Heterogenous lamellar microstructure design to resist ductile-to-brittle transition of body-centered cubic structural metals

The ductile-to-brittle transition (DBT) is an important characteristic that restricts the use of body-centered cubic metals as structural materials at low temperatures. This study shows a design of heterogenous lamellar microstructure (HLM), which is implemented in a low-density steel to promote the occurrence of delamination parallel to the direction of the main crack propagation, and result in delamination toughening effect. When the Charpy impact test temperature was reduced from room temperature to −196°C, the impact absorbed energy displayed a linear decline. The design of HLM brings a new pathway for developing economical BCC-structured metal plates with reliable low-temperature toughness. GRAPHICAL ABSTRACT IMPACT STATEMENT This is the first time of utilizing ‘delamination toughening’ to fully suppress the ductile-to-brittle transition in a body-centered cubic metal plate.


Introduction
The pursuit of enhancing toughness continues to be of significant interest remains a perennial objective in the realm of structural materials research, given its profound implications for the safety, design and performance of construction and equipment [1][2][3].Metals with bodycentered cubic (BCC) structure, which a majority of steels exhibit and refractory metals, exhibit a distinctive phenomenon known as the ductile-to-brittle transition (DBT) [3,4].The DBT phenomenon is characterized by the fact that as the temperature decreases, the materials abruptly shift from highly ductile fracture to almost completely brittle fracture, which results in the decrease of toughness within a small temperature range [4].The significance of DBT lies in the sudden transformation of materials into brittle fracture, introducing a high level of uncertainty in the fracture behavior.The median of the temperature range DBT occurs is referred to as the ductile-to-brittle transition temperature (DBTT).To avoid the occurrence of brittle fracture, materials susceptible to DBT must have a precise understanding of their DBTT and must adhere rigorously to usage above the DBTT.Significant effort has been made to reduce the DBTT of materials, enabling them to function effectively in low temperature environment [5][6][7][8][9][10].
In essence, DBT represents a competition between crack extension and dislocation source operation [11].Notably, studied have revealed that the ratio of screw dislocation velocity to edge dislocation velocity serves as an important factor influencing the DBT of BCC metals [12].A more intuitive way to understand DBT is through the Yoffee diagram, an idealized representation illustrating the relative likelihood of ductile or brittle fracture at the tip of a preexisting crack [13].As the intensity of the applied stress increases towards failure, the peak stress at the crack tip initially reaches one of the two thresholds: the effective yield stress which triggers significant plastic deformation, or the brittle fracture stress, leading to crack propagation in a brittle mode.The Yoffee diagram also proposes two approaches to mitigate DBT: increasing the brittle fracture stress or reducing the effective yield stress [3].The most efficient method for enhancing brittle fracture stress involves effective grain refinement, specifically, increasing the density of effective boundaries [13][14][15][16][17].
Another approach to suppress DBT is to lower the effective yield stress [18][19][20].An effective method to achieve this is by designing the microstructure so that it spontaneously delaminates in the stress field at the crack tip, effectively dividing itself into a laminate of thin sheets and reducing the effective stress at the crack tip [14,[21][22][23][24][25].An important instance of using delamination to enhance toughness was reported by Kimura et al. [26].Through a novel thermal-mechanical treatment, namely 'tempforming', they fabricated an 1800 MPa grade ultrahigh-strength steel bar with remarkable performance in low-temperature toughness.Within the temperature range of −60 to −20°C, this steel exhibits an inverse temperature dependence of toughness, where the Charpy v-notch impact energy greatly increases, and the samples do not completely fracture.The principle of this phenomenon lies in the impact process, where delamination perpendicular to the propagation direction of the main crack occurs, leading to the tilt and arrest of the main crack.It is worth noting that the material is barshaped.If the material is in the form of a plate and the notched direction is on the lateral side of plate, then the delamination planes are generally parallel to the rolling surface, i.e. parallel to main crack propagation direction, which cannot trigger the mechanism of crack arrest.However, the delamination parallel to main crack propagation direction can still play a role in improving toughness through another method.The mechanism involves delamination, i.e. dividing the sample into thin sheets and reducing the effective stress at the crack tip [24,27,28].Especially for pipeline steels, different studies noted that the additional plasticity introduced by delamination during the fracture process leads to an effective increase in toughness [29].Liu et al. found a more significant effect that fracture resistance can be greatly improved by activating delamination toughening coupled with transformation induced plasticity in an ultra-strong steel [30].These studies inspired us to consider and propose a new approach to resist DBT in BCC metals.
In the present study, a heterogenous lamellar micro structure (HLM) [30,31] design was applied to a lowdensity steel to promote delamination and suppress the occurrence of DBT.A series of Charpy impact tests were conducted as a function of temperature in the range of ambient temperature to −196°C to illustrate the change in toughness with temperature.The morphology of main crack and layered cracks were observed to analyze the toughening effects of delamination.The microstructure of experimental steel was also characterized to reveal the mechanism of delamination occurrence.

Materials and methods
An aluminum-bearing low density steel (Fe-1.5Mn-1.96Al-0.23C-0.98Cr,wt.%) was selected in this study because it has a relatively wide ferrite-austenite dual phase region.The experimental steel was melted in a 50 kg vacuum induction furnace.A HLM consisting of ferrite and tempered martensite was obtained by homogenizing 120 × 120× Length mm 3 ingots at 1200°C for 120 min, two-stage hot rolling (first rolling stage: 1100-1060°C and second rolling stage: 950-850°C) to 10 mm thick plate, quenching in water, annealing at 950°C for 60 min, and tempering at 750°C for 60 min.The yield strength of 496 MPa could be achieved in this steel, tensile strength was 641 MPa and the total elongation was 38.6%.A detailed study on mechanical properties and microstructure of experimental steel has been published previously [32].The experimental steel was subsequently machined to Charpy v-notched impact specimens along the rolling direction; the notch was placed in the lateral direction of plate.The DBT curve was obtained via a JBW-450H microcomputer-controlled oscillometer impact tester in accordance with the ISO 1481:2009 standard.The measurements were carried out from room temperature to −196°C at intervals of 20°C.The model of the cooling equipment was DWC-196Y, and the cooling medium was liquid nitrogen.At each test temperature the measurements were repeated three times.The morphology of fracture surface and layered cracks was observed with a ZEISS ULTRA-55 field emission scanning electron microscope (SEM).The microstructure was characterized with a Leica DM183 optical microscope.A TESCAN MAIA3 field emission SEM equipped with an Oxford Instrument Symmetry electron backscatter diffraction (EBSD) camera was employed to analyze and statistically characterize the boundaries of the experimental steels.The testing step size of EBSD was set to 0.2 µm.Additionally, the precipitates were analyzed with a TALOS F200X G2 transmission electron microscope (TEM).

Results and discussion
Figure 1 shows results of Charpy impact tests of experimental steel at different temperatures.Interestingly, the impact energy showed an almost linear decreasing trend (R 2 = 0.985) with decreasing temperature, which exhibits significant differences from the typical S-shaped BDT curve of BCC metals (Figure 1a) [10,33].The linear relationship of impact energy with testing temperature was only observed in steels with face-centered cubic (FCC) structure [34].To understand this phenomenon in detail, the load-displacement curve obtained from the instrumented impact test was studied.Several load-displacement curves at typical temperatures are presented in Figure 1b.Notably, as the test temperature decreases from 0 to −140°C, the load-displacement curve compresses along the displacement axis, with the curve's shape exhibiting minimal change.This is primarily manifested in the gradual decrease of load after reaching its peak, as opposed to a steep decline observed at −160°C.Remarkably, the maximum load (F max ) not only remained constant but increased with the decrease of test temperature.The impact energy was further dissected into two components: crack initiation energy (E i ) and crack propagation energy (E p ), as indicated in the schematic diagram in Figure 1b.The trends of crack initiation energy and crack propagation energy with testing temperature are presented in Figure 1c.Generally, the fraction of crack propagation energy was significantly greater than the crack initiation energy, particularly at higher test temperatures.As the temperature decreases, the crack propagation energy exhibits two plateaus with a decrease in temperature from −60 to −120°C.Concurrently, within this temperature range of crack initiation energy decline, crack propagation energy experiences a plateau.Consequently, the combined behavior of the two curves forms an approximately linear downward trend.
Figure 2 shows the fracture morphology of Charpy impact samples.Figure 2a provides a macroscopic view of the fracture surface at different temperatures.A prominent feature of all the layered cracks is that they are almost perpendicular to the fracture surface and specifically aligned parallel to the main crack propagation direction.The orientation relationship of main cracks, layered cracks and specimen is shown in Figure 2b.At higher test temperatures, the layered cracks are relatively short.Examining the microscopic view of the fracture surface reveals a high density of dimples, including some larger ones (Figure 2c), indicating significant plastic deformation before fracture.As the test temperature decreases, some of the layered cracks increase in length and penetrate the sample.Although the macroscopic deformation diminishes, some dimples, along with cleavage fracture, exists even at extremely low temperature test of −140°C, as shown in Figure 2d.This suggests the continued contribution of plastic deformation to impact energy at such a low temperature, and is attributed to the toughening effect of delamination.Delamination can effectively divide the specimen into multiple thin slices, resulting in reducing the stress triaxiality [35] at the crack tip, and thus enhancing the plastic work during crack propagation.
Figure 2e shows the cross-section of tested Charpy impact samples.Two main types of layered cracks are identified: the crack passes through ferrite grains and the crack propagates along the phase boundary of ferrite and tempered martensite.Both types of layered cracks could be found in all the samples, with no significant difference in their respective fraction.Additionally, a statistical analysis of the length of layered cracks (Figure 2f) revealed that the number, maximum length and the total length of layered cracks exhibit an initial increase followed by a decrease trend as temperature decreases, which indicates that layering behavior is initially promoted during the material embrittlement process, and the delamination can be inversely suppress material embrittlement.During this process, the DBT was greatly suppressed.When the temperature further decrease ( < 160°C), the delamination behavior can hardly be triggered, thus, the effect of delamination toughening also disappears accordingly.It worth noting that the strongest delamination appeared approximately at −60 to −120°C, which corresponds exactly to the plateau of crack propagation energy in Figure 1c.This confirms the influence of delamination on toughening during the crack propagation process.
Although the reason of delamination can be attributed to fact that the specimen withstands three-dimensional stress under impact loading, the principal stress value must be significantly greater than the lateral stress.Therefore, the material must possess weak or brittle bonding in the corresponding direction [36], i.e. plate thickness direction in present study.This is real characteristic of HLM design.Figure 3a displays the optical micrograph of the experimental steel, showing a microstructure composed of laminar ferrite (soft phase) and tempered martensite (hard phase).The above important observations are that the layered cracks always pass along the ferrite and tempered martensite boundaries or pass through the ferrite grains.The particles along the interface must be the main cause of embrittlement at the ferrite-martensite interface, as shown in Figure 3c.Energy spectrum analysis identifies these particles as Fe-Cr-C precipitates.And further TEM diffraction  analysis indicates that these precipitates are mainly of M3C structure (Figure 3e).Previous studies have indicated that such precipitates can weaken the grain boundary [36][37][38][39][40] and lead to delamination [41], explaining the formation of layered cracks along the ferrite-martensite interface, as illustrated in Figure 2e.Moreover, it can also be clearly observed that the layered cracks propagate within the ferrite layers, but seldom within the tempered martensite layers.Besides the role of texture, the distribution of large angle grain boundaries also plays a crucial role in promoting layered crack propagation along specific planes.An EBSD analysis allows for the characterization of different types of boundaries (misorientation angle > 5°) in the experimental steel.The boundaries within the martensite exhibit a much higher density than ferrite-ferrite and ferrite-martensite boundaries (Figure 3b), with a greater proportion of boundaries within martensite having large misorientation angles (Figure 3f).Large-angle grain boundaries are acknowledged for their significant inhibitory effect on brittle crack propagation [42,43].Consequently, the distribution of large angle grain boundaries within the present experimental steel promotes layered crack propagation along ferrite layers while limiting penetration into tempered martensite layers.
The HLM design provides toughening via another way involving delamination plasticizing mechanism, which differs from the delamination arresting crack mechanism proposed by Kimura et al. [43].As shown in Figure 4, when the main crack forms in the impact sample, the tip of main crack is subjected to three-dimensional stress.The anisotropic fracture of HLM results in prioritizing delamination fracture in the layers of ferrite, and the layered cracks are parallel to main crack propagation direction.On the one hand, the layered cracks provide additional absorbed energy during the impact fracture process, while on the other hand, and even more importantly, the sample splits into several sheets by layered cracks, which decreases the intensity of stress triaxiality of the main crack tip.A lower stress triaxiality results in greater plastic deformation of the sample during fracture [14], as the high stress triaxiality limits the ability of dislocations to move [44].Moreover, the prototype of steel with HLM studied in this research is low alloyed and completely composed of BCC phases.Consequently, this study obviously differs from existing top studies of 'delamination toughening' [30,43] in more than one aspect, including alloy systems, heterogeneous phase selections, microstructure control concepts and toughening mechanisms.This study also indicates utilizing HLM design to promote 'delamination toughening' is a new and unique approach to enhance low temperature toughness and restrict DBT in BCC metals with simpler compositional profiles.

Conclusions
In conclusion, a low-density steel with a heterogenous lamellar microstructure composed of ferrite and tempered martensite demonstrated a distinctive near-linear trend of impact energy with test temperature without a typical DBT behavior.This peculiar behavior can be ascribed to the occurrence of delamination parallel to the direction of the main crack propagation, which divides the impact sample into multiple thin sheets, thereby reducing the stress triaxiality at the crack tip and enhances plasticity.The utilization of a heterogenous lamellar microstructure design is envisaged as an innovative approach to alleviate DBT in BCC metals.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Figure 1 .
Figure 1.Charpy impact test results of the experimental steel from room temperature to −196°C, (a) curves of impact absorbed energy with different testing temperatures, (b) load-displacement curves at typical test temperatures and (c) curves of crack initiation energy and crack propagation energy with different test temperatures.

Figure 2 .
Figure 2. The Charpy impact samples of experimental steel.(a) the macroscopic view of the fracture surface of samples tested from 25 to −196°C, (b) Schematic diagram of main cracks and layered cracks (RD: rolling direction, ND: normal direction, TD: transverse direction), (c) and (d) microscopy of fracture surface of samples tested at 0 and −140°C, respectively, (e) macroscopic and microscopic observation of layered cracks on cross-section (M: martensitic, F: ferrite), (f) statistical results of layered crack length.

Figure 3 .
Figure 3. (a) Optical observation of microstructure of experimental steel.(b) Classification of boundaries with misorientation angle greater than 5°(blue lines: internal boundaries of martensite, green lines: boundaries between ferrite grains, red lines: boundaries between ferrite and martensite), (c)-(e) Transmission electron micrographs of precipitate morphology, bright-field TEM image and selected area electron diffraction in the experimental steel, (f) the misorientation distribution of different types of boundaries.