Study of fractography of ferritic ductile iron at different temperatures and loading conditions

This study characterizes the fractography of ferritic ductile iron under various loads, including tensile, fatigue and bending, and impact conditions. The results indicate that ductile fracture is the primary mechanism observed during tensile testing at room temperature. The fractography resulting from fatigue testing exhibits characteristics similar to cleavage fracture, and explains the formation of fatigue striations caused by the joint effects of dislocation slip and oxidation under crack tip stress. Under impact testing, the main fracture mechanism transitions from ductile to brittle with decreasing temperature. At high temperatures, fractography is mainly characterized by elliptical dimples with graphite nodules at the center that deform along the stress direction. In the ductile-brittle transition temperature range, a mixed fracture mechanism involving both dimples and cleavage patterns is observed. At low temperatures, the fracture mechanism is cleavage fracture, cleavage fracture is mainly caused by the deformation twin, inducing crack nucleation. These findings further validate D.O.Frenandino’s quantitative analysis method for determining the main crack propagation direction of ductile fracture and brittle fracture. By employing larger statistical datasets, it is shown that this method yields high accuracy in determining the main crack propagation direction of ferritic ductile iron, thereby promoting its application as a general method for impact fractography analysis.


Introduction
Ferritic ductile iron (FDI) is one of the ductile iron (DI), which has good ductility, low-temperature impact properties, and low production cost.FDI has been used to replace low-carbon steel, widely used in exhaust manifolds, spent fuel casks, and hollow rotor shafts of wind turbines and other components [1][2][3].
The application range of ductile iron components is gradually increasing, and the accident rate is also gradually increasing.Therefore, it is very important to analyze the failure of FDI parts, determine the fracture reasons, and put forward corresponding measures to ensure the safe operation of parts.Some characteristics of the fracture surface of metal materials can be used to determine the main fracture mechanism.There are four common failure modes of metal materials, which are cleavage fracture, ductile fracture, intergranular fracture, and fatigue [4,5].The mechanical properties and fracture mechanism of ductile iron at different conditions have mainly been investigated [6][7][8][9][10][11][12].In the in situ experiment, Di Cocco [13] observed the surface of tensile specimen by scanning electron microscope, and found that the main damage mechanism of FDI is the nucleation and propagation of graphite nodule cracks caused by the decrease of solubility of γ phase carbon ('Onion-like' mechanism).Martinez [14] analyzed that the working mode of FDI under tensile and bending load at room temperature is ductile tearing with microvoid aggregation.Iacoviello [15] analyzed the fatigue crack fractography of four different ductile iron, and the graphite elements are not only voids embedded in the matrix, but also attractors of fatigue cracks.Zhang [16] systematically studied the effect of Ni content on the microstructure and low temperature impact fracture behavior of FDI.Due to the optimal combination of grain boundary character distribution and dislocation density, FDI with Ni content of 0.75% has the best impact properties at 40 °C.Frenandino [17] proposed a method to determine the macroscopic direction of main crack propagation under ductile and brittle fracture modes for FDI under impact and bending loads, which has good approximation in identifying the main crack propagation direction.Extracting valid information from the fractography is difficult, which leads to a large number of missing key diagnostic elements in failure analysis.The root cause of fracture is also little.These reasons have become a limitation in the popularization and use of materials.
This study explores the influence of different types of loads (tensile, fatigue and bending, impact), loading rates, and test temperatures on the fractography of FDI, obtains the main fracture mechanism by characterizing and analyzing the fractography, and explores the reasons for the formation of these fracture characteristics.In addition, the relationship between the main crack propagation direction and fractography features under impact load is investigated, with a deeper understanding of the complex fracture process and broadening the fracture information of FDI.

Materials
The metallographic and elemental analysis samples used in this study were taken from the center of the FDI spent fuel cask which is the same location as the mechanical specimens.The composition of the samples was analyzed by SparkCCD 7000 spectrometer, and the results are listed in table 1.All samples are etched with 4% (Nital) by standard polishing and etching methods.The metallographic specimens of the etching samples are shown in figure 1. Acoording to the ISO 945-1:2019 [18], the nodularity, the graphite nodule size, nodule count, ferrite content,and ferritic grain size were determined by using the OLYMPUS optical microscope and Image Pro Plus software, as shown in table 2.

Mechanical testing
The specimens used for the mechanical properties test were taken from the center of the FDI spent fuel cask, as shown in figure 2 (a), and the orientation in the length direction of the specimens was parallel to the axis of the FDI spent fuel cask.j5 mm bar specimens were used for tensile testing [19](ASTM E8/E8M), 10 × 10 × 55 mm Charpy V-notch impact specimens were used for impact test [20] (ASTM E23), and the single-edge notch bending (SEN(B)) specimen were used for fracture toughness test [21] (ASTM E1820), with specific dimensions  The tensile tests were carried out using the universal testing machine of CMT4204(20kN), and the strain rate and the test temperature were 0.00025 s −1 and 20 °C.
The impact tests were performed by the pendulum impact machine of WANCE-PIT752H, the impact velocity of the pendulum was 5.24 m s −1 , the maximum impact energy was 750 J, and the test temperatures were −80 °C,−60 °C,−40 °C,−20 °C, 0 °C, 20 °C, 40 °C, 60 °C and 80 °C.
The fatigue pre-cracks of SEN(B) specimens were performed using the INSTRON 8802-100KN machine.The fatigue pre-crack parameters were f=0.1 Hz, R=0.1, the maximum stress intensity factor K fmax =26 MPa•m 1/2 , and the number of cycles were 10000∼20000.In the second stage, crack propagation is induced by displacementcontrolled loading, and the loading rate of the indenter is 1.0 mm/min.Fracture toughness tests were carried out with SEN(B) specimens at −40 °C.

Results and discussion
3.1.Tensile fractography and mechanism of fracture analysis Figure 3 depicts the fractography resulting from slow monotonic tensile loading at 20 °C. Figure 3(a) reveals an irregular rugged topography, while figure 3(b) highlights a prominent ductile fracture mode, which is commonly referred to as tearing and microvoid coalescence.In figure 3(c), the fractography primarily consists of equiaxed round dimples centered on graphite nodules, with significant plastic deformation apparent in the ferrite matrix around these nodules.The behavior of the ferrite matrix can be attributed to the coalescence of growing microvoids resulting from the interaction between graphite nodules and other defects in the ferrite matrix, as well as microvoids in the final solidification zone.In figure 3(d), sharp boundaries and lips are observed between the ferrite matrix and graphite nodules, which are caused by significant plastic deformation in the final stage of coalescence, just prior to fracture.

Fatigue and bending fractography and crack propagation analysis
The fractography resulting from the J IC test reveals two distinct zones.Zone 1 corresponds to the stable crack propagation zone of the SEN(B) specimen under monotonic bending load, while Zone 2 corresponds to the fatigue pre-crack zone.Fatigue pre-crack was initiated after machining and sawing the notch, and cyclic loading was applied using three-point bending equipment in accordance with the ASTM E1820 standard for fatigue precrack [21].
Figure 4(a) presents the macroscopic fractography of the SEN (B) specimen.At −40 °C, the fracture surface of the fatigue pre-crack zone (Zone 2) appears brighter than that of the stable crack propagation zone (Zone 1), and the boundary between these two zones is clearly visible,as shown in figure 4(b).Image pro plus and PS were used to mark and calculate the perimeter and radius of the graphite nodule on the fractography of zone1 and zone2.According to the table 3, the radius and perimeter of the graphite nodule at the fractography under bending load were larger, the ferrite matrix around the graphite nodule produces greater plastic deformation, and the fracture surface is more rough.Oval dimples are observed around each graphite nodule, as shown in  The crack propagation within Zone 2 is characterized by a river pattern that bears a resemblance to cleavage characteristics, as shown in figure 5(a).Nonetheless, upon closer observation, it is not indicative of a cleavage fracture.The fractography in Zone 2 exhibits step-shaped striations and facets of varying sizes, shapes, and orientations.Fatigue striations emerge during stage II (Paris stage) of the fatigue crack propagation, where the plastic zone size at the crack tip increases and spans multiple grains.The crack propagation alternates between two slip systems, and two 'wave' fatigue striations appear in the fractography in different directions, as shown in figure 5(b).The direction of crack propagation (white arrow) is often perpendicular to the fatigue striations.Figures 5(b), (c), and (d) list several fatigue striations exhibiting varying orientations within Zone 2, with an average distance of 0.54 ± 0.02 μm between fatigue striations.
Neumann [22] proposed a model which has a comprehensive explanation for the fatigue striations observed on the ferrite matrix of the fatigue fractography.In this model, fatigue striations are primarily attributed to the slip of dislocations at the crack tip.However, slip may not occur consistently at the fatigue crack tip due to the influence of crystal structure and microscopic defects present within the crystal.
The crack tip represents a region of stress concentration within the specimen.Elastic deformation takes place in other regions of the specimen, whereas plastic deformation occurs exclusively at the crack tip, where a large number of dislocation slips exist.During the cyclic load increase process, tensile stress emerges near the crack tip, causing the slip surface adjacent to the crack tip to alternate motion and resulting in crack tip stretching.As the slip progresses, the crack tip becomes blunt, as shown in stage A of figure 6. Upon decreasing the fatigue load, the residual stress accumulated during the rising stage of the fatigue load is released, thereby placing the crack tip in a compressive stress state.Despite some slip surfaces being oxidized in the prior slip process, other slip surfaces may slip in the opposite direction under the influence of compressive stress, restoring the sharpness of the crack tip, as shown in stage B of figure 6.Under the influence of cyclic fatigue stress, the crack tip undergoes this process repeatedly, resulting in continuous crack growth.

Impact fractography and crack initiation and propagation analysis
The impact test was performed under rapid monotonic loading, and the fractography was analyzed to explore the fracture mechanisms at different test temperatures using SEM.The impact data for FDI at various temperatures are presented in table 4.

3.3.1.
Fractography and crack initiation and propagation analysis at high temperatures (80 °C,60 °C,40 °C,20 °C) Figure 7 illustrates the impact fractography obtained from temperatures ranging from 80 °C to 20 °C.As the test temperature decreases, the macroscopic morphology of the impact specimen gradually transitions from gray to bright, with a corresponding downward trend in the impact toughness value.SEM images of the impact fractography reveal that the fracture mechanism within the test temperature range is a primarily ductile fracture, with graphite nodules serving as the center and forming dimples that correspond to each graphite nodule.Additionally, some dimples without graphite nodules are observed, indicating that some graphite nodules detached during impact.Within the area surrounding the graphite nodule, secondary cracks are observed, as shown in figure 7(c).As the test temperature decreases, the depth of the dimples and the number of graphite nodules exposed on the fracture surface decrease.There are smooth cleavage surfaces and river patterns emerge at 20 °C, indicating the onset of brittle fracture.
Taking the impact fracture at 60 °C as an example, Stress concentration occurs at the V-notch, and voids are observed around the graphite nodules near the notch, as shown in figure 8(a).These voids are created due to debonding between graphite nodules and the matrix, as well as the plastic deformation of the matrix under external load, and they function as stress concentrators.The void acts as the source of the crack, and the cracks continue to propagate into ferrite grains according to the direction of graphite nodular strain.On the void wall around the graphite nodules, 'snake slip' features are observed, as shown in figure 8(c).This phenomenon is the result of multiple slip systems in ferritic grains with different orientations sliding and intersecting each other.Numerous deformed graphite nodules and voids are observed at 60 °C, and smaller dimples exist on the ferritic tearing ridge surrounding the graphite nodules, as shown in figure 8(d).Dislocation loops tend to accumulate around graphite nodules.When subjected to type I stress, these dislocations will begin to move, and cause the dislocation loops to migrate towards the graphite nodules.As the elastic strain energy accumulated by the dislocation loop becomes sufficient to overcome the interface bonding force between the graphite nodules and the matrix, a new surface will form, leading to the creation of microvoids and the detachment of the interface between the graphite nodules and the ferrite matrix.The voids surrounding the graphite nodules will expand and grow into larger voids due to continued dislocation slip, as shown in figure 9(a).
Microcracks formed near the stress concentration notch tend to preferentially develop into the main crack.Meanwhile, some microcracks at the front of the main crack initiate and connect, expanding into larger microcracks.These microcracks will connect with the front of the main crack in a rapid jump mode along a path with low energy consumption and low resistance.All these processes are facilitated through dislocation slip, as shown in figure 9(b).Even if the cracks are formed, the concentrated stress cannot be entirely released, which leads to the fracture of the slip bands due to the excessive stress within the matrix.The boundaries of the graphite nodules are connected to each other until they ultimately break, as shown in figure 9(c).

3.3.2.
Fractography and crack propagation analysis at ductile and brittle transition temperature range(0 °C,−20 °C) Figure 10(a) is the impact fractography at 0 °C.Elliptical dimples have ductile fracture features and graphite nodules have undergone deformation along the direction of crack propagation.Additionally, a significant number of cleavage planes and river patterns can be observed, and the proportion of cleavage fracture increases.The impact energy obtained at −20 °C is approximately half of that obtained at high temperatures (20 °C∼80 °C).The proportion of cleavage fracture and ductile fracture is equal, indicating that −20 °C is the    ductile-brittle transition temperature.As compared to the fractography at 0 °C, the depth of dimples around the deformed graphite nodules decreases, as shown in figure 10.
In the case of impact fractography at −20 °C, deformation twins are observed on the smooth cleavage plane, and twinning occurs in the form of twin shear when dislocation motion is impeded.On the cleavage plane where twins are present, a cleavage tongue is observed, as shown in figure 11(b).Furthermore, the cleavage section of the fracture displays river patterns, which are formed by a series of cleavage steps along the direction of crack propagation, as shown in figure 11(c).

Fractography and crack initiation analysis at low temperatures(−40 °C,−60 °C,−80 °C)
The fracture mechanism is primarily cleavage fracture at −40 °C, with only a few instances of ductile fracture, as shown in figure 12(a 2 ).The morphology of the cleavage surface indicates cracks propagate through grain boundaries at −60 °C and −80 °C.The cleavage fracture zone of FDI can reflect the grain size, where grain boundaries with different orientations impede the propagation of cleavage cracks, leading to the formation of numerous irregular steps.The fractography is dominated by smooth cleavage surfaces and some broken graphite nodules at −60 °C and −80 °C, as shown in figure 12(b 2 ) and (c 2 ).
Taking the impact fractography at −60 °C as an example, when the crack front passes through the small angle grain boundary, the river pattern simply changes direction and continues to flow in the adjacent grains, as shown in Zone A of figure 13(a).This phenomenon can be attributed to the formation of small angle grain boundaries, which is the vertical arrangement of edge dislocations of the same signs, and they do not alter the cleavage steps when cracks intersect with them.
As shown in Zone A of figure 13(a), cleavage steps can be also observed on the cleavage plane on the right side of the small angle grain boundary, which is considered to be the result of the interaction between crack and screw dislocation during crack propagation.When the crack front propagates along the cleavage plane, it encounters a screw dislocation b s with cleavage planes perpendicular to each other.After the crack passes through the dislocation b s , a step with the same size as the b s of the dislocation will be generated.The orientation of the step varies with the direction of crack propagation, as shown in figure 13(b).On the macroscopic scale, the cleavage step can be regarded as the accumulation of several small steps.The formation of steps requires a certain amount of energy, leading to the merging of river patterns and the convergence of tributaries towards the mainstream.The direction of river propagation is aligned with crack propagation, and the fracture initiation area can be found upstream.The smallest observable height of a river pattern step is 0.5 μm in figure 13(a).
Deformation twins can be also observed on the smooth cleavage surface of the impact fractography at −60 °C, as shown in Zone B of figure 13(a).The twins exhibit mutually perpendicular cleavage tongues, which are caused by cleavage cracks turning into initial twins, already formed twins, or cracks on the interface between twins and matrix during twin formation.As shown in figure 13(c), the cleavage plane of the ferrite matrix is {100} crystallographic plane family, while the twin interface is {112} crystallographic plane family.When the crack propagates on the {100} crystallographic plane and meets the {112} twin interface, the direction of crack propagation changes from 〈110〉 to 〈111〉 along the {112} twin interface until point C, where it then breaks along CD.Concurrently, the main crack also propagates along the DE direction from both sides of the twin, forming a cleavage tongue.The angle between the (100) crystallographic plane and the (112) crystallographic plane is 35°1 6′, thus cleavage tongues in FDI often occur at the position of 35°16′ with respect to the cleavage plane.Barik [23] prove the theory to explain the formation of the cleavage tongue of ferrite iron.
Mutual vertical twins are closely related to mutual vertical cleavage tongues.The  twin plane {112}, as shown in figure 14(a).The cleavage fracture mechanism of FDI at low temperatures is mainly 'crack initiation caused by vertical and crossed twins', crack initiation caused by twins is the primary cause of cleavage fracture [5].Zhang [24] derived that the critical condition for cleavage fracture is determined by the critical size of the deformation twins.Based on Griffith criterion, the critical twin size P c can be expressed as follows:  3.4.Relationship between main crack propagation direction and the fractography features under impact load 3.4.1.Brittle failure method As described in section 3.3, brittle cleavage fracture is the dominant failure mechanism when the samples are impacted at a lower temperature(<−40 °C), and the fractography exhibits numerous river patterns that are aligned with the direction of main crack propagation.Fernandino [25] proposed a quantitative analysis method for the fractography of DI that involves determining the direction of main crack propagation by analyzing the direction of river patterns.The river patterns with varying orientations eventually converge into a major river pattern, local river patterns are consistent with the direction of local crack propagation.
To mitigate the influence of the edge effect, the SEM images of the samples are taken from the center of the sample.In contrast to Fernandino's work [25], five images were randomly selected from the blue frame of the impact fractography shown in figure 15, which increased the sample counts and made the results more representative.The main crack propagation direction was set as the reference axis(0°), and the local crack propagation direction along the cleavage plane is shown using local propagation vectors (each of the local vectors meets at the point where two river patterns join).The angle α formed by the local vectors relative to the reference axis direction was then compared, as shown in figure 15.
The relative frequency distribution histogram of the local angle is drawn, as shown in figure 16.The relative frequency histogram has 18 intervals from −90°to 90°.Table 5 lists the results (mean, median, and mode) and counts of characteristic statistical datasets at different temperatures.In all cases, the mean (M) of the data set is between 1.8°and 2.5°, the median (Me) is between 3.1°and 5.6°, and the mode value (Mo) at different temperatures is −1.5°∼2.5°.Based on the results presented in figure 16, it can be observed that the relative frequency of the angle of 0°to 10°is the highest among the different temperatures examined.These findings  suggest that local crack propagation is inclined to align with the main crack propagation direction (0°).Furthermore, the method utilized in this study demonstrates a high accuracy in predicting the main crack propagation direction of unknown fractography.

Ductile failure method
As indicated in section 3.3, the fracture mechanism of the sample is predominantly characterized by ductile fracture when subjected to impact at higher temperatures (>20 °C), which results in the formation of numerous elliptical dimples around the graphite nodules.The plastic deformation occurring around the graphite nodules is influenced by both the loading conditions and the direction of crack propagation.Therefore, a novel approach is proposed for identifying the direction of crack propagation by measuring the plastic deformation of the matrix surrounding the graphite nodules.
Similar to the quantitative analysis of cleavage fracture at low temperatures, the present study characterizes dimples observed on the fractography of SEM images.Specifically, images from random locations of the fracture center of samples are selected.For each graphite nodule and its corresponding dimples in the SEM images, an equivalent ellipse is manually defined, and the most appropriate contour for deformation is then determined [26].Firstly, the angle β between the long axis of the ellipse and the main crack propagation direction is  measured to examine the relationship between the deformation direction of the dimple around the graphite nodule and the main crack propagation direction.Then, the Y-axis length parallel to the main crack propagation direction and the X-axis length perpendicular to the main crack propagation direction is measured, as shown in figure 17.The proposed method is time-consuming and requires a significant amount of patience.Observations indicate that plastic deformation becomes easier at higher temperatures, leading to a greater number of exposed graphite nodules and counted dimples on the impact fracture surface.Table 6 presents the measured Y/X ratios and angles (β) at four temperatures on the upper shelf.A relative frequency histogram of Y/X values and a cumulative distribution function of Y/X values are constructed at different temperatures, as shown in figure 18.
The statistical analysis of dimples with Y/X values obtained at different test temperatures reveals that these dimples tend to deform along the main stress direction, which corresponds to the main crack propagation direction.It is supported by the fact that dimples with Y/X values greater than 1 account for approximately 65% of all observed dimples across all fractography, as shown in table 5.Moreover, the average value of the included angle β is relatively small and consistent across different temperatures, ranging from −3.8°to 1.8°.In view of the reference axis(0°) being set as the direction of main crack propagation, the low β values demonstrate that this method has high accuracy in determining the direction of main crack propagation.Therefore, this method can be utilized to accurately identify the main crack propagation direction on unknown fractography.It is worth noting that the main crack propagation direction is independent of the setting of the reference axis, the main crack propagation direction is unique within a small error range.The experimental results of FDI obtained in this study are similar to those reported for FDI and austempered ductile iron [17,26], which further supports the use of this method as a general testing approach for determining the main crack propagation direction on fractography.

Conclusion
The fractography under tensile, fatigue and bending, and impact loads at different temperatures are characterized and compared, and the following conclusions can be obtained.
(1) Under slow monotonic loading (tensile), the fracture mechanism observed at room temperature is a ductile fracture, which is characterized by tearing and the coalescence of microvoids.
(2) Fatigue striations are observed in certain areas of the fractography under fatigue loading.The coarse slip model of fatigue explains the formation of fatigue striations, which accounts for the combined effects of dislocation slip and oxidation at the crack tip.
(3) The fracture mechanism is mainly the growth and coalescence of voids centered on graphite nodules under high-temperature impact.In the ductile-brittle transition temperature range, the fracture mechanism is a mixed mode of dimple and cleavage.Under low-temperature impact, the fracture mechanism is mainly cleavage of twins, deformation twins induce crack nucleation.
(4) The quantitative analysis methods proposed by Fernandino are used for brittle fracture and ductile fracture, and have high accuracy for determining the main crack propagation direction of fractography under impact load.

Figure 2 .
Figure 2. Specimens used for mechanical testing (a) Sampling plan (b) Specimen size.

Figure 6 .
Figure 6.Mechanism of fatigue striation propagation of Paris stage.

Figure 8 .
Figure 8. Fractography near V-notch at 60 °C (a) SEM image near V-notch (b) Void around graphite-ferritic matrix (c) Slip line in dimple (d) Small dimples of tearing ridge.

Figure 9 .
Figure 9. Mechanism of ductile fracture (a) Mechanism of Voids growth (b) Mechanism of Void coalescence (c) Model for ductile fracture.
[100] line and [−110] line at the bottom of the vertical cleavage tongues represent the intersection of the main cleavage plane {100} and the

Figure 13 .
Figure 13.Fractography and mechanism of fracture features at −60 °C (a) Fractography (b) Formation mechanism of cleavage step (c) Formation mechanism of twins.

Figure 14 .
Figure 14.(a) Cleavage fracture mechanism of vertical and crossed twins (b) Deformation twins.

Figure 15 .
Figure 15.Method used for the analysis of brittle fracture of the direction of macroscopic crack propagation.

Figure 17 .
Figure 17.Method used for the analysis of ductile fracture of the direction of macroscopic crack propagation .
as shown in figure 2(b).The mechanical properties obtained under different conditions are the average value of at least three tests.

Table 3 .
Statistical results of graphite nodule of the SEN(B) sample.
SEN(B) sample areaMean perimeter of the graphite nodule/μmThe mode of the perimeter of the graphite nodule/μmThe mean radius of the graphite nodule/μmThe mode of the radius of the graphite nodule/μm Zone 1 (Bend)

Table 4 .
Impact data at different temperature.

Table 5 .
Datasets of cleavage planes obtained at different temperatures.