Conditions for the occurrence of acicular ferrite transformation in HSLA steels
- 2k Downloads
For the class of steels collectively known as high strength low alloy (HSLA), an acicular ferrite (AF) microstructure produces an excellent combination of strength and toughness. The conditions for the occurrence of the AF transformation are, however, still unclear, especially the effects of austenite deformation and continuous cooling. In this research, a commercial HSLA steel was used and subjected to deformation via plane strain compression with strains ranging from 0 to 0.5 and continuous cooling at rates between 5 and 50 °C s−1. Based on the results obtained from optical microscopy, scanning electron microscopy and electron backscattering diffraction mapping, the introduction of intragranular nucleation sites and the suppression of bainitic ferrite (BF) laths lengthening were identified as the two key requirements for the occurrence of AF transformation. Austenite deformation is critical to meet these two conditions as it introduces a high density of dislocations that act as intragranular nucleation sites and deformation substructures, which suppress the lengthening of BF laths through the mechanism of mechanical stabilisation of austenite. However, the suppression effect of austenite deformation is only observed under relatively slow cooling rates or high transformation temperatures, i.e., conditions where the driving force for advancing the transformation interface is not sufficient to overcome the austenite deformation substructures.
In HSLA steels, acicular ferrite (AF) is defined as a highly substructured, non-equiaxed ferrite that forms on continuous cooling . Despite the fact that the strengths of AF dominant microstructures are very close to those of BF dominant microstructures [2, 3, 4], these microstructures are named as AF separately because AF microstructures are relatively chaotic arranged and irregularly shaped in contrast to the highly organised and aligned bainitic ferrite (BF) microstructure which contains packets of parallel laths .
It should be pointed out that, besides HSLA steels, the term “acicular ferrite” has also been used to describe a particular microstructure in steel welds. In steel welds, AF laths nucleate on non-metallic inclusions which are point nucleation sties. In steel welds, it was found that AF laths emanate from these point nucleation sites and grow in many different directions , which is very different from the irregular grain shape and chaotic arrangement of AF in HSLA steels. Therefore, from the point of view of present authors, AF in steel welds and AF in HSLA steels are two different microstructures and the latter is the focus of this research.
AF dominant microstructures have a greater potential to improve the strength , toughness [5, 6], H2S resistance  and fatigue behaviour  compared to a ferrite-pearlite microstructure through grain refinement, increased dislocation density and easily controlled precipitation . Furthermore, the toughness associated with an AF microstructure is considerably higher than that of BF, due to a higher density of high angle grain boundaries (HAGBs), which are capable of arresting cracks or deflecting their propagation directions .
Given the good property combination intrinsic with an AF microstructure, it is not surprising that HSLA steels with AF dominant microstructures have already been successfully applied in several major pipeline projects [10, 11]. However, some aspects about the AF transformation are still unclear, especially as related to its formation conditions, including the effects of austenite deformation and subsequent continuous cooling. For example, increasing austenite deformation [4, 9, 12, 13] has been shown to increase the AF fraction, while ‘AF-like’ microstructures are produced when continuously cooled without austenite deformation [14, 15, 16, 17]. These observations are contradicted from research [18, 19, 20] which indicates that even after significant amounts of austenite deformation, the transformation product still consists of parallel BF laths and the typical AF microstructure is absent.
As for the influence of cooling rates, their effect on AF transformation is also under debate. For example, AF was found in the final microstructure with cooling rates below 10 °C s−1 , but BF has been shown to be dominant at a slower cooling rate (5 °C s−1) , and as the cooling rate rises to 35 °C s−1, the fraction of intragranularly nucleated AF laths is increased, due to the decreased transformation onset temperature and thus the increased thermodynamic driving force, which is supposed to promote the intragranular nucleation as expected by nucleation theory .
In order to make clear the conditions for the occurrence of AF in HSLA steels, in this research, a commercial HSLA steel which has been successfully utilised to build a API X80 pipeline project [22, 23] was utilised and subjected to austenite deformation and accelerated cooling with different deformation and cooling parameters. The transformed microstructures were characterised by optical microscopy (OM), scanning electron microscopy (SEM) and electron back-scattered diffraction (EBSD). Based on the characteristics of microstructural evolution, the conditions for AF transformation were proposed and the underlying mechanisms were elucidated.
To simulate the industrial rolling process, plane strain compression (PSC) tests were adopted and conducted on a HSLA steel with the following composition in wt%: 0.045C-1.43Mn-0.14Si-0.09Nb-0.21Cr-0.12Ni-0.21Cu-0.01Ti with a balance of Fe.
In order to fully dissolve Nb carbide or carbonitride precipitates and shorten the preheating time during PSC tests, all specimens were heated to 1250 °C, held for 30 min with argon atmosphere protection, and water quenched to room temperature before PSC testing.
Samples with different cooling rates and strain2 = 0 were used to investigate the effect of cooling rates without austenite deformation, and samples with different amounts of strain2 (0–0.5) and the same cooling rate (10 °C s−1) were used to investigate the effect of austenite deformation. Finally, samples with different cooling rate and the same strain2 = 0.5 were utilised to investigate the effect of cooling rates after a given amount of austenite deformation.
Specimens for metallographic observation were cut on the rolling direction–normal direction (RD-ND) plane and ground and polished carefully using standard procedures. A 2% nital etchant was used to reveal the transformed microstructures and saturated aqueous picric acid solution was used to reveal the prior-austenite grain boundaries (PAGBs). OM and SEM observations were carried out on Nikon Eclipse LV150 and JEOL JSM 6400, respectively.
To prepare specimens for EBSD mapping, after grinding and polishing, the specimens were then polished with a colloidal silica suspension for an additional 2 min. EBSD data acquisition was performed on the RD-ND plane of specimens in a FEI Sirion Field Emission Gun Scanning Electron Microscope with a step size of 0.2 μm and accelerating voltage of 20 kV. To reduce the mis-indexing of phases in this complex microstructure, α iron was chosen as the only matching unit. Following a recommended method proposed in research , random noise points of the raw data were removed and non-indexed points were filled with the common orientations of their neighbours using HKL Channel 5 Tango software.
Besides microscopy, mechanical properties of the transformed products were characterised by Vickers microhardness with a 500 g applied load and 20 s dwelling time on a Mitutoyo hardness testing machine. Ten indentations were measured from each specimen.
Before the illustration of the transformed microstructure, it is necessary to describe the phase discrimination strategy used in this research. Firstly, phases transformed at different temperatures or with different transformation mechanisms have different defect densities and thus different microhardness. In research , microhardness values for different phases were provided for an ultra-low-carbon steel, HV140, HV175, HV220 and HV315 for polygonal ferrite (PF), quasi-polygonal ferrite (QF), BF and martensite, respectively. Since AF has similar strength to that of BF [2, 3, 4], the microhardness values of AF and BF should be comparable. So, there is a distinct difference between the microhardness values of PF/QF and AF/BF and Vickers microhardness values were used to determine whether the transformed microstructure is PF/QF or AF/BF dominant. Another way to determine whether a microstructure is PF/QF or AF/BF dominant depends on the distribution of disorientation angles. Since PF/QF grains usually do not have a near K-S OR with the prior-austenite grains they grow into, the disorientation angle distribution of PF/QF grains will be quite random and a high density of boundaries with disorientation angle between 21° and 47° exist [26, 27]. While for displacive transformations, variants following a near K-S OR within the same prior-austenite grain do not have disorientation angles between 21° and 47°, so for displacive transformation products, e.g. AF and BF, the frequencies of disorientation angles between 21° and 47° should be significantly lower than others as reported in research [4, 27]. Therefore, the distribution of disorientation angles could also be used to differentiate PF/QF and AF/BF dominant microstructures. Secondly, AF and BF can be readily discriminated from each other by their morphological characteristics, chaotically arranged and irregularly shaped for AF and parallel and aligned for BF. It is also found that AF microstructures usually have a high density of high angle grain boundaries (HAGBs) [27, 28], while BF microstructures transformed at high temperatures or with slow cooling rates are lack of HAGBs . In the end, it should be pointed out that the above-mentioned method only determines the dominant phases of the transformed microstructure and it is extremely difficult to determine whether a specific ferrite grain is QF, AF or granular bainite (GB) by its morphology [28, 30]. If a grain-level phase identification is required for HSLA steels, EBSD-based method can be resorted to [27, 28].
Additionally, throughout all the micrographs, fine martensite/retained austenite (M/A) constituents can be found within the matrix, dark under OM observation and white under SEM observation . The formation of this microphase can be attributed to the partitioning of carbon from ferrite laths to the surrounding austenite during transformation . The shape of the M/A constituents also changes with the increased strain2, from elongated rods in Fig. 4a, b to small equiaxed particles in Fig. 4c, d.
Microstructure under SEM
Since BF dominates in the microstructures cooled at 5–50 °C s−1 without austenite deformation, and the transition between AF and BF mainly happens with austenite deformation, the following microstructure characterisation in this research mainly focuses on the samples with austenite deformation.
Following noise reduction, a small area of each data set was used to plot an inverse pole figure (IPF) coloured orientation map and a corresponding boundary map. For a statistical analysis of the boundary interception length, each whole data set was used.
Although refined AF microstructures are dominant in Fig. 8c, d, some coarse packets identified by black arrows can still be seen, indicating the existence of BF laths, which corresponds well with the microstructures shown in Fig. 6c, d.
According to the microstructures in Figs. 8 and 9, the differences in austenite deformation and cooling rates leads to the changes in microstructure refinement. Disorientation threshold values of 4° and 15° are typical threshold values to define the grain size for strengthening and toughening, respectively [34, 35]. Low-angle disorientation (4°) grain size is one of the factors that controls the yield strength of steel as boundaries with disorientation angles above this value are expected to be able to oppose dislocation movement, while HAGBs with disorientation angles higher than 15° provide effective barriers to cleavage fracture [34, 36].
In Fig. 10a, the grain sizes following both disorientation threshold values decreases with the rise of strain2 and the grain size range is very similar to which found in Ref. . Furthermore, the differences between samples with strain2 = 0.1–0.5 are statistically significant as there are no overlaps of their 95% CL error bars. The error bars also become shorter as strain2 increases, which means the grain size distribution becomes more centralised. Although a monotonic relationship exists in Fig. 10a, it is more complicated for samples with different cooling rates. In Fig. 10b, grain sizes following both disorientation threshold values decreases with cooling rate increasing from 5 to 10 °C s−1. However, the mean linear intercept lengths defined by 15˚ increases with cooling rate rising from 10 to 50 °C s−1, while the intercept length defined by 4° first increases and then decreases. The changes between samples with cooling rates of 10 and 20 °C s−1 are not statistically significant because of the large overlaps.
Acicular ferrite transformation mechanism
The transformation mechanism of AF is still controversial. By different researchers, it was suggested as being similar to various common microstructures in steels, including granular bainite (GB) , quasi-polygonal ferrite (QF) [16, 38, 39], Widmanstätten ferrite (WF)  and bainitic ferrite (BF) [2, 3]. Differently, a mixture of these microstructures was regarded as AF in research [35, 41].
However, in our opinion, AF should be classified as a type of bainitic transformation product. This judgement is supported by many characteristics associated with the AF transformation and microstructure. For example, in an isothermal holding test , a transition from BF to AF occurs, which means that AF and BF transformations can proceed under the same thermodynamic conditions. Furthermore, when the AF transformation is nearly finished after long isothermal holding at 600 °C, the final volume fraction of AF was only 51% , corresponding well to the incomplete transformation phenomenon of the bainitic transformation . As for the microstructure characteristics, the lack of boundaries with disorientation angles between 21° and 47° in the AF microstructures [4, 27] suggests that AF holds a near Kurdjumov–Sachs (K–S) orientation relationship with the parent austenite, which means that the AF transformation is displacive. Additionally, it was found in many TEM observations that AF laths contain a higher density of dislocations than those associated with reconstructive transformation products [12, 42, 43, 44], and the AF microstructure consists of several parallel subunits with a size less than 1 μm [12, 42] resembling the sheaf morphology of bainitic ferrite . In summary, AF shows similar transformation behaviours and microstructure characteristics to those of bainite; therefore, it is reasonable to say that the AF transformation mechanism is in fact bainitic, which is also agreed by many other researchers [2, 3, 12, 44, 45, 46].
Transition from BF to AF
Although the AF transformation mechanism is bainitic, AF has a quite different morphology compared with that of BF. The formation of AF can be explained by the existence of intragranular nucleation sites introduced by the deformation of the austenite.
The nucleation process of bainite relates to the spontaneous dissociation of specific dislocations in the parent austenite , which was originally proposed for martensite nucleation by Olson and Cohen . During nucleation, existing dislocations dissociate on the closest packed planes, forming groups of stacking faults, and these stacking fault groups become unstable and undergo a spontaneous transformation during cooling. The dislocations involved are glissile, so that this mechanism does not require the diffusion of lattice atoms, which corresponds well to the low transformation temperature of bainite and martensite at which thermal activation is in short supply . Despite the similar nucleation mechanism between martensite and bainite, carbon partitioning is required during bainite nucleation to guarantee a reduction of free energy, due to the higher transformation temperature of bainite .
That martensite readily nucleates at densely stacked dislocation arrays has been widely accepted [48, 49, 50, 51], and the potency of this nucleation site was proposed to be determined by the dislocation quantity in the array and the nucleus requires a critical quantity of dislocations inside to be sufficiently potent . Therefore, besides austenite grain boundaries, where densely stacked dislocation arrays can be accommodated, intragranular dislocation arrays introduced by austenite deformation can also act as nucleation sites for the martensitic transformation. Furthermore, the strain field of a martensite lath may interact with the strain field of the intragranular nucleation site, a highly deformed region, so that the free energy barrier for the nucleation can decrease [53, 54]. More convincingly, martensite blocks were found developing from sub-grain boundaries introduced by austenite deformation through in situ laser scanning confocal microscopy technique .
Since the AF transformation mechanism is bainitic, the nucleation of AF should be similar to that of martensite with the additional requirement of carbon partitioning as illustrated above. It is reasonable to assume that densely stacked dislocation arrays introduced intragranularly by austenite deformation can act as nucleation sites for AF. Actually, the nucleation of AF has been found at various deformation substructures, like deformation bands [9, 13, 56, 57] and dislocation cell walls . Therefore, despite a wide range of cooling rates used (5–50 °C s−1) in this research, for samples undeformed (strain2 = 0) or slightly deformed (strain2 = 0.1), laths nucleated primarily on PAGBs where densely stacked dislocation arrays can be accommodated, and these laths finally developed into BF packets with a parallel morphology as shown in Figs. 3 and 4a, b. On the other hand, when the austenite is heavily deformed, a high density of dislocations were introduced into the austenite grains, and these dislocations acted as intragranular nucleation sites, leading to the intragranularly nucleated laths which finally developed into the AF dominant microstructures as shown in Figs. 4c, d and 5a–c. Therefore, as the strain2 increased from 0 to 0.5, a transition from BF to AF occurred, Fig. 4, highlighting the fact that austenite deformation is required for the AF transformation in HSLA steels.
AF laths nucleated intragranularly at deformation substructures do not develop into a parallel morphology on a large scale because the development of a parallel morphology is stifled by impingement between adjacently nucleated AF laths . Actually, several adjacent parallel AF laths form a packet on a small scale and different packets have different lengthening directions . AF packets will impinge with other packets formed at nearby nucleation sites, which prevents the formation of a general parallel morphology and brings about the irregular morphology of AF .
The transition from boundary nucleated BF to intragranularly nucleated AF leads to the increase in HAGB density, Fig. 8, and the reduction of effective grain size, Fig. 10a. During the bainite transformation, strong variant selection occurs and under relatively slow cooling rates or high transformation temperatures, variant selection mainly favours the variant pairs with low disorientation angles between them , resulting in the low HAGB density, Fig. 8e–f, and large effective grain size, Fig. 10a. However, for AF dominant microstructures, austenite deformation increases the intragranular nucleation sites, therefore, more laths nucleate directly at intragranular nucleation sites instead of sympathetically nucleating in the form of variant pairs. The formation of these intragranularly nucleated AF laths can weaken the variant selection mechanism , and thus more types of variants and thus boundaries can be generated in each austenite grain. Consequently, increasing the fraction of intragranularly nucleated AF by austenite deformation, higher density of HAGBs and smaller effective grain size can be obtained.
Transition from AF to BF
However, meeting the condition of introducing intragranular nucleation sites, the AF transformation does not necessarily happen as shown in Fig. 7f with a cooling rate of 50 °C s−1. For this sample, a high density of intragranular nucleation sites had already been introduced into austenite by deformation (strain2 = 0.5), but the transformation product still mainly consists of a typical BF dominant microstructure with a parallel morphology. In other words, intragranular nucleation sites are unable to induce the AF transformation at this high cooling rate. Therefore, there must be another condition needed for the occurrence of AF transformation.
In Fig. 12a where BF is the dominant microstructure, parallel BF laths forming packets with similar colours nucleated on PAGBs and extended across the whole grain or impinged with other BF packets. However, in Fig. 11a, parallel BF laths nucleated on PAGBs and extended into PA2 were stifled and the remaining austenite transformed into irregularly arranged AF laths, which corresponds very well to the characteristics observed in Fig. 7b, c. It is interesting to note that the more irregular the transformed microstructure, the more spreading of the α iron orientations in the pole figures. There are some possible reasons for the spread of α iron orientations. Under slow cooling rates, there may be some PF/QF grains formed in the final microstructure which do not necessarily follow the K–S orientation relationship with the parent austenite that they grow into. The spreading of α iron orientations may also reflect the austenite orientation spread induced by deformation and the existence of more types of variants promoted by intragranular nucleation.
Based on the above microstructure characteristics, in addition to the requirement of introducing intragranular nucleation sites, another condition for the occurrence of the AF transformation can be hypothesised as below. Since austenite grain boundaries are very potent nucleation sites, BF blocks firstly nucleate on the austenite grain boundaries during cooling. If these BF blocks develop across the whole austenite grain or impinges with other boundary nucleated BF blocks, a BF dominant microstructure is formed. On the other hand, if these boundary nucleated BF blocks are stifled prematurely by certain mechanisms and cannot take up the whole parent austenite grain, AF transformation at intragranular nucleation sites happens during the subsequent cooling process. Therefore, suppression of BF laths lengthening should be another requirement for the occurrence of AF transformation.
The possible mechanism suppressing the lengthening of BF laths is also related to the austenite deformation. Displacive transformations involve the coordinated movement of lattice atoms, and such motion of a glissile interface can be suppressed if it encounters strong defects such as austenite grain boundaries or deformation substructures . Less strong defects such as isolated dislocations also retard the progress of such product interfaces but can often be incorporated into the transformation product lattice. These mechanisms indicate that austenite deformation can retard or suppress the advance of displacive transformation interfaces. This retardation of displacive transformations by austenite plastic deformation is called mechanical stabilisation of austenite  and has been found for all of the displacive transformations in steels, including Widmanstätten ferrite [63, 64, 65], martensite [66, 67] and bainite [68, 69, 70].
Based on these driving stress results, the lowest suppression-effective dislocation densities in austenite were calculated using Eq. (1) with an austenite shear modulus of 8 × 1010 Pa, Poisson’s ratio 0.27 and Burgers vector 2.52 × 10−10 m . However, the dislocation density of deformed austenite is difficult to measure. Therefore, the dislocation densities were roughly converted to boundary disorientation angles according to the relationship bρ 1/2 ≈ θ where θ is the boundary disorientation angle.
The lowest suppression-effective dislocation densities and corresponding boundary disorientation angles at different temperatures are shown in Fig. 13b. As the temperature increases, for example, from 460 to 550 °C, the lowest suppression-effective boundary disorientation angle increases, from 0.9° to 2.0°. Regarding the practical sub-grain disorientation angles formed during austenite deformation, in research , PSC tests were conducted on a Ni–29.5%Fe–0.01%C–0.02%Mn model alloy with an initial grain size of ~ 50 μm, deformation temperature of 900 °C, strain rate of 1 s−1 and strains ranging from 0.2 to 0.7, which are very similar to the experimental conditions in present study. In research , the relationship between the average disorientation angle (θ av) and strain (ε) at 900 °C was described by a power law equation, θ av = 1.77ε 0.23. Using this relationship, with a strain of 0.5 in this research, the average sub-grain disorientation angle is 1.5°. As for the disorientation angle distribution, according to the scaling hypothesis [74, 75], the distribution functions of the disorientation angles normalised by the average disorientation angles are the same and insensitive to material and strain volume. Based on the distribution scaled shown in research , the highest disorientation angle density appears at ~ 0.7θ av , which equals 1.1° with θ av = 1.5°.
Therefore, in this study with strain2 = 0.5, lowering the transformation temperature by increasing the cooling rate, the lowest suppression-effective boundary disorientation angle increase, Fig. 13b, and the possibility of the BF transformation interface being stopped by deformation-induced sub-grain boundaries is reduced, resulting in the transition from AF to BF with cooling rate increasing from 10 to 50 °C s−1. In other words, with the same strain, it is the transformation temperature that determines whether deformation substructures can suppress BF laths lengthening and give room for AF to develop.
It is clear that under high cooling rates or low transformation temperatures, intragranular nucleation sites are unable to induce the AF transformation and the transformed microstructure is BF dominant. For BF microstructures, austenite deformation can result in a strong variant selection that BF variants with habit planes parallel to the active slip planes during austenite deformation are favoured . This kind of variant selection can be found in PA2 and PA3 in Fig. 12a. The traces of the lath boundaries on the RD-ND plane are around 32° away from the RD and based on the Schmid factor analysis for the main ideal texture component of austenite during plane strain compression , the traces of active slip planes in the RD-ND plane are 19.5°–45° away from the RD. Therefore, for deformed austenite, with the cooling rate rising from 10 to 50 °C s−1, due to the increased BF fraction and the strong BF variant selection resulted from the austenite deformation, the effective grain size become increasingly large as shown in Fig. 10b.
Although the BF laths nucleated on PAGBs need to be suppressed to give room for AF to develop, parallel BF laths as shown in Fig. 7b, c do not appear on each PAGB. This is because the deformation strain is not evenly distributed within each austenite grain. Normally the strain at the austenite grain boundaries is higher than that in the grain interior, and correspondingly, the disorientation angles of sub-grain boundaries near austenite grain boundary will be higher than that in the grain interior. Therefore, PAGB-nucleated BF laths are more likely to be stopped near the austenite grain boundaries without forming large BF packets.
Even in AF dominant microstructures, there are still some long BF laths across the whole parent austenite grains. This can be attributed to the unevenly distributed deformation strains between different parent austenite grains. The differences of deformation-induced substructures between austenite grains with different orientations were reported in research .
Although increasing the transformation temperature will promote the mechanical stabilisation effect of austenite deformation and suppress the BF laths more effectively, the maximum possible transformation volume of AF and BF will be reduced with the rise of the transformation temperature and finally reaches zero at the bainite transformation start temperature according to the incomplete transformation phenomenon of bainitic transformation . Therefore, the continuous cooling rate and the cooling interrupt temperature need to be selected carefully to stop the lengthening of BF laths and simultaneously get a high volume fraction of AF microstructure and continuous cooling is more suitable than isothermal holding to get a high volume fraction of AF.
PF transformation happened before both the AF and the BF can consume the nucleation sites on PAGBs and more importantly, the solid solution atoms partitioned from PF grains can stabilise their neighbouring austenite and retard bainite nucleation . Moreover, PF grains formed intragranularly can also suppress the BF laths. Therefore, the formation of PF grains is beneficial to AF transformation. But the existence of PF grains will undoubtedly lower the strength. Despite the beneficial effect on the AF transformation, the formation of PF grains is not a prerequisite for AF transformation. In research  cooling from austenite deformation temperature 850 °C to isothermal holding temperature 500–600 °C at a rate of 75 °C s−1, AF transformation still occurred without the aid of PF transformation. Similarly, in this research, it can be seen from Fig. 14 that with a cooling rate of 100 °C s−1 to reduce the interference of PF transformation, AF still is the dominant microstructure. In other words, PF/QF transformation is not a prerequisite for the AF transformation.
Without austenite deformation, the transformed microstructure gradually changes from granular bainite to lath bainite as the cooling rate increased from 5 to 50 °C s−1. Both of them nucleate on austenite grain boundaries and acicular ferrite can hardly be found in the transformed microstructures.
Under the same cooling rate of 10 °C s−1, bainitic ferrite (BF) dominates the microstructure when the austenite is undeformed or slightly deformed, and increasing the austenite deformation, the fraction of acicular ferrite (AF) becomes larger and a transition from BF to AF happens. In the meanwhile, the transformed microstructure is refined with the effective grain size reducing from 5.8 μm to 2.9 μm.
Subjected to the same austenite strain, the fraction of AF decreases and BF becomes more and more dominant with cooling rates increasing from 10 to 50 °C s−1, which means a transition from AF to BF occurs. At the same time, the transformed microstructure is coarsened with the effective grain size increasing from 2.9 to 3.5 μm.
Based on the microstructure evolution characteristics, the introduction of intragranular nucleation sites and the suppression of BF laths nucleated on austenite grain boundaries are found as two conditions should be fulfilled for the occurrence of acicular ferrite transformation. Increasing austenite strain can introduce a higher density of dislocations into the austenite, which can act as intragranular nucleation sites. Besides, the deformation substructures can also suppress the lengthening of BF laths through the mechanism of mechanical stabilisation of austenite.
Suppression of BF laths lengthening can only happen under relatively slow cooling rates or higher transformation temperatures. This is because a high cooling rate results in a low transformation temperature which increases the driving stress for transformation interfaces to overcome austenite deformation substructures and finally leads to a BF dominant microstructure.
The authors are grateful for the financial support provided by Companhia Brasileira de Metalurgia e Mineração (CBMM) and the China Scholarship Council (CSC).
Compliance with ethical standards
Conflict of interest
The authors declare that they have no conflict of interest.
- 1.Smith Y, Coldren A, Cryderman R (1972) Manganese–molybdenum–niobium acicular ferrite steels with high strength and toughness. In: Toward improved ductility and toughness. Climax Molybdenum Company (Japan) Ltd, Kyoto, pp 119–142Google Scholar
- 3.Cryderman R, Coldren A, Bell J, Grozier J (1969) Controlled-cooled structural steels modified with columbium, molybdenum and boron. ASM Trans Q 62(3):561–574Google Scholar
- 5.Bhadeshia HKDH (1992) Bainite in steels. Inst. of Metals, LondonGoogle Scholar
- 10.Janzen TS, Horner WN (1998) Alliance pipeline—a design shift in long distance gas transmission. In: Proceedings of the international pipeline conference. ASME, New YorkGoogle Scholar
- 11.Grimpe F, Meimeth S, Heckmann C, Liessem A, Gehrke A (2005) Development, production and application of heavy plates in grades up to X120. In: 1st international conference “Super-High Strength Steels”, Rome, Italy, pp 2–4Google Scholar
- 21.Shewmon PG (1969) Transformations in metals. McGraw-Hill, New YorkGoogle Scholar
- 22.Hulka K, Gray J (2001) High temperature processing of line pipe steels. In: Proceedings of the international symposium niobium, pp 587–612Google Scholar
- 23.Stalheim DG (2005) The use of high temperature processing (HTP) steel for high strength oil and gas transmission pipeline applications. Iron Steel 40(11):699–704Google Scholar
- 37.Fang H, Wang J, Yang Z, Li C, Bo X, Zheng Y (1999) Bainite transformation. Science Press, BeijingGoogle Scholar
- 38.Araki T, Kozasu I, Tankechi H, Shibata K, Enomoto M, Tamehiro H (1992) Atlas for bainitic microstructures. ISIJ, TokyoGoogle Scholar
- 39.Araki T, Shibata K, Enomoto M (1991) Reviewed concept on the microstructural identification and terminology of low carbon ferrous bainites. Mater Sci Forum 56:275–280Google Scholar
- 41.DeVito CPaA (1979) High-strength microalloyed pipe steels resistant to hydrogen-induced failures. In: MiCon 78: optimization of processing, properties, and service performance through microstructural control 672:53Google Scholar
- 42.Shang C, Hu L, Yang S, Wang X, Zhao Y, He X (2005) Forming and controlling of the acicular ferrite in low carbon microalloyed steel. Acta Metall Sin 41(5):471–476Google Scholar
- 53.Magee CL (1970) The nucleation of martensite. In: Aaronson HI, Zackay VF (eds) Phase Transformations. ASM, Metals Park, pp 115–156Google Scholar
- 54.Christian J (1969) The mechanism of phase transformations in crystalline solids. Institute of Metals, LondonGoogle Scholar
- 66.Leslie WC, Miller RL (1964) The stabilization of austenite by closely spaced boundaries. ASM Trans Q 57(4):972–979Google Scholar
- 67.Machlin ES, Cohen M (1951) Burst phenomenon in the martensitic transformation. Trans AIME 191(9):746–754Google Scholar
- 71.Honeycombe RWK (1969) Plastic deformation of metals. Cambridge University Press, CambridgeGoogle Scholar
- 76.Miyamoto G, Iwata N, Takayama N, Furuhara T (2012) Variant selection of lath martensite and bainite transformation in low carbon steel by ausforming. J Alloys Compd 577(1):528–532Google Scholar
- 77.Babu S, Bhadeshia H (1991) Mechanism of the transition from bainite to acicular ferrite. Mater Trans Jpn Inst Met 32(8):679–688Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.