitation. Studies of dynamic precipitation kinetics in Nb and V steels43) indicate typical “C” curve behaviour (Figure 21) with the time for completion of precipitation being dependent on temperature. For a steel containing 0.035%Nb, and a steel containing 0.11%V and 0.006%N, precipitation kinetics were similar at 875°C, but the time required for completion of precipitation was approximately 10 mins. The time for the tensile test under the conditions observed, and this suggests that precipitation was incomplete during the time taken for the test,
and that the equilibrium volume fractions were not achieved. In recent work, has attempted to relate predicted dynamic precipitation kinetics to the hot ductility curve of microalloyed steels with some success. it  shows a good relationship for Nb steels between Tn, the predicted temperature at which the rate of dynamic precipitation is a maximum, and the temperature for a reduction of area value of 50%. However, the detailed
agreement between precipitation model and experimental results in this work does depend on factors such as the hold time prior to testing in the hot tensile test, and the total level of interstitial elements. The temperature at which ductility begins to recover also appears to be influenced by the volume fraction of precipitates.  It shows how the calculated equilibrium precipitate solution temperature varies with the temperature at which reduction in area recovers to 75%, for the same data that was used to construct Figure 20. There is a general trend for steels with higher solubility temperatures to have higher ductility recovery temperatures, although the
solution temperature is always higher than the ductility recovery temperature. The work  shows a similar trend for Nb steels. As Nb is less soluble in austenite than V, this may explain the tendency for the Nb steels to have wider ductility troughs. As well as precipitate volume fraction, there is clear evidence that the size of Nb(CN) precipitates influence the likelihood of transverse crack formation, and it has been shown that
finer precipitates produce increased levels of cracking,. Similarly, there is evidence that in hot ductility tests, that finer precipitates result in lower ductility,There are several observations that Nb(CN) precipitates observed in hot tensile tests tend to be finer than V(CN) precipitates.26,41) The reason for this may lie in the greater amounts of V in solution, which will promote particle coarsening, and also the higher diffusion coefficient of V in austenite compared with Nb. Thus if it is assumed that precipitate volume fractions for Nb and V steels are similar, the results in Figure 20 may be explained by differences in precipitate sizes, with the V steels having coarser precipitates.
4.4 Summary

Hot ductility is strongly influenced by composition, and the addition of Nb is especially detrimental to hot ductility, extending the ductility trough to higher temperatures, and deepening the ductility trough. The effect of V is much less severe and at N levels of 0.005%, V additions of 0.1% can be made with very little detrimental effect on hot ductility. As N is increased in V steels ductility deteriorates, but even at 0.11%V and 0.01%N, the ductility of a 0.028%Nb steel is still worse. Also, V additions to an Nb containing steel appear to slightly improve hot ductility. The differing effects of V and Nb may be explained by the generally
coarser precipitates observed in V containing steels. The effects of Ti are complex, and depend on the Ti:N ratio.

5. STRATEGIES FOR DEALING WITH TRANSVERSE CRACKS

5.1 Introduction

From the previous sections, it is apparent that that if stresses and strains are introduced during the continuous casting process over certain critical temperature ranges for which ductility is low, transverse surface cracks can occur in continuously cast products. There are many steps, which can be taken to minimise the likelihood that these cracks will form, but it may not be possible to completely eliminate them, in which case some form of slab repair operation is required. The following sections discuss methods to minimise cracking, and the implications of slab repair prior to final rolling.

5.2 Techniques for Crack Minimisation

5.2.1 Control of Composition
It is evident from the above discussion that composition, and particularly the use of microalloying elements, can strongly influence transverse cracking by their influence on hot ductility. This suggests that to minimise cracking a steel composition should be chosen which maximises hot ductility bearing in mind the final product requirements. The following guidelines should help to maximise hot ductility and minimise transverse cracking:
Choose C and alloy additions to avoid peritectic solidification, and particularly avoid 0.1- 0.13%C
Minimise Nb
Use V or V/N combinations to replace Nb, Minimise Al,Minimise N , Make V additions to Nb steels
Consider Ti additions

5.2.2 Machine Operation

Mould Heat Transfer
Thermal stresses and surface structure can be influenced by heat transfer in the mould. It is important to consider the type of mould powder used, particularly viscosity, and ensure good and uniform fluxing.

Mould Oscillation

The importance of oscillation marks to transverse crack formation has already been mentioned, and the depth of these oscillation marks can be reduced by increasing the mould oscillation frequency and/or decreasing the stroke to reduce the heal time. Increasing the oscillation frequency has been shown to reduce the incidence of transverse cracking.Deep, irregular oscillation marks may also be formed due to poor mould level control and other factors.

Secondary Cooling
The secondary cooling strategy is very important to minimise transverse cracking. In the previous sections, it has been shown that there is a wide ductility trough associated with microalloyed steels, and if slab straightening is carried out within this ductility trough, transverse cracking can result. If slab straightening is carried out at temperatures either above or below this temperature range, cracking should be minimised. Both these different cooling strategies (“soft” cooling and “hard” cooling) have been used on various machines around the world, with some success in reducing cracking.11,12) When a “soft” cooling strategy is used, it is important to keep the entire cross section of the slab above the critical temperature,including the slab corners, which are typically colder than the broad face. This has encouraged the installation of devices to maintain high temperatures in the slab corner region in some plants. A steep temperature gradient through the slab thickness is also desirable using this cooling strategy, to minimise the penetration of surface cracks which may form in cold spots. For “hard” cooling strategies, it is important to maintain all cooling nozzles; blocked cooling nozzles may lead to localised regions of the slab having temperatures within the critical range. A “hard” cooling strategy may also lead to subsurface crack formation:11) the distance between these cracks and the slab surface must be such that they are not exposed during subsequent reheating operations. However, “hard” cooling practices may increase thermal stresses. It should be noted that the use of these cooling strategies requires a knowledge of the temperature range over which low ductility exists, and this temperature range may not necessarily correspond to those obtained in a hot tensile test, A variant on the use of different cooling strategies which has been used to minimise crack formation during the rolling of hot charged slabs is the use of slab quenching.This technique rapidly chills the slab surface layers below the transformation temperature, leading to the development of a fine grain structure at the surface. This fine grain structure then restricts the formation and propagation of cracks, which may have formed otherwise during
the subsequent rolling process.Non-uniform secondary cooling can promote thermal stresses, and hence lead to cracking.This requires good nozzle design and maintenance, and preferably the use of air-mist cooling.

Mechanical Stresses

Mechanical stresses can be introduced by poor alignment of the components of the machine, and from many other sources, but of most relevance to transverse cracking is the straightening operation. Straightening may be carried out over a single point, or by multipoint straightening. The effect of these two types of straightening operation on transverse cracking are not clear, but there are reasons to suppose that multi-point bending will not improve hot ductility: strain rate will be reduced, which will reduce hot ductility; more time will be allowed for dynamic precipitation; and at least for Nb containing steels, stress relaxation between each bending point seems to be minimal.

5.3 Repair of Slab/Final Product

If slab cracking cannot be avoided by any of the above methods, the only options to avoid losses in the final product are to repair the continuously cast product prior to rolling. In the most extreme case, slab repair could involve machine scarfing of top and bottom broad faces, removing several mm from each face, together with scarfing of the narrow faces. After machine scarfing, it may be necessary to inspect the slabs, and remove any remaining defects by hand scarfing.These slab repair operations are expensive and time consuming. A recent example from a European plant suggested that machine scarfing of top and bottom broad faces, followed by inspection and hand scarfing of any remaining defects added a cost of $600,000 to a production volume of 100,000 tonnes. This figure will of course vary from plant to plant, but is still likely to be a considerable sum. Nevertheless, this additional cost is preferable to having to reject the final product. In the above example, prior to the introduction of scarfing, costs due to a 10% rejection level in the final product were approximately $1,600,000.

6. CONCLUSIONS

1. The only type of defect in continuously cast products reported to be influenced by microalloying elements are transverse surface cracks.
2. Nb additions are reported to have a strong influence in promoting transverse cracks, but V additions at low levels of N do not lead to cracking. Ti additions in themselves do not appear to produce transverse cracks, and Ti additions to Nb steels reduce transverse cracking.
3. Transverse cracks can occur at the slab broad face, narrow face or corner, and can be several mm in depth. They follow austenite grain boundaries, and the fracture surfaces are covered with particles such as MnS and AlN.
4. Transverse cracks are believed to initiate in the mould, but propagate during the
straightening process.
5. Several regions of low ductility exist when steels are tested at high temperatures under conditions simulating those experienced during continuous casting. Nb, and to a much lesser extent V, extend the depth and width of this ductility trough. The ductility trough is associated with the precipitation of microalloy carbonitrides and the transformation from austenite to ferrite.
6. To minimise transverse cracking, careful attention should be given to the choice of steel composition: Nb should be minimised, and consideration given to replacing Nb with V and N.
7. Machine operating conditions should be optimised, and in particular the straightening temperature should be chosen to be outside the region of low ductility.